SPRINGER BRIEFS IN APPLIED SCIENCES AND

TECHNOLOGY  COMPUTATIONAL INTELLIGENCE

Amit Joshi
Nilanjan Dey
K. C. Santosh   Editors

Intelligent
Systems and
Methods
to Combat
Covid-19

123
SpringerBriefs in Applied Sciences
and Technology

Computational Intelligence

Series Editor
Janusz Kacprzyk, Systems Research Institute, Polish Academy of Sciences,
Warsaw, Poland
SpringerBriefs in Computational Intelligence are a series of slim high-quality
publications encompassing the entire spectrum of Computational Intelligence.
Featuring compact volumes of 50 to 125 pages (approximately 20,000–45,000
words), Briefs are shorter than a conventional book but longer than a journal article.
Thus Briefs serve as timely, concise tools for students, researchers, and
professionals.

More information about this subseries at http://www.springer.com/series/10618

Amit Joshi Nilanjan Dey K. C. Santosh
• •

Editors

Intelligent Systems
and Methods to Combat
Covid-19

123
Editors
Amit Joshi Nilanjan Dey
Global Knowledge Research Foundation Department of Information Technology
Ahmedabad, Gujarat, India Techno International New Town
Kolkata, West Bengal, India
K. C. Santosh
Department of Computer Science
University of South Dakota
Vermillion, SD, USA

ISSN 2191-530X ISSN 2191-5318 (electronic)

SpringerBriefs in Applied Sciences and Technology
ISSN 2625-3704 ISSN 2625-3712 (electronic)
SpringerBriefs in Computational Intelligence
ISBN 978-981-15-6571-7 ISBN 978-981-15-6572-4 (eBook)
https://doi.org/10.1007/978-981-15-6572-4

© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature
Singapore Pte Ltd. 2020
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by similar
or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt from
the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the
authors or the editors give a warranty, expressed or implied, with respect to the material contained
herein or for any errors or omissions that may have been made. The publisher remains neutral with regard
to jurisdictional claims in published maps and institutional affiliations.

This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd.
The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721,
Singapore
Preface

Since December 2019, the novel coronavirus disease (COVID-19) pandemic

(originated from Wuhan, China) has caused serious threats to humanity across the
globe. Even though its mortality has not been determined yet, its spreading rate is
exponential (dated May 04, 2020).
In this book, several different intelligent systems and methods are proposed to
prevent further COVID-19 spreading. The intelligent systems and methods, in this
book, include artificial intelligence, machine learning, computer vision, signal
processing, pattern recognition, and robotics, just to name a few. To make it clear,
the book is not limited to detect/screen COVID-19-positive cases using one type of
data, such as radiological image data. They also include how data analytics-based
tools can help predict/project further pandemic, where socioeconomic impacts are
also taken into account. As COVID-19 has a lot of uncertainties that are primarily
due to multiple factors in collecting data, several issues/challenges, such as social
distancing (including opportunities), are discussed. As the robustness of AI-driven
tools depends on how big the data collection is, we also discuss unavoidable issues
in using apps, where data collection, its privacy, and security issues are crucial. To
combat COVID-19, not limited to just data analytics-based tools for projection
using time series data and pattern analysis tools for unusual pattern discovery
(anomaly detection) in image data, AI-enabled robotics and the possible usages are
addressed.
In a nutshell, this book primarily aims to provide a variety of approaches for a
large audience: Computing and Methodologies, Health Sciences to Business
Analytics.
In Chapter “Data Analytics: COVID-19 Prediction Using Multimodal Data”,
authors address the use of multimodal data for COVID-19 predictions, where data
analytics tools are of primary interest. Such predictions will be useful to accom-
modate future possible threats.
In Chapter “COVID-19 Apps: Privacy and Security Concerns”, author addresses
the usefulness of apps to track and predict the possible events/threats that come
from COVID-19. It considers the use of smartphone-based apps to track peoples’
movement, for example. While making such apps available, authors highlight a few

v
vi Preface

but major issues, such as privacy and security at the time they collect data. In brief,
the chapter explores different apps that were developed aiming to combat
COVID-19 and the related personal data privacy concerns that arise in the
post-coronavirus era.
In Chapter “Coronavirus Outbreak: Multi-Objective Prediction and Optimization”,
authors provide the insight into the mathematical perception about the coronavirus
outbreak and analyze them from prediction and optimization points of view. Very
specially, constraints, objectives, and measures are identified, and their association
and dependencies are analyzed. In their study, authors observed that the coronavirus
outbreak is the constrained multi-objective prediction and optimization problem,
and it is time-variant.
In Chapter “AI-Enabled Framework to Prevent COVID-19 from Further
Spreading”, authors discuss the use of AI-enabled (AIE) framework to prevent
further COVID-19 spreading. Authors mainly consider AI-enabled framework,
where algorithms such as machine learning (ML) and devices such as Global
Positioning System (GPS) can be used to design an innovative automation system.
Within this context, authors designed an experimental setup, where AIE framework
is designed by considering the scenario of urban and rural regions of the state of
Haryana, India.
In Chapter “Artificial Intelligence-Enabled Robotic Drones for COVID-19
Outbreak”, authors address the use of AI-enabled robotics to prevent further
COVID-19 outbreak spreading. Authors consider the use of drones and robots
equipped with IoT devices to collect data so that appropriate actions can be made.
With IoT, the power of AI and edge computing is considered, where authors
explain big data analytics and deployment of suitable tools/techniques across edge
and cloud. In brief, the chapter presents how current AI-enabled robotic drone
applications and network connectivity are used to improve their performance and
increase efficiency in various situations to fight COVID-19. It provides an in-depth
review of the literature in addition to possible future research in edge intelligence,
AI-enabled robotics, and intelligent networks.
In Chapter “Understanding and Analysis of Enhanced COVID-19 Chest X-Ray
Images”, authors address the use of radiological image data to check/observe
COVID-19 positive cases. In their study, authors consider chest X-ray images.
Authors clearly highlight the importance of image enhancement for AI-driven tools.
In Chapter “Deep Learning-Based COVID-19 Diagnosis and Trend Predictions”,
authors address the use of deep neural networks to predict and diagnose COVID-19
positive cases. In short, the primary motivation is to adopt deep learning
(DL) models to check whether it can support clinical decisions by using radiological
image data, i.e., computed tomography (CT) scans. Besides, interestingly, authors
extend the standard SEIR model, where the parameters are learned from DL model.
Such a modified SEIR model can effectively predict the transmission trend of
COVID-19 and can be used for short-term trend prediction of the epidemic.
In Chapter “COVID-19: Loose Ends”, authors address the possible loose ends
due to COVID-19. With limited data, technologies that are built upon artificial
intelligence (AI), robotics (R), and IoT (I), i.e., ARI, and their predictions/forecasts
Preface vii

could deviate from what they are. Based on the data available, it is possible to have
unwanted outcomes. In short, the chapter discusses the role of ARI and loose ends
in their implementation, where three major aspects are listed: AI algorithms in
analysis and prediction, the use of robotics in control and prevention of the pan-
demic, and the role of IoT for the patient monitoring system (PMS).
In Chapter “Social Distancing and Artificial Intelligence—Understanding the
Duality in the Times of COVID-19”, authors discuss the impact of social distancing
during COVID-19 and the proper use of AI tools and techniques. This chapter
studies the impact of AI-driven tools on a variety of data that are collected when
social distancing is considered.
In Chapter “Post-COVID-19 and Business Analytics”, considering post
COVID-19, authors highlighted the use of AI tools and techniques on a variety of
business models. They enlisted a few major global issues that can assist the poli-
cymakers to consider developing a business model to bounce back the world
economy after this crisis is over. In particular, the chapter enhances the under-
standing of stakeholders of business about the importance of application of the AI
in businesses in a volatile market in post-COVID-19 period.

Ahmedabad, India Amit Joshi

Kolkata, India Nilanjan Dey
Vermillion, USA K. C. Santosh
Contents

Data Analytics: COVID-19 Prediction Using Multimodal Data . . . . . . . . 1

Parikshit N. Mahalle, Nilesh P. Sable, Namita P. Mahalle,
and Gitanjali R. Shinde
COVID-19 Apps: Privacy and Security Concerns . . . . . . . . . . . . . . . . . . 11
Surekha Borra
Coronavirus Outbreak: Multi-Objective Prediction and Optimization . . . 19
Nileshsingh V. Thakur
AI-Enabled Framework to Prevent COVID-19
from Further Spreading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29
Dalip and Deepika
Artificial Intelligence-Enabled Robotic Drones
for COVID-19 Outbreak . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
Dharm Singh Jat and Charu Singh
Understanding and Analysis of Enhanced COVID-19 Chest
X-Ray Images . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
M. C. Hanumantharaju, V. N. Manjunath Aradhya,
and G. Hemantha Kumar
Deep Learning-Based COVID-19 Diagnosis and Trend Predictions . . . . . 57
Juanying Xie, Mingzhao Wang, and Ran Liu
COVID-19: Loose Ends . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65
Minakshi Pradeep Atre
Social Distancing and Artificial Intelligence—Understanding
the Duality in the Times of COVID-19 . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Deepti Gupta, Amit Mahajan, and Swati Gupta
Post-COVID-19 and Business Analytics . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Monomita Nandy and Suman Lodh

ix
About the Editors

Dr. Amit Joshi is currently the Director of Global Knowledge Research

Foundation, also an Entrepreneur & Researcher who has completed his Masters and
research in the areas of cloud computing and cryptography in medical imaging. Dr.
Joshi has an experience of around 10 years in academic and industry in prestigious
organizations. Dr. Joshi is an active member of ACM, IEEE, CSI, AMIE, IACSIT,
Singapore, IDES, ACEEE, NPA, and many other professional societies. Currently,
Dr. Joshi is the International Chair of InterYIT at International Federation of
Information Processing (IFIP, Austria). He has presented and published more than
50 papers in national and international journals/conferences of IEEE and ACM. Dr.
Joshi has also edited more than 40 books which are published by Springer, ACM,
and other reputed publishers. Dr. Joshi has also organized more than 50 national
and international conferences and programs in association with ACM, Springer, and
IEEE to name a few across different including India, UK, Europe, USA, Canada,
Thailand, Egypt, and many more.

Dr. Nilanjan Dey is an Assistant Professor at the Department of Information

Technology at Techno International New Town (Formerly known as Techno India
College of Technology), Kolkata, India. He is a Visiting Fellow at the University of
Reading, UK, and was an honorary Visiting Scientist at Global Biomedical
Technologies Inc., CA, USA (2012–2015). He was awarded his Ph.D. from
Jadavpur University in 2015. He has authored/edited more than 75 books with
Springer, Elsevier, Wiley, and CRC Press and published more than 300
peer-reviewed research papers. He is the Editor-in-Chief of the International Journal
of Ambient Computing and Intelligence, IGI Global. He is the Series Co-Editor of
Springer Tracts in Nature-Inspired Computing, Springer Nature; Series Co-Editor
of Advances in Ubiquitous Sensing Applications for Healthcare, Elsevier; and
Series Editor of Computational Intelligence in Engineering Problem Solving and
Intelligent Signal Processing and Data Analysis, CRC.

xi
xii About the Editors

Dr. K. C. Santosh (IEEE Senior Member) is the Chair and Associate Professor of
the Department of Computer Science at the University of South Dakota (USD).
Before joining USD, Dr. Santosh worked as a Research Fellow at the US National
Library of Medicine (NLM), National Institutes of Health (NIH). He worked as a
Postdoctoral Research Scientist at the LORIA Research Centre, Universite de
Lorraine, in direct collaboration with ITESOFT, France. He also served as a
Research Scientist at the INRIA Nancy Grand Est Research Centre, France, where
he has received his Ph.D. diploma in Computer Science. Dr. Santosh has published
more than 65 peer-reviewed research articles, 100 conference proceedings, and 7
book chapters. He has authored 4 books, and edited 3 books, 11 journal issues, and
4 conference proceedings. He is currently Editor-In-Chief of IJSIP and an Associate
Editor for several journals, such as International Journal of Machine Learning and
Cybernetics and IEEE Access. He has also chaired more than 10 international
conference events. His research projects have been funded by multiple agencies,
including the SDCRGP, Department of Education (DOE), and the National Science
Foundation (NSF). Dr. Santosh is the proud recipient of the Presidents Research
Excellence Award (USD, 2019) and an award from the Department of Health &
Human Services (2014).
Data Analytics: COVID-19 Prediction
Using Multimodal Data

Parikshit N. Mahalle, Nilesh P. Sable, Namita P. Mahalle,

and Gitanjali R. Shinde

Abstract Globally, there is massive uptake and explosion of data, and the challenge
is to address issues like scale, pace, velocity, variety, volume, and complexity of
this big data. Considering the recent epidemic in China, modeling of COVID-19
epidemic for cumulative number of infected cases using data available in early phase
was big challenge. Being COVID-19 pandemic during very short time span, it is
very important to analyze the trend of these spread and infected cases. This chapter
presents medical perspective of COVID-19 toward epidemiological triad and the
study of state of the art. The main aim of this chapter is to present different predictive
analytics techniques available for trend analysis, different models and algorithms, and
their comparison. Finally, this chapter concludes with the prediction of COVID-19
using Prophet algorithm indicating more faster spread in short term. These predictions
will be useful to government and healthcare communities to initiate appropriate
measures to control this outbreak in time.

Keywords COVID-19 · Predictive analytics · Machine learning · Prediction ·

Pandemic

P. N. Mahalle (B) · G. R. Shinde

Department of Computer Engineering, STES’S Smt. Kashibai Navale College of Engineering,
Pune, Maharashtra 411041, India
e-mail: [email protected]
G. R. Shinde
e-mail: [email protected]
N. P. Sable
Department of Computer Engineering, JSPM’s Imperial College of Engineering and Research,
Wagholi, Pune, Maharashtra 412207, India
e-mail: [email protected]
N. P. Mahalle
Pathology Department, Deenanath Mangeshkar Hospital and Research Center, Pune, Maharashtra
411004, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 1
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_1
2 P. N. Mahalle et al.

1 Introduction

A novel human corona virus was originated from China on December 2019, causing
a severe potentially fatal respiratory syndrome (COVID-19). The symptoms of
COVID-19 may or may not be visual in infected individual; hence, spread rate can
be faster as individual himself not aware of the infection [1]. Despite the continuous
efforts, the virus has managed to spread in most of the territories in the world; World
Health Organization (WHO) has announced COVID-19 as pandemic [2, 3]. Countries
throughout the world working cooperatively and openly with one another and coming
together as a united front in regards of efforts to bring this situation under control
using available information and communication technology (ICT). ICT needs to be
critically used to bring the situation under control, and predictive analytics can be
empowered using ICT services, tools, and applications. ICT can empower epidemi-
ological study to find the determinants, occurrence, and distribution of health and
disease in a defined population in terms of COVID-19.
As study [4] shows that 5–80% of people are tested positive for SARS-COV-2
may be asymptomatic. Predictive analytics using ICT plays an important role as
some asymptomatic cases will become symptomatic over a period of time. Artificial
intelligence (AI) can be beneficial tool to fight against pandemic like COVID-19. AI
models can be used for estimating and predicting spread rate; AI was also used in the
past pandemics like Zika-virus in 2015. Due to accurate and fast predictions, spread
rate can be minimized by taking necessary precautionary action before the time.
In nutshell, taking into consideration the current scenario a sad reality of the
COVID-19 pandemic is that many people have been infected. As per the daily situ-
ation report of WHO as on April 09, 2020, the COVID-19 transmission scenario
reports 1,436,198 confirmed cases with 85,522 deaths globally. The main contri-
bution of this chapter is comparison of various predictive analytics models and
algorithms and their applications to appropriate use cases. This study recommends
prophet machine learning algorithm for prediction due to various reasons which are
discussed in Sect. 4 of this chapter. Another contribution of this study is to present
various avenues to initiate high-quality research in biomedical science along with
integrative approach of predictive analytics and mathematical modeling to control
outbreak of any pandemic. In the view of above-mentioned-related issues, we should
also promote ecumenical and interfaith collaboration and peaceful coexistence during
the COVID-19 pandemic.
The main objectives of this chapter are as follows:
1. To understand medical perspective of COVID-19 toward epidemiological triad;
2. To analyze state of the art for different approaches and models used for forecasting
and prediction;
3. To understand various predictive analytics models and algorithm as well as their
comparison with respect to the use cases;
4. To study the performance of Prophet Algorithm for the prediction of COVID-19.
The remainder of this chapter is organized as below. Section 2 presents medical
perspective of COVID-19 in terms of its origin, most infected underline age group,
Data Analytics: COVID-19 Prediction Using Multimodal Data 3

and transmission. Section 3 discusses the analysis of different studies available in the
literature for predicting COVID-19. Section 4 presents various predictive analytics
models, algorithms, and their comparison. Finally, Sect. 5 concludes the chapter with
future outlook.

2 Medical Perspectives

The emergence of SARS-CoV-2 is confirmed from Wuhan’s Huanan Seafood market,

China, but specific animal source still remains uncertain. There is uncertainty
regarding origin of SARS-CoV-2 [5]. The situation with SARS-CoV-2 is developing
faster with the numbers of infected cases and death is increasing exponentially. The
unprecedented control measures taken have been effective in preventing spreading
of SARS-CoV-2. Still, there is continued rise in number of cases with infection of
SARS-CoV-2. Hence, it is essential to identify that the increase is due to infected
cases before lockdown, due to community transmission; hospital acquired infection
or spread within family. This should be determined experimentally, which may help
in revealing the actual numbers of infected patients and asymptomatic carriers.
Many studies have confirmed transmission among human of SARS-CoV-2 [6, 7],
but mechanism of transmission and pathogenesis in spreading in humans remains
to be fully explored? During transmission from human to human, whether the
pathogenicity of this virus is decreased with the increase in rate of transmission?
If the transmission of this virus is declined, the outbreak may eventually end. Never-
theless, if there is continuous and effective transmission, SARS-CoV-2 will develop
into an additional human coronavirus which is community acquired. It is difficult
to recognize and take further actions in patient with undefined and mild symptoms.
Studying a group of asymptomatic infected cases, and following them for their clin-
ical presentation, titers of antibody and viral loads, will help in understanding about
the number of subjects have symptoms later, whether viral shedding is actually less
robust and how frequently asymptomatic carriers can transmit virus further. A study
reported that asymptomatic infection is high (15.8%) in children under 10 years [8].
COVID-19 can be spread through respiratory droplets or due to close contact
with the infected patients. SARS-CoV-2 was isolated from fecal samples of infected
patients, which support the significance of feco-oral route in the transmission of
SARS-CoV-2, but a WHO-China joint commission report has denied this route of
transmission [9]. However, the likelihood of transmission of SARS-CoV-2 through
human waste, contaminated water, aerosols, and air conditioners cannot be under-
estimated; this may have happened in case of Diamond Princess cruise ship, where
there was widespread COVID-19 infection [10]. Still, to confirm the role of feco-oral
transmission of SARS-CoV-2, further studies will be required. Severe cases caused
by the infection of SARS-CoV-2 may develop neurological, respiratory, gastroin-
testinal, and hepatic complications leading to mortality. Many studies have reported
low sense of smell and taste as a manifestation of COVID-19 [11], but whether this
4 P. N. Mahalle et al.

is a unique feature of COVID-19 is uncertain. Till date, we do not have definite anti-
viral drug or vaccine for SARS-CoV-2. However, screening of new drug molecules
may prove beneficial in treating COVID-19, which will have therapeutic effect.
Globally, there has been lot of progress in monitoring and control of disease spread.
It is evident that there are lot of uncertainties and questions regarding transmission
mechanism, asymptomatic or subclinical patient’s virus shedding, origin of virus,
virus pathogenesis, treatment, symptoms, etc. This highlights the need of integrative
approach of predictive analytics and mathematical modeling with biological science,
which may help government to take appropriate measures and method for future
preparedness in fighting against this outbreak. In spite of rapid progression in research
toward this outbreak, most of the studies are unable to suggest and guide effective
measures to control this current situation. However, more high-quality research in
biomedical science along with predictive analytics and mathematical modeling is
warranted to manage public health crisis in short and long terms.

3 Related Works

As per the Italy official release, there are total 27,980 infected cases and 2158 deaths
of people who were positive of COVID-19 [12]. Due to rapid spread of COVID-
19, in short time many studies have been carried out for prediction of trend and its
impact. This section briefs about all such recent studies which are essentially related
to predictive analytics. Giordano et al. [12] propose epidemic prediction model that
compares infected density and the level of symptoms. Authors have proposed a
SIDARTHE Model which helps to redefine reproduction number, and simulation
results also show that the proposed model gives accurate results after comparing
the findings with real data on the COVID-19 epidemic in Italy. Bannister-Tyrrell
et al. [13] presented an interesting study to establish the correlation of temperature
and evidence of COVID-19 in Europe. Authors claim that the seasonal variation
essentially in the temperature greatly impacts the spread of COVID-19. Study states
that higher average temperature is potential candidate to limit the spread of COVID-
19. Russo et al. [14] presented a mechanism to find the first day of infections and
predictions of COVID-19 in Italy. Depending upon proposed work, authors are able
to estimate that the actual count of exposed cases of COVID-19. Volpert et al. [15]
nicely presented the effect of quarantine model on the spread of coronavirus infection
using data analytics. The main aim of this work is to present the assessment of placed
quarantine mechanism using mathematical modeling.
Weber et al. [16] presented the trend analysis of COVID-19 pandemic in China
using globally accepted SIR model in this study. The dataset used in this study is
taken from Johns Hopkins University site for analysis, and it is found that epidemic
was contained in China. The basic aim of the study presented by Zhang et al. [17]
is to provide control measures to be considered internationally for global control
of this pandemic. The time frame of dataset is from 3 to 10 February, 2020, and
authors presented a time series model to predict number of infected cases and the
Data Analytics: COVID-19 Prediction Using Multimodal Data 5

turning point where the spread is at peak. Feasibility analysis of controlling COVID-
19 spread by isolating infected cases and quarantine is presented by Hellewell et al.
[18]. The proposed probabilistic model presented in this study considered varied
scenarios like initial infections, basic reoccur number, and probability of contacts
traced and rate of clinical infections. The results show that, in epidemic situation,
isolation of infected people and contact tracing is not sufficient to minimize the rate
of spread. Modeling of COVID-19 epidemic in China for cumulative number of
infected cases using data available in early phase based on differential equation is
presented by Liu et al. [19]. Simulation results show that if the restrictions would
have been applied one week before, then there would have been significant reduction
in the number of infected cases.
Various ML models are discussed in the literature however for better accuracy deep
learning models can be used for better predictions [20–22]. Furthermore, predictions
can be more accurate using active learning models in this multitudinal and multi-
modal data used for predictions instead of single type of data [23]. In addition to
this, early forecasting of COVID-19 from small dataset is presented by Fong et al.
[24]. Fong et al. [25] have also proposed to use Composite Monte-Carlo simulation
forecasting method for helping government to initiate critical actions and decisions
to control spread of novel coronavirus. Experimental results using deep learning-
based composite Monte-Carlo with fuzzy rule induction show that decision makers
are benefited more in the form of better fitted Monte-Carlo outputs.
All the studies discussed above are centric toward prediction and forecasting of
COVID-19 based on short-term data available on this pandemic. Literature shows that
various mathematical and stochastic theory-based approaches are used for estimation
and prediction of spread rate of COVID-19. Most of the studies are giving expected
predictions. There are so many predictive analytics models, such as Susceptible-
Infection-Recovered (SIR) [26] and Hospital Impact Model for Epidemics (CHIME)
[27] which has been working from decades. The SIR models work best in the case
where data is not dynamic. In COVID-19, there is frequent change in data; hence,
learning model can be suitable for analysis of pandemic data like COVID-19. Predic-
tion of number of hospitals and facilities, i.e., hospital beds and ventilators, is also
equally important. In the view of this, predictive healthcare team developed COVID-
19 CHIME model at Penn Medicine. These predictions can help hospitals to be
prepared for worst-case scenarios.

4 Predictive Analytics

Predictive analytics is specialized branch of data analytics for making better predic-
tions using past data and using analysis techniques which include statistical and
learning methods. Discovery of patterns in input data and anticipating what is likely
to happen is the main objective of predictive analytics. Statistical analysis, predictive
modeling, and machine learning are three main pillars of predictive analytics. The
6 P. N. Mahalle et al.

main capabilities of predictive analytics are statistical analysis, predictive modeling,

linear regression, and logistic model.
Selection of appropriate predictive model and algorithm decide how efficiently
we can make the better insights and useful decisions. Use case like hospital inter-
ested in prediction of number of patients likely to be admitted in intensive care unit
in next seven days and prediction of fraud transaction for online banking provider
might require different predictive analytics model than for predicting defaulter appli-
cant for loan provider and predicting number of COVID-19 infected patient in next
10 days. Selection of appropriate predictive model is based on what predictive
question would you like to address and how optimization can be carried out using
predictive algorithms. The major pillars of predictive analytics are listed below:
1. Predictive Analytics Models;
2. Predictive Analytics Algorithms.

4.1 Predictive Analytics Models

Classification models are best for decision problems where the answer is merely
Yes or No. This model classifies data into multiple categories using past data, and
the prediction of fraud transaction for online banking provider will come into this
model. Clustering model arranges data into multiple logical groups based on some
common attributes. An interesting use case for this model might be grouping of
students into different logical buckets based on marks, city they come from in order
to decide the distribution of amount of effort for improving performance. Forecast
model is another most popular predictive model and mainly applied to the use case
where past numerical data is available to predict the value performance metrics or
new value using learning from past data. As stated earlier, forecasting number of
COVID-19 infected patient in next 10 days will fit into this model. When dataset
contains inconsistent data records, outlier models are most useful as these models
can identify these inconsistent entries. Finding strange records in insurance claim
can be solved by this model. Time series model are used for short-term predictions
using data points collected from the past in time domain (i.e., based on time as input
parameter). Collecting short-term data from China epidemic and predicting the same
for India can be solved using these models.

4.2 Predictive Analytics Algorithms

Predictive analytics algorithms are either based on machine learning or deep learning.
Machine learning algorithms are used when there is a need of classification or clus-
tering for prediction, decision, or analysis. These algorithms are more suitable for
structured data and can be linear or nonlinear in nature. Deep learning algorithms
are subset of machine learning algorithms and more useful when there is a need of
Data Analytics: COVID-19 Prediction Using Multimodal Data 7

identification or to recognize something. These algorithms are more useful to bigger

data like audio, video, and images where machine learning algorithms start under-
performing. The predictive analytics is mainly driven by learning techniques, and
there are wide ranges of applications for disease prediction in healthcare commu-
nity [28, 29]. Random forest algorithm is based on decision trees and used for both
classification and regression purposes. This algorithm is more suitable for big data
and uses bagging to avoid the errors. This model can address over fitting more
effectively. Gradient boosted model is ensemble model of decision tress and used
for classification. This model uses incremental model by building one tree at each
time by correcting errors made by previously trained tree. In contrast, in random
forest there is no relation among trees. K-means algorithm works on unlabeled data
and places new incoming data into logical groups based on some common feature.
Consider the COVID-19 example where clusters are formed of various patients based
on some severity of infection. K-mean model is useful to put new incoming patient
into appropriate cluster. This method is extremely useful in this growing pandemic
of COVID-19 due to large number of cases. Prediction of mortality and spread rate
plays very important role in pandemic disease like COVID-19, as based on this
prediction precautionary measures can be taken by public, government, and heath-
care systems. We have used FBProphet [30] algorithm for training the model and
predicting number of infected cases in next three months. We agree that there are
many machine learning algorithms present in the literature. However, this study
recommends Prophet Algorithm for better prediction because it is mainly an open-
source algorithm giving more accurate prediction. As we are aware that in sudden
pandemic likes COVID-19, adequate data is not available due to various reasons like
duration and lack of required parameters for better prediction, prophet algorithm
enables better forecast and does not require dataset training in time series methods.
The key features of this algorithm are it works more accurate for time series data
and mainly used for prediction and capacity planning. Dataset can be referred from
widely accepted sources like John Hopkins University and WHO. In this study, the
dataset is referred from Kaggle where the statistics for this COVID-19 pandemic is
given in the form of features like state/province, country, latitude, longitude, date,
confirmed infected, deaths, and recovered. There are eight fields in the dataset and
another feature of Prophet is it does not require splitting of dataset wherein for
fitting it takes whole dataset for accurate results. Figures 1 and 2 show the short-term
prediction of number of infected cases.
Figures 1 and 2 show the prediction of spread of COVID-19. The numbers of
confirmed cases of COVID-19 within respective duration are presented in the graph,
X-axis presents the duration and Y-axis shows the number of COVID-19 confirmed
cases. ML model is trained for prediction based on the worldwide dataset retrieved
from Github. Predictions are shown in Fig. 1 and it shows that the confirmed COVID-
19 infected cases would be 1.6 million and 2.3 million by the end of May and June,
respectively, and hence can be concluded that with increasing duration spread of
COVID-19 increasing and government should initiate appropriate control measures
in time to regulate this pandemic.
8 P. N. Mahalle et al.

Fig. 1 Prediction of confirmed cases till end of May

Fig. 2 Prediction of confirmed cases till end of June

5 Conclusion

Due to pandemic of Coronavirus and COVID-19, all countries are looking toward
mitigation plan to control the spread with the help of some modeling techniques. This
research work aims to understand the complete medical perspective of this COVID-
19 pandemic and how predictive analytics will empower the predictions. Analysis
Data Analytics: COVID-19 Prediction Using Multimodal Data 9

of various predictive analytics methods available in the literature is presented in this

chapter. We have also discussed and presented the comparative analysis of various
predictive analytics models and algorithm by suggesting more appropriate use cases
for application. Out study indicates that there is a need of thorough assessment of
these predictive analytics algorithm based on the type of question to be answered.
Application of prophet predictive analytics algorithm on Kaggle dataset, its predic-
tions are also presented in this chapter. Simulation result of this model shows that the
confirmed COVID-19 infected cases would be 1.6 million and 2.3 million by the end
of May and June, respectively. We hope that these predictions will be also helpful to
pharmaceutical companies to manufacture drugs in faster rate.

References

1. The Novel Corona Virus Pneumonia Emergency Response Epidemiology Team. (2020). The
epidemiological characteristics of an outbreak of 2019 novel corona virus diseases (COVID-
19)—China, 2020. China CDC Weekly, 2, 113–122.
2. WHO. (2020). Novel corona virus (2019-nCoV) situation report—39. Cited March 2020. Avail-
able from: https://www.who.int/docs/default-source/coronaviruse/situationreports/20200228-
sitrep-39-covid-9.pdf.
3. WHO. (2020). Novel corona virus (2019-nCoV) situation report—52. Cited March 2020. Avail-
able from: https://www.who.int/docs/default-source/coronaviruse/situation-reports/20200312-
sitrep-52-covid-19.pdf.
4. The center of evidence-based medicine develops, promotes and disseminates better evidence
for healthcare CEBM-University of OXFORD report. (2020). Cited April 6, 2020. https://www.
cebm.net/covid-19/covid-19-what-proportion-are-asymptomatic/.
5. Singhal, T. (2020). A review of coronavirus disease-2019 (COVID-19). Indian Journal of
Pediatrics, 87(4), 281–286.
6. Rothe, C., Schunk, M., Sothmann, P., Bretzel, G., Froeschl, G., Wallrauch, C., et al. (2020).
Transmission of 2019-nCoV infection from an asymptomatic contact in Germany. New England
Journal of Medicine, 382(10), 970–971.
7. Chan, J. F. W., Yuan, S., Kok, K. H., To, K. K. W., Chu, H., Yang, J., et al. (2020). A familial
cluster of pneumonia associated with the 2019 novel coronavirus indicating person-to-person
transmission: A study of a family cluster. Lancet, 395(10223), 514–523.
8. Xiaoxia, L., Liqiong, Z., Hui, D., Jingjing, Z., Yuan, L., Jingyu, Q., et al. (2020). SARS-CoV-
2 infection in children. New England Journal of Medicine. https://doi.org/10.1056/NEJMc2
005073.
9. https://www.who.int/docs/default-source/coronaviruse/who-china-joint-mission-on-covid-
19-final-report.pdf.
10. Moriarty, L. F., Plucinski, M. M., Marston, B. J., Kurbatova, E. V., Knust, B., Murray, E.
L., et al. (2020). Public health responses to COVID-19 outbreaks on cruise ships-worldwide.
MMWR Morbidity and Mortality Weekly Report, 69, 347–352.
11. Russell, B., Moss, C., Rigg, A., Hopkins, C., Papa, S., & Van Hemelrijck, M. (2020). Anosmia
and ageusia are emerging as symptoms in patients with COVID-19: What does the current
evidence say? Ecancer Medical Science, 14, ed98.
12. Giordano, G., Blanchini, F., Bruno, R., Colaneri, P., Di Filippo, A., Di Matteo, A., & Colaneri,
M. (2020). A SIDARTHE model of COVID-19 epidemic in Italy. arXiv preprint arXiv:2003.
09861.
13. Bannister-Tyrrell, M., Meyer, A., Faverjon, C., & Cameron, A. (2020). Preliminary evidence
that higher temperatures are associated with lower incidence of COVID-19, for cases reported
globally up to 29th February 2020. medRxiv.
10 P. N. Mahalle et al.

14. Russo, L., Anastassopoulou, C., Tsakris, A., Bifulco, G. N., Campana, E. F., Toraldo, G.,
& Siettos, C. (2020). Tracing DAY-ZERO and forecasting the fade out of the COVID-
19 outbreak in Lombardy, Italy: A compartmental modelling and numerical optimization
approach. medRxiv.
15. Volpert, V., Banerjee, M., & Petrovskii, S. (2020). On a quarantine model of coronavirus
infection and data analysis. Mathematical Modelling of Natural Phenomena, 15, 24.
16. Weber, A., Ianelli, F., & Goncalves, S. (2020). Trend analysis of the COVID-19 pandemic in
China and the rest of the world. arXiv preprint arXiv:2003.09032.
17. Zhang, F., Zhang, J., Cao, M., & Hui, C. (2020). A simple ecological model captures the
transmission pattern of the coronavirus COVID-19 outbreak in China. medRxiv.
18. Hellewell, J., Abbott, S., Gimma, A., Bosse, N. I., Jarvis, C. I., Russell, T. W., & Flasche, S.
(2020). Feasibility of controlling COVID-19 outbreaks by isolation of cases and contacts. The
Lancet Global Health.
19. Liu, Z., Magal, P., Seydi, O., & Webb, G. (2020). Predicting the cumulative number of cases
for the COVID-19 epidemic in China from early data. arXiv preprint arXiv:2002.12298.
20. Mukherjee, H., Ghosh, S., Dhar, A., Obaidullah, S. M., Santosh, K. C., & Roy, K. (2020).
Shallow convolutional neural network for COVID-19 outbreak screening using chest X-rays.
https://doi.org/10.36227/techrxiv.12156522.v1.
21. Rajinikanth, V., Dey, N., Raj, A. N. J., Hassanien, A. E., Santosh, K. C., & Raja, N. (2020).
Harmony-search and Otsu based system for coronavirus disease (COVID-19) detection using
lung CT scan images. arXiv preprint arXiv:2004.03431.
22. Das, D., Santosh, K. C., & Pal, U. (2020). Truncated inception net: COVID-19 outbreak
screening using chest X-rays. https://doi.org/10.21203/rs.3.rs-20795/v1.
23. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: Need of active learning
and cross-population train/test models on multitudinal/multimodal data. Journal of Medical
Systems, 44, 93. https://doi.org/10.1007/s10916-020-01562-1.
24. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Finding an accurate
early forecasting model from small dataset: A case of 2019-nCoV novel coronavirus outbreak.
arXiv preprint arXiv:2003.10776.
25. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing, 106282.
26. Teles, P. (2020). Predicting the evolution of SARS-COVID-2 in Portugal using an adapted SIR
model previously used in South Korea for the MERS outbreak. arXiv preprint arXiv:2003.
10047.
27. http://predictivehealthcare.pennmedicine.org/2020/03/14/accouncing-chime.html.
28. Chen, M., Hao, Y., Hwang, K., Wang, L., & Wang, L. (2017). Disease prediction by machine
learning over big data from healthcare communities. IEEE Access, 5, 8869–8879.
29. Shinde, G. R., Kalamkar, A. B., Mahalle, P. N., Dey, N., Chaki, J., & Hassanien, A. E. (2020).
Forecasting models for coronavirus (COVID-19): A survey of the state-of-the-art. TechRxiv.
Preprint. https://doi.org/10.36227/techrxiv.12101547.v1.
30. https://facebook.github.io/prophet/docs/quick_start.html.
COVID-19 Apps: Privacy and Security
Concerns

Surekha Borra

Abstract Today, with the rapid spread of COVID-19, many governments and start-
ups are coming forward to develop smartphone apps that trace where we all are, whom
we met and for how long, with a goal of interrupting new chains by informing poten-
tially exposed people. These new platforms make use of anonymous use of Bluetooth
technology and GPS, enabled either on smartphones or armbands in order to prepare
maps corresponding to quarantine monitoring, contact tracing, movement tracking,
social distancing and density reports. With different apps for different countries, one
thing most of the apps facilitate is tracking. To save lives during an extraordinary
crisis, many governments are willing to overlook privacy implications. Keeping in
view that the sensitive data being collected is not exclusive to public health organi-
zations and governments, this chapter explores different apps that were developed
aiming to combat COVID-19, and the related personal data privacy concerns that
arise in the post-coronavirus era.

Keywords Apps · Bluetooth · COVID-19 · GPS · Privacy · Security

1 Introduction

While the researchers around the world are busy developing COVID-19 related AI-
driven tools [1, 2], forecasting methods [3–5], screening [6] and image-assisted
decision support systems [7], COVID-19 mobile apps are being developed for a
variety of reasons, ranging from quarantine monitoring, contact tracing, movement
tracking, social distancing and density reports. All these mobile apps with or without
the consent of the user collect user personal information, including location histories
and stores the data on the third-party servers, which might lead to serious cyberthreats
and associated fears. Hence, reporting the available apps and analysing the side effects
including security and privacy concerns is the need of the hour.

S. Borra (B)
Department of Electronics and Communication Engineering, K.S. Institute of Technology,
Bengaluru, Karnataka, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 11
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_2
12 S. Borra

While the privacy compromises have prevented more open take-up of such tech-
nologies, China used phone tracking for contact tracing successfully apart from
voluntary registrations. China made it mandatory for its citizens to install apps on
their mobiles, to monitor their movements, and to give instructions [8]. Tools [9]
that can detect if someone is not wearing a mask are also in use. These tools can
even recognize individuals from just their eyes when wearing masks. China also used
drones to monitor sections of the population [10].
Singapore’s Trace Together app [11] shares ‘proximity information’ to Ministry
of Health when a user device with the app installed is in the Bluetooth range of a
carrier device via Bluetooth, for follow-up actions. Quarantine breaches are moni-
tored by real-time tracking the carriers in Israel using ‘anti-terror’ technology. Some
governments are using GPS phone tracking, Bluetooth technology and credit card
records, to track carriers’ movements. While location tracking information from
mobile companies is being collected by Austria, Germany and Italy, countries like
Hong Kong, Singapore, Berlin, South Korea, Australia and Taiwan are using location
monitoring systems to map the COVID-19 carriers. To fight the spread of COVID-
19, the UK planned to use the databases of mobile network trade association, the
GSMA. The Pan-European Privacy-Preserving Proximity Tracing (PEPP-PT) [12],
developed by European Union, tracks virus spread by tracking the smartphones.
Google and Facebook have recently collaborated with the USA in collecting and
monitoring the location history based on smartphone apps.

2 COVID-19 Apps

The conventional way of quarantine monitoring and contract tracing is by humans,

which is very much time-consuming and resource intensive. Usage of personal digital
devices, installation of software apps, aggregation of Bluetooth, GPS and mobile
signals is must in the current scenario. All these technologies assist the authorities in
updating the real-time data automatically in a central server at a fast rate, at different
scales. The data includes new COVID-19 cases, proximity contacts, social distancing
and quarantine status along with the location histories. A variety of apps based on
different technologies were developed with respect to COVID-19 in a span of just
4 months by different government authorities, private companies and researchers,
some of which are listed in Tables 1 and 2.
The social distance enforcing apps warns the people if they get too close to others
or if they spend too much time outside or away from their homes, during the lock-
down. The quarantine apps are mandated by the governments like India, Poland,
Taiwan. These ask for sharing the GPS location and selfies, to ensure those quaran-
tined at home are not breaking the rules. Switching off their phone or venturing to
leave their houses triggers the alert system, police visits, further calls and messages
are sent to the concerned person to ascertain their whereabouts. To trace the move-
ments of quarantined and infected people, their contacts and to score the likelihood
of infections, the authorities need to know where they had been and who else had
COVID-19 Apps: Privacy and Security Concerns 13

Table 1 COVID-19 apps

App category Name of the app
Information Coronavírus—SUS, Pakistan’s National Action
Plan for COVID-19, TGN Emergències, Atman,
iVH HIT, Coronavirus Bolivia, Hamro Swasthya,
Apple COVID-19, COVID-19!, Coronavirus
Australia
Self-testing or self-assessment of symptoms Coronavírus—SUS, MySejahtera, Atman,
COVID-19 Armenia, Apple COVID-19,
Coronavirus Australia, STOP COVID19 CAT
Share experiences of the patients CoronaReport—COVID-19 reports for social
science
Registering health data and follow up Coronavirus UY, MySejahtera, Kenya COVID-19
Tracker, Karantinas, GVA Coronavirus,
HEALTHLYNKED COVID-19 Tracker,
PatientSphere for COVID19
COVID-19 test result ALHOSN UAE
Monitor symptoms and health information HSE COVID-19, PatientSphere for COVID19,
STOP COVID19 CAT
Quarantine BeAware Bahrain, Kenya COVID-19 Tracker,
Karantinas, StayHome App, HSE COVID-19
Contact tracing BeAware Bahrain, Kenya COVID-19 Tracker,
TraceCovid, TraceTogether

been there. Many countries are leveraging cell phone location data from cell phone
towers, marketing-style databases, Google, Apple and Facebook Apps to track the
spread of the virus.

3 Advantages and Concerns of COVID-19 Apps

With an end goal of reducing the spread of COVID-19, mass information collection
strategies are as of now being put to utilize. Conducting interviews with patients in
which they detail where they had been and with whom they met in the weeks going
before they were tested positive is complex and time taking task, for the govern-
ments in this crucial time. Some countries have published comprehensive digital
maps and lists of confirmed cases, their movements, and their travel history, by
means of personal interviews with patients and surveillance videos. Some govern-
ments [13] have published the pre-quarantine movements of people well before they
were diagnosed with the virus, gender, age, occupation, address, and where they
travelled recently including restaurants, gyms and hospitals [14]. Most of the coun-
tries are actively tracking population movements by the CCTV footage, credit card
records, cell phone signals and mobile location data to help combat coronavirus. The
advantages of such technologies are listed below:
14 S. Borra

Table 2 COVID-19 apps developed in India

App category Name of the app
Information Jaano (service providers for daily needs), GoK
Direct—Kerala
Quarantine COVID-19 Quarantine Monitor Tamil Nadu,
MahaKavach, GCC—Corona Monitoring, UP
Self-Quarantine App, Corona Mukt Himachal,
Quarantine Watch, COPE Odisha, COVID-19
West Bengal Government, SMC COVID-19
Tracker, CoBuddy—Covid19 Tool, Covid
Locator, COVID CARE
F/B to administration Jaano, COVID19 Feedback (f/b on treatment),
CoBuddy—Covid19 Tool
Request for COVID-19 test Niramaya
Call support COVA Punjab
Reporting symptoms GCC—Corona Monitoring, Uttarakhand CV 19
Tracking System
Self-testing or self-assessment of symptoms Aarogya Setu, Kavach, Niramaya, COVA
Punjab, Test Yourself Goa, GCC—Corona
Monitoring, Test Yourself Puducherry
Contact tracing Aarogya Setu, Tracetogether, SMC COVID-19
Tracker
Tracking (Geofencing) (GPS) MahaKavach, CoBuddy—Covid19 Tool, Covid
Locator, COVID CARE, Corona Watch (14 days
of movement history)
Lockdown pass KSP Clear Pass Checker, CG Covid-19 ePass

• Eases countrywide lockdowns.

• Speeds up the preventive actions.
• Limits the spread of the virus.
• Manages and tracks virus carriers in a quick way.
• Tracks those at a risk of infection effectively.
• Works automatically and accurately.
• Populates the anonymized map data.
• Enables better management of the evolving situation.
• Enforces government directives with less manpower.
• Estimates asymptomatic carriers’ population.
• Helps in identifying where to target critical medical resources.
• Enables researchers and scientists to learn how long the virus survives on a surface.
• Communicates information to the citizens at fast rate.
Apps like Google collect data from those who have turned their GPS ON. While
some mobile networks do not give options to the users for their consent to collect
their location data, most of the apps ask user permission to track their location history.
Some sneaky apps track the locations without the knowledge. Therefore, though the
COVID-19 Apps: Privacy and Security Concerns 15

smartphone surveillance might seem like a good solution to tracking the COVID-19
spread, it is far from guaranteed to work and have many data protection and privacy
issues and concerns [15–18], some of which are listed below:
• Enable unwanted corporate or government surveillance.
• Apps that collect locations feed this data to marketing companies and are paid to
do so.
• Some apps use the information to serve the user more relevant advertisements and
content.
• Data more often serves private profit such as advertising.
• Inherent loss of privacy, highly invasive.
• Location tracking apps reveal sensitive information about everything from
political dissent to journalists’ sources to extramarital affairs.
• Sacrifice of freedom still causes anxiety, despite the clear public interest.
• May sometimes do more harm than good.
• Apps might sometimes sow unnecessary alarm or confusion.
• Possibility of privacy threats from contacts, snoopers and the authorities.
• Announcing the infected people identities stirs up public shaming and rumour-
mongering.
• Incorrect information might encourage risky behaviour leads to false sense of
security.
• Apps might only provide a very crude picture of the spread.
• It may be challenging to get people install such apps and report their status.
• Once the privacy back door opens, there are chances that it remains permanent.
• Fine-grained location tracking is complex.
• Accuracy may often be less precise.
• Complex database systems involving tagging and monitoring citizens.

4 Discussions and Recommendations

Smartphone apps that model and monitor the spread of diseases based on Bluetooth
and other wireless technology were first developed by Jon Crowcroft and Eiko Yoneki
at Cambridge University in 2011 [19]. Fluphone [20] asks the people to report the flu-
like symptoms, for monitoring the person who got sick, and to let to know who had
contacted that person. However, because only 1% of people used that app, tracking
and reporting the transmissions was not as expected. With the current COVID-apps
and technologies, the concerns are:
• How to reduce trade-offs among the speed of implementation, effectiveness and
privacy.
• How well such a system would work and how is it being utilized.
• Whether such systems to be developed and deployed country by country or agree
a cross-border solution.
• Whether the apps better deal with this kind of situation next time around.
16 S. Borra

• Whether more than half the population would trade personal privacy to avoid
further lockdowns.
Though there are several implications of utilizing location histories data, here are
some recommendations for the authorities and the app developers:
• Governments should ensure trust and privacy principles.
• Keep location data private.
• Avoid surveillance.
• Encryption of information can be done before saving and publication.
• Government should take measures to explain to the citizens, how the apps work,
what it gathers, for how long, when it is disposed and for what purpose data is
utilized.
• Apps should opt-in and should not last beyond a time of crisis.
• No data records should link back to an individual.
• Any personal data should not leave a device.
• Data itself, though, must be uniquely identifiable.

5 Conclusions

Though the Google’s mobility reports and mobile networks data sources are avail-
able, these are not used for contact tracing until recently. With the occurrence of
new COVID-19 cases developing continuously, the government authorities started
harnessing apps and the cell phone location data bases, for monitoring isolation,
social distancing and behavioural changes in the war against the spread of the new
coronavirus. However, unless majority of population uses the app, the tracking and
reporting of the transmissions will not be successful. Practicing digital distancing,
restricting the access permissions to apps, confirming if the app is by legitimate
developer and downloading the app from trusted mobile app stores can assist in
dealing with potential security issues that may arise in future.

References

1. Santosh, K., Das, D., & Pal, U. (2020). Truncated inception net: COVID-19 outbreak screening
using chest X-rays.
2. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: Need of active learning
and cross-population train/test models on multitudinal/multimodal data. Journal of Medical
Systems, 44(5), 1–5.
3. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Finding an accurate
early forecasting model from small dataset: A case of 2019-nCoV novel coronavirus outbreak.
arXiv preprint arXiv:2003.10776.
4. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing, 106282.
COVID-19 Apps: Privacy and Security Concerns 17

5. Mahalle, P., Kalamkar, A. B., Dey, N., Chaki, J., & Shinde, G. R. (2020). Forecasting models
for coronavirus (COVID-19): A survey of the state-of-the-art.
6. Mukherjee, H., Ghosh, S., Dhar, A., Obaidullah, S., Santosh, K. C., & Roy, K. (2020). Shallow
convolutional neural network for COVID-19 outbreak screening using chest X-rays.
7. Rajinikanth, V., Dey, N., Raj, A. N. J., Hassanien, A. E., Santosh, K. C., & Raja, N. (2020).
Harmony-search and Otsu based system for coronavirus disease (COVID-19) detection using
lung CT scan images. arXiv preprint arXiv:2004.03431.
8. https://www.nytimes.com/2020/03/01/business/china-coronavirus-surveillance.html.
9. https://www.bbc.com/news/technology-51717164.
10. https://www.forbes.com/sites/zakdoffman/2020/03/05/meet-the-coronavirus-spy-drones-that-
make-sure-you-stay-home/#1319d4ee1669.
11. Help speed up contact tracing with TraceTogether. (2020, March). Singapore Government Blog.
[Online]. Available: https://www.gov.sg/article/help-speed-up-contact-tracing-with-tracetoge
ther.
12. PEPP-PT. https://www.pepp-pt.org.
13. http://ncov.mohw.go.kr/en/.
14. Kim, M. J., & Denyer, S. (2020, March). A travel log of the times in South
Korea: Mapping the movements of coronavirus carriers. The Washington Post.
[Online]. Available: https://www.washingtonpost.com/world/asiapacific/coronavirus-south-
koreatracking-apps/2020/03/13/2bed568e-5fac-11eaac50-18701e14e06dstory.html.
15. Cho, H., Ippolito, D., & Yu, Y. W. (2020). Contact tracing mobile apps for COVID-19: Privacy
considerations and related trade-offs. arXiv preprint arXiv:2003.11511.
16. Berke, A., Bakker, M., Vepakomma, P., Raskar, R., Larson, K., & Pentland, A. (2020). Assessing
disease exposure risk with location histories and protecting privacy: A cryptographic approach
in response to a global pandemic. arXiv preprint arXiv:2003.14412.
17. Demirag, D., & Ayday, E. (2020). Tracking and controlling the spread of a virus in a privacy-
preserving way. arXiv preprint arXiv:2003.13073.
18. Bell, J., Butler, D., Hicks, C., & Crowcroft, J. (2020). TraceSecure: Towards privacy preserving
contact tracing. arXiv preprint arXiv:2004.04059.
19. https://www.wired.com/story/phones-track-spread-covid19-good-idea/.
20. https://www.cl.cam.ac.uk/research/srg/netos/projects/archive/fluphone/.
Coronavirus Outbreak: Multi-Objective
Prediction and Optimization

Nileshsingh V. Thakur

Abstract Coronavirus disease-19 (COVID-19) is the name given by World Health

Organization (WHO) and the cause of this pandemic is severe acute respiratory
syndrome-related coronavirus (SARS-CoV-2). This pandemic was started in China
and later got spread throughout the world. As of now, it has affected around 213
countries, areas or territories and the world has eyes on it. This chapter provides the
insight on mathematical perception about the coronavirus outbreak and is analyzed
from prediction and optimization point of view. Constraints, objectives and measures
are identified and their association and dependencies are analyzed from mathemat-
ical point of view. Four levels and weighting factors are used for the identified
constraints. Presented mathematical formulation can evaluate the count of posi-
tive cases, mortality, recovery cases, transmission rate and prediction of peak time
period. It is observed that the coronavirus outbreak is the constrained multi-objective
prediction and optimization problem and also it is time variant.

Keywords COVID-19 · Constrained optimization · Multi-objective prediction ·

Multi-objective optimization · Mathematical problem

1 Introduction

World Health Organization (WHO) named the present outbreak of a coronavirus-

associated acute respiratory disease as coronavirus disease-19 (COVID-19) [1]. The
Coronaviridae Study Group (CSG) of the International Committee on Taxonomy of
Viruses (ICTV) has classified this virus and tentatively named 2019-nCoV, within
the Coronaviridae [2]. This virus has been recognized as severe acute respiratory
syndrome coronaviruses (SARS-CoVs) of the species and designated as severe acute
respiratory syndrome-related coronavirus (SARS-CoV-2) by CSG. It is reported that
the virus might be bat origin [4]. The genetic features and some clinical findings

N. V. Thakur (B)
Computer Science and Engineering, Prof Ram Meghe College of Engineering and Management,
Badnera, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 19
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_3
20 N. V. Thakur

of the infection have been reported recently [5–7]. Spread at international level via
commercial air travel had been assessed [8].
Uniqueness of COVID-19 can be described by three characteristics—(1) Time
delay between real cases and daily reported cases, due to two week incubation
period [9], (2) Epidemic trend depends on the available medical resources, preven-
tive and corrective measures and the efficiency of confirmation of positive cases,
(3) Quarantine measures are implemented by most of the countries to hold no risk
of contagion. Due to the outbreak and control-related distinctiveness of COVID-
19 from existing contagious diseases, the existing epidemic models cannot be
purposeful to portray the observed data directly. So, it is difficult to describe COVID-
19 with Susceptible-Infectious-Recovered (SIR), Susceptible-Exposed-Infectious-
Recovered (SEIR), etc. epidemic models [10, 11].
Recently, various research works are reported to detection and prediction of
COVID-19 [12–17]. In [12], the authors have proposed the Truncated Inception Net
deep learning model which is based on convolutional neural network. The authors
have carried out the experimentation on six data sets which are composed of chest
X-rays images. In [13], the authors have demonstrated the use of X-rays images for
the mass screening to find the COVID-19 positive cases through the light-weight
CNN-based shallow architecture which works with less parameters. In [14], authors
have used CT scan images to locate the infected region and infection level of severity
in the lungs. They have employed the use of Harmony-Search-Optimization and Otsu
thresholding for the image enhancement.
The importance of artificial intelligence-driven tools for the prediction of spread
rate of COVID-19 is elaborated in [15]. The author has mentioned about the need
of cross-population models which should be based on active learning and it should
employ the multitudinal and multimodal data. In [16], the authors have proposed the
methodology based on polynomial neural network with corrective feedback for the
prediction of spread of COVID-19. This methodology is designed with the collection
of five optimized forecasting models. The use of deterministic and non-deterministic
input data for the prediction of epidemic spread through the stochastic model based
on Monte-Carlo model with the use of deep learning network and fuzzy rule induction
is described in [17].
As COVID-19 is distinct from the existing contagious diseases, it is difficult to
describe coronavirus outbreak by using SIR and SEIR epidemic models. Hence,
there is a need of mathematical analysis of coronavirus outbreak with additional
parameters. This chapter proposes the mathematical formulation which is based
on the preventive measures, corrective measures and identified constraints. This
mathematical formulation can evaluate-positive cases count (P), mortality count
due to coronavirus (M), transmission rate of coronavirus (T ), coronavirus recovery
cases count (R) and prediction of peak time period (PP). Four levels and weighting
factor is considered for constraint parameters. This chapter describes the coronavirus
outbreak as the constrained multi-objective prediction and optimization problem.
Relevant mathematical analysis and scope of utilizing artificial intelligence and data
science are discussed.
Coronavirus Outbreak: Multi-Objective Prediction … 21

Remaining part of this chapter is organized as follows: Symptoms, prevention

and impingement related to coronavirus are briefly described in Sect. 2. Constraints,
objectives and measures are identified in view of mathematical perception which
is discussed in Sect. 3. Discussion on constrained multi-objective prediction and
optimization is provided in Sect. 4. Section 5 elaborates the general idea to use the
artificial intelligence and data science followed by conclusion in Sect. 6.

2 Coronavirus: Symptoms, Prevention, Impingement

In general, this virus affects different people in different ways. It is observed that the
symptoms are mild to moderate and most of the infected people recover without any
special treatment. Coronavirus symptoms and prevention measures are suggested by
WHO [1].
Common symptoms include—fever, tiredness and dry cough. Other symptoms
include—shortness of breath, aches and pains, sore throat and rare symptoms include
diarrhea, nausea or a runny nose. It is intolerant in people who have underlying
medical conditions and those over 60 years old, due to a higher risk of developing
severe disease and death.
General preventive measures are—Hand wash—Wash hands on regular basis
with soap and water, or clean with alcohol-based hand rub; Keep away—Maintain
at least 1 meter distance from coughing or sneezing fellow; Avoid face touch—
Evade face touching; Face mask—Use mask to cover the face which will help to
cover mouth and nose when coughing or sneezing; Avoid outing—Stay home if not
well; Lungs caring—Avoid smoking and other activities that weaken the lungs; and
Distancing—Staying away from large groups of people.

2.1 Impingement

On December 31, 2019, the WHO China Country Office was informed of cases
of pneumonia unknown etiology (unknown cause) detected in Wuhan City, Hubei
Province of China. Till January 3, 2020, a total of 44 case-patients with pneumonia
of unknown etiology were reported to WHO by the national authorities in China.
The Chinese authorities identified a new type of coronavirus, which was isolated
on 7 January 2020. On 13 January 2020, the Ministry of Public Health, Thailand
reported the first imported case of lab-confirmed novel coronavirus (2019-nCoV)
from Wuhan, Hubei Province, China. Spread of this virus is reported at Japan and
Republic of Korea, also. Patient cases in Thailand, Japan and Republic of Korea were
exported from Wuhan City, China [1].
As of now, COVID-19 grabbed the whole world which includes 213 countries,
areas or territories with 1848439 confirmed cases and 117217 confirmed deaths [18].
The first case in India was reported on 30 January 2020. Since then, to 14 April 2020,
22 N. V. Thakur

Fig. 1 Statistics of COVID-19 cases in India from March 14 to April 14, 2020

10815 confirmed cases, 1190 recovered cases, 9272 active cases and 353 confirmed
deaths are reported [19]. Figure 1 shows the statistics of COVID-19 cases in India
from 14 March 2020 to 14 April 2020.

3 Coronavirus Outbreak—Mathematical Perceptions

Due to non-availability of proper medical treatment, coronavirus eruption becomes a

tedious task for the government and medical practitioners. Engineering academicians
and researchers may contribute to this pandemic by providing mathematical analysis
and formulation for the avoidance, prediction, and minimization of spread of COVID-
19. Apart from the medical treatment, some typical measures, namely—lockdown,
social distancing, home quarantine, isolation, etc. are already adopted, prominently,
by China and India. Prominent and concrete mathematical solution have not yet
arrived which may perfectly predict and optimize this pandemic. To explore the role
of mathematics, some basic mathematical formulation is required.

3.1 Constraints

The aim of any government or healthcare unit of any country is to minimize the impact
of COVID-19 pandemic. This impact may be related to the parameters, namely—
count of coronavirus positive cases, count of mortality due to coronavirus, social
and economical disruption, and mental stress, etc. One has to minimize all these
parameters, but, with the existing constraints and some other imposed constraints
due to the developed panic of COVID-19. As no drug exists to overcome the COVID-
19 and it is a contagious disease, some other measures have to be considered and
implemented to avoid the spread of this pandemic. As the incubation period of
Coronavirus Outbreak: Multi-Objective Prediction … 23

Table 1 Constraint parameters with their levels and weight factor

Constraint parameters Level (weight factor λ)
Demographic trend (DT) Kid (10) Young (10) Middle (10) Old (10)
Gravity of existing disease (GD) Poor (5) Average (8) Moderate (9) Critical (10)
Population count (PC) Low (5) Average (8) Moderate (9) High (10)
Density of population (DP) Low (5) Average (8) Moderate (9) High (10)
Living style (LS) Poor (5) Average (8) Moderate (9) High (10)
Mental stress (MS) Low (5) Average (8) Moderate (9) High (10)
People connectivity (CO) Low (5) Average (8) Moderate (9) High (10)
Stage of spread (SS) Onset (5) Local (8) Community (9) Peak (10)

this disease is of 14 days, it is difficult to guess or predict the number of affected

cases. Identified constraint parameters with their levels and weight factor are given
in Table 1. To predict and optimize the coronavirus outbreak, existing and imposed
constraints have to be considered.
Weight factor (λ) value range is from 5 to 10. COVID-19 is a contagious disease
and no drug is available, the probability of becoming coronavirus affected is 0.5 and
as the highest weight factor is 10, therefore, the lowest weight factor is considered
as 5. Different variants of combinations of these constraints can be used for the
mathematical formulation. These constraints weight factor may vary with respect to
the locality.

3.2 Objectives and Measures Adopted

Currently, sole aim in COVID-19 pandemic is to have minimum number of coro-

navirus positive cases. Objectives are identified from mathematical analysis point
of view and are divided into two categories—Minimization of-coronavirus positive
cases count (P), mortality count due to coronavirus (M), transmission rate of coron-
avirus (T), social and economical disruption (S), and Maximization of—coronavirus
recovery cases count (R). These objectives have to be optimized to conquer coro-
navirus effect. Apart from this, another objective is-prediction of peak time period
(PP), i.e., the prediction of spread of COVID-19 with the imposition of identified
constraints and accordingly proper preventive and corrective measures should be
implemented. Therefore, from mathematical point of view, coronavirus outbreak is
the multi-objective prediction and optimization problem with imposed constraints.
Following measures are adopted in China and India to minimize the panic of
coronavirus outbreak. These measures are divided into two categories—Preventive
measures (θ ) include—personal care (PR), social distancing (SD), lockdown (LD),
and house-to-house inspection (HH), and Corrective measures (η) include—home
quarantine (HQ), institutional quarantine (IQ), medical treatment (MT).
24 N. V. Thakur

4 Constrained Multi-Objective Prediction

and Optimization

The coronavirus outbreak appeared to be nonlinear in nature and time variant. To

do the prediction and optimization, bottom-up approach has to be incorporated, i.e.,
specific to general. Therefore, local prediction and local optimization have to be
carried out first to find the global prediction and global optimization. The complete
geographic region has to be divided into different sub-regions and then the local
evaluation should be carried out. The process of evaluation of proposed mathematical
formulation is shown in Fig. 2 where E-Evaluation, P-Prediction and O-Optimization.
With reference to Table 1, total 48 constraints can be formulated. Preventive
measures adopted affect few of the identified constraints, for example, personal
care—affects the constraint GD, LS, MS, CO and social distancing, lockdown, house-
to-house inspection—affects the constraint MS, CO. Corrective measures adopted
affect few of the identified constraints, for example, home quarantine, institutional
quarantine and medical treatment—affects the constraint GD, LS, MS and CO. Other
constraints DT, PC, DP and SS remain as it is. Presented mathematical analysis is
related to the local evaluation. We can apply it to any local region of the country.
Every local region mostly consists of all the types of level for the constraint DT.
If so, then based on the percentage evaluation of level, the weight factor value is
applicable. Main objectives of this mathematical analysis are—P, M, T, S and R.
Out of these, presently, we are considering only P, M, T, R and excluding S as it
involves many uncontrolled parameters. P, M, and T are related to the minimization
and R is related to the maximization. Identified objectives are associated with other
objectives and measures. Identified constraints DT, GD, PC, DP, LS, MS, CO and SS
are directly or indirectly associated with P, M, T and R. Objective association with
other objectives, measures and constraint parameters is given in Table 2. Associated
objectives, measures and constraints sequence is based on relevance, more relevant
appears first.

Fig. 2 Process of evaluation of proposed mathematical formulation

Coronavirus Outbreak: Multi-Objective Prediction … 25

Table 2 Objective association—objectives, measures and constraint parameters

Objectives—association Type Association with constraints
with other objectives, Direct (ϕ) Indirect (φ)
preventive measure (θ)
and corrective measure
(η)
Positive cases (P)-T and Minimization CO, SS DP, PC, GD, DT, LS,
PR, SD, LD, HH, HQ, IQ MS
Mortality (M)-P and MT, Minimization CO, GD, DT, SS, DP, PC, LS
HQ, IQ MS
Transmission rate Minimization SS, CO, DP, PC, GD, DT, MS
(T)-PR, SD, LD, HH, LS
HQ, IQ
Recovery (R)-P, M and Maximization CO, SS DP, PC, GD, DT, LS,
MT MS
Prediction (PP)-P, M, T, Depends on P, M, T, R All constraints All constraints
R and PR, SD, LD, HH,
HQ, IQ

All the computations have to be performed with respect to the time t. Preventive
measure (θ ) and corrective measure (η) values are in percentage. T, α, β can be eval-
uated by using moving average of previous days data or cumulative data. Equation 1
through Eq. 8 can be used for the computations. All relevant objective functions, that
is, minimization of P, M, T and Maximization of R are subject to the constraints and
adopted measures.

P(t) = T(t) · P(t−1) , M(t) = α(t) · M(t−1) , R(t) = β(t) · R(t−1) (1)
⎡
T(t) + T(t+1) 1 
n=2
1 
m=6
P(t+1) = ⎣ + λi + λj
2 10 · n i=1, i∈ϕ 10 · m j=1, j∈φ
⎤

k=4
θk 
l=2
ηl ⎦
− − P(t) (2)
k=1, k∈θ
100 · k l=1, l∈η 100 · l
⎡
α(t) + α(t+1) 1 
n=5
M(t+1) = ⎣ + λi
2 10 · n i=1, i∈ϕ
⎤
1 
m=3 
l=3
ηl ⎦
+ λj − M(t) (3)
10 · m j=1, j∈φ l=1, l∈η
100 · l
26 N. V. Thakur
⎡
1 
n=5
1 
m=3
T(t+1) =⎣ λi + λj
10 · n i=1, i∈ϕ 10 · m j=1, j∈φ
⎤

k=4
θk 
l=2
ηl ⎦
− − (4)
k=1, k∈θ
100 · k l=1, l∈η 100 · l
⎡ ⎤
1 
n=5
1 
m=3 
l=3
ηl ⎦
α(t+1) =⎣ λi + λj − (5)
10 · n i=1, i∈ϕ 10 · m j=1, j∈φ l=1, l∈η
100 · l
⎡ ⎤
β(t) + β(t+1) 1 
n=2
1 
m=6  ηl ⎦
R(t+1) =⎣ + λi + λj − R(t)
2 10 · n i=1, i∈ϕ 10 · m j=1, j∈φ l=1, l∈η
100
(6)
⎡ ⎤
1 
n=2
1 
m=6  ηl ⎦
β(t+1) = ⎣ λi + λj − (7)
10 · n i=1, i∈ϕ 10 · m j=1, j∈φ l=1, l∈η
100
 
P P(t+n) = T(t+1) · P(t+1) + α(t+1) · M(t+1) + β(t+1) · R(t+1) (t + n) (8)

5 Scope to Use Artificial Intelligence and Data Science

Tools based on artificial intelligence can be used to analyze the coronavirus outbreak.
It can also be used to predict the nature of spread. These tools, in general, require
reasonable amount of data with variety. Presently, the availability of coronavirus
outbreak data is the key concern. Existing artificial intelligence-based models are
not capable enough to analyze and predict about the coronavirus outbreak. As the
coronavirus outbreak is nonlinear in nature and varying with time, so the developed
artificial intelligence-based model should be based on the multitudinal and multi-
modal data. To make the model or system more robust, one can explore the concepts
of data science to generate the knowledge. Based on this knowledge representation,
the model or system can be evolved.

6 Conclusion

As incubation period of coronavirus is of two weeks, it is difficult to predict anything

initially. Rather, the preventive measures can be adopted for the avoidance of spread.
Presented mathematical formulation is carried out on the basis of identified preventive
Coronavirus Outbreak: Multi-Objective Prediction … 27

and corrective measures and the constraints. Association and dependencies of these
measures and constraints are also explored from the evaluation of positive cases,
mortality, transmission rate, recovery and prediction of peak time of spread. This
mathematical formulation can be applied with certain time intervals to evaluate,
predict and optimize the mentioned objectives with some empirical estimation of the
constraint parameters and their weight factor values. Presently, the evaluation has not
been carried out, but the data from the reliable sources like John Hopkins University
can be used for the evaluation. Very soon the evaluation of proposed mathematical
formulation will be carried out. This leads for the scope of improvement in the
presented mathematical formulation.

References

1. World Health Organization. Coronavirus. World Health Organization, cited April 15, 2020.
Available: https://www.who.int/health-topics/coronavirus
2. International Committee on Taxonomy of Viruses, cited April 15, 2020. Available: https://talk.
ictvonline.org/
3. Gorbalenya, A. E., Baker, S. C., Baric, R. S., et al. (2020). The species Severe acute respira-
tory syndrome-related coronavirus: classifying 2019-nCoV and naming it SARS-CoV-2. Nat
Microbiol, 5, 536–544. https://doi.org/10.1038/s41564-020-0695-z.
4. Zhou, P., Yang, X. L., Wang, X. G., Hu, B., Zhang, L., Zhang, W., et al. (2020). A pneumonia
outbreak associated with a new coronavirus of probable bat origin. Nature, 579, 270–273.
https://doi.org/10.1038/s41586-020-2012-7.
5. Huang, C., Wang, Y., Li, X., Ren, L., Zhao, J., Hu, Y., et al. (2020). Clinical features of patients
infected with 2019 novel coronavirus in Wuhan, China. The Lancet, 395, 497–506. https://doi.
org/10.1016/S0140-6736(20)30183-5.
6. Chan, J. F., Yuan, S., Kok, K. H., To, K. K., Chu, H., Yang, J., et al. (2020). A familial
cluster of pneumonia associated with the 2019 novel coronavirus indicating personto-person
transmission: a study of a family cluster. The Lancet, 395, 514–523. https://doi.org/10.1016/
S0140-6736(20)30154-9.
7. Zhu N, Zhang D, Wang W, Li X, Yang B, Song J, et al. (2020) A novel coronavirus from
patients with pneumonia in China, 2019. N Engl J Med: 727-733. https://doi.org/10.1056/NEJ
Moa2001017
8. Bogoch II, Watts A, Thomas-Bachli A, Huber C, Kraemer MUG, Khan K (2020) Pneumonia
of unknown etiology in Wuhan, China: Potential for International Spread Via Commercial Air
Travel. J Travel Med: 1-3. https://doi.org/10.1093/jtm/taaa008
9. Q&A on coronaviruses (COVID-19), cited April 8, 2020. Available: https://www.who.int/
news-room/q-a-detail/q-a-coronaviruses
10. Stehlé, J., Voirin, N., Barrat, A., Cattuto, C., Colizza, V., Isella, L., et al. (2011). Simulation
of an SEIR infectious disease model on the dynamic contact network of conference attendees.
BMC Medicine, 9, 87.
11. Fisman, D., Khoo, E., & Tuite, A. (2014). Early epidemic dynamics of the west african 2014
ebola outbreak: estimates derived with a simple two-parameter model. PLoS Curr., 6, 6.
12. Das D, Santosh KC, Pal U (2020) Truncated Inception Net: COVID-19 Outbreak Screening
using Chest X-rays, DOI:https://doi.org/10.21203/rs.3.rs-20795/v1
13. Mukherjee H, Ghosh S, Dhar A, Obaidullah SM, Santosh KC, Roy K (2020) Shallow Convo-
lutional Neural Network for COVID-19 Outbreak Screening using Chest X-rays https://doi.
org/10.36227/techrxiv.12156522.v1
28 N. V. Thakur

14. Rajinikanth V, Dey N, Raj ANJ, Hassanien AE, Santosh KC, Raja N (2020) Harmony-Search
and Otsu based System for Coronavirus Disease (COVID-19) Detection using Lung CT Scan
Images. arXiv preprint arXiv:2004.03431.
15. Santosh, K. C. (2020). AI-Driven Tools for Coronavirus Outbreak: Need of Active Learning
and Cross-Population Train/Test Models on Multitudinal/Multimodal Data. Journal of Medical
Systems, 44, 93. https://doi.org/10.1007/s10916-020-01562-1.
16. Fong SJ, Li G, Dey N, Crespo RG, Herrera-Viedma E (2020) Finding an accurate early fore-
casting model from small dataset: A case of 2019-ncov novel coronavirus outbreak. arXiv
preprint arXiv:2003.10776.
17. Fong SJ, Li G, Dey N, Crespo RG, Herrera-Viedma E (2020) Composite Monte Carlo decision
making under high uncertainty of novel coronavirus epidemic using hybridized deep learning
and fuzzy rule induction. Applied Soft Computing, 106282.
18. Coronavirus disease (COVID-19) outbreak situation, cited April 14, 2020. Available: https://
www.who.int/emergencies/diseases/novel-coronavirus-2019
19. COVID-19 India, cited April 14, 2020. Available: https://www.mohfw.gov.in/
AI-Enabled Framework to Prevent
COVID-19 from Further Spreading

Dalip and Deepika

Abstract The first case of COVID-19 was detected in Wuhan the city of China
and now it has been spread over more than 100 countries. Due to this epidemic
situation, the World Health Organization declares emergency in various cities and
countries which causes an outbreak for COVID-19 that leads to control this virus
spread. AI plays a significant role to control the COVID-19 a pandemic disease.
This chapter explains an automated Artificial Intelligence Enabled (AIE) framework
which is designed to control the further spread of the novel coronavirus. Artificial
intelligence (AI) is used to address about the current disease coronavirus if applied in
an innovative way. The main motive of this chapter suggested that AI is merged with
the latest technologies like machine learning (ML) and Global Positioning System
(GPS) which are used to design an innovative automation system which is used to
control the further spreading of coronavirus. For experimental results, the concept of
AIE framework is designed by considering the scenario of urban and rural regions
of the state of Haryana(India). The comparative analysis of AIE with traditional
frameworks is shown in Table 1 which defines the novelty and efficiency of designed
framework. Finally on the basis of this innovative AIE-designed automation system,
the 98 and 97% accuracy is achieved in urban and rural regions respectively. Finally,
we can conclude that this automated system is fully utilized and works efficiently to
control and prevent the further spreading of this novel epidemic coronavirus.

Keywords Artificial intelligence · COVID-19 · Coronavirus · Automation

system · GPS · Regions · Locations · Machine learning

Dalip (B) · Deepika

M.M. Institute of Computer Technology and Business Management, Maharishi Markandeshwar
(Deemed to be University), Mullana–Ambala, Haryana, India
e-mail: [email protected]
Deepika
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 29
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_4
30 Dalip and Deepika

1 Introduction

This novel coronavirus (COVID-19) threats a human life across all over the world.
As per the existing data analytics, total number [6] of COVID-19 cases are 2,192,974,
total deaths are 147,376, total recovered are 554,676, total active are 1,490,922 and
total serious critical are 56,566 by the time of writing this chapter. It also brings
huge economic losses along with the serious threats to human lives and proves to
be an infectious disease. A group of experts is working toward the development of
vaccines against COVID-19 but no vaccine is founded for coronavirus till date. This
epidemic virus is spread from the Wuhan which acts as an origin of this pandemic
disease and now spread almost in the entire globe over more than 100 countries.
China takes the first initiative to implement preventive measures to control this virus
as this country has strong medical background and has best medical facilities. There
are different technologies like artificial intelligence (AI), Internet of Things (IoT) and
data science that are used to track and fight against this disease [1]. The innovative
applications can be designed with the help of AI to track and fight against COVID-19
such as location tracking and geo-fencing are used to track the infected peoples. By
analyzing news reports, social media platforms and government documents AI can
be used to identify, track and forecast outbreaks. AI designed models are used to
help medical professional to a greater extent as huge chances are there to get them
infected from this disease as they are intact contact with the infected persons. Another
technology is drones which helps the medical workers to protect themselves from
this infected disease in order to provide the medical supplies to their patients in a
safest way without going in contact with the infected person. A robotics application
of an AI can protect our homes, buildings, streets and cities from this virus and this
disease cannot harm robots so they are used to accomplished many tasks such as
cleaning, sterilizing, delivering food and provide medical facilities to the patients to
avoid the human to human interaction. Robots are also utilized by the police officials
in an effective way in order to protect them from this virus by avoiding the regular
public interaction. So with these types of AI systems the police persons, medicated
staff and sweepers can protect themselves. The IoT-based patient care system can
help to measure their temperatures and upload the measurements with their mobile
devices to the cloud for further analysis. Due to the lack of AI-enabled automated
models, therefore an AIE automated framework is designed by using the existing
technologies which helps us to control and prevent the further spread of this novel
epidemic coronavirus. The proposed framework automatically calculates the fuzzy
weight and threshold values for decision making.
The rest of the chapter is categorized as follows. Sect. 2 discussed about the
related work and elaborates the contribution of different researchers over COVID-
19. Motivation and contribution are explained in Sect. 3. A novel methodology of
the proposed framework is shown in Sect. 4. The results and discussion are analysis
in Sect. 5. At last, the conclusion is present in Sect. 6.
AI-Enabled Framework to Prevent COVID-19 from Further Spreading 31

2 Background

The World Health Organization (WHO) [1] has been declared coronavirus an
epidemic contagious disease that transmit from human to human very rapidly which
becomes a danger for human life in all the age groups. So, if the rate of transmission
between human to human is reduced like 0.25 then we can control this virus at some
extent [7]. This virus spread rapidly across various countries and among the whole
world and also proves to be a threat for the global health and for the economy as
well. This virus named as novel coronavirus disease COVID-2019 [1] which states
a total number of 80,868 confirmed cases and 3,101 deaths [6] in Chinese main-
land by National Health Commission of China by March 8, 2020. In the early stage
of this virus, no one thinks about this epidemic situation comes and this will rise
to an emergency in China. To outbreak this pandemic, the Chinese officials take
some preventive and strict actions to restrict their all resources especially traveling,
business and schools. The lockdown and curfew need to be implemented to save
the human lives and to outbreak this pandemic. It becomes a difficult task to know
about the time span of this outbreak of coronavirus. For this, some automated models
are need to be designed with the help of some existing models and latest technolo-
gies. With the help of existing studies, various models of explainable AI are there
that explains how people define, generate, select, present and evaluate explanations
[2]. The disease spread simulation model and contact tracking system are helpful
to prevent and control this infectious disease [3]. These contact tracking system
are helpful for the medical [5] professionals to identify the infected and high-risk
peoples from this disease and to locate and then isolate them so that they cannot
further spread this virus among other human beings [4]. The other innovative tech-
niques like composite Monte-Carlo model (CMC) uses GROOMS methodology [7],
deep learning and various soft computing algorithms and techniques like Fuzzy Rule
Induction, BFGS and PNN that provides efficient results for decision support. As
a result, the data collected by the agencies of China government using the CMC
methodology on this epidemic disease COVID-19 are superior than their earlier
methods. The forecasting model [8] uses data mining and machine learning concepts
which uses Polynomial Neural Network with Corrective Feedback (PNN + CF)
which calculate the prediction time with lower error rates and becomes a best tech-
nique to determine the critical times of this disease outbreak. The forecasting model
is one of the best alternatives to deal with the future epidemics. It is important to
predict the lifetime of this virus in early stages so that preventive measures should be
taken to spread this virus like schools, commercial, etc. are closed. Artificial intel-
ligence (AI) are also deployed with machine learning algorithms, test models and
test cases which are used for data analysis and decision-making processes sets a new
dimensions in healthcare. AI-driven tools are implemented during the time of data
collection where active learning [9] is to be deployed with it. It means AI tools are
used to determine this disease outbreak forecast time which spread all over the globe
and is useful to prevent this coronavirus further spread.
32 Dalip and Deepika

To prevent and control this epidemic disease, an AI-based contact tracing systems
are suggested. The motivation and contributions of the AIE framework are explained
in the next section.

3 Motivation and Contributions

The existing studies present different research work on COVID-19 such as compara-
tive analysis, innovative techniques, models and AI-based decision-making systems,
etc. The challenge of designing an efficient framework is that there are lack of AI
designed enabled automation systems for coronavirus which inspires us to design the
Artificial Intelligence Enabled (AIE) automated framework that helps to control and
prevents COVID-19 from further spreading. The proposed methodology is based on
face and full body detection that prevent coronavirus spread by tracing the camera
locations in both the urban and rural regions; also send the alerts and notifications at
authorized center on rigid time.

4 A Novel Methodology

A novel framework is designed which helps to detect the COVID-19 infected persons
from both the rural and urban regions on different times and locations by using
preventive measures. The proposed system automatically divides the regions into
three clusters or zones based on their fuzzy weights and also set the threshold values
to the movement attributes. The movement attributes are those which are useful to
decide the parameter for movement in lockdown situations. The zones categorization
is done on the basis of more, less and not affected with COVID-19 as they are
considered as RED, ORANGE and GREEN zones, respectively. The color label of
these zones will depend on their fuzzy weights which will automatically update.
The fuzzy weights assigned in these zones on the basis of the occurrence of
COVID-19 causes. The RED zone is known as hot spotted zones; there will be
no movement of public and services in between and full lockdown applicable. In
ORANGE zones, there is partial movement of public and services with some condi-
tions. In GREEN zones, there is a full movement of public and services under
government guidelines. The full body and face detection module is implemented
for screening the peoples and automatically detect persons those who are not follow
the lockdown rules. The screening of the public is done on their gestures. If the
gestures of the detected peoples are normal (no gestures of cough, cold and fever)
than screening test will be negative and no action will be taken against those but if
the gestures of the detected peoples are abnormal (gestures of cough, cold and fever)
than screening test result will be positive and designed system send their location
automatically at the authorized centers and government can take necessary action
for their quarantine. There are three preventive measures which are to be taken for
AI-Enabled Framework to Prevent COVID-19 from Further Spreading 33

Table 1 Comparative
Framework Traditional Novel AIE
analysis between traditional
components framework framework
framework and novel AIE
framework for uniqueness Automation Partially Fully
Qualitative division No Yes
Coverage area Urban only Both urban and
rural
Fuzzy weight No Yes
assignment
Dynamically infected No Yes
zone division

COVID-19 are social distancing, avoid social gathering and to wear face mask. The
designed system automatically detects the person who violates these safety measures.
If the safety measures conditions are fulfilled than no action will be taken, otherwise
it sends the location to the authorized centers. It also sends alert notifications on time
to the village authorities regarding those persons who disobey the safety measures
given by the government. The uniqueness of the proposed system is shown in Table 1.
For comparative analysis, the different parameters are taken. The AIE framework is a
fully automated system which provides the accurate qualitative information on time
about the infected person such as “more infected”, “infected”, “less infected” and
“not infected” depends on their fuzzy score. It also tracks and locates the specific
information of the infected people on the map and also provides the history of recently
contacted persons with infected one. Geo-fencing provision is also used here which
informs the authorized center and users also if any infected person leaves the isolation
ward and try to enter in the fencing zone. India is primarily an agricultural country.
During lockdown situations, the government of any country cannot stop the agricul-
ture work for a long time. Therefore, there are more chances of COVID-19 spread
in rural regions.
The proposed framework is also very helpful in rural regions. The system
can monitor the safety measures against COVID-19 using drone technology. The
designed system sends the alerts to the authorized centers in case of framers and
laborers violate the safety measures against COVID-19. So that necessary action
would be taken by authority on time. The application software AIE is implemented
in Python programming language using Haar Cascades classifiers.

5 Results and Discussion

The designed system is tested over both the regions (Urban and Rural) on different
locations, different dates and times. On x-axis, the first, second and third rows indicate
locations, starting time and ending time, respectively.
34 Dalip and Deepika

Fig. 1 Artificial Intelligence Enabled (AIE) automated framework

The given calculations are obtained up to writing this chapter during the lockdown
period for the best results. Figure 2 shows the total number of cases detected for the
different safety measures against COVID-19 on different locations and time periods
automatically in urban regions. The 98% rate of accuracy is achieved by the AIE
model in urban regions as shown in Fig. 2. Similarly, Fig. 3 showcases the total
no. of infected cases for different safety measures automatically against COVID-19
on different locations and the time period in the rural region. The rate of accuracy
measured by the AIE model is 97% in rural regions.

6 Conclusion

In this chapter, the automated AIE framework is designed on the basis of some
existing models. After the analysis and study of some relevant and important findings
from the existing research using the AI technology against COVID-19 which have
been provided some insight into how this work can be used in an explainable designed
framework, it has been quite cleared from Table 1 some unique components of the
designed framework make it unique and more efficient. The AIE model can replicate
its role in helping to control against the spread of COVID-19 pandemic in an efficient
way with (97–98) % rate of accuracy.
AI-Enabled Framework to Prevent COVID-19 from Further Spreading 35

Fig. 2 Accuracy of AIE model in urban regions at different locations

Fig. 3 Accuracy of AIE model in rural regions at different locations

References

1. Zhou, M., Zhang, X., & Qu, J. (2020). Coronavirus disease 2019 (COVID-19): A clinical update.
Frontiers in Medicine. https://doi.org/10.1007/s11684-020-0767-8.
2. Miller, T. (2019). Explanation in artificial intelligence: Insights from the social sciences. Artificial
Intelligence, 267, 1–38.
3. Tsui, K., Wong, Z. S., Goldsman, D., & Edesess, M. (2013). Tracking infectious disease spread
for global pandemic containment. IEEE Intelligent Systems, 28(6), 60–64.
4. Chen, H., Yang, B., Pei, H., & Liu, J. (2018). Next generation technology for epidemic prevention
and control: Data-driven contact tracking. IEEE Access, 1–1. https://doi.org/10.1109/access.
2018.2882915.
5. Lupia, T., Scabini, S., Pinna, S. M., Di Perri, G., De Rosa, F. G., & Corcione, S. (2020). 2019
novel coronavirus (2019-nCoV) outbreak: A new challenge. Journal of Global Antimicrobial
Resistance, 21, 22–27.
36 Dalip and Deepika

6. https://www.worldometers.info/coronavirus/#countries.
7. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing. https://doi.org/10.1016/j.asoc.
2020.106282.
8. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Finding an accurate
early forecasting model from small dataset: A case of 2019-nCoV novel coronavirus outbreak.
International Journal of Interactive Multimedia and Artificial Intelligence, Special Issue on Soft
Computing, 6(1), 132–140. https://doi.org/10.9781/ijimai.2020.02.002.
9. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: Need of active learning and
cross-population train/test models on multitudinal/multimodal data. Journal of Medical Systems.
https://doi.org/10.1007/s10916-020-01562-1.
Artificial Intelligence-Enabled Robotic
Drones for COVID-19 Outbreak

Dharm Singh Jat and Charu Singh

Abstract Artificial intelligence (AI) can help to address coronavirus if applied

creatively. Artificial intelligence training models to deal with COVID-19 are having
challenges as historical data is still not available. Drones and robots equipped with IoT
devices provide raw data that needs computing analysis to make that data meaningful
and actionable without human involvement. The power of AI and edge computing lies
in its ability to process a massive amount of data at breakneck speed and improving
efficiency. It enables big data analytics and deployment of algorithm and transmission
of data across edge and cloud. This chapter presents how current artificial intelli-
gence enables robotic drone applications and network connectivity is used to improve
their performance and increase efficiency in various situations to fight COVID-19.
Further, it provides an in-depth review and analysis of literature on related work on
COVID-19 outbreak. It also gives the necessary background for future research in
edge intelligence, AI-enabled robotic drone and intelligent networks.

Keywords Artificial intelligence · COVID-19 · 5G · Edge computing · Robotic

drone

1 Introduction

Artificial Intelligence (AI) is the fourth industrial revolution technology in the current
and developing environment. People from the various fields are trying to implement
AI to solve their problems. Artificial intelligence has not yet been impactful to fight
COVID-19 outbreak due to massive and noisy data. It is up to us to identify new and
innovative ways to leverage what AI can do. Artificial intelligence-enabled drones

D. S. Jat (B)
Namibia University of Science and Technology, Windhoek, Namibia
e-mail: [email protected]
C. Singh
University of Southern California, Los Angeles, CA, USA
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 37
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_5
38 D. S. Jat and C. Singh

can be useful to fight COVID-19 by the monitoring system and can process the
captured data and report back in real time.
Self-directed drones pose many challenges like lack of wireless bandwidth, limited
processing capacity, non-availability of sufficient power, accuracy and latency for
real-time applications [1]. In this study, the experimental results describe that the
integration of edge computing devices and drones processing power can save the
essential wireless bandwidth requirement and therefore increase throughput, improve
scalability and decrease latency for sensitive application applications [1]. Artificial
intelligence software can be useful to guide drone for collecting the raw data and
process it into actionable insights for emergency teams including doctor, police, etc.,
during COVID-19 or similar situation [2].
The study presents a framework that determines the optimal location of the distri-
bution centre in a disaster-affected region for low-cost drones for emergency supply
and services. The drones may not be useful for densely and sparsely population
centres due to drone routing limitations. Further study also do not consider power
constraints and assume fully charged battery will be available for drones during
distribution [3]. A study related to edge intelligence provides insights for possible
future research in AI at the edge [4]. The rapid advancement of communication
technologies and Internet of things devices (IoTs) generates a billion of data bytes
at the network edge. Therefore, there is a high demand for integrated AI and edge
computing to process and analyse the data at the edge. Further, the study describes
artificial intelligence on edge as well as intelligence-enabled edge computing. The
edge intelligence is in the early stage and has attracted more and more researchers,
organisations and industries to get involved in research in this area [4].

2 Motivation and Contributions

Experts recognised a novel coronavirus (nCoV) in December 2019 in China and

governments and authorities all over the world are facing challenges to control
COVID-19 outbreak [5]. A study states that there is a need to predict the epidemic
outbreak for timely decision and measures, including closing universities, schools,
international and interstate borders and suspending crowded community services
and customers [6]. For post-COVID-19 outbreak, the suitable training models for
AI-driven tools were discussed and also mentioned that many machine learning
(ML) algorithms are useful for decision-making related to medical procedures and
treatments and to identify infested COVID-19 cases [7].
More researches recently published on chest X-rays for detection of infested
COVID-19 cases. One of the research suggested the method based on convolutional
neural network for automatic chest X-rays for detection of infested COVID-19 [8].
A study proposed an image-assisted system to extract information from lung CT
scans of COVID-19 infected cases. It will also assist the pulmonologist in helping in
developing the best treatment plan [9]. X-ray imaging techniques are cheaper than CT
Artificial Intelligence-Enabled Robotic Drones for COVID-19 … 39

scan systems. Therefore, a study suggested Truncated Inception Net deep learning
model is used to detect infested COVID-19 positive cases using chest X-rays [10].
Most of the researches have been conducted for COVID-19 outbreak which
focuses on chest X-rays detection of infested COVID-19 cases [5–10].
According to the researchers’ best knowledge, a few published literature exists
on artificial intelligence-enabled robotic drones technology; therefore, there is a gap
in research for drones technology and communication for COVID-19 outbreak.
This study describes how artificial intelligence-enabled robotic drones are being
used to improve their performance and increase efficiency in various situations during
and post-COVID-19 outbreak. In COVID-19 outbreak, the intelligent drone can be
used to recognise people of interest from the large gathering. Moreover, the problem
of power and computing can be solved by implementing the edge computing devices
so that drones can offload their data for pre-processing and receive desired action to
be taken.
In summary, this chapter provides an in-depth review of the literature on
artificial intelligence-enabled robotic drone used to fight COVID-19 and other
similar scenarios. The chapter also describes AI-enabled intelligent networks, future
generation networks (5G) and intelligent edge.

3 Related Work on COVID-19 Outbreak

These days, drone technology is being used in various applications, including natural
disaster relief, defence, security, construction, agriculture, etc. Individuals and small
organisations cannot afford as such systems are bulky and expensive. The AI-enabled
drone can be used for the monitoring system as they can now process what they record
and report back in real time. AI can be useful to guide the drone to capture raw data
and provide it for police, doctors and other emergency teams for the action to be taken.
The AI-powered drone can be used to identify the hotspot areas during COVID-19
outbreak.
More research work is needed in the area related technology to make drones a
valuable and reliable source of information for various organisations. Drones and
other sensor-equipped devices provide raw data which requires computing analysis
to make that data meaningful and actionable without human involvement. Edge
computing devices and IoT applications enable analytics, machine learning, and
seamless deployment of algorithm and transmission of data across the edge and the
cloud [11].
40 D. S. Jat and C. Singh

3.1 Robotic Technology

Intelligent robots are progressively being used in various fields. The study suggested
optimal methods for the control of mobile robots with the help of artificial intelligence
technology. Further, the study implemented this AI technology for traffic control
robot modelling and robotic sprinkler system in the field of irrigation. This developed
mechanism can also be used for the algorithm design of mobile robots in other
applications [12]. A police robot has been used to monitor areas of Tunisia’s capital,
Tunis, to make sure that people are following a coronavirus lockdown. This robot
can help to reduce physical contact between police and residents/patients suffering
from COVID-19 [13].
China is progressively using robotic technology and artificial intelligence (AI)
during these exceptional times. This technology is used in the form of disinfecting
robots, thermal camera-equipped drones, etc., to fight coronavirus. During the peak
outbreak of the disease in China, smart helmets, disinfecting robots, and advanced
facial recognition software are being used to fight against COVID-19 [14]. Many
Chinese companies have innovated automated technologies for contact-free delivery,
spraying sanitizers and executing necessary diagnostic tasks to reduce the spread of
the infection. In Shenzhen city, MicroMultiCopter drones are also being deployed to
transport medical samples and also to perform thermal imaging [14].

3.2 Drones Against COVID-19

Nowadays, mini-unmanned aerial vehicles (UAVs) called drones are being used
in many application related to agriculture, defence, communications, surveillance,
public services, photography due to several advantages like low cost, low power
requirement, easy to transport, etc. These drones can also provide real-time activities
by transmitting real-time video streaming and can be used to pose severe outbreak to
community security like situational awareness information for COVID-19 outbreak.
The main problems faced by implementing drones in public residence are a direct
physical attack on building, property, people and other infrastructure. Additionally,
the research suggested a deep learning-based method to identify the possible illegal
drones in real time in the area of interest [15]. The study proposed a model for drone
crowdsourcing and demonstrated the use of drone in a fire situation at the island.
In such a situation, with the help of drone and devices, people can be located who
are stuck. Drones are having limited resources and can carry limited weight and
lightweight battery only. Some applications also require multiple drones, which are
working in a coordinated manner. Therefore, the combination of additional resources
(edge devices) beyond the drone capability and multiple drown can overcome these
limitations [16].
The study presents the application of drones in precision agriculture farming. The
farmer can control the overall area of the attacked plants or observe the activities
Artificial Intelligence-Enabled Robotic Drones for COVID-19 … 41

of parasites moving among plants by drones equipped with the camera [17]. In
agriculture, plant disease can be identified in early stage by using drones together
with camera and the global positioning system (GPS) device. This early detection
will help in stopping the infection from the crop. By implementing the Internet of
things (IoT) devices and drone together, real-time data can be gathered and analysed
with the help of image processing algorithm which performs the crop classification
and disease detection [18].
A similar drone was deployed to sanitise the hotspots of COVID-19 by spraying
a 1% sodium hypochlorite solution as a disinfectant over the locality by the South
Delhi Municipal Corporation (SMDC) in India [19]. To control the spread of the
virus, Delhi is under lockdown, and no public transport, including private buses, taxis
and autorickshaws are allowed to ply on the roads. Delhi police make use of drones
to monitor the situation in various areas, during the complete lockdown in Delhi
[20]. Networked mobile robotics integrated with cameras are using in unmanned
aerial vehicles (UAVs) and remotely steered aircraft systems to combat COVID-
19 in improving responses to humanitarian emergencies, health monitoring and
to detect the infectious and respiratory conditions which include monitoring body
temperatures, heartbeats and respiratory rates [21].
Apart from robots and drones, China has also deployed many sophisticated surveil-
lance mechanisms to keep a check on infested individuals and impose quarantines.
Across China facial recognition cameras are ubiquitous. Some Chinese companies
have introduced AI-enabled temperature detection systems and also to trace the
people without masks [14].

3.3 AI-Enabled Intelligent Networks

Today 5G provides an economical solution for collecting and analysing data for
real-time applications. AI techniques can improve automation to manage robotic
drones, IoT connected devices and solve the problems that arise from 5G use cases
and network complexity. Today’s current mobile networks are providing services
to the robotic drones in the low-altitude airspace only. 5G networks are capable of
providing efficient mobile wireless connectivity for large-scale drone placements for
various applications [22]. The integration of IoT and 5G technologies, operation and
maintenance of communication networks, and the various industrial applications are
facing many challenges. Artificial intelligence (AI) technology-based models have
capabilities such as high computing massive data analysis, for network operation,
management and maintenance in the future generation of wireless networks (5G) era
[23].
For a real-time sensitive application like robotic drones application for COVID-19
outbreak, there is a need for reliable, low latency communications and ubiquitous
connectivity with robotic drones. 5G next-generation wireless networks promise
enhanced mobile broadband (eMBB), ultra-reliable low latency communications
42 D. S. Jat and C. Singh

(URLLC) and massive machine-type communications (mMTC) in a real-time appli-

cation and dynamic environment. Data-driven functions and communication in 5G
wireless networks can be empowered by using the concepts of machine learning
across the wireless edge and core infrastructure [24].
Machine learning which is the subset of artificial intelligence (AI) has done a
remarkable job in the communication technologies and can help telecom industries
to optimise their investment and helpful in planning and smart slicing of 5G network.
In the deployment of 5G wireless networks, AI-enabled edge is essential to minimise
the cost. Strategy analytics predicts that the integration of artificial intelligence (AI)
and edge computing technologies will be beneficial to deliver 5G connectivity as
promised to improve the return on investment (ROI) [25].

3.4 5G: Coronavirus Outbreak

In the future, there will be a massive demand for high definition video and high resolu-
tion for mobile communication. The requirement for mobile broadband connectivity
will continue to increase as the things around us become more connected.
The availability of lower bandwidth in 2G communication systems did not fulfil
the need for immediate requirement of mobile Internet users. This led to a demand
for new upgraded 3G standards, which advanced to provide fast data rate services
and more bandwidth available for voice communication. Further, the 4G mobile
communication system was developed to provide a high data rate service for multi-
media communication. Groupe Speciale Mobile (GSM) later called as Global System
for Mobile Telecommunication has provided mobile communications since 1991 to
the world. In 2019, according to the GSM Association (GSMA) Intelligence, 4G
long-term evolution (LTE) is also essential mobile network technology to support
higher peak bit rates worldwide, and more than three billion number of connections
worldwide.
Therefore, 1G provides only voice communication, 2G improves voice quality and
provides a text messaging, 3G offers integrated voice, and affordable mobile Internet
and 4G provides high data rate transmission for mobile multimedia communication.
With the anticipated growth of the Internet of things (IoTs) and smart devices in
future, there will be more users, more diverse range of device types than ever before.
Moreover, an emergency like COVID-19 outbreak will require low latency, high
bandwidth, improved reliability and long battery life for devices like drones, robots,
etc. All 4G LTE evolution will not be enough to handle this new wave of heteroge-
neous big data traffic. Therefore, there is a need for the upgraded new generation
(5G) radio system and network architecture for the availability of high data rate
broadband network and ultra-robust and low latency connectivity to connect people
of everything. IoT applications in various fields are likely to be the main driver for
further growth of WiFi and cellular communication. Therefore, IoTs applications
will enable the devices like drones, robot, health devices etc., to interact and share
Artificial Intelligence-Enabled Robotic Drones for COVID-19 … 43

data in real time. Nowadays, the network connects human, but in future 5G systems
will be about people of things and become more and more connected every day.
Deploying wired broadband networks requires high investment, and it is not viable
for least densely populated countries like Namibia. The estimated population density
of Namibia is 2.6 persons per km2 [26]. Fixed wireless access (FWA) networks can
be used for more cost-effective and high-speed broadband networks installation.
WiMAX or LTE fixed wireless access networks are using for last-mile connec-
tivity for many years, but the speed is slow in comparison with the wired high-speed
broadband network. 5G communication also called as millimetre-wave communi-
cation is being considered for the use of above 6 GHz and the frequency bands are
mentioned as the millimetre band (mm-band). Now by using 5G, the speed of FWA
networks is comparable to the fibre network [27]. For rural area FWA networks,
using below 6 GHz spectrums band can be used for broader coverage by applying
massive MIMO technologies.
These days’ people also believe that 5G radiation is the cause of coronavirus
outbreak. Keri Hilson thinks that the cause of the coronavirus outbreak is due to
5G radiation [28]. The assumption behind this is that the Wuhan is the first city in
China with 5G connectivity. However, at the time of COVID-19 spread, Wuhan was
not the only city to have 5G connectivity, several other cities were also having 5G
connectivity. People also assume that 5G harms the immune system of human beings.
5G is not accelerating the spread of the new coronavirus as there is no scientific
evidence to support the negative impact on the immune system [29]. The statement
related to the spread of the new coronavirus is incorrect as 5G wireless connectivity
is based on radio frequency and therefore does not produce any type of viruses.

3.5 5G: COVID-19 Affected Africa

People also believe that COVID-19 has not affected Africa as it is “Not a 5G region”
[28]. This statement is not valid, and 5G wireless networks are not the cause that
Africa has not been affected by COVID-19. According to the report released by the
World Health Organisation (WHO), Regional Office for Africa reported that COVID-
19 is spreading nearly in every country. Around 17,000 COVID-19 positive cases
and 900 deaths across the Africa continent as on 16 April 2020 were reported. South
Africa is the country which has a severe outbreak of new coronavirus, while cases
are still increasing in West and Central Africa [30].
African countries took preventive measures at the very initial stage of COVID-19
spread [31]. Appropriate precautionary measures were taken to deal with COVID-19
outbreak on time as soon as WHO declared COVID-19 as a global pandemic. The
Republic of Namibia also took the appropriate precautionary measure to deal with
COVID-19 as soon as two confirmed cases of COVID-19 were found on Namibian
soil. President of Namibia immediately declared a State Emergency and took the
measures which aim at curbing the spread of the disease and mainly to prevent the
occurrence of local transmission [32].
44 D. S. Jat and C. Singh

4 Conclusion

This chapter provides an in-depth review and analysis of literature on related work
on COVID-19 outbreak, robotic technology, drones used to fight COVID-19, AI-
enabled intelligent networks, 5G: coronavirus outbreak and 5G: COVID-19 affected
Africa. Artificial intelligence has not yet been found impactful to fight COVID-19
due to big and noisy data. Also, there is a lack of historical data on which to train
AI models to fight COVID-19. Artificial intelligence and edge computing devices,
including robots and drones, can contribute to the fight against COVID-19: early
warning systems for disasters and social control. Massive and noisy data can also
be filtered at the edge by implementing edge computing devices. The statement
related to the spread of the new coronavirus is due to 5G technology is incorrect
because it is based on radio frequency and does not create viruses. Also, there is no
scientific evidence to support that 5G harms the immune system. To conclude, this
chapter describes AI-enabled intelligent networks, future generation networks (5G)
and intelligent edge.

References

1. Wang, J., et al. (2018). Bandwidth-efficient live video analytics for drones via edge computing.
In Proceedings of 2018 3rd ACM/IEEE Symposium on Edge Computing SEC 2018 (pp. 159–
173).
2. International Telecommunication Union. (2019). Artificial intelligence for good. [Online].
Available: https://www.itu.int/en/mediacentre/backgrounders/Pages/artificial-intelligence-for-
good.aspx.
3. Chowdhury, S., Emelogu, A., Marufuzzaman, M., Nurre, S. G., & Bian, L. (2017). Drones
for disaster response and relief operations: A continuous approximation model. International
Journal of Production Economics, 188, 167–184.
4. Deng, S., Zhao, H., Fang, W., Yin, J., Dustdar, S., & Zomaya, A. Y. (2020). Edge intelligence:
The confluence of edge computing and artificial intelligence. IEEE Internet of Things Journal,
PP(c), 1.
5. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing, 106282.
6. Fong, S. J., Li, G., Dey, N., Gonzalez-Crespo, R., & Herrera-Viedma, E. (2020). Finding an
accurate early forecasting model from small dataset: A case of 2019-nCoV novel coronavirus
outbreak. International Journal of Interactive Multimedia and Artificial Intelligence, 6(1), 132.
7. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: Need of active learning
and cross-population train/test models on multitudinal/multimodal data. Journal of Medical
Systems, 44(5), 1–5.
8. Mukherjee, H., Ghosh, S., Dhar, A., & Obaidullah, S. (2019). Shallow convolutional neural
network for COVID-19 outbreak screening using chest, 1–10.
9. Rajinikanth, V., Dey, N., Raj, A. N. J., Hassanien, A. E., Santosh, K. C., & Raja, N. S. M. (2020).
Harmony-search and Otsu based system for coronavirus disease (COVID-19) detection using
lung CT scan images.
Artificial Intelligence-Enabled Robotic Drones for COVID-19 … 45

10. Das, U. P. D., & Santosh, K. C. (2020). Truncated inception net: COVID-19 outbreak screening
using chest X-rays. [Online]. Available: https://doi.org/10.21203/rs.3.rs-20795/v1.
11. Guttman, C. (2019). Drone innovation turns to edge computing. 2019 Nutanix, Inc.
[Online]. Available: https://www.nutanix.com/theforecastbynutanix/technology/drone-innova
tion-turns-to-edge-computing%0A.
12. Irshat, K., Petr, R., & Irina, R. (2018). The selecting of artificial intelligence technology for
control of mobile robots. In International Multi-Conference on Industrial Engineering and
Modern Technologies. FarEastCon 2018 (pp. 1–4).
13. Jawad, R. (2020). Coronavirus: Tunisia deploys police robot on lockdown patrol. [Online].
Available: https://www.bbc.com/news/world-africa-52148639.
14. Jakhar, P. (2020). Coronavirus: China’s tech fights back. [Online]. Available: https://www.bbc.
com/news/technology-51717164.
15. Chen, H., Wang, Z., & Zhang, L. (2020). Collaborative spectrum sensing for illegal drone
detection: A deep learning-based image classification perspective. China Communications,
17(2), 81–92.
16. Alwateer, M., Loke, S. W., & Fernando, N. (2019). Enabling drone services: Drone
crowdsourcing and drone scripting. IEEE Access, 7, 110035–110049.
17. Potrino, G., Palmieri, N., Antonello, V., & Serianni, A. (2018). Drones support in precision
agriculture for fighting against parasites. In 2018 26th Telecommunications Forum, TELFOR
2018—Proceedings (pp. 1–4).
18. Kitpo, N., & Inoue, M. (2018). Early rice disease detection and position mapping system using
drone and IoT architecture. In Proceedings—12th SEATUC Symposium SEATUC 2018 (p. 4).
19. IANS. (2020, April 2). Drone deployed to sanitize Nizamuddin Markaz, the ‘epicentre’ of
COVID-19 outbreak. The New Indian Express.
20. Delhi police uses drone to monitor the situation amid lockdown. (2020, April 1). ABP Live.
[Online]. Available: https://news.abplive.com/topic/drone.
21. Cozzens, T. (2020, March 26). Pandemic drones to monitor, detect those with COVID-19.
[Online]. Available: https://www.gpsworld.com/draganfly-camera-and-uav-expertise-to-help-
diagnose-coronavirus/#comments.
22. Takacs, A., Lin, X., Hayes, S., & Tejedor, E. (2018). Drones and networks: Ensuring safe and
secure operations. Ericsson White Paper, 14.
23. ZTE uSmartNet network intelligent solution. ZTE Corporation. [Online]. Available: https://
www.zte.com.cn/global/solutions/201905201711/201905201746/201906170947.
24. Chen, M., Challita, U., Saad, W., Yin, C., & Debbah, M. (2019). Artificial neural networks-
based machine learning for wireless networks: A tutorial. IEEE Communications Surveys &
Tutorials, 21(4), 3039–3071.
25. de Grimaldo, S. W. (2018). AI + edge computing essential in the 5G era (pp. 1–9).
26. Census projected population. The Government of Namibia. [Online]. Available: https://gov.
na/population.
27. Samsung. (2018). 5G fixed wireless access. Samsung Electronica Co., Ltd. [Online]. Avail-
able: https://images.samsung.com/is/content/samsung/p5/global/business/networks/insights/
white-paper/samsung-5g-fwa/white-paper_samsung-5g-fixed-wireless-access.pdf.
28. Cedric “BIG CED” Thornton. Keri Hilson believes the coronavirus outbreak may have been
caused by 5G radiation. [Online]. Available: https://www.blackenterprise.com/keri-hilson-bel
ieves-the-coronavirus-outbreak-may-have-been-caused-by-5g-radiation/.
29. Milan, A. (2020). Coronavirus UK: Why do people think 5G is responsible for the COVID-19
pandemic? Metro, UK. [Online]. Available: https://metro.co.uk/2020/04/02/coronavirus-uk-
people-think-5g-responsible-covid-19-pandemic-12500019/?ito=cbshare.
30. World Health Organisation. WHO, WFP and AU deliver critical supplies as COVID-19
accelerates in West and Central Africa. [Online]. Available: http://whotogo-whoafroccmas
ter.newsweaver.com/JournalEnglishNewsletter/10dge3vjomny48iiujdam4?email=true&lang=
en&a=11&p=56844493.
46 D. S. Jat and C. Singh

31. African countries move from COVID-19 readiness to response as many confirm cases. World
Health Organization. [Online]. Available: https://www.afro.who.int/health-topics/coronavirus-
covid-19.
32. President of the Republic of Namibia. (2020, March 28). State of emergency—COVID-19
regulations: Namibian constitution, government gazette of the Republic of Namibia (No. 7159).
Windhoek, Republic of Namibia.
Understanding and Analysis of Enhanced
COVID-19 Chest X-Ray Images

M. C. Hanumantharaju, V. N. Manjunath Aradhya,

and G. Hemantha Kumar

Abstract The 2019 coronavirus disease (COVID-19) with its origin in China has
spread rapidly to other nations and infected millions of people. In this context, this
paper proposes the development of algorithm that enhances the details of images and
assists the doctors in knowing the exact location of affected area. The proposed tech-
nique improvises the most popular image enhancement algorithm, namely, multiscale
retinex and adjusts the parameters to intensify the details of chest X-ray/CT images of
COVID-19 patients. Multiscale retinex (MSR) is human perception-related enhance-
ment algorithm which improves intensity, contrast, and sharpness in medical image
through dynamic range compression. The proposed scheme improves the details
of images and validates the resulting images using novel metric called wavelet
energy. The proposed study is evaluated on images of COVID-19 patients have
been obtained from the open-source GitHub repository. Considering the experi-
mental result presented and performance metric, the proposed algorithm has provided
important details to doctors in making right decision.

Keywords COVID-19 · Image enhancement · Improved multiscale retinex ·

Wavelet energy · X-ray/CT images

M. C. Hanumantharaju
Department of Electronics and Communication Engineering, BMS Institute of Technology and
Management, Bengaluru, India
e-mail: [email protected]
V. N. Manjunath Aradhya (B)
Department of Computer Applications, JSS Science and Technology University, Mysuru 570006,
India
e-mail: [email protected]
G. Hemantha Kumar
Department of Studies in Computer Science, University of Mysore, Mysuru 570006, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 47
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_6
48 M. C. Hanumantharaju et al.

1 Introduction

Globally, nearly 3 million confirmed cases of COVID 2019 have been reported to
World Health Organization (WHO) and more than 2,10,000 deaths as of April 28,
2020. We are seeing alarming acceleration in several countries and spread of virus
to rural areas and can see clusters of cases and community spread in more than 16
countries. Pneumonia, headache, fever, dry cough, and breathing difficulties are the
typical symptoms of the patients infected from this disease. Respiratory problem
owing to alveolar damage (can be seen in chest CT images) and even death may be
the result of this disease. With this outbreak, the question is with different testing
methods and its efficiency at this level. With respect to the various tests available,
radiological imaging can also help in diagnosing the COVID efficiently. Recently,
it is noted that among radiological imaging data, assessment of disease through
chest imaging is of great use in understanding and analysis of COVID-19 [1, 2].
Initial chest imaging reported abnormal findings in most of the patients of COVID
2019 [3]. It is also suggested that, imaging technique such as CT or X-ray is quite
recommended for follow up in individuals who are recovering from COVID 2019
and can be effectively detected. Automated artificial intelligence models/tools are a
must in order to categorize COVID-19 positive cases to non-positive and these tools
can help in providing accurate results [2, 4]. Medical imaging system with deep
architectures including shape and spatial relation features have been reported in the
literature [5]. Interesting research works have been reported in recent past on COVID
2019 [6]. Image enhancement certainly assists physicians in detecting and diagnosis
anomaly/abnormality in automated medical imaging systems and few recent works
on image enhancements can be seen in [4, 7–13]. Variants of MSR algorithms can
be seen in the literature recent past. Our previous works on retinex-based algorithm
are reported in the literature. Using a popular particle swarm optimization (PSO) for
color image enhancement based on MSR [14], medical image enhancement based
on multi-rate sampling is addressed in [15]. Image quality metric plays an important
role in design and evaluation process of imaging systems. In this connection, accurate
objective criterion for assessing the performance of image enhancement algorithms
is a concern.
The main contributions of this paper are: (i) we propose an improvised version of
retinex algorithm for better analysis of COVID-19 chest X-ray/CT images. (ii) The
resultant of the proposed algorithm may be used in diagnosis and assist the doctors
and radiologists in detecting the spread of virus in chest and lower intestine part of
human body effectively. (iii) Concept of wavelet energy is introduced for evalua-
tion purpose. The remaining part of the paper is as follows: proposed methodology
followed by experiment and critical analysis is shown in Sects. 2 and 3, respectively.
Paper concludes in Sect. 4.
Understanding and Analysis of Enhanced COVID-19 Chest … 49

2 Proposed Methodology

The process of conversion from RGB to HSV space is much essential and resolves
the gray world violation problem that exists in the traditional retinex algorithm.
The HSV domain processing also avoids color shifting issues because the value or
intensity channel is separated from chrominance component. The proposed improved
multiscale retinex (MSR) algorithm is applied on the value component by preserving
hue and saturation. Retinex is human perception based enhancement algorithm that
improves intensity, contrast, and sharpness in medical image through dynamic range
compression. The parameter of retinex algorithm varies adaptively depending on
the intensity component and image depth. The improved image is validated using a
wavelet energy metric [16] that reveal approximate and detail components. Approxi-
mate coefficients provide global enhanced information and detail coefficients convey
sharpness of the image. The proposed framework exploits envelop-based retinex
method [17] to obtain the illumination. In traditional retinex algorithm [18], illu-
mination estimation is achieved by applying several Gaussians surround functions
and produces the resultant image by processing the reflectance in the logarithmic
domain.

2.1 Illumination Estimation

The proposed method assumes spatial smoothness in the original image and need
of envelop is based on the technique proposed by Kimmel [19]. The computational
complexity in the traditional retinex algorithm is mainly due to the log domain
processing. This complication has been overcome here by processing each pixel
directly instead of using logarithm. With the need of piecewise spatial smoothness
in the original image, we concatenate the weighting function with the envelope and
restrict the performance of the illumination around the boundaries and corners of an
image. Because of spatial smoothness assumption in the original image, the derivative
of illumination has to be minimized. In addition, we restrict that the illumination is
close in order to achieve low reflectance. Based on these procedures, we model the
cost operator and minimize its penalty using Eq. (1)
  
2
F(L(x, y)) = ∇ L(x, y)2 + αL(x, y) − I (x, y) dxdy (1)

where ∇ specify first-order differential operator and . operation retains posi-
tive values. Subsequently, we use cost-minimizing operation in descent gradient
framework. Further, the iteration framework is formulated as given by Eq. (2)

L j (x, y) = L j−1 (x, y) − β · G (2)

50 M. C. Hanumantharaju et al.

where L j (x, y) and L j−1 (x, y) denotes intensity images at jth step and (j − 1)th
step; β signify the step size, G denote gradient operator of F(L(x, y))
The gradient operator defined in [19] is given by

G = −L + α(L − 1) (3)

G ≈ −L(x, y) ∗ K lap (x, y) + α · (L(x, y) − I (x, y)) (4)

Here  the second-order Laplacian differential operator that is predicted as a

convolution operation with the spatial filter.
Lastly, we formulate the equation to obtain the illumination.
 
L j (x, y) = max w(∇ I ) · I (x, y) + (1 − w(∇ I )) · L j (x, y) · I (x, y) (5)

w0 , if ∇ I > Th
w(∇ I ) =  ∇ I 2 (6)
w0 · Th , Otherwise

∇ I ≈ I (x, y) ∗ H (x, y) +

I (x, y) ∗ H T (x, y)
(7)

where Th is the threshold, H(x, y) indicates pyramid operator, H T (x, y) indicates

the transpose of H(x, y) used to make use of robust edge.

2.2 Estimation of Reflectance

In this work, estimation of reflection is realized by dividing the value and resulting
lighting. The brightness adaptive change and depth image improvement are achieved
by the following equations:

L k (1+ N )
y

Γ (y) = N (8)
N
  
1 1
β(r ) = exp g − (9)
1 + e−b·logr 2

where k ranges from 0 to 1, b ranges from 0 to 10 and g varies between 1 and 10; N
value is 255.
Understanding and Analysis of Enhanced COVID-19 Chest … 51

2.3 Validation of Enhanced Using Wavelet Energy (WE)

Metric

To evaluate the amount of improvement of the proposed framework, we employed

WE as a metric. As the detailed coefficients increases, the depth of the image is
better. The enhanced component of the image is verified using WE metric. The
Daubechies wavelet transform is employed in the WE computation. The depth of the
image appears to be better if the detailed coefficients are large than the input image.
Approximate WE coefficients are used to improve overall information of the image.
Higher the approximate coefficients than the input image, globally the enhanced
image appears to be better. More information about WE can be seen in [16, 20].

3 Experiment Results and Comparative Analysis

In this part, we show the experiments and comparative analysis with standard well-
known algorithm. We have conducted a number of experiments on more than 100
images from variety of COVID-19 database publically available in [21]. As it was
found from the experimental results that proposed scheme reconstructs the images
with improved quality, more visual details and improved appearance. The proposed
enhancement scheme is mainly used by doctors, radiologists, and researchers in order
to make right decisions. Figure 1 shows the experimental results obtained for various
X-ray/CT scan images. Figure 1a shows the original X-ray/CT images of COVID
2019 patients of resolution 1024 × 768. Figure 1b shows the results obtained using
NASA’s multiscale retinex algorithm [18]. Figure 1c are the results of the proposed
improved multiscale retinex algorithm. Corresponding histograms are also seen. The
efficacy of the proposed algorithm is also tested and provided satisfactory results for
CT scan images. In our experiments, we also opted CT scan images since these
types of images assists in exact location of coronavirus spread. From the results, it
is quite evident that the histogram plot of the enhanced image are tends toward the
brightness area. Although histogram-based enhanced image analysis is subjective
technique, objective techniques such as peak signal to noise ratio (PSNR), contrast
enhancement performance (CEP), luminance enhancement performance (LEP), and
wavelet energy (WE) are employed in this work. Table 1 shows the performance
comparison for the set of images considered, which is based on PSNR, CEP, and
LEP. From Table 1 it can be seen that the proposed algorithm is able to produce
better results compared to existing MSRCR method. Table 2 shows the performance
comparison applied on all four set of images (shown in Fig. 1) considered based on
approximate and detailed wavelet energy. WE metric shows that the proposed method
not only improves overall contrast of an image but also increases inner details in an
image.
52 M. C. Hanumantharaju et al.

Fig. 1 Results for various images and its corresponding histogram. a Column 1—original images.
b Column 2—multiscale retinex algorithm [18]. c Column 3—proposed algorithm

4 Discussion and Conclusion

Chest X-ray images are widely used to detect the COVID-19 positive cases mainly
due to quick, easy, and accessible in all places. Radiologists recommend chest X-ray
for the cases especially for the patients with non-transportable or difficult to move.
The proposed adaptive retinex framework offers detailed enhanced results for both
chest X-ray and CT scan images. Our scheme employs HSV domain-based retinex to
separate chromatic component from intensity. The adaptive improved retinex method
applied on value component of MSR outperforms existing method. The images with
and without image enhancement scheme produce a huge difference in the results.
Understanding and Analysis of Enhanced COVID-19 Chest … 53

Fig. 1 (continued)

Table 1 Performance comparison based on PSNR, CEP, and LEP

Test images MSRCR Proposed improved MSRCR
PSNR CEP LEP PSNR CEP LEP
1. 36.77 0.118 0.132 36.95 0.133 0.131
2. 32.83 0.172 0.167 32.82 0.171 0.177
3. 35.90 0.397 0.174 35.94 0.399 0.173
4. 34.34 0.134 0.128 34.36 0.141 0.134
54 M. C. Hanumantharaju et al.

Table 2 Performance comparison on approximate and detailed wavelet energy

Test images Org. image MSRCR Proposed improved
MSRCR
Approx. Detail Approx. Detail Approx. Detail
1. 99.34 0.339 99.50 0.327 99.57 0.409
2. 99.62 0.252 99.69 0.302 99.59 0.392
3. 99.41 0.387 99.10 0.390 99.71 0.411
4. 99.67 0.283 99.97 0.112 99.81 0.390

This is evident from the experimental result presented. With the image enhancement
algorithm applied, the visual details are visible to the naked eye. The doctors can
locate any abnormalities by verifying each corner of the cells. The proposed image
enhancement scheme is a big boon to the doctors, radiologists, and researchers in
making right decisions to identify COVID-19 cases.

References

1. Hosseiny, M., Kooraki, S., Gholamrezanezhad, A., Reddy, S., & Myers, L. (2020). Radiology
perspective of coronavirus disease 2019 (COVID-19): Lessons from severe acute respiratory
syndrome and middle east respiratory syndrome. American Journal of Roentgenology AJR,
214, 1–5.
2. Mukherjee, H., Ghosh, S., Dhar, A., Obaidullah, S. K. M., Santosh, K. C., & Roy, K. Shallow
convolutional neural network for COVID-19 outbreak screening using chest X-rays. https://
doi.org/10.36227/techrxiv.12156522.v1.
3. Chung, M., Bernheim, A., Mei, X., et al. (4 February 2020). CT imaging features of 2019 novel
coronavirus (2019-nCoV). Radiology (In Press).
4. Fernandes, S. L., Rajinikanth, V., & Kadry, S. (2019). A hybrid framework to evaluate breast
abnormality using infrared thermal images. IEEE Consumer Electronics Magazine, 8(5), 31–
36. https://doi.org/10.1109/MCE.2019.2923926.
5. Gomez, P., Semmler, M., Schutzenberger, A., Bohr, C., & Dollinger, M. (2019). Low-light
image enhancement of high-speed endoscopic videos using a convolutional neural network.
Medical and Biological Engineering and Computing, 57(7), 1451–1463.
6. https://www.who.int/emergencies/diseases/novel-coronavirus-2019/global-research-on-
novel-coronavirus-2019-ncov.
7. Hashemi, S., Kiani, S., Noroozi, N., & Moghaddam, M. E. (2010). An image contrast enhance-
ment method based on genetic algorithm. Pattern Recognition Letter, 31(13), 1816–1824.
8. Hanumantharaju, M. C., Manjunath Aradhya, V. N., Ravishankar, M., & Mamatha, A. (2012).
A particle swarm optimization method for tuning the parameters of multiscale retinex based
color image enhancement. In Proceedings of the International Conference on Advances in
Computing, Communications and Informatics, pp. 721–727.
9. Zhang, R., Huang, Y., & Zhen, Z. (2011). A ultrasound liver image enhancement algorithm
based on multi-scale Retinex theory. In 5th International Conference on Bioinformatics and
Biomedical Engineering, (ICBBE), pp. 1–3.
10. Dey, N., et al. (2019). Social-group-optimization based tumor evaluation tool for clinical brain
MRI of Flair/diffusion-weighted modality. Biocybernetics and Biomedical Engineering, 39(3),
843–856. https://doi.org/10.1016/j.bbe.2019.07.005.
Understanding and Analysis of Enhanced COVID-19 Chest … 55

11. Satapathy, S. C., Raja, N. S. M., Rajinikanth, V., Ashour, A. S., & Dey, N. (2018). Multi-level
image thresholding using Otsu and chaotic bat algorithm. Neural Computing and Applications,
29(12), 1285–1307. https://doi.org/10.1007/s00521-016-2645-5.
12. Rundo, L., Tangherloni, A., Nobile, M., Militello, C., Besozzi, D., Mauri, G., et al. (2019).
MedGA: A novel evolutionary method for image enhancement in medical imaging systems.
Expert Systems with Applications, 119, 387–399.
13. Jobson, D. J., Rahman, Z., & Woodell, G. A. (1997). Properties and performance of a
center/surround retinex. IEEE Transactions on Image Processing, 6(3), 451–462.
14. Hanumantharaju, M. C., Ravishankar, M., Rameshbabu, D. R., & Aradhya, M. (2014). A new
framework for Retinex-based colour image enhancement using particle swarm optimization.
International Journal of Swarm Intelligence, 1(2), 133–155.
15. Setty, S., Srinath, N., & Hanumantharaju, M. (2013). Development of multiscale Retinex algo-
rithm for medical image enhancement based on multi-rate sampling. In Proceedings of Interna-
tional conference on Signal Processing, Image Processing and Pattern Recognition (ICSIPR),
pp. 1–6.
16. Hanumantharaju, M. C., Ravishankar, M., & Rameshbabu, D. R. (2014). Natural color image
enhancement based on modified multiscale Retinex algorithm and performance evaluation
using wavelet energy. In Recent Advances in Intelligent Informatics. Advances in Intelligent
Systems and Computing, vol. 235. Springer.
17. Shen, C. T., & Hwang, W. L. (2009). Color image enhancement using Retinex with robust
envelop. In Proceedings of International Conference on Image Processing (ICIP-2009).
18. Jobson, D. J., Rahman, Z. U., & Woodell, G. A. (1997). A multiscale retinex for bridging the
gap between color images and the human observation of scenes. IEEE Transactions on Image
processing, 6(7), 965-976.
19. Kimmel, R., Elad, M., Shaked, D., Keshet, R., & Sobel, I. (2003). A variational framework for
Retinex. International Journal on Computer Vision, 52(1), 7–23.
20. Hanumantharaju, M. C., Ravishankar, M., Babu, D. R. R., & Aradhya, V. N. M. (2011). An
efficient metric for evaluating the quality of color image enhancement. In Indian International
Conference on Artificial Intelligence, 2011, pp. 1016–1026.
21. https://github.com/ieee8023/covid-chestxray-dataset.
Deep Learning-Based COVID-19
Diagnosis and Trend Predictions

Juanying Xie, Mingzhao Wang, and Ran Liu

Abstract During the Chinese Spring Festival travel rush in 2020, a new type of
pneumonia disease, named COVID-19 subsequently broke out in Wuhan, Hubei
province, China. The COVID-19 was quickly spreading in China and emerged nearly
all over the world. In this chapter, our motivation is to adopt the deep learning
techniques to help clinic doctors to diagnose the patients of COVID-19 and predict
the trend of COVID-19. To realize our motivation, we on the one hand adopt deep
learning techniques to analyse CT images of patients. The transfer learning and data
augmentation techniques are adopted for the lacking of samples in our obtained CT
image data set. We build a model by designing and training a new deep network
to help clinic doctors to make an appropriate diagnose decision. On the other hand,
according to the spreading characteristics of COVID-19 and the controlling measures
adopted by Chinese government, we propose to modify the classic SEIR (susceptible-
exposed-infectious-recovered) model and establish a new SEIR dynamics model with
considering the infectiousness of the people in the latent period and the quarantine
period. The appropriate parameters of our modified SEIR model are learned by
using deep learning techniques. Our proposed deep learning-based diagnosis for
COVID-19 can help medicine doctors to make an appropriate diagnostic decision.
Our modified SEIR model can effectively predict the transmission trend of COVID-
19 and can be used for short-term trend prediction of the epidemic.

Keywords COVID-19 · Artificial intelligence · Deep learning · CT images · SEIR

model

J. Xie (B) · R. Liu

School of Computer Science, Shaanxi Normal University, Xi’an 710119, China
e-mail: [email protected]
R. Liu
e-mail: [email protected]
M. Wang
College of Life Sciences, Shaanxi Normal University, Xi’an 710119, China
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 57
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_7
58 J. Xie et al.

1 Introduction

It was reported that there was an atypical pneumonia with unknown aetiology
appeared in Wuhan, the capital city of Hubei province in China, in December 2019.
Subsequently, it was detected that this atypical pneumonia was a new coronavirus [1].
The new coronavirus was first named as 2019 novel coronavirus (2019-nCOV) by
World Health Organization (WHO) on January 20, 2020, and the pneumonia caused
by it was named as COVID-19 on February 11, 2020 [2]. Patients with COVID-19
generally have symptoms similar to SARS patients, such as cough, fever, dyspnea,
acute respiratory syndrome and kidney failure. [3, 4]. In addition, chest computed
tomography (CT) images showed that there were diffuse and small ground glass
nodules in the bilateral lungs of COVID-19 patients [5]. It is reported that the natural
host of 2019-nCOV is likely to be bats [6, 7], but the intermediate host is still
unknown [8]. Moreover, the evidence shows that it can lead the infections from
person-to-person [9, 10]. Although we do not know where the 2019-nCOV comes
from, it is the indisputable fact that the COVID-19 has spreaded to nearly all of the
countries in the world and has become a very serious global pandemic. As of April
15, 2020, the accumulative number of infections in the worldwide has exceeded
1.95 million, and the number of deaths is about 130,000. This epidemic has already
brought a major impact on the world’s public health and the economy of the world.
This 2019-nCOV just happened before the Chinese Spring Festival. The huge
personnel flow advanced the spread of it to other regions in China. In order to stop
the propagation of the epidemic and minimize its impact on the public health and the
national economy, the Chinese government quickly launched the “Level I response
to major public health emergencies”, and ordered to block the Wuhan city on January
23, 2020, and cut off all transportations between Wuhan and any other cities in China.
Meanwhile, China government called on all of Chinese to stay at home, and wearing
masks when he/she must leave home to go to public places, and suspending some
intra-city public transportations, closing entertainment venues and banning public
gatherings, so as to block the spreading of this serious infectious disease COVID-19.
In addition, there are many artificial intelligence (AI) experts devoted themselves
to do related researches about COVID-19, published or issued many related papers.
Although the situations in China were much better after two months, the COVID-
19 was still serious in many other countries. The medicine doctors are not enough
compared to the rising up COVID-19 patients. So, it needs us, all of AI researchers,
to use AI technologies to do some work about the COVID-19, and to help humankind
to overcome the crisis.
In this chapter, we will present our two studies about COVID-19. The first is
that our doing analysis to the CT images of patients by using deep learning tech-
niques, while using the transfer learning and data augmentation technologies to solve
the problem of over-fitting due to the small number of samples in the data set we
obtained. We modified the RetinaNet networks, so as to learn the high-level abstract
features of CT images of patients. The second is that we present a new modified
SEIR dynamics model with considering the infectivity of an individual in latent
Deep Learning-Based COVID-19 Diagnosis and Trend Predictions 59

period and the quarantine period. Those appropriate parameters of the new modified
SEIR model are learned by using deep learning techniques. Our motivation is to help
clinic doctors make appropriate diagnostic decisions, and help whoever related to
make efficient prediction to the short-term trend for the COVID-19 epidemic, so as
to provide decision-makers a basis to prevent and control the new infectious disease.

2 Related Work

Deep learning is a branch of machine learning. Its aim lies in establishing while
simulating the human brain neural networks to do analysis and learn. It imitates the
ways of human brains to analyse images, sounds, texts and various kinds of data.
The essential of deep learning is artificial deep neural network, and its structure is a
multi-layer perceptron with multi-hidden layers. The advantages of deep learning are
that it can automatically extract features and get a high-level abstract representation
of the data. It has been used in many fields, including early screening and diagnosis
of medical images [11].
Due to the strengths of deep learning, it is believed by researchers that this
technology will play invaluable roles in controlling and help diagnose COVID-
19. Therefore, many researchers have used deep learning to study COVID-19‚ so
as to help humankind to overcome the crisis. The deep learning-based model for
detecting COVID-19 using the high-resolution computed tomography was published
by researchers from Wuhan University People’s Hospital and Wuhan EndoAngel
Medical Technology Company on medRxiv platform [12]. To validate the constructed
deep learning model, this study collected and processed 46,096 anonymous CT
images of 106 inpatients in Wuhan University People’s Hospital. The results showed
that the developed deep learning-based model had got a comparable performance
to that of radiologists and significantly reduced the work stress of radiologists.
A customized deep convolutional neural network was proposed for detecting the
COVID-19 cases from chest X-ray (CXR) images by the researchers from the Univer-
sity of Waterloo [13]. The findings of this study showed that the potent promising
results could be achieved in detecting COVID-19 with chest radiographs. Mukherjee
et al. [14] proposed a lightweight convolutional neural network (CNN)-tailored
shallow architecture to detect COVID-19 positive cases using chest X-rays auto-
matically from non-COVID ones. They obtained the accuracy of 96.92%, sensitivity
of 0.942 and AUC of 0.9869 by using 260 chest X-rays, and the false positive rate was
0 for 130 COVID-19 positive cases. Considering the CXR images can be obtained
easier than CT images. Das et al. [15] proposed a deep learning-based CNN model
to filter COVID-19 positive CXRs from non-COVID ones. They concluded that
they achieved the best results among the existing AI-driven tools in detecting out
COVID-19 by using CXRs. Rajinikanth et al. [16] proposed an image-assisted system
to extract COVID-19 infected sections from lung CT images, so as to assist clinic
doctors to detect out COVID-19 patients and help to plan an appropriate treatment
process.
60 J. Xie et al.

The SEIR model is a common disease transmission dynamics model [17]. Since
SEIR model takes the latent into account compared to the SIR model, it has been
applied to the analysis and prediction of many major infectious diseases, such
as severe acute respiratory syndrome (SARS) [18] and Middle East respiratory
syndrome (MERS) [19] and human immunodeficiency virus (HIV) [20]. At present,
the SEIR model is also used by many researchers in predicting and analysing the
development trend of the COVID-19 epidemic. Based on the transmission charac-
teristics of 2019-nCOV, Tang et al. [21] brought forward the deterministic SEIR
compartmental transmission dynamics model with a comprehensive consideration
of the clinical progression, close contact and isolation of the disease and other inter-
ventions, and predicted that the epidemic peak would be around February 5. Cao
et al. [22] incorporated the characteristics of COVID-19 incubation patients with
infectious ability into the SEIR model and proposed a modified SEIR dynamics
model that jointly took the infectivity of the individuals in the incubation period
and the influence from the isolation intervention into account. This modified SEIR
model parameters were solved by using the Euler integral method. As a result, it
can effectively simulate and forecast the development of the epidemic situation of
Hubei province. Yang et al. [23] integrated population migration data and COVID-
19 epidemiological data into the SEIR model to forecast the development trend of
COVID-19, and they also used the SARS data in 2003 to train a long short-term
memory (LSTM) model to forecast the number of new infections.

3 Deep Learning for COVID-19 Diagnosis

From the aforementioned studies in Sect. 2, we can see that deep learning techniques
have been adopted to detect the COVID-19 patients by learning the CT images.
Compared with reverse transcription polymerase chain reaction (RT-PCR) detection
method, the deep learning-based CT image examination is timely, accurate, and
with high positive rate. Furthermore, the range of lung lesions is closely related
to clinical symptoms. Therefore, the CT image-based diagnosis is a preferring one
for diagnosing COVID-19. It was reported that the most important way to make
the accurate diagnosis is to read CT images for clinic doctors during COVID-19
epidemic in China, and the decisions by analysing the CT images were with high
positive rate. For patients’ CT images, a deep learning model aims to learn the
abstract mapping between the raw data and the expected clinical outcomes. If there
are enough samples, after many training iterations, the expected output of the final
model will infinitely approximate the real labels. We show the working process of
the deep learning techniques in Fig. 1.
On account of the privacy protection, CT images of patients are not always open
and accessed freely. Even though the current COVID-CT data set is open, the number
of samples contained in the data set is insignificant, and it is still difficult to achieve
the amount of data required by the deep learning model. This greatly impedes the
accuracy of detecting COVID-19 patients by the CT images using deep learning
Deep Learning-Based COVID-19 Diagnosis and Trend Predictions 61

Fig. 1 Schematic workflow

of a deep learning model

techniques. Because training deep learning model on such a small data set is very
easy to fall into over-fitting, which makes the model perform well on the training data,
but bad on the test data, such that the generalization performance of model is poor. In
addition, there are no existing principles for designing network structures for specific
tasks. Fortunately, there are three solutions to these aforementioned problems:
I. It is well-known that transfer learning is the most commonly used method in
deep learning which uses large amounts of data from related fields to assist
model training and learning. For example, a large number of CXR images can
be used to pretrain a deep convolutional neural network, and then fine-tune the
trained network on the COVID-CT data set, which can achieve twice the result
with half the effort.
II. Create new images from limited training data and add them to the original
training set. The methods to create new images include expanding the original
images by flipping, rotating and translating. This method is called data augmen-
tation and is the most common technique in deep learning to solve over-fitting
problems caused by the lacking of training data.
III. Combine the above two methods. That is, use the augmented data to fine-tune
the pretrain model.

4 Prediction Model About COVID-19 Outbreak

The SEIR epidemic model is a classic disease transmission model based on complex
networks. It has been used to model and analyse the infectious diseases such as SARS,
MERS and HIV. The SEIR model classifies people into four categories: susceptible
(S), exposed (E), infectious (I) and recovered (R). Figure 2 displays the processes to
establish the classic SEIR model.
Considering the quarantine measures taken by the Chinese government for
preventing and controlling COVID-19, we added the quarantined susceptible (S q )
and the quarantined exposed (E q ) categories to the classic SEIR model. The quaran-
tined individuals are divided into S q and E q according to whether they are infected
62 J. Xie et al.

Fig. 2 Classic SEIR epidemic dynamics model, where α is the infection probability of the suscep-
tible, β1 is the probability of the exposed being transformed into the infected and β2 is the recovery
probability

or not. The quarantined susceptible ones become the susceptible (S) again after
being released and the quarantined exposed ones become exposed (E) category after
releasing quarantine if they do not show symptoms of infection during the quaran-
tine period. In addition, we consider that the exposed individuals also have got the
ability to infect the virus to the susceptible individuals. Our modified SEIR epidemic
dynamic model is shown in Fig. 3, where α1 is the infection probability of the
susceptible by the infected, and α2 is the infection probability of the susceptible by
the exposed. The appropriate parameters of our modified SEIR model are obtained
using deep learning techniques.
We proposed a two-stage SEIR model to forecast this COVID-19 prevalence in
Shaanxi province before and after the response to the “level I response to major
public health emergencies” on January 25, 2020. In the first stage, due to insuffi-
cient knowledge about COVID-19, Shaanxi province did not take any prevention
and control measures against COVID-19 before January 25, 2020. So, we use the
classic SEIR model to make the prediction for this COVID-19 development trend.
In the second phase, after January 25, 2020, Shaanxi province implemented various
measures such as restricted travel, wearing masks and isolating contacts. Therefore,
we adopted the revised SEIR model to predict the COVID-19 epidemic situation
for this stage. Meanwhile, we also calculated the basic reproduction number of two
stages by changing the value of the parameters to simulate the development trend of
COVID-19 infection before and after taking prevention and control measures. The
study results show that the measures such as home prevention and control quarantine
and centralized treatment taken by Chinese government have greatly suppressed the
widespread spread of the COVID-19 epidemic, and also gained valuable time and
reference for other countries to control the COVID-19 epidemic.

Fig. 3 Our proposed SEIR epidemic dynamics model to simulate COVID-19

Deep Learning-Based COVID-19 Diagnosis and Trend Predictions 63

5 Conclusions

In our study about COVID-19, we first adopt the deep learning methods to analyse
and identify the COVID-19 patients by CT images, where we combined the transfer
learning and data augmentation technologies to solve the over-fitting problem due
to the small size of data set. The deep learning-based COVID-19 studies can help
medicine doctors quickly obtain diagnostic results and make appropriate decisions
in a short time. Then taking into account the transmission characteristics of 2019-
nCOV virus and the epidemic prevention measures in China, we established a new
modified SEIR epidemic dynamics model. A two-stage SEIR epidemic dynamics
model was proposed based on this new model to analyse the transmission pattern of
COVID-19 in Shaanxi province. The appropriate parameters for our two-stage SEIR
epidemic model are learned by using the deep learning technologies. Our destination
is to help humankind to overcome the COVID-19 pandemic crisis. We hope that
our study could help clinic doctors to make the correct diagnosis to the patients of
COVID-19 as fast as possible and provide some references about prevention and
control measures for other countries in the world.

Acknowledgements This work is supported by the NSFC under Grant No. 61673251, and by the
NKRDPC under Grant No. 2016YFC0901900, and by the Fundamental Research Funds for the
Central Universities under Grant No. GK201806013, and by the Innovation Funds for Graduate
Programs in Shaanxi Normal University under Grant Nos. 2016CSY009 and 2018TS078.

References

1. Li, Q., Guan, X., & Wu, P. (2020). Early transmission dynamics in Wuhan, China, of novel
coronavirus–infected pneumonia. New England Journal of Medicine.
2. Xinhuanet.com (2020) WHO named the novel coronavirus pneumonia as COVID-19. Available
from: http://www.xinhuanet.com/world/2020-02/12/c_1125561389.htm.
3. Chen, Nanshan, Zhou, Min, & Dong, Xuan. (2020). Epidemiological and clinical characteristics
of 99 cases of 2019 novel coronavirus pneumonia in Wuhan, China: a descriptive study. The
Lancet, 395(10223), 507–513.
4. Guan, W., Ni, Z., & Hu, Y. (2020). Clinical characteristics of 2019 novel coronavirus infection
in China. MedRxiv.
5. Qu, J., Yang, R., Song, L., & Kamel, I. R. (2020). Atypical lung feature on chest CT in a lung
adenocarcinoma cancer patient infected with COVID-19. Annals of Oncology.
6. Xu, X., Chen, P., & Wang, J. (2020). Evolution of the novel corona-virus from the ongoing
Wuhan outbreak and modeling of its spike protein for risk of human transmission. Science
China-life Sciences, 63(3), 457–460.
7. Zhou, P., Yang, X., & Wang, X. (2020). A pneumonia outbreak associated with a new
coronavirus of probable bat origin. Nature, 579(7798), 270–273.
8. Cyranoski, David. (2020). Mystery deepens over animal source of coronavirus. Nature,
579(7797), 18–19.
9. Chan, J. F. W., Yuan, S., & Kok, K. H. (2020). A familial cluster of pneumonia associated
with the 2019 novel coronavirus indicating person-to-person transmission: a study of a family
cluster. The Lancet, 395(10223), 514–523.
64 J. Xie et al.

10. Rothe, C., Schunk, M., & Sothmann, P. (2020). Transmission of 2019-nCoV infection from an
asymptomatic contact in Germany. New England Journal of Medicine, 382(10), 970–971.
11. Xie, J., Liu, R., Luttrell IV, J., & Zhang, C. (2019). Deep learning based analysis of
histopathological images of breast cancer. Frontiers in Genetics, 10, 80.
12. Chen, J., Wu, L., & Zhang, J. (2020). Deep learning-based model for detecting 2019
novel coronavirus pneumonia on high-resolution computed tomography: A prospective study.
MedRxi.
13. Wang, L., & Wong, A. (2020). COVID-Net: A tailored deep convolutional neural network
design for detection of COVID-19 cases from chest radiography images. ArXiv preprint arXiv:
2003.09871.
14. Mukherjee, H., Ghosh, S., Dhar, A., Obaidullah, M. S., Santosh, K. C., & Roy, K. (2020).
Shallow convolutional neural network for COVID-19 outbreak screening using chest X-rays.
Techrxiv, 12156522, v1.
15. Das, D., Santosh, K. C., & Pal, U. (2020). Truncated inception net: COVID-19 outbreak
screening using chest X-rays. Research Square. https://doi.org/10.21203/rs.3.rs-20795/v1.
16. Rajinikanth, V., Dey, N., Raj, A. N. J., Hassanien, A. E., Santosh, K. C., & Raja, N. S. R. (2020).
Harmony-search and Otsu based system for coronavirus disease (COVID-19) detection using
lung CT scan images. ArXiv preprint arXiv:2004.03431.
17. Miller, W. (2002). Mathematical approaches for emerging and reemerging infectious diseases:
Models, methods, and theory. New York, NY: Springer.
18. Dye, C., & Gay, N. (2003). Modeling the SARS epidemic. Science, 300(5627), 1884–1885.
19. Eifan, S. A., Nour, I., & Hanif, A. (2017). A pandemic risk assessment of Middle East respiratory
syndrome coronavirus (MERS-CoV) in Saudi Arabia. Saudi Journal of Biological Sciences,
24(7), 1631–1638.
20. Zakary, O., Elmouki, I., & Rachik, M. (2016). On the impact of awareness programs in
HIV/AIDS prevention: An SIR model with optimal control. International Journal of Computer
Applications, 133(9), 1–6.
21. Tang, B., Wang, X., & Li, Q. (2020). Estimation of the transmission risk of the 2019-nCoV
and its implication for public health interventions. Journal of Clinical Medicine, 9(2), 462.
22. Cao, S., Feng, P., Shi, P. (2020). Study on the epidemic development of corona virus disease-19
(COVID-19) in Hubei province by a modified SEIR model, Journal of Zhejiang University,
49(2):178–184.
23. Yang, Z., Zeng, Z., Wang, K., Wong, S. S., Liang, W., Zanin, M. et al. (2020). Modified
SEIR and AI prediction of the epidemics trend of COVID-19 in China under public health
interventions. Journal of Thoracic Disease, 12(3):165–174.
COVID-19: Loose Ends

Minakshi Pradeep Atre

Abstract The sudden outburst of the COVID-19 has hit the world badly. Avoid,
control, and monitor (ACM) is the need of time! With limited expert manpower in
COVID-19, the technologies like artificial intelligence (AI), robotics (R), and IOT
(I), i.e., (ARI) would help in avoidance of further spread and control of disease
transmission. Prediction and analysis tools are effectively implemented only when
sufficient data is available. Though an early stage of any pandemic has to deal with
the scarcity of data, an early detection and prediction are equally important steps
to fight COVID-19. But a complete solution to the pandemic is impossible because
of the loose ends like availability of data and “dependent” development of ARI
technology. Since December 2019, there has been a continuous up-scaling in analysis
and prediction algorithms because of more data getting available in terms of features
and more number of the cases across the world. This chapter will discuss the role of
ARI and loose ends in their implementation. It is focused on three major aspects: AI
algorithms in analysis and prediction, the use of robotics in control and prevention
of the pandemic and the role of IOT for the patient monitoring system (PMS). This
discussion will provide an evolutionary path of the algorithms. The accuracy rate for
diagnosis and prediction has been increased because of the various novel approaches
by researchers. They are trying to overcome the loopholes and are tying the loose
ends with the advent of more and more data!

Keywords COVID-19 · AI · IOT · Robotics · Avoid-Control-Monitor (ACM) ·

Loose ends · Patient monitoring systems (PMS)

1 Introduction

The outbreak of COVID-19 (COrona VIrus Disease-2019) happened first in Wuhan,

China in December 2019 [12] and then rapidly spread across the world. This was
later named as nCov, an abbreviation for “novel Corona.” The scientists have not

M. P. Atre (B)
Department of E&TC, PVG’s COET, Pune, SPPU, Pune, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 65
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_8
66 M. P. Atre

developed any vaccinations till date, though they could find the reason immediately.
To deal with such a pandemic, often a teamwork of the scientists, technologists,
engineers, and mathematicians is needed. The technology is growing fast and helping
the life sciences to avoid, control, and monitor such diseases. The AI or machine
learning algorithms can be exploited to predict and gain control over this pandemic.
The algorithms would help to understand and model the spread factor of the nCov
beforehand and analyze the case fatality ratio (CFR) [1, 2], based on different COVID-
19 parameters.
As said earlier, a complete IOT-based system or an AI algorithm cannot be devel-
oped unless the loose ends are tied up. The first limitation is data: the larger datasets
would help in more accurate detection and prediction of the pandemic. The stan-
dards and protocols must be defined for the exchange of patients’ data, respecting
the privacy of a patient. The second limitation is benchmark datasets and clin-
ical study [7]. COVID-19 is novel coronavirus and being in early stage there is no
benchmark dataset available. The other aspect is the completely dynamic nature of
the disease. It changes quickly from time to time, person to person, culture to culture,
and place to place [8]. The third limitation is accessing the information through
IOT devices for patient monitoring systems (PMS). It includes the mobile phones
and wearable devices like smart watches. There is a need for standardization of
protocols to encourage communication between devices and across systems without
compromising data safety and preventing data oversight [11]. These three main loose
ends could be addressed with one question unanswered, what would happen if there is
second transmission cycle? Because all the analysis techniques have one underlying
assumption of no second cycle of COVID-19 transmission.
This chapter is divided into three sections. Section 2 discusses the parameters of
COVID-19 while Sect. 3 discusses the modeling techniques for different phases of
COVID-19 and loose ends in modeling techniques. And Sect. 4 illustrates the role of
artificial intelligence (AI) and Internet of things (IoT) in patient monitoring system.

2 Discussion of Parameters of COVID-19

There are three aspects of COVID-19 parameters. The first is the set of person
dependent parameters, the second is the set of global parameters, and the third
is COVID-19 symptoms. The symptoms would depend on these two parameters.
The person dependent parameters will include diet, age, vaccinations, travel history,
and medical history (prevailing diseases like blood pressure, diabetes, asthma). The
chapter enlists the symptoms as: loss of smell (major), with other symptoms as,
fatigue, fever, cough, diarrhea, and shortness of breath [4]. The global parameters
would include geographical location, atmospheric temperature, and miscellaneous
parameters (like few people believed health getting affected by 5G). Figure 1 depicts
the parameters and symptoms of COVID-19.
Figure 2 shows the number of nCov cases exhibiting the combinational symptoms.
This reference paper claims that the very first dominant symptom is loss of smell. The
COVID-19: Loose Ends 67

Person Dependent Parameters

• diet
• age
• vaccina ons
• travel history
• Medical history

Global Parameters
• geographical loca on
• atmospheric temperature

Symptoms
• loss of smell
• fever
• fa gue
• cough
• diarrhea
• shortness of breath

Fig. 1 COVID-19: key aspects

reference says diarrhea and shortness of breath appear later than the loss of smell.
The challenge is to identify how fast the symptoms develop and is there a particular
sequence of these symptoms. The modeling techniques completely depend on such
critical parameters. The phases of disease will be identified depending on the % value
of these symptoms. The critical phase includes the pneumonia infection in the lungs.

3 Modeling Techniques for Different Phases of COVID-19

Figure 3 depicts the phases of this pandemic. It is divided into different phases:
symptomatic phase [3], confirmation phase after the incubation period, medication
phase and recovery period for the COVID-19.
A model can be developed for the symptomatic and incubation phases once
the COVID-19 symptoms are identified for variety of patients. Sources like WHO,
NHS(UK), and other governmental medical organizations could be used to obtain
68 M. P. Atre

Fig. 2 Tracking symptoms Source COVID symptom tracker team, UK

SymptomaƟc MedicaƟon Recovery

ConfirmaƟon
Phase Phase Period

IncubaƟon
Period

Fig. 3 Depiction of the phases of corona (COVID-19)

the data required for the fast and accurate implementation of the models. The symp-
tomatic phase along with the incubation period is modeled by Stephen Lauer et al. [3]
by using log-normal distribution. They estimated the mean incubation period for the
collected database of 181 cases. They also tried to overcome the possible bias on the
result by the patients with normal cough and fever symptoms. The incubation period
estimation is also verified with other parametric distributions like Weibull, Gamma,
and Erlang distributions. What affects these results are (1) the correct discrimination
of normal cough and fever patients from nCov patients and (2) the exact onset time
COVID-19: Loose Ends 69

for incubation estimation! Algorithms can be developed to distinguish the normal

and nCov coughs. The loose ends, that need to tied up!!
There are two categories of the confirmation of COVID-19: local and global. A
local confirmation is the confirmation of the COVID-19 patient’s stage of health
while the global confirmation can be done on the large scale, i.e., the phase of the
spread of the epidemic which counts the number of patients per country. Phase-1,
phase-2, phase-3, phase-4, and phase-5 are the global stages of this epidemic [2].
An exponential graph depicts the phase of the epidemic. A country-wise count is
the global confirmed number of cases. For the correct local and global COVID-19
confirmation, the data must be regularly and immediately updated. This contributes to
the spread factor analysis of COVID-19 and the graph plotting of virus transmission.
Data (number of patients) updating on global level seems to be a limitation!
The confirmation phase has been implemented by using various AI-driven tools.
It is modeled using one of the predictive analysis techniques like decision tree tech-
niques in [10]. This was also used previously in prediction of SARS epidemic. The
researchers implemented decision tree with gain ratio and Gini index for the predic-
tion analysis. The proposed work by authors in [7] aims to extract and evaluate the
coronavirus (COVID-19) caused pneumonia infection in lung using CT scans. An
image-assisted system will be implemented to extract COVID-19 infected sections
from lung CT scans with the final expected view, named as “coronal view.”
After the confirmation phase, medication phase is the most challenging phase
at this point of time where the doctors are using combinational medicines for the
recovery. With sudden arrival and rapid spread of COVID-19, the death rate was
really high. The researchers were trying to analyze the COVID-19 transmission
graph. The global data was expressed with the term “case fatality ratio (CFR).” It is
calculated by dividing the number of known deaths by the number of known cases
[1]. The loophole was the accuracy to identify the nCov cases. The spread factor was
not possible to be calculated because of the complex dependency of the symptomatic
phase. It has dependency on many parameters as: how the virus is transmitted, how
much is the incubation period, had the infected come in contact with others, and the
most important, the reproductive number R0 of the virus [2]. (The ongoing research
on coronavirus has not confirmed with R0 parameter.) These are the loose ends, need
to tie up.
The last phase of recovery has again complex dependency but similar to confir-
mation phase modeling, a binary tree will be helpful with sufficient amount of data.
Figure 4 below summarizes the modeling techniques for the different phases of the
COVID-19.
Apart from the analysis and prediction of COVID-19 at different phases, the trans-
mission of disease also needs to be modeled and a transmission structure needs to
be developed as a precautionary measure. The researchers [9] used auto-encoders
with latent variables and clustering algorithms to group the provinces for investi-
gating the transmission across that province. Another novel forecasting model has
been developed in an early phase of COVID-19 despite the limited data by the
researchers in the article [5]. A polynomial neural network with corrective feedback
(PNN + cf) was developed for the forecasting of COVID-19 transmission.
70 M. P. Atre

SymptomaƟc MedicaƟon Recovery

ConfirmaƟon
Phase Phase Period
•Parametric •Decision Tree •Probability •EsƟmaƟon
DistribuƟon •CT scan of Model Model or
funcƟons like Lungs •Decision Tree •PredicƟve
Log-normal, model-binary
Weibull, tree
Erlang and
Gamma

Fig. 4 Deploying the COVID-19 model

The authors in [6] state that “Considering the spread rate of COVID-19 across the
globe, AI-driven tools are expected to work as cross-population train/test models.”
The researchers in the article [8] also used Composite Monte-Carlo (CMC) approach
toward forecasting of COVID-19 under high uncertainty by extrapolating the avail-
able data. The researchers in the above articles have used different AI-driven tools
to overcome the scarcity of the data with one underlying assumption of no second
transmission.

4 Role of AI, Robotics, and IOT

An insight into the symptoms of COVID-19 and the loose ends in the modeling
process have been discussed in previous two sections. This section discusses how,
the three major areas of technology, AI, IOT, robotics can be exploited in the entire
process of ACM.
The different phases of the COVID-19 could be modeled successfully by tying
up the loose ends mentioned earlier, by using the artificial intelligence or machine
learning algorithms. Calculation of spread factor, the probability of infection, the
estimation of recovery could be solved using machine learning algorithms. The most
important factor to track the travel history of the person is through his mobile GPS
and use it to estimate the spread factor.
As discussed in Sect. 3, AI algorithms (models) can be categorized as: prediction
model to estimate the number of patients and if the person is at the risk of COVID-19
in incubation period, and recovery model to estimate the chances of survival.
By modeling the features of the COVID-19, the control mechanisms can be further
developed using drones, the robotics.
The drones could be used for (1) spraying disinfectants, (2) surveillance, i.e.,
detection of mobility and movement of people and alert message to control room,
wait for further instructions, and spray disinfectant/hit and alarm, (3) identification
of cluster and intimation to control room. A fully equipped drone system can be
COVID-19: Loose Ends 71

designed to control and avoid the spread of COVID-19. The three pillars of science
and technology to win over COVID-19:
AI: for prediction modeling
Robotics: sanitization drones as precautionary measures.
IOT: IOT-based systems for monitoring the quarantined persons.
Internet of things is another major technology which can be deployed on two
major fronts, first in isolation homes on individual scale and the second front is on
large scale for monitoring and tracking of quarantined person.
In isolation homes, IOT can be used for:
(1) maintaining the optimum room temperature for speedy recovery (using ACs
and radiators)
(2) keeping the spirits of the patient high by playing upbeat music, encouraging
audiobooks and suggesting inspiring movies
(3) keeping the patient updated about COVID-19
(4) ensuring the patient follows hygiene etiquette
(5) reminding him/her about taking medications, if any
(6) providing excellent communication facilities to make sure he can talk to his
friends, relatives, etc.
The second front tasks majorly include
(1) Monitoring quarantine people (2) location tracking of a quarantined person, and
(3) sending an alert message to local authority if he crosses the boundary. The second
task is of utmost importance because of scarcity of space for keeping the quarantined
people. In such cases, they can be monitored and tracked silently. And if they disobey
by crossing the pre-defined boundary, an alert message can be sent to local authority
or control room.
To illustrate the role of AI and IOT, a patient monitoring system (PMS) is proposed
here in Fig. 5. It presents the conceptual block schematic of the PMS with all the
facilities required for controlling and curing the COVID-19. Further, the parameters
discussed in Sect. 1.2 can also be monitored using different sensors. The frequency
of monitoring can be selected to be every 2 h or more frequently depending on the
stage of COVID-19.
The use of IoT devices reminds the author of this chapter about the suggestions
given in the article, [11]. The author strongly feels that an increased data sharing
practice must be enforced in the urban health sector while abiding to the dimensions of
privacy and security due to the sensitive nature of information. A smart city initiative
has been taken up by many countries, and based on this, artificial intelligence with
wearable IOT devices can help to trace and track the quarantined person or a patient.
72 M. P. Atre

Fig. 5 Conceptual block schematic of the patient monitoring system

5 Summary

COVID-19 has caused huge damage to the society in terms of number of casualties
and economic activity. In order to minimize its effects, an efficient use of science
and modern technology such as AI, ML, and IOT must be employed in tandem. The
chapter summarizes the loose ends that the scientific and technological community
need to address further. It highlights the role of AI-ML algorithms and use of IOT
systems in combating the ongoing deadly pandemic. The automation in diagnosis
is the need of time because of limited medical experts and that is possible because
of ARI technologies. The implementation of PMS system proposed here will ensure
quicker control with minimal use of manpower to counter COVID-19.

Acknowledgements I would like to thank my family members, Aditya and Pradeep, for their
valuable inputs while writing this chapter.

References

1. Ghani, A. C., et al. (2005). Methods for estimating the case fatality ratio for a novel, emerging
infectious disease. American Journal of Epidemiology, 162(5), 479–486.
2. Battegay, M., et al. (2020). 2019-novel Coronavirus (2019-nCoV): Estimating the case fatality
rate–a word of caution. Swiss Medical Weekly, 150, 0506.
COVID-19: Loose Ends 73

3. Lauer, S. A., et al. (2020). The incubation period of coronavirus disease 2019 (COVID-19) from
publicly reported confirmed cases: estimation and application. Annals of Internal Medicine,
172(9), 577–582.
4. Mayor, S. (2020). Covid-19: Researchers launch app to track spread of symptoms in the UK.
5. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Finding an accurate
early forecasting model from small dataset: A case of 2019-ncov novel coronavirus outbreak.
arXiv preprint arXiv:2003.10776.
6. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: need of active learning
and cross-population train/test models on multitudinal/multimodal data. Journal of Medical
Systems, 44(5), 1–5.
7. Rajinikanth, V., Dey, N., Raj, A. N. J., Hassanien, A. E., Santosh, K. C., & Raja, N. (2020).
Harmony-Search and Otsu based System for Coronavirus Disease (COVID-19) Detection using
Lung CT Scan Images. arXiv preprint arXiv:2004.03431.
8. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing, 106282.
9. Hu, Z., Ge, Q., Jin, L., & Xiong, M. (2020). Artificial intelligence forecasting of covid-19 in
china. arXiv preprint arXiv:2002.07112.
10. Jiang, X., Coffee, M., Bari, A., Wang, J., Jiang, X., Huang, J., … & Wu, Z. (2020). Towards
an artificial intelligence framework for data-driven prediction of coronavirus clinical severity.
CMC-Computers, Materials and Continua, 63(1), 537–551.
11. Allam, Z., Dey, G., & Jones, D. S. (2020). Artificial intelligence (AI) provided early detec-
tion of the coronavirus (COVID-19) in China and will influence future Urban health policy
internationally. AI, 1(2), 156–165.
12. www.who.int.
Social Distancing and Artificial
Intelligence—Understanding the Duality
in the Times of COVID-19

Deepti Gupta, Amit Mahajan, and Swati Gupta

Abstract It is a well-established fact that ‘the human beings are the social beings.’
But many a times in a specific situations or circumstances, human beings are forced to
adopt and practice social distancing. With the recent outbreak of the pandemic caused
by novel coronavirus (COVID-19), people around the world have been advised by
the authorities to practice social distancing. This chapter is a modest attempt to
study the impact of artificial intelligence on social distancing. AI has contributed
and has helped people in different ways in order to maintain social distance. The
current situation as well as the advancement of AI in dealing with it has sociological,
economic, political, cultural and environmental consequences on the lives of the
people all over the world. The present chapter attempts to study the impact of artificial
intelligence on maintaining social distancing. The data for understanding the impact
on the lives of people shall be collected from the secondary sources as collection of
primary data is not possible in this time of coronavirus. The analysis has been done
through different case studies related to various technologies which with the help of
AI aid in facilitating social distancing and in turn curbing the menace of coronavirus.

Keywords Social distancing · Artificial intelligence · COVID-19 · Community ·

Technology

D. Gupta (B)
Department of Sociology, Government MAM College, Cluster University of Jammu, B.R.
Ambedkar Road, Jammu 180001, India
e-mail: [email protected]
A. Mahajan
Centre for IT Enabled Services and Management, University of Jammu, B.R. Ambedkar Road,
180001 Jammu, India
e-mail: [email protected]
S. Gupta
Directorate of School Education, Government of Jammu and Kashmir, Jammu 180005, India
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 75
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_9
76 D. Gupta et al.

1 Introduction

The beginning of the New Year 2020 started with the news about the spread of
a cluster of mysterious, suspected pneumonia cases in China with its epicenter in
Wuhan, the capital of Hubei province. By the first week of 2020, new coronavirus
was identified as the cause of the pneumonia. Since then, it has spread to almost
all the countries of the world and has resulted into infecting more than 2.1 million
people and leaving at least 142,000 dead. Countries, all over the world are fighting
hard to overcome this pandemic.
World Health Organization (WHO) on March 11, 2020 (almost two months after
its outbreak) announced the outbreak of coronavirus illness as a pandemic, which had
affected more than 118,000 people in over 110 countries and territories around the
world and there were chances as well as risk of further spread around the globe [3].
WHO described it as not just public health crises, but a crisis that would touch every
sector, so recommended that every individual and every sector should be included in
order to fight it. Soon after declaring it a pandemic, WHO, issued certain guidelines
to be followed by the people, in order to avoid it from spreading further. One of the
major advices was to maintain ‘social distancing,’ which was rephrased as ‘physical
distancing’ after some time which emphasized the need for physical-not social-
distancing from others. This was changed so as to make people remain connected
with each other while maintaining physical distances. For the purpose of this chapter
phrase ‘social distancing’ is used, as the physical distances are influencing the social
relationships that the people have with each other through the process of face to face
verbal and nonverbal interactions.

2 Social Distancing and Its Relevance During the Spread

of COVID-19

From the varied experiences of people around the world of this novel coronavirus,
it has been concluded that the only available remedy present to the world is social
distancing, i.e., avoiding the physical contact with anyone, who can be the probable
carrier of this virus. Scientists around the globe are trying hard to device measures
to combat this disease which is taking toll on the health system and resulting into
number of deaths. Till the medicines or vaccines are designed for this harmful virus,
the only measure available for human society is to limit its spread through social
distancing.
According to Webster’s Dictionary, the word ‘social distance’ was first used in
the year 1824 and was referred to as the level of acceptance or rejection among
individuals of different groups. But now after the outbreak of this disease related to
coronavirus, social distancing is considered as avoiding a close contact with others
so as to prevent the spread of this contagious disease [10]. Social distancing has been
enforced in many parts of the world through lockdowns. These enforced lockdowns
Social Distancing and Artificial Intelligence—Understanding … 77

have slowed down the spread and flattened the curve of this disease. For example,
the enforced lockdown by China helped it to curb the spread. China was the first
country to put a lockdown into effect in a bid to curb the spread of coronavirus
across the nation on January 23, 2020. At that time, the whole world was stunned by
the decision taken by China but now after the span of about three months the whole
world is under lockdown, i.e., practicing social distancing.
Distancing from the society to protect oneself from getting ill has many problems
associated with it. Humans have been the most social of all living beings and our
survival depends upon others in the society. Governments are taking all necessary
steps to make this lockdown comfortable for the members of the community. If we try
to figure out the factors which are responsible for making these lockdowns successful
and curbing the spread to the community, then the foremost factor in facilitating this
social distancing is the contribution of technology.

3 Artificial Intelligence in the Times of COVID-19

The technologies related to artificial intelligence are considered as the most potential
tool which can facilitate the process of defeating this coronavirus pandemic. Artifi-
cial intelligence or AI is the use of computer science programming to imitate human
thought and action by analyzing data and surroundings, solving or anticipating prob-
lems and learning or self-teaching to adapt to a variety of tasks. The term was coined
in 1955 by John McCarthy, as ‘the science and engineering of making intelligent
machines’ [9].
Since time, we have witnessed the contribution of AI in different fields of our
life. Though the machines have not yet taken over our lives but still they are
affecting our lives, our work and our entertainment to the vast extent. There are
numerous examples and applications for Artificial intelligence in use today ranging
from personal assistants like Alexa and Siri which are voice-powered, to more
underlying and fundamental technologies such as behavioral algorithms, sugges-
tive searches and autonomously powered self-driving vehicles boasting powerful
predictive capabilities [1].
For the purpose of the present study, AI is understood as machine learning (ML)
which is capable of using big data-based models. Natural language processing (NLP),
and computer vision application teaches the computers to use this data, which
is further processed and used for pattern recognition, explanation, and prediction
purposes. For the present study, it has been proved that these functions of the artifi-
cial intelligence can be used for the purpose of recognition (diagnose), prediction, and
explanation (treatment) of coronavirus infections. The knowledge generated through
AI can also be used to manage socio-economic impacts of this pandemic.
78 D. Gupta et al.

4 Impact of AI on Social Distancing

There are many ways through which AI can help in curbing the spread of coronavirus.
As we all know that the most efficient way of preventing the spread of this pandemic
is social distancing, AI has contributed in developing various tools and technologies
which assist the process of social distancing.
Firstly, AI can help in prediction of the spread of these types of viruses. By taking
into consideration the data related to the pattern of spread of coronavirus across
the world, AI can predict the hotspots. Governments across the nations can use
such information to prevent the places becoming hotspots of coronavirus. Thus, the
authorities can give early warnings and alerts to people to maintain social distancing
and take precautions in those predicted hotspots.
Secondly, AI can help in the detection of the disease by identifying the symptoms.
There are various apps being made across nations to identify and detect people with
symptom, which in turn can help in curbing the spread of the disease. Those with
symptom, detected with the help of AI, can maintain social distance from rest of the
community by self-isolating them and can consult the medical professionals if the
symptoms become severe. Thus, it also facilitates the process of tracking the contacts
or those with symptoms to prevent it from further spread.
Thirdly, AI can facilitate the responses after the disease has been identified. By
analyzing the pattern of symptoms and the spread of virus, AI can be used in devel-
oping the mode of treatment as well as vaccines and medicines which are most likely
to be effective for curing the patients. Machine learning can, thus, result in opti-
mizing the disease management after the outbreak of the disease. The data related
to the disease with the help of AI can be used for diagnosis and prognosis purposes.
The treatments and cures which are being developed are the outcome of the use of
AI tool and technologies.
Lastly, AI can also assist in initiating a response after the outbreak has been
contained. It may guide the governments and health agencies to device out a plan
to prevent such type of outbreaks in future. Health policies, plans and programs can
be initiated by taking into considerations the finding of machine learning. Thus, AI
has the potential to assist in the prediction, guidance, controlling the menace of this
disease.
There are many success stories whereby, the communities with the help of various
applications, which are AI based, have been able to curtail the spread of this infectious
virus. In this chapter, the further discussion is on those applications/technologies
which are studied as case studies.
Social Distancing and Artificial Intelligence—Understanding … 79

5 Case Studies Related to Application of Artificial

Intelligence

5.1 Case Study 1

Aarogya Setu Mobile App was launched on April 2, 2020 and was developed by
National Informatics Centre. With the help of this application, the Indian govern-
ment aims to help connect essential health services with the people while India is
under lockdown until May 3, 2020. The description of this app is that it is aimed
at augmenting the initiatives of the Government of India, particularly the Depart-
ment of Health, in proactively reaching out to and informing the users of the app
regarding risks, best practices and relevant advisories pertaining to the containment
of COVID-19.
This mobile application which was launched by the Ministry of Electronics and IT,
Government of India, with the objective is to make people aware of coronavirus and
COVID-19. This application, which is available in eleven Indian languages including
English, uses the location as well as Bluetooth data of the user to notify him/her about
the location of the person near them, who is under quarantine or tested positive. It
enables people to identify and access their risk of contracting coronavirus infection.
It alerts the person if he/she comes in close proximity of, even unknowingly, those
who have tested positive for COVID-19. It also includes the detailed information on
how to self-isolate in case one develops symptoms and when one needs medical help.
Thus, the app calculates the risks and predicts on the basis of people’s interaction
with others, using Bluetooth technologies, algorithm and artificial intelligence.

5.2 Case Study 2

The success story of Taiwan, a small island in close proximity to China, is another
example of the use of AI in combating COVID-19 and curtailing its spread. Early
action, centralized command system, frequent-transparent communication, recourse
allocation along with integration of big data and technology are the contributing
factors in the success of Taiwan.
Taiwan’s government combined the data from the National Health Insurance
Administration and Immigration Agency and identified 14-day travel history of
patients. Along with that, they used the data from citizen’s household registration
system and foreigner’s entry cards to identify individuals, who were at high risk, self-
quarantined them and monitored them through their cell phones. This whole data was
utilized through artificial intelligence to provide information about patient’s travel
histories to all the hospitals, clinics, etc. across the country and also to update the
people about the hotspots or risky areas which they need to avoid. Thus, AI helped
in maintaining social distancing as well as disease prevention.
80 D. Gupta et al.

5.3 Case Study 3

Another success story is that of Republic of Korea, who have been able to flatten
the curve of spread of coronavirus with the help of using the technologies related
to artificial intelligence. They developed the testing kits using AI within weeks of
the outbreak. This widespread testing, with the capacity of 15,000 tests per day,
targeted the high-risk groups. Smart Quarantine information system was developed
using information from Ministry of Justice, the Ministry of Foreign Affairs, airline
companies and major telephone telecommunication companies. This information
was then processed using AI platforms and then transferred to the medical staff to
have full access of the history of the places that the infected person had traveled in
the recent past, to identify and isolate or treat them in a timely manner.
Mobile phone technology data was used for self-health check and to receive
information regarding precautions, symptoms and the procedure to follow, if the
person develops any coronavirus symptom. In order to trace the contacts of an infected
person, they processed the data gathered from mobile phones, credit cards transaction
records and CCTV footage. This further helped them to publish detailed maps of
movements of infected people by using AI.
Artificial intelligence-based tools and technologies were being used and it helped
in initiating the process of efficient and timely diagnosis which further helped them
in classifying those affected by the coronavirus into four groups. For example, they
have been able to develop a hand-held chest X-ray camera which can be used for
scanning the chest of the patient in just three seconds. Furthermore, mobile apps were
used for information sharing related to healthcare staff, volunteers, testing stations,
point of purchase for available masks, etc. Robots using AI technologies were also
manufactured and were being used by the authorities to provide information to the
people on different ways of responses to this infectious virus.

6 Conclusions

There are many ways to tackle the contagious pandemic but the best one is social
distancing or self-isolation, and the technologies related to artificial intelligence
have contributed in facilitating it, for the control of this disease. The role of AI in
combating this infection can be understood at various levels, i.e., from predicting
the spread in the hotspots to assisting the prevention by facilitating social distancing,
from guiding in discovering the treatment to planning the policies so that such type
of pandemic can be avoided in the future. There are numerous examples which depict
the contribution of AI in our daily lives but during this period of turmoil. AI-related
tools and technologies have contributed much and the experts are working hard to
device measures to further strengthen its contribution. There is an AI-based detector
which act as a tool that issues an alert when anyone is less than the desired distance
from a colleague at a workplace. Another AI company in USA unveiled a new feature
Social Distancing and Artificial Intelligence—Understanding … 81

in its interface for construction site workers, which was able to point out the indicators
related to safety, productivity and quality, that include worker proximity and their
use of personal protective equipment. Such examples along with many more have
proved the potential use of AI in developing a solution for combating this infectious
illness.
Social distancing, be it in the form of self-isolation, quarantine or at the workplace,
is essential for the prevention of the spread of COVID-19 and is being facilitated by
the tools and technologies related to AI.

References

1. Adams, R. L. (2017). 10 Powerful examples of artificial intelligence in use today. Forbes, Jan
10, 2017. www.forbes.com.
2. Chen, S. (2020). Taiwan sets example for world on how to fight Coronavirus. ABC NEWS,
March 13, 2020. https://abcnews.go.com/Health/taiwan-sets-world-fight-coronavirus/story?
id=69552462.
3. Ducharme, J. (2020). https://time.com/5791661/who-coronavirus-pandemic-declaration/.
4. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Finding an accurate
early forecasting model from small dataset: A case of 2019-ncov novel coronavirus outbreak.
arXiv preprint arXiv:2003.10776.
5. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing, 106282.
6. ITU News. (2020). COVID-19: How Korea is using innovative technology and AI to flatten the
curve. ITU News, April @ 2020. www.News.itu.int/covid-19-how-korea-is-using-innovative-
technology-andai-to-flatten-the-curve/.
7. Leo, K. (2020). Coronavirus: Covid-19 detecting apps face teething problems. BBC NEWS,
April 8, 2020. www.bbc.com/news/amp/technology-52215290.
8. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: Need of active learning
and cross-population train/test models on multitudinal/multimodal data. Journal of Medical
Systems, 44, 93. https://doi.org/10.1007/s10916-020-01562-1.
9. Sraders, A. (2019). What is artificial intelligence? Examples and news in 2019. TheStreet,
January 3, 2019. www.Thestreet.com.
10. Webster’s Dictionary. (2020). https://www.merriam-webster.com/dictionary/social%20distanc
e#h1.
Post-COVID-19 and Business Analytics

Monomita Nandy and Suman Lodh

Abstract This paper highlights the way companies can apply artificial intelligence
(AI) in the post-COVID-19 period. We show that how the AI can be advantageous
to develop an inclusive model and apply to the businesses of various sizes. The
recommendation can be beneficial for academic researchers to identify several ways
to overcome the obstacles that companies may face in post-COVID-19 period. The
paper also addresses few major global issues, which can assist the policy makers
to consider developing a business model to bounce back the world economy after
this crisis is over. Overall, this paper enhances the understanding of stakeholders of
business about the importance of application of the AI in businesses in a volatile
market in post-COVID-19 period.

Keywords Artificial intelligence · Crisis · COVID-19 · Artificial neural network ·

Coronavirus · Economic development

1 Introduction

The current COVID-19 outbreak, started in December 2019 in Wuhan city of China,
brings an extraordinary crippling impact on the world economy. During this unprece-
dented socio-economic crisis for business, it is too early to recommend a business
model for companies that can be useful when the world is out of the COVID-19
pandemic. Based on the existing literature on financial crisis or similar exogenous
shocks, researchers have started predicting the effect of COVID-19 on world financial
markets and direct or indirect impact on economic development (e.g. [1, 8]). After
the failure of the Lehman Brothers in 2008, a strand of literature has evolved and

M. Nandy (B)
Brunel Business School, Brunel University London, Kingston Lane, Uxbridge UB8 3PH, UK
e-mail: [email protected]
S. Lodh
Middlesex Business School, Middlesex University London, The Burroughs, Hendon, London
NW4 4BT, UK
e-mail: [email protected]

© The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2020 83
A. Joshi et al. (eds.), Intelligent Systems and Methods to Combat Covid-19,
SpringerBriefs in Computational Intelligence,
https://doi.org/10.1007/978-981-15-6572-4_10
84 M. Nandy and S. Lodh

started discussing the ‘Space Economy’ that focuses on the application of advanced
technologies such as the artificial intelligence (AI). Extant studies on the AI show
the applicability and effectiveness of the AI in restructuring and reorganisation of
economies and financial markets across the world (see [13, 14]). To initiate the
economic development and reduction of inequality in allocation of resources for
development of stakeholders, application of this technology receives an immense
importance in academia and practice. Based on the above discussion, the objective
of this paper is to identify the scope of application of the AI by companies in the
post-COVID-19 crisis period as there is a lack of detailed studies on the impact of the
use of the AI to overcome a pandemic shock similar to the one we are experiencing at
the beginning of the year 2020. To the best of our knowledge, this original research
is the first study to highlight the possibility of application of the AI by companies in
the COVID-19 recovery phase.

2 Advantages of Using the AI After the End of COVID-19

Crisis

Companies can maximise the value of their concerns by minimising the operating
cost. Porter [19] argues that companies apply their sustainable models to take compet-
itive advantages over their peers. One of the main challenges faced by companies
during the last decade is to deal with big data created by high information flow through
the Internet. To address this challenge, companies start utilising the AI to improve the
world economy [15]. Similar to big companies, small- and medium-size enterprises
with the government interventions allow themselves to think in a creative manner.
In addition, these companies introduce several disruptive changes in their businesses
by applying the AI. The development of such infrastructure by big, medium and
small companies positively impacts the rate of unemployment, GDP and inflation, to
name a few, of many countries [17]. Moreover, an application of the super-intelligent
system creates new scopes for business of various sizes and allows the transmission
of required information within a few nanoseconds. As a result, the development of
the economy becomes visible because companies, mostly in advanced countries, of
all sizes can apply this sophisticated and efficient business model based on advance
technology such as the AI. In fact, the analysis of big data allows the companies to
reduce the percentage of error in their business model. Furthermore, the global coor-
dination and participation has increased because of the application of these advanced
technologies as knowledge and research and development (R&D) begin to transmit
widely from one country to another.
Competition among peers in the same industry or between big and small busi-
nesses influences innovation to find a sustainable business model. The AI-based
models allow companies to reach rural or underdeveloped areas by introducing user-
friendly technologies in day-to-day life. For example, a digital-biological converter
can make a number of copies of flu vaccines remotely in the absence of human to
Post-COVID-19 and Business Analytics 85

support the local health system [2]. Thus, various sectors such as health, transport,
manufacturing and agriculture contribute to the development of the country-level
economy, which in turn affects the world economy. During the financial crisis in
2008, the application of the AI by companies remains largely limited. However,
recently, due to rapid advancement of technologies, companies are trying to apply
a composite Monte Carlo decision-making process in the highly uncertain post-
coronavirus period [5, 6]. In contrast, before applying the AI-based models to recover
the economy from the current crisis, companies need to consider the unprecedented
damage caused by the novel coronavirus which is not comparable with the previous
financial crises, e.g. collapse of the Lehman Brothers.

3 Application of AI for Global Development

in Post-COVID-19

One of the major global challenges for the past decades is to keep the global warming
stay below 2 °C to reduce the risk of biodiversity. The rapid climate change is posing a
high risk to the livelihoods of human being and to animal kingdom. A huge number
of studies indicate that failure to conserve wildlife can create a threat to human
being (https://www.eauc.org.uk/6998). In addition, the modern business activities
are causing damage to environment, and this in turn can allow the harmful virus to
find shelter in human being. Thus, the wildlife conservation needs to take care of a
large database associated to business, which is difficult to obtain manually. In other
words, companies need to identify the habitats to protect and then create wildlife
corridors as these corridors have immense biological significance. Let us take the
case of Montana and Idaho in the USA. The conservation scientists of wild animal
are using the AI-assisted system to track and record the movements of wild animals.
As a result, to reduce the risks associated with biodiversity, company can use the AI-
embedded technologies and can continue to work on maintaining the climate change
in the post-COVID-19 pandemic period.
The immense use of the AI can be observed in the healthcare sector, which is
a big challenge for all countries around the world. During the recent pandemic,
we observe the use of AI-driven tools and importance of active learning and cross-
population test models [22]. For instance, robots can disinfect hospitals to support
the health workers, 3D printers can manufacture the personal protective equipment
(PPE) for health workers in hospitals and care homes, smartphone enabled-tracking
system can identify the close contact of a person with another infected person, etc.
are few to mention. Similar to the COVID-19 pandemic, we can also observe an
application of the AI by healthcare business in the past decades. For example, the AI
system of IBM Watson Health, in partnership with Barrow Neurological Institute,
has been used to manage in analysis of many studies to conclude about the genes
related to Amyotrophic Lateral Sclerosis (ALS) disease. In addition, remote treatment
without risking the life of health carer is only possible by the application of advanced
86 M. Nandy and S. Lodh

technology. Consequently, once we come out of the crisis, it is important for the
companies to analyse a huge amount of information available from every affected
country through their AI-based forecasting model. This can mitigate the risk of
encountering the similar pandemic in the future.
In the recent days, we observe a huge investment in developed and developing
countries by the public and private businesses for clean energy (Bloomberg NEF).
In the post-COVID-19 era, the business can start utilizing the invested resources
and generate more units of clean energy (or green energy) with the help of AI-
based technology. For example, quantum computing can create plasma reaction in a
nuclear fusion reactor which can reduce fuel-based energy and produce clean energy.
Companies can also focus on assisting large businesses to find a technology-enhanced
way of controlling the cost associated with the cooling system in the big data centre.
Deep mind is one of the examples of cost-saving smarter energy used by big business,
like Google.
The non-self-improving AI can create philosophical zombies (p-zombies), where
we can observe a dead subjectivity. By combining the AI with the existing technolo-
gies, companies can address complicated problems through biological or artificial
neural networks [4] or they can use the AI that do not self-improve even after inter-
acting with the government structures. Industry can focus more on a short period for
accurate early forecasting model using small dataset to check the suitability of an
application of the AI [5, 6]. If businesses can get knowledge about how to reduce
the cost of application of AI, how to merge the AI to time [3] and how to control
various parameters of global issues by application of the AI [28]. As a result, the
global control system can introduce a limited super-intelligence for the benefit of the
society [7].

4 An Experiment on Application of the AI on Forecasting

Cryptocurrency

Let us take a practical application of the AI on a real-time data. In this section, we show
the use of artificial intelligence, especially an application of artificial neural network
(ANN) in time-series forecasting. Similar to the function of human brains, the ANN is
composed of a large number of highly interconnected processing elements. Currently,
the application of neural network is considered as one of the most sophisticated
methods for natural language processing and computer data visioning. For instance,
in a study on bankruptcy prediction of banks and firms, Kumar and Ravi [11] find
that ANN algorithm can outperform many single or hybrid classical forecasting
techniques such as ARIMA and GARCH. In this brief experiment, we use a combi-
nation of well-known neural network algorithms such as long short-term memory
(LSTM), time-lagged neural network (TLNN), feed-forward neural network (FNN)
and seasonal artificial neural network (SANN) to forecast a sample (time-series) of
Post-COVID-19 and Business Analytics 87

Fig. 1 Closing price of Bitcoin

cryptocurrency closing price1 for the year 2014–2020 (till April 28, 2020). Figure 1
shows the time-series graph of a daily closing price data for bitcoin retrieved from
Kraken.
To keep our analysis simple, we calculate the monthly average closing price in
each year from the daily 2298 observations. We use 25% of this information as the
test data and 75% data as the training data. In this training method, all the above-
mentioned four models try to recognise the regularities and patterns in the input data,
learn from historical data and then provide us generalised forecast values, based on
the known previous information. Thus, the method is self-adaptive and non-linear in
nature. So, it overcomes a priori assumptions of statistical distribution of the data.
Based on the optimal parameters—such as root mean square errors (RMSE), our
experiment suggests that LSTM model is a better method to forecast the bitcoin
price movement. Our finding is reported in Fig. 22 , and it indicates a downward
trend of price of cryptocurrency since January of 2020. However, the model can be
a complex one in practice if we consider transaction costs and other financial or
environmental exogenous shocks such as economic lockdown due to COVID-19.
Note that, our above-mentioned experiment is to show the applicability of ANN
rather than drawing conclusion from the results for policy makers.

5 Challenges of Using the AI After the End of COVID-19

Crisis

The AI opens up a new chapter in the world economy. But, many studies, e.g. Roubini
[20] and Stiglitz [25] raise serious negative implications of the application of AI in
the World Economic Forum 2015 (WEF 2015). They mention that there is a need of
huge investment of money and R&D to invest in the AI-embedded robots that can

1 The data is obtained from www.cryptydatadownload.com.

2 The other figures are reported in the Appendix.
88 M. Nandy and S. Lodh

Fig. 2 forecast values of closing price

only be borne by large companies only. Thus, there is a limited opportunity to include
both small and large businesses in the same model which may not be sustainable in
the growing economy. Extant studies argue that a huge job loss can slow down the
progress of the economy [12]. Popper [18] reports that the volatility can be higher
in the economy when companies are able to use alternative digital money such as
cryptocurrency. So, the lack of opportunities to small businesses can increase a higher
gap in performance between public and private sectors or small and large companies.
This may reduce the efficiency and accuracy in big data analysis and development of
a business model applicable to all companies. The privilege of a group of companies
with application of the AI may limit the development of the world economy. In
addition, there can be a certain catastrophic AI risk. The challenges around the AI-
safety or its alignment (see [21]) can be a huge concern for the companies, especially
in the post-coronavirus crisis as there may be a lack of job force in practice.
As it is hard to be definite about the future uncertainties, companies can depend
on forward looking taxonomy. One of the popular taxonomies, stated by Sotala and
Yampolskiy [24], defines the dangerous impact of friendly AI [27]. For example,
a bio-hacking of business can use AI to understand the published genomes, which
might create a multi-pandemic [26] and such business model can create neural inter-
faces to affect the human brains adversely [10]. Thus, it still remains a puzzle as
to what extent companies can apply the AI safely and effectively once the world
economy gradually comes out of the COVID-19 pandemic.
Post-COVID-19 and Business Analytics 89

6 Conclusion

In this forward-looking, constructive paper, we identify few challenges and yet major
advantages that any business can take advantage in using the AI in the post-COVID-
19 period. However, we acknowledge that immense challenges are waiting for us and
the policy makers, around the world, should come together to solve these issues. One
major challenge to the policy makers is to identify how to implement ethical practices
in business to transmit data securely to get it analysed by AI-based technology for
the benefit of the society. The local and international decision makers need to take
responsibility to share their expertise to educate the mass about technology, which can
reduce the risk of job loss. Moreover, by creating “Artificial Intelligence Marketing”
[16], the restructuring of the world economic development is possible when the
regulators allow the business to use the AI to enhance production-led profitability and
to reduce the associated risk through innovative methods. Following the predictions
of other studies, we expect that AI-led business can outperform all tasks of human as
early as 2023 after the world economy manages to come out of COVID-19 pandemic
(see [9, 23]). In a nutshell, the applications of the AI in the post-COVID-19 era
can allow the individual and business communities to work together for the world
development at a rapid rate by outweighing the negative aspect of using technology
in society.

Appendix: Actual and Predicted Values from ANN Models

90 M. Nandy and S. Lodh

References

1. Bin, M., Cheung, P., Crisostomi, E., Ferrearo, P., Lhachemi, H., Murray-Smith, R., Myant, C.,
Parsini, T., Shorten, R., Stein, S., Stone, L. (2020). On fast multi-shot Covid-19 interventions
for post lock-down mitigation. https://arxiv.org/pdf/2003.09930.
2. Boles, K.S., Kannan, K., Gill, J., et al. (2017). Digital-to-biological converter for on-demand
production of biologics. Nature Biotechnology, 35, 672–675.
3. Bostrom, N. (2003). Astronomical waste: The opportunity cost of delayed technological
development. Utilitas, 15, 308–314.
4. Christiano, P. (2016). Prosaic AI alignment. https://ai-alignment.com/prosaic-acontrol-b95964
4d79c2.
5. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Composite Monte
Carlo decision making under high uncertainty of novel coronavirus epidemic using hybridized
deep learning and fuzzy rule induction. Applied Soft Computing, 106282.
6. Fong, S. J., Li, G., Dey, N., Crespo, R. G., & Herrera-Viedma, E. (2020). Finding an accurate
early forecasting model from small dataset: A case of 2019-ncov novel coronavirus outbreak.
arXiv preprint arXiv:2003.10776.
7. Goertzel, B. (2012). Should humanity build a global ai nanny to delay the singularity until it’s
better understood? Journal of Consciousness Studies, 19(1–2), 96–111.
8. Goodell, J. W. (2020). COVID-19 and finance (2020) Agendas for future research. Finance
Research Letters https://doi.org/10.1016/j.frl.2020.101512.
9. Grace, K., Salvatier, J., Dafoe, A., et al. (2017). When will AI exceed human performance?
Evidence from AI experts. arXiv:1705.08807 [cs.AI].
10. Hines N (2016) Neural implants could let hackers hijack your brain. In: Inverse. https://www.
inverse.com/article/19148-neural-implants-could-let-hackers-hijack-your-brain.
11. Kumar, P. R., & Ravi, V. (2007). Predictions in banks and firms via statistical and intelligent
techniques—A review. European Journal of Operational Research, 180(1), 1–28.
12. Lalive, R. (2007). Unemployment benefits, unemployment duration, and postunemployment
jobs: A regression discontinuity approach. The American Economic Review, 97(2), 108–112.
13. Lewis, Maureen. (2001). The economics of epidemics. Georgetown Journal of International
Affairs, 2, 25.
14. Madhav, N., Oppenheim, B., Gallivan, M., Mulembakani, P., Rubin, E., & Wolfe, N. (2017).
Pandemics: Risks, impacts, and mitigation. In: Disease control priorities: Improving health and
reducing poverty, 3rd edn. The International Bank for Reconstruction and Development/The
World Bank.
15. Manyika, J., Chui, M., Bughin, J., Dobbs, R., Bisson, P., & Marrs, A. (2013). Disruptive
technologies: Advances that will transform life, business, and the global economy (Vol. 180).
San Francisco, CA: McKinsey Global Institute.
16. Murray, S. H., & Keevil, A. A. (August 21, 2014). System and method for interactive virtual
banking, Patent, No: US 20140236740 A1.
17. Öztürk, B., & Mrkaic, M. (2014). SMEs’ access to finance in the Euro area: What helps or
hampers? IMF Working Paper, WP/14/78. International Monetary Fund, European Department.
18. Popper, N. (29April 2015). Can Bitcoin conquer Argentina? New York Times Magazine. http://
www.nytimes.com/2015/05/03/magazine/howbitcoin-is-disrupting-argentinas-economy.
html?smid=fbnytimes&smtyp=cur&bicmp=AD&bicmlukp=WT.mc_id&bicmst=140923272
2000&bicmet=1419773522000&_r=0.
19. Porter, M. E. (1985) Competitive advantage: Creating and sustaining superior performance.
Free Press, New York, NY.
20. Roubini, N. (2014, December 08). Rise of the machines: Downfall of the Economy. http://www.
roubinisedge.com/nouriel-unplugged/rise-of-themachines-downfall-of-the-economy.
21. Russell, S. (2017). 3 principles for creating safer AI. https://www.youtube.com/watch?v=EBK-
a94IFHY.
Post-COVID-19 and Business Analytics 91

22. Santosh, K. C. (2020). AI-driven tools for coronavirus outbreak: Need of active learning
and cross-population train/test models on multitudinal/multimodal data. Journal of Medical
Systems, 44, 93. https://doi.org/10.1007/s10916-020-01562-1.
23. Shakirov, V. (2016). Review of state-of-the-arts in artificial intelligence with application to AI
safety problem. ArXiv Prepr ArXiv160504232.
24. Sotala, K., & Yampolskiy, R. (2014). Responses to catastrophic AGI risk: A survey. Physica
Scripta, 90, 018001.
25. Stiglitz, J. E. (2014, November). Unemployment and innovation. Working Paper 20670 (Vol.
1050, p. 3). National Bureau of Economic Research. http://www.nber.org/papers/w20670.
26. Turchin, A., Green, B., Denkenberger, D. (2017). Multiple simultaneous pandemics as most
dangerous global catastrophic risk connected with bioweapons and synthetic biology. Rev
Health Secur.
27. Yudkowsky, E. (2003). HUMOR: Friendly AI critical failure table. http://www.sl4.org/archive/
0310/7163.html.
28. Yudkowsky, E. (2008). Artificial intelligence as a positive and negative factor in global risk,
in global catastrophic risks. Oxford: Oxford University Press.