Received 30 July 2024, accepted 25 August 2024, date of publication 28 August 2024, date of current version 10 September 2024.

Digital Object Identifier 10.1109/ACCESS.2024.3450970

Impact of Artificial Intelligence in Nursing for

Geriatric Clinical Care for Chronic Diseases:
A Systematic Literature Review
MAHDIEH POODINEH MOGHADAM 1 , ZABIH ALLAH MOGHADAM2 ,
MOHAMMAD REZA CHALAK QAZANI 3 , PAWEŁ PŁAWIAK 4,5 ,
AND ROOHALLAH ALIZADEHSANI 6 , (Member, IEEE)
1 Department of Nursing, Faculty of Nursing and Midwifery, Zabol University of Medical Sciences, Zabol 98616-15881, Iran
2 Department of Computer Engineering, School of Technical and Engineering, Birjand Branch, Islamic Azad University, Birjand 14778-93855, Iran
3 Faculty of Computing and Information Technology, Sohar University, Sohar 311, Oman
4 Department of Computer Science, Faculty of Computer Science and Telecommunications, Cracow University of Technology, 31-155 Kraków, Poland
5 Institute of Theoretical and Applied Informatics, Polish Academy of Sciences, 44-100 Gliwice, Poland
6 Institute for Intelligent Systems Research and Innovation (IISRI), Deakin University, Waurn Ponds, VIC 3216, Australia

Corresponding author: Paweł Pławiak ([email protected])

ABSTRACT Nurses are essential in managing the healthcare of older adults, particularly those over 65, who
often face multiple chronic conditions. This group requires comprehensive physical, mental, and functional
care. Recent advancements in artificial intelligence (AI) have significantly improved nursing capabilities
by enabling real-time health monitoring, thus bolstering the early detection and prevention of severe health
issues. Despite these advancements, the current systematic literature predominantly focuses on machine
learning (ML) applications for a limited set of chronic diseases, often overlooking the extensive capabilities
of deep learning (DL) technologies. Additionally, these reviews cover a narrow spectrum of studies,
potentially needing broader insights and developments in the field. To address these shortcomings, our study
conducts a systematic literature review of ML and DL applications in geriatric care for chronic disease
management. We meticulously analyzed peer-reviewed articles published from 2014 to 2024, concentrating
on AI technologies in elderly care. This review included 76 selected articles from leading publishers
such as Elsevier, Springer, IEEE, MDPI, Wiley, Taylor & Francis, Nature, Cambridge University Press,
Oxford University Press, and arXiv, which we categorized into three main groups: Neurological disorders
(27 articles), Mental Health disorders (22 articles), and Physical/physiological disorders (27 articles). Our
findings reveal that Random Forest, logistic regression, and convolutional neural network (CNN) are the most
frequently used AI techniques, typically evaluated by accuracy metrics and the area under the curve (AUC).
The findings indicate that although AI applications in geriatric care are promising, they require significant
enhancements in technology and methodology to improve accuracy and reliability. Future research should
focus on developing advanced AI tools, integrating cutting-edge deep learning models and comprehensive
datasets to refine diagnostics and treatment protocols for chronic diseases in the elderly, ultimately enhancing
patient outcomes.

INDEX TERMS Nurse, older patients, chronic disease, artificial intelligence, machine learning, deep
learning.

I. INTRODUCTION
The United States Census Bureau projects that the population
The associate editor coordinating the review of this manuscript and of older adults will reach 88.5 million by 2050 [1], [2]. This
approving it for publication was Moussa Ayyash . demographic frequently faces numerous concurrent health
2024 The Authors. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.
For more information, see https://creativecommons.org/licenses/by-nc-nd/4.0/
VOLUME 12, 2024 122557
 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

challenges that can result in deteriorating health outcomes, a nursing context. We evaluated research published between
increased disability, and greater reliance on institutional care. 2014 and 2024 that utilized AI algorithms to assist in manag-
Rapid and accurate diagnosis is challenging due to older ing chronic conditions in older adults. Our extensive review
patients’ complex medical histories and symptoms. Figure 1 led us to select 76 articles from renowned publishers such
illustrates key issues impacting these patients and their as Elsevier, Springer, IEEE, MDPI, Wiley, Taylor & Fran-
caregivers [3]. cis, Nature, Cambridge University Press, Oxford University
Comprehensive assessments of older patients require thor- Press, and arXiv. We organized these articles into three main
ough and time-consuming efforts that demand collaboration categories based on the health issues they address: Neu-
across various healthcare disciplines. Nurses play a critical rological disorders (27 articles), Mental Health disorders
role in the ongoing care and monitoring of older patients. (22 articles), and Physical/physiological disorders (27 arti-
However, they often encounter difficulties in fully assessing cles). In summary, this systematic review showcases:
individuals with multiple health issues during routine visits. • An in-depth analysis of ML and DL methods to evaluate
These challenges are exacerbated by the increasing burden the effectiveness of algorithms across 76 studies span-
of clinical paperwork, the use of outdated technologies, and ning the last 11 years.
a shortage of nurses specialized in geriatric care, all of • We also explore future research directions in automated
which can lead to suboptimal care and possibly missed or detection techniques, shedding light on the changing
incomplete diagnoses [4]. Physicians complement the work scope of chronic disease management.
of nurses by typically leading the initial assessments and • The review highlights research on identifying chronic
diagnosis. They make key medical decisions using detailed conditions using various data types, including clinical
clinical assessments, laboratory results, and diagnostic tools records, imaging, and biological samples.
that nurses may not utilize extensively [5]. Nevertheless, • The review identified 18 unique diseases and recognized
there are significant areas of shared responsibilities. For 9 distinct algorithms.
instance, both nurses and physicians are actively involved in This study is structured into five key sections. Section II
the ongoing assessment and management of chronic condi- presents related works, while Section III outlines the method-
tions prevalent among older patients, collaborating to monitor ology adopted for the selection of research papers. Section IV
treatment effectiveness and adapt care plans as needed based provides an overview of current machine learning and deep
on changes in patient conditions. This teamwork enables learning research. Section V offers a discussion of the studies
the swift and accurate exchange of patient observations and examined. Finally, Section VI concludes the article.
treatment responses, which is crucial for improving medical
outcomes [6]. II. LITERATURE REVIEW
Artificial intelligence (AI) in healthcare can enhance pre- A. THREATS TO VALIDITY
vention, diagnosis, and treatment [7], [8], [9]. AI applications This section outlines potential biases and limitations inherent
are already making strides in areas such as robotic-assisted in the reviewed studies within the domain of chronic disease
disease management [10], [11], cancer detection [12], [13], management using AI. Several factors may compromise the
[14], and improving patient safety [15], [16]. Nurses are using validity of these reviews:
AI to guide decision-making, streamline drug development, • Narrow disease focus: Numerous systematic reviews
and monitor patient care more effectively [17]. In some target a limited array of chronic diseases, potentially
instances, AI has started to exceed human capabilities, espe- introducing bias by failing to represent the broader spec-
cially in handling large datasets and identifying patterns that trum of conditions that afflict older adults. This narrow
might elude human observers [18]. For instance, AI systems focus restricts the generalizability of the findings and
have shown superior accuracy in diagnosing diseases from may overlook diseases where AI could have a substantial
imaging data at a faster rate than human radiologists [19]. impact.
As another example, AI algorithms have successfully pre- • Technological limitations: There is a notable predom-
dicted disease outbreaks by analyzing patterns in social media inance of traditional machine learning techniques in
data, which human analysts may miss due to the sheer volume the literature, often at the expense of exploring more
and complexity of information [20]. advanced deep learning technologies. This technological
Recent systematic reviews in chronic disease management bias can skew the outcomes of the reviews, leading to
have made essential contributions [21], [22] but encounter an underestimation of the potential benefits that newer,
significant limitations. Firstly, they tend to focus on a narrow more sophisticated AI models could provide. These
range of chronic diseases. Secondly, while they frequently advanced models can analyze more complex data sets
discuss machine learning applications, they often overlook and identify nuanced patterns that traditional meth-
the potential of deep learning techniques. Additionally, these ods might miss, significantly improving diagnosing and
reviews typically analyze a limited selection of articles. managing complex chronic conditions.
To address these issues, our study conducts a comprehen- • Selective article analysis: Often, these reviews con-
sive literature review on applying AI in senior care within clude a restricted selection of scholarly articles. Such

122558 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

FIGURE 1. Visual representation of the needs of older people and the challenges nurses face.

selection bias can result in an incomplete assessment of uses a CNN to analyze spectrogram images from elec-
the current state of AI in healthcare, potentially missing troencephalography (EEG) data. This system is designed to
out on critical studies that offer a more comprehensive identify multiple neurological disorders, including autism,
understanding of the field’s capabilities and limitations. epilepsy, Parkinson’s disease, and schizophrenia, as well as
This selective approach may ignore pivotal research that to differentiate these from healthy controls. Additionally,
could challenge existing conclusions or highlight new Mujahid et al. [25] introduced an ensemble method that com-
opportunities for AI in healthcare. bines VGG16 and EfficientNet for diagnosing Alzheimer’s
• Outdated data: The rapid advancement in AI technolo- disease from magnetic resonance imaging (MRI) images.
gies means that some reviews might incorporate studies They used an adaptive synthetic oversampling technique to
that need to be updated at the time of publication. This balance the highly imbalanced dataset, thereby enhancing the
inclusion can lead to conclusions that may need to accu- training of the model and improving diagnostic accuracy by
rately reflect the current capabilities and challenges of learning complex patterns from the combined outputs of both
AI applications in healthcare, rendering some findings models.
obsolete. Several review articles have focused on chronic diseases.
To address these threats, a meticulous approach is nec- Woodman and Mangoni [26] explored the growing role of
essary. It involves systematically selecting and evaluating a machine learning in healthcare, driven by increased access
wide range of studies that cover various diseases and integrate to global health data. Their review details the taxonomy of
the latest advancements in AI technology with intense learn- machine learning algorithms, outlining the functions, capa-
ing. This comprehensive approach ensures a more balanced bilities, and specific applications in geriatric medicine. They
and current view of the field, promoting a deeper understand- stress the importance of educating clinicians about AI to
ing of how AI can effectively contribute to chronic disease encourage its adoption. They also discuss the challenges
management in diverse healthcare settings. of using clinically approved, yet often less interpretable,
machine learning tools in patient care. They also high-
B. COMPREHENSIVE REVIEW OF EXISTING RESEARCH light the need to develop explainable machine learning to
Chronic diseases are common in the elderly and are char- build trust and utility in clinical settings. Hamaker et al.
acterized by their complex and multifaceted progression. [27] conducted a systematic review by searching Medline
In geriatric care, accurately predicting disease trajectories and and Embase for studies analyzing the effects of geriatric
potential acute events early on can significantly enhance man- assessments on treatment decisions, non-oncologic inter-
agement strategies. This allows for preventive actions and ventions, communication, and outcomes in older cancer
personalized treatment plans. Researchers have employed patients. Their review assessed the impact of different assess-
various methods to diagnose chronic diseases, focusing pri- ment types on treatment planning, the implementation of
marily on neurological disorders, mental health disorders, interventions, patient-doctor communication, and overall
and physical disorders [23]. For example, Tawhid et al. [24] treatment outcomes for this demographic. Cai et al. [28] con-
developed a computer-aided diagnosis (CAD) system that ducted a comprehensive literature search through databases

VOLUME 12, 2024 122559

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

such as PubMed, EMBASE, PsycINFO, and Web of Sci- applications. The review aims to expand the research focus of
ence to explore the global prevalence of depression among medical decision support systems, which have traditionally
older adults and its influencing factors. They utilized a centered on single disease prediction, by exploring various
random-effects model to address variances in demographic aspects to encourage future research on predicting multi-
and clinical characteristics across studies, enabling the cal- ple diseases. Armand et al. [34] investigated the integration
culation of the pooled prevalence of depression and its 95% of AI in nutrition as part of the broader automation and
confidence interval. Correia et al. [29] utilized a systematic digitalization efforts of Industry 4.0. They systematically
review and meta-analysis to assess the content and effec- reviewed the application of AI, ML, and DL in nutrition
tiveness of therapeutic patient education (TPE) programs for sciences using a hybrid approach of systematic literature
chronic disorders. They accessed databases like Web of Sci- review (SLR) and PRISMA guidelines. The search covered
ence, MEDLINE, CINAHL, PsycINFO, and COCHRANE major databases, followed by meticulous study selection
up to August 2019, employing a rigorously tested search and methodological quality assessments. This comprehen-
strategy focusing on patient education, chronic diseases, sive review discusses AI’s role in smart nutrition, dietary
study designs, and outcomes. After a detailed screening assessments, food recognition, disease prevention modeling,
process, two reviewers extracted both qualitative and quan- and monitoring, outlining the current capabilities and future
titative data from randomized controlled trials concerning challenges in the field.
TPE interventions. They also created a taxonomy related Table 1 summarizes the review articles on chronic diseases.
to curriculum skills and intervention delivery techniques to To address challenges, including a narrow disease focus,
assist in data extraction. Collado-Mateo et al. [30] performed technological limitations, selective article analysis, and out-
a systematic review following the preferred reporting items dated data, our study undertakes a systematic literature review
for systematic reviews and meta-analyses (PRISMA) guide- of ML and DL applications in geriatric care for managing
lines to investigate adherence to physical exercise among chronic diseases. We thoroughly examined peer-reviewed
chronic and older adults. They reviewed relevant literature articles published from 2014 to 2024, focusing on AI tech-
and selected fifty-five articles to identify critical factors nologies in elderly care. This review includes 76 selected
commonly suggested to improve exercise adherence. The articles from leading publishers, categorized into three main
results were categorized based on the target population and groups: 27 articles on neurological disorders, 22 on mental
participant characteristics to determine the most prevalent health disorders, and 27 on physical/physiological disor-
factors affecting adherence. Kumar et al. [31] performed a ders. After reviewing the selected articles, we conclude that
comprehensive survey on using artificial intelligence tech- while AI applications in geriatric care show promise, they
niques, from machine learning to deep learning, in healthcare. still require significant technological and methodological
This survey included a systematic literature review up to improvements to enhance accuracy and reliability. Future
October 2020 from Web of Science, Scopus, Google Scholar, research should aim to develop advanced AI tools and inte-
PubMed, Excerpta Medical Database, and Psychology Infor- grate cutting-edge deep learning models and comprehensive
mation. It focused on diagnosing diseases like Alzheimer’s, datasets to refine diagnostics and treatment protocols for
cancer, and diabetes using AI techniques. The review fol- chronic diseases in the elderly, ultimately improving patient
lowed Preferred Reporting Items for Systematic Reviews outcomes.
and Meta-Analysis guidelines to select studies that applied
AI for early disease prediction, involving medical imag- III. REVIEW METHOD
ing datasets, feature extraction, and classification processes. To comprehensively select research articles for our study,
Sawad et al. [32] systematically reviewed conversational we meticulously designed a search strategy that utilized
agents in healthcare, particularly for managing chronic con- highly pertinent keywords, including ‘‘chronic disease,’’
ditions. This review, conducted in February 2021, focused on ‘‘elderly patients,’’ ‘‘older adults,’’ ‘‘machine learning,’’ and
the agents’ communication technologies, evaluation metrics, ‘‘deep learning.’’ Our goal was to identify papers that specif-
and AI methods. The researchers used databases such as ically address the intersection of these topics. We conducted
PubMed Medline, EMBASE, PsycINFO, CINAHL, Web of a thorough search across several well-established digital
Science, and ACM Digital Library to gather studies. They databases known for their robust collections of academic
included studies where conversational agents were tested research. These databases comprised Elsevier, Springer,
with human users in prevention, treatment, or rehabilita- IEEE, MDPI, and Wiley, Taylor & Francis, Nature, Cam-
tion contexts targeting consumers, caregivers, or healthcare bridge University Press, Oxford University Press, and arXiv.
professionals. Merabet et al. [33] conducted a systematic We confined our search to works published in English to
literature review using AI-based clinical decision support maintain consistency in our analysis and to ensure that we
systems (CDSS) and the internet of medical things (IoMT) to could accurately assess the quality and relevance of the
predict multiple diseases. This study highlights the improve- content.
ments in diagnostics, therapy, and prognosis offered by AI, Figure 2 in our documentation offers an insightful break-
mainly through ML and DL, enhanced by IoMT technologies down of the studies related to ML and DL that emerged from
that link networked biomedical devices with software our search parameters. This visual representation provides

122560 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 1. Comparison of systematic review studies focusing on AI in the context of healthcare, chronic conditions, and geriatric care.

VOLUME 12, 2024 122561

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 1. (Continued.) Comparison of systematic review studies focusing on AI in the context of healthcare, chronic conditions, and geriatric care.

a clear overview of the distribution and frequency of publi- unrelated, leaving 68 for full-text review. This step resulted
cations across various databases and categorizations, which in 12 studies being included after 56 were rejected. For
is pivotal for understanding the current research landscape Taylor & Francis, we initially reviewed 151 studies, of which
in this field. We refined our search methodology multiple 101 were rejected at the screening stage. The remaining
times to guarantee a comprehensive and inclusive search 50 underwent detailed full-text assessment, which resulted
result. This iterative process was critical to encompass a broad in the exclusion of 45 studies, leaving 5 that met our inclu-
spectrum of potentially relevant scholarly works and filter out sion criteria. Nature provided 125 studies, with 75 discarded
extraneous content that did not meet our specific criteria. during initial screening. Of the 50 that progressed to full-
Table 2 presents a concise summary of the findings text review, 46 were subsequently excluded, culminating in
from our search queries. It enumerates the volume of ini- 4 studies deemed suitable for inclusion. From the Cambridge
tially retrieved documents, offering a snapshot of the wealth University Press database, we initially considered 90 stud-
of literature available on the given subjects. Each of the ies, rejecting 60 at the screening stage. The 30 reviewed
1532 documents identified through this process was subjected in greater depth resulted in 28 rejections, with 2 studies
to a rigorous review. This involved an initial screening based ultimately included. The Oxford University Press presented
on titles and abstracts, followed by a more in-depth evaluation 98 studies for review. Initial screening led to the exclusion
of the full text where necessary. This meticulous review of 58 studies, and of the remaining 40 that underwent full-
aimed to ensure that the final selection of articles was highly text assessment, 38 were rejected, resulting in 2 studies being
pertinent to our research objectives, allowing us to draw from included. Finally, from arXiv, we screened 57 studies, elim-
a pool of high-caliber. These insightful studies contribute inating 37 initially. The 20 studies that moved to full-text
to the understanding of AI applications in geriatric care. review experienced 18 rejections, with 2 studies meeting our
Our review process started with evaluating 397 studies from inclusion criteria.
the Elsevier database, focusing on their titles and abstracts.
This initial step led us to exclude 216 documents for being TABLE 2. Overview of research articles found in the explored databases.
off-topic, while 181 were selected for a more detailed full-
text examination. Of these, 167 did not meet our criteria
upon closer inspection, leaving 14 studies to be included
in our review. From the Springer database, we scanned
148 studies’ titles and abstracts, which resulted in 98 being
discarded as irrelevant and 50 moving forward for full-text
review. After a thorough examination, 34 were excluded,
and 16 were deemed suitable for inclusion. We looked at
215 studies in the IEEE database, eliminating 136 after the
initial screening and taking 79 further for full-text assess-
ment. This process led to 11 studies being included and
68 rejected. The MDPI database presented 104 studies for
initial review. Post-screening, 63 were found irrelevant, and
41 underwent full-text review, resulting in 8 studies being While reviewing titles and abstracts, we first assessed each
included after 33 were rejected. Within the Wiley database, study’s relevance to the predefined themes of our research,
147 studies were initially considered; 79 were dismissed as which focus on the application of ML and DL in geriatric

122562 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

FIGURE 2. Data on ML and DL studies. a) Number of publications by each database, b), Number of publications by year,
c) Proportion of open access vs. close access publications, c) Publication types classified as articles and conference papers.

VOLUME 12, 2024 122563

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

care for managing chronic diseases in older adults. Studies progression severity, making it a critical tool in manag-
that did not align with these themes were immediately con- ing and researching chronic conditions [42], [43].
sidered off-topic. Additionally, we evaluated the clarity and • K-Nearest Neighbor (KNN) [44]: The KNN algorithm
completeness of the abstracts to determine if they provided classifies data points based on proximity, enabling it
enough information to justify a full review. In the review of to dynamically adjust to changes in input data, which
the complete texts, we outlined our criteria for inclusion in is common in continuous patient monitoring [45]. This
more detail. This process involved assessing the robustness flexibility is crucial for tracking disease progression over
of the study’s methodology, the significance of its findings, time or adjusting treatment plans based on new data,
and its overall contribution to the field. We stipulated that, at a offering a direct route to personalized medicine [46].
minimum, a manuscript must demonstrate a sound method- • Logistic Regression (LR) [47]: While noted for its
ological framework and present relevant, clearly articulated binary classification capabilities, logistic regression can
results to be considered for inclusion. This thorough approach also be extended to multiclass classification to handle
ensures that only studies of substantial merit and relevance multiple outcomes, which is often necessary for com-
are included in our review, thus maintaining the integrity and plex chronic diseases manifest in various forms. It can
depth of our research analysis. also incorporate interaction terms to explore how dif-
In all, 76 studies were identified in the explored databases. ferent risk factors interact to impact disease likelihood,
Figure 3 presents the outcomes of the search for research providing a deeper understanding of disease mecha-
articles. nisms [48], [49].
• Decision Tree (DT) [50]: An added advantage of deci-
IV. OVERVIEW OF ML AND DL METHODS sion trees in chronic disease management is their
The potential application of ML techniques in diagnosing interpretability, which is critical for clinical decision-
chronic diseases has garnered significant interest due to their making [51], [52]. Clinicians can follow the paths in
ability to uncover complex patterns in medical data. This the tree to understand the decision process, which aids
has led to a more nuanced understanding of disease pro- in patient communication and educational efforts about
gression and patient outcomes. As ML algorithms become how lifestyle or demographic factors might influence
more sophisticated, their predictive accuracy and diagnos- their health outcomes [53].
tic capabilities improve, offering a promising toolset for • Naive Bayes (NB) [54]: In addition to its efficiency,
healthcare professionals. These advancements could lead to Naive Bayes handles missing data effectively, an essen-
earlier detection of chronic conditions, personalized treat- tial feature given the frequent incompleteness of medical
ment plans, and better management strategies tailored to records [55], [56]. It can still perform well with par-
individual patient profiles, ultimately enhancing patient care tial data, making it particularly useful in real-world
and potentially reducing the burden of chronic diseases on clinical environments where perfect datasets are
healthcare systems. Below is a list of renowned ML algo- rare [57].
rithms extensively used in diagnosing chronic conditions. • Extreme Gradient Boosting (XGBoost) [58]: One of
Their inclusion reflects their widespread adoption and proven XGBoost’s notable features is its ability to handle dif-
success across numerous studies. We discussed these specific ferent types of data issues, such as missing values and
algorithms due to their prevalence and effectiveness in the various types of variable importance. It also allows for
field, as detailed in our literature survey (refer to Table 3). fine-tuning of parameters to avoid overfitting, which
• Support Vector Machine (SVM) [35]: Beyond establish- is particularly beneficial in chronic disease modeling,
ing a hyperplane, SVM is particularly adept at managing where the risk of overfitting can skew critical predictions
non-linear relationships in data using kernel functions. and decision-making [59].
This allows the algorithm to model complex medical
Advancements in hardware capabilities, mainly graphics
phenomena that are not linearly separable [36], [37].
processing units, along with decreasing costs per unit, are key
Its robustness in handling noise and outliers makes it
factors contributing to the surge in the popularity of DL [60],
invaluable in clinical settings where data anomalies are
shared, thus ensuring reliable biomarker detection and [61]. This growth has been further supported by advance-
patient categorization even under varied clinical condi- ments in machine learning and data processing research [62],
tions [38]. [63], [64], [65], coupled with an expansion in the availability
• Random Forest (RF) [39]: This algorithm reduces
of training datasets. DL has been widely adopted across
overfitting through its ensemble method and provides various fields, such as computer vision [9], [61], [63], [66],
essential insights into feature importance. In chronic dis- natural language processing [67], [68], and speech recogni-
ease applications, understanding which variables most tion [69], [70], employing diverse deep learning frameworks
significantly impact disease outcomes can guide clin- like ANN, CNN, and RNN:
ical practice and research [40], [41]. The ability to • ANNs are computational models inspired by the human
perform both classification and regression makes Ran- nervous system, consisting of interconnected layers of
dom Forest versatile for predicting disease presence and neurons, including input, output, and hidden layers.

122564 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

FIGURE 3. Breakdown of the selection process for research articles.

Each layer processes the output of the previous one to insights, they should not be considered fully developed diag-
identify progressively complex patterns in the data [71], nostic or classification systems for age-related disease.
[72]. In chronic illness studies, ANNs can be partic- In reviewing the research presented in Table 3, it is evi-
ularly useful for modeling complex, non-linear rela- dent that e-health and m-Health technologies are integrated
tionships between various health indicators, aiding in into various studies. However, m-Health is featured in a
predicting disease progression and patient outcomes more limited number of studies. For instance, m-Health
[73], [74]. is employed in research on COPD ([79]), where wearable
• CNNs are specialized for processing visual imagery, sensors monitor respiratory rates and activity levels. Sim-
including videos and images. They consist of distinct ilarly, in studies on depression ([89]), mood-tracking apps
layers that perform specific operations, such as convo- are used to collect self-reported mood scores and anxiety
lution, pooling, and activation functions [52], [68], [75]. levels. Research on autism spectrum disorder ([154], [160],
CNNs are highly effective in medical imaging tasks, [161]) also incorporates m-Health technologies, using behav-
such as analyzing X-rays, MRI scans, and CT images, ioral tracking apps to monitor behavioral patterns and social
to detect and diagnose chronic conditions like can- interactions. Additionally, asthma research ([164]) involves
cer, cardiovascular diseases, and neurological disorders. using mobile devices and sensors to track respiratory rates
Their ability to automatically learn spatial hierarchies of and lung function. One reason for the limited use of m-Health
features makes them ideal for detecting subtle anomalies technologies in these studies may be the challenges of ensur-
in medical images [76], [77]. ing data accuracy and reliability. Mobile health technologies
• RNNs are unique in their ability to consider previ- often rely on self-reported data or wearable sensors, which
ous and current inputs, synthesizing information from can be prone to user error, non-compliance, and variability
past and present contexts to generate outputs [78]. This in sensor accuracy. Additionally, integrating m-Health tech-
suits them, particularly for time-series analysis and nologies requires robust data security measures to protect
sequential data, which is common in chronic illness sensitive health information, which can be a significant bar-
monitoring. RNNs can be used to predict disease tra- rier for some research initiatives. Furthermore, implementing
jectories, manage patient health records, and analyze m-Health solutions may necessitate substantial investment in
longitudinal health data to provide insights into disease technology and training, limiting their adoption in resource-
progression and the effectiveness of treatment plans constrained settings.
[173], [174], [175]. On the other hand, e-health technologies are widely uti-
Table 3 presents the ML and DL models for the early lized across many studies, particularly in the form of clinical
identification of chronic illnesses. Most of the research we data and imaging technologies. Studies on Alzheimer’s dis-
reviewed approached chronic diseases in older adults fixedly, ease ([82], [87], [93], [102]), heart disease ([90]), schizophre-
applying existing algorithms to historical data. These models nia ([92], [100]), and multiple sclerosis ([148], [149])
were neither assessed against a set procedural benchmark frequently employ e-Health tools such as electronic health
(such as a clinical gold standard) nor validated with live data records (EHRs) and clinical imaging systems. These e-health
over time. Consequently, while these studies offer valuable tools facilitate collecting, storing, and analyzing vast amounts

VOLUME 12, 2024 122565

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 3. List of papers considered for chronic diseases. Self indicates that the data were gathered directly by the researcher or author of the paper
rather than being sourced from any external database or previous research.

122566 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 3. (Continued.) List of papers considered for chronic diseases. Self indicates that the data were gathered directly by the researcher or author of
the paper rather than being sourced from any external database or previous research.

VOLUME 12, 2024 122567

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 3. (Continued.) List of papers considered for chronic diseases. Self indicates that the data were gathered directly by the researcher or author of
the paper rather than being sourced from any external database or previous research.

122568 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 3. (Continued.) List of papers considered for chronic diseases. Self indicates that the data were gathered directly by the researcher or author of
the paper rather than being sourced from any external database or previous research.

VOLUME 12, 2024 122569

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 3. (Continued.) List of papers considered for chronic diseases. Self indicates that the data were gathered directly by the researcher or author of
the paper rather than being sourced from any external database or previous research.

122570 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 3. (Continued.) List of papers considered for chronic diseases. Self indicates that the data were gathered directly by the researcher or author of
the paper rather than being sourced from any external database or previous research.

of health data, enabling researchers to draw more accurate [160], [161]), allowing continuous monitoring of behavioral
and comprehensive conclusions. EHRs provide a central- patterns and social interactions, which can be challenging to
ized and standardized way to access patient data, which can assess in traditional clinical environments. Moreover, wear-
improve the efficiency of data collection and reduce the able devices and sensors are used as assistive technologies to
likelihood of errors. Additionally, clinical imaging technolo- provide continuous health monitoring and support for patients
gies, such as MRI and CT scans, offer high-resolution and with chronic conditions. For example, in COPD research
detailed visual data crucial for diagnosing and monitoring ([79]), wearable sensors help monitor respiratory rates and
disease progression. The widespread use of e-health tech- physical activity, providing real-time data that can be used to
nologies can be attributed to their established presence in adjust treatment plans dynamically. In asthma studies ([164]),
clinical settings and the robust infrastructure supporting their mobile sensors track respiratory function, offering immediate
use. Many healthcare institutions already have EHR systems, feedback and enabling patients to manage their conditions
making it easier for researchers to access and utilize this data. more effectively.
Moreover, the high standardization and regulation of e-health The software and tools often include advanced machine
technologies ensures data quality and consistency, which is learning frameworks such as TensorFlow or PyTorch, which
essential for rigorous scientific research. Integrating e-health facilitate the complex computations required for deep learn-
tools into clinical workflows also allows for seamless data ing applications. For instance, CNNs used for imaging data
sharing and collaboration among healthcare professionals, often leverage these frameworks to process and analyze large
further enhancing their utility in research. datasets efficiently. Data preprocessing, especially in imaging
Assistive technologies are also a crucial component in sev- studies where normalization and augmentation are critical,
eral studies, particularly in managing and assessing chronic is crucial in preparing data for analysis and significantly
diseases. Cognitive assessment tools are prominently featured impacts the outcomes. Imaging data from sources such as
in Alzheimer’s research ([82], [144]), aiding in the evalua- MRI scans and fundus autofluorescence images undergo
tion of cognitive decline through digital platforms that can preprocessing steps like normalization, resizing, and augmen-
perform complex cognitive testing remotely or in clinical tation to enhance the model’s robustness and performance.
settings. Speech analysis tools are used in studies on cognitive For clinical data, which includes electronic health records
decline ([106]), where they analyze patients’ speech patterns and biochemistry data, preprocessing might involve cleaning
to detect changes that may indicate the progression of neu- the data, handling missing values, and standardizing the data
rological conditions. Additionally, behavioral tracking apps format. These steps ensure high consistency in the data fed
are employed in autism spectrum disorder research ([154], into machine learning models. Clinical data models often

VOLUME 12, 2024 122571

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

employ algorithms such as LR, RF, and SVM, implemented States and Europe. For example, collaborative research
using tools like Scikit-learn, a versatile machine learning between American and European universities has led to
library in Python. Questionnaire and survey data, common advances in predictive analytics for patient management.
in studies addressing mental health and aging, are processed Furthermore, cooperative projects between countries like
to convert categorical data into numerical formats suitable Germany and the UK drive AI algorithm innovations that
for machine learning algorithms. Techniques such as one-hot enhance diagnostic precision and patient outcomes in elderly
encoding and normalization are typically applied here. Mod- care. These international collaborations push the envelope on
els like XGBoost and generalized estimating equations are AI technology development and ensure that these advance-
trained on these preprocessed datasets using cross-validation ments are seamlessly integrated into various healthcare
to ensure they generalize well to new, unseen data. Signal systems, offering scalable and adaptable solutions for a global
data, such as EEG or CO2 waveforms, often requires feature audience.
extraction and noise reduction before model training. This
preprocessing step can involve Fourier or wavelet transforms A. CHARACTERISTIC OF THE STUDIES
to convert signals into a more interpretable format. Tools like Figure 4 categorizes the research approaches used in the
MATLAB or Python libraries such as SciPy are commonly studies examined, focusing on identifying chronic diseases
used. The extracted features are then used to train models using ML and DL techniques. It highlights that RL is the
such as SVMs or neural networks. Gait data and genomic most frequently studied algorithm within the ML category,
data involve unique preprocessing steps as well. Gait data which dominates the research with 65 studies, followed by
from wearable sensors or instrumented treadmills must often LR, SVM, RF, and NB. This indicates a strong preference
be filtered to remove noise and extract meaningful features for RL in the current research landscape. On the DL front,
such as stride length and gait cycle. Genomic data, like which is less represented with 11 studies, CNN is the leading
nucleotide sequences, may involve alignment and annotation method, with RNN, LSTM, and MLP being considerably less
processes using bioinformatics tools before being fed into explored. While various techniques are being applied to the
machine learning models. Social media data, which includes challenge of detecting chronic diseases, there is a concentra-
posts and related metadata, typically undergoes text prepro- tion of research on specific algorithms, particularly RL and
cessing steps such as tokenization, stemming, and removal CNN. This could suggest that these methods are currently
of stop words. Natural language processing (NLP) libraries seen as more promising or suitable for the complexities of
like NLTK or SpaCy prepare the data for models such as chronic disease identification. However, the lower represen-
NB or SVM. The training and validation processes employed tation of DL methods might also indicate a potential area
rigorous methodologies, utilizing cross-validation and split- for further exploration, as advancements in DL could offer
ting data into distinct training and testing sets to ensure that new insights and improvements in the detection and diagnosis
models are accurate and generalizable. Cross-validation helps of chronic diseases. It also underscores a possible research
to tune hyperparameters and avoid overfitting, ensuring the opportunity to expand the application of less-used methods
models perform well on new data. The metrics used for that might address the problem from different angles or offer
evaluating model performance—predominantly accuracy and improvements over more commonly used techniques.
AUC—are chosen based on the specific requirements of the The distribution of research studies by health condition
health condition being addressed, ensuring that the models in Table 4 underscores significant disparities in research
provide practical and clinically relevant results. For example, focus and funding, influenced by disease prevalence, societal
high AUC values are crucial in diagnostic applications where impact, and the potential for medical advancement. Neuro-
the trade-off between sensitivity and specificity impacts logical disorders, including Alzheimer’s, Parkinson’s, and
clinical decisions. multiple sclerosis, garner a substantial portion of research
Based on Table 3, international contributions to AI in with 27 publications, indicative of the high level of scientific
healthcare show significant diversity. The United States leads and medical community interest. This is mainly due to the
with 38 articles, underscoring its role in pioneering AI-driven severe impact these diseases have on quality of life and their
healthcare solutions. Germany and the United Kingdom fol- generally irreversible progression. The complexity of neuro-
low, with seven articles reflecting their deep involvement in logical disorders, coupled with their increasing prevalence
developing technologies that could enhance diagnostic accu- in an aging population, necessitates ongoing research into
racy and treatment reliability. Notably, contributions from advanced diagnostics and therapeutic strategies. This intense
China and South Korea, with 5 and 4 articles, respectively, focus also mirrors the urgent need for breakthroughs in treat-
indicate a growing interest and capability in AI across Asia. ment options and management practices that can significantly
Japan, Taiwan, Sweden, Malaysia, and Italy, each with three alter disease outcomes and patient quality of life. In contrast,
articles, demonstrate a wide range of innovative efforts that mental health disorders accounted for by 22 publications
span from East Asia to Europe. also reveal a strong research interest driven by the grow-
Regarding collaboration, the most effective partnerships ing recognition of mental health’s critical impact on overall
in AI-driven healthcare occur among academic institutions, health and societal well- being. Disorders such as depression,
healthcare providers, and private enterprises in the United autism spectrum disorders, and schizophrenia have seen a

122572 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

FIGURE 4. The most common methods utilized for identifying chronic diseases in the research surveyed. The numbers enclosed in
brackets reflect the number of research papers.

significant rise in global prevalence, prompting a surge in a direct insight into the patient’s health status, treatment
research activities aimed at understanding their complex history, and biological markers, all essential for personal-
etiologies and developing more effective, personalized treat- ized treatment plans and longitudinal studies. Conversely,
ment plans. The research in this area reflects a broader the relatively lower utilization of sensor data (1.3%) and
shift in healthcare priorities towards mental well-being, gait data (1.3%) can be attributed to the niche applications
acknowledging mental health’s profound role in achieving and the specific types of chronic diseases they are relevant
overall health security and social stability. Physical and for, such as multiple sclerosis, where movement metrics
physiological disorders, a significant research focus with are critical. Additionally, the complexity and the required
27 publications, cover a broad spectrum of conditions primar- infrastructure to collect and analyze such data accurately
ily affecting the aging population and those with substantial might limit their broader application in the field. Social
lifestyle impacts. Conditions such as coronary artery disease, media and genomic data, accounting for 2.6% and 3.9%,
COPD, and osteoarthritis necessitate extensive research due respectively, represent emerging fields in chronic disease
to their widespread prevalence and the considerable health- research. Social media offers a unique perspective through
care resources they command. The focus on these diseases patient-reported experiences and outcomes, but its subjec-
is strategic, aimed at enhancing life expectancy and quality tive nature and potential privacy concerns might hinder its
of life through improved treatment protocols and preventive broader acceptance. Genomic data, though increasingly cru-
measures, highlighting the significant societal and economic cial for understanding genetic predispositions and tailoring
impacts these conditions hold. Overall, allocating research treatments, requires high-tech equipment and specialized
across these health conditions illustrates a strategic approach knowledge, contributing to its lesser usage than more tradi-
to addressing the most pressing health challenges. The tional data types.
emphasis on diseases with high societal burdens and com- Table 5 outlines the computational demands of various arti-
plex management needs reflects a targeted effort to optimize ficial intelligence models utilized in chronic disease research.
healthcare outcomes through focused and sustained research It categorizes them based on their space complexity and
investments. This distribution informs about current scien- the computational expense involved during the training and
tific endeavors in chronic disease management and indicates prediction phases. In this table, the variable n represents
potential shifts in health priorities and resource allocation the number of samples in the dataset, which is a common
essential for future healthcare strategies. factor in complexity calculations, affecting how algorithms
Figure 5 showcases the various data types used in chronic iterate over data points. The variable d denotes the number of
disease research, illustrating the preference for specific features or dimensions in the dataset, influencing operations
methodologies based on their practical application, reliability, across data features. For SVM, v indicates the number of
and the depth of insights they provide. Imaging data tops support vectors crucial for defining the model’s decision
the chart at 35.5%, highlighting its indispensable role in boundaries, and s refers to the size of each support vector
the field due to its non-invasive nature and ability to offer in memory. In ensemble methods like Random Forest and
precise, real-time visual insights into the body’s internal state. XGBoost, t signifies the number of trees, which directly
Such technologies as MRI and CT scans are crucial for impacts both training and prediction complexity due to the
diagnosing and monitoring disease progression and assessing ensemble nature of these models. The variable m varies in
treatment efficacy, making them a cornerstone in chronic meaning; in SVMs, it can reflect the output space’s dimen-
disease research. Clinical data, which includes detailed med- sionality, while in CNNs, it often pertains to the number of
ical records and laboratory test results, represents 26.3% of feature maps. The number of neurons or layers in neural
the usage. This data type is fundamental because it provides networks is captured by q, and h details the number of units

VOLUME 12, 2024 122573

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 4. Distribution of research studies by health condition category.

FIGURE 5. Data types used in chronic disease research.

in a hidden layer, which is relevant to both ANNs and RNNs. processes in each input sequence. Understanding these vari-
For CNNs, k represents the number of filters, which is crucial ables is essential for assessing an algorithm’s feasibility and
for understanding the processing depth of these networks. efficiency concerning data size, features, and computational
The variable f may relate to the storage requirements for capacity.
feature maps in CNNs, whereas p in RNNs indicates the The information in Table 5 reveals that SVM demands
sequence length, reflecting how many time steps the network considerable computational resources. The training time

122574 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

TABLE 5. Computational complexities of AI models in chronic disease research.

complexity can rise exponentially with data size, making classification of the presence or absence of disease. Recall,
SVMs less ideal for large-scale datasets commonly found in employed in 14 publications, is acknowledged for its role in
chronic disease studies. This is compounded by their space minimizing false negatives, which is paramount in a clini-
complexity, which is influenced by the number of support cal setting where missing a positive case can have serious
vectors and the dimensionality of the data. Despite these consequences. Precision, cited in 12 studies, is valued for
challenges, SVMs remain valuable for smaller, well-defined its measure of result relevancy, ensuring that the identified
datasets where the high dimensionality needs to be effec- cases are indeed correct and reducing the rate of false pos-
tively managed. On the other hand, ensemble methods like itives. With 68 publications using it, the AUC is the most
RF and XGBoost demonstrate their robustness in handling frequently applied metric. The AUC’s significance lies in
large and complex datasets, a typical scenario in chronic its comprehensive evaluation of a model’s performance at
disease research. These models, characterized by their deep various thresholds, providing a singular measure of effective-
integration of multiple decision trees, show a significant ness that encompasses both the true positive and negative
computational cost that scales with the number and depth rates. The F-measure, employed in 18 publications, offers
of each tree. However, their ability to improve predictive a balanced view by combining precision and recall into a
accuracy and control over-fitting makes them indispensable, single metric. This is especially useful in scenarios where
especially in scenarios where predictive reliability is critical. both false positives and false negatives carry significant con-
KNN and LR present lower prediction complexities, indi- sequences, ensuring a harmonized assessment of the model’s
cating faster operational capabilities during the deployment accuracy in predicting true positive and actual negative cases.
phase. However, KNN’s significant space complexity, due to Its utilization in a smaller subset of studies may indicate a
its need to store the entire dataset, limits its practicality in more targeted approach where both errors are equally critical
large-scale applications. While more scalable in its training to the research outcomes. These metrics together form a
phase, Logistic Regression offers a balance between com- multifaceted evaluation framework essential for developing
putational efficiency and predictive performance, making it and validating AI tools in diagnosing and managing chronic
a go-to model for baseline assessments in many studies. diseases. The balance of these metrics ensures that AI mod-
Furthermore, advanced deep learning models such as ANN, els are accurate, clinically reliable, and relevant, leading to
CNN, and RNN exhibit considerable complexities both in improvements in patient care and health outcomes.
space and time. These models require extensive computa-
tional resources due to their intricate architectures involving
multiple layers and numerous parameters. However, their B. FEATURES
ability to handle unstructured and complex data types, like The role of features in the diagnosis of chronic diseases is of
medical imaging and sequential patient data, makes them paramount importance, particularly within the disciplines of
particularly useful in extracting nuanced patterns and deliv- ML and DL. In the case of chronic diseases, features could
ering high-accuracy predictions crucial for diagnosing and include a wide array of data points, from basic demographic
managing chronic diseases. information like age and gender to complex biological mark-
The selection of evaluation criteria is a critical step in ers such as gene expressions or protein levels. These features
chronic disease research, as it directly impacts the inter- are vital because they are the fundamental inputs that feed
pretation and applicability of AI model outcomes. Figure 6 into predictive models. The quality and relevance of these
describes how frequently various metrics are utilized to assess features directly impact the model’s ability to learn from data.
AI performance in this field. The dominance of accuracy, For example, in a dataset concerning heart disease, relevant
used in 63 studies, highlights its status as a fundamen- features might include blood pressure readings, cholesterol
tal metric, suggesting that researchers prioritize the correct levels, smoking status, and exercise frequency. Each feature

VOLUME 12, 2024 122575

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

FIGURE 6. Metrics used in chronic disease research.

captures a different aspect of a patient’s health profile that genetic markers can signal the propensity for conditions
could indicate their risk for heart disease. like hypertrophic cardiomyopathy [182]. The analysis
In ML and DL models, the feature extraction process of genetic markers is a complex field in which machine
involves selecting data pieces that are most likely to reveal learning can significantly benefit, as it can handle large
patterns related to the onset and progression of chronic dis- datasets and uncover patterns that may not be readily
eases. The significance of this process lies in its ability to visible to human researchers [63].
transform raw data into a format that predictive models can • Imaging data: Imaging data from MRI, CT scans, or X-
use to discern complex relationships within the data that rays provide a non-invasive window into the body’s
may not be immediately apparent to human observers [176]. internal structures [9], [183]. In neurological disorders,
Once the features are identified and extracted, ML algorithms imaging can reveal the size and volume of brain regions,
can be trained to detect patterns and anomalies that signify which may shrink or show lesions in diseases like Multi-
the presence of a disease. DL models, which are particu- ple Sclerosis (MS) or Alzheimer’s. MRI scans can show
larly adept at handling vast amounts of unstructured data, the plaques and tangles associated with Alzheimer’s,
can identify even subtler patterns within complex features, while MS is characterized by brain and spinal cord
such as those derived from medical imaging. The predictive lesions, which can be detected via these imaging tech-
power of these models hinges on the careful selection and niques. Machine learning models trained on this data can
processing of features. High-quality features allow models help radiologists and neurologists identify these signs
to achieve greater accuracy, leading to earlier and more reli- earlier than ever, which is crucial because earlier diag-
able detection of chronic diseases. This can inform treatment nosis can lead to more effective management of these
decisions and lead to better patient outcomes. For instance, conditions. They can also monitor changes, providing
early detection of a chronic condition like diabetes can lead insight into disease progression and treatment effective-
to interventions that prevent complications, reduce the burden ness [61].
on healthcare systems, and ultimately save lives [177]. • Speech patterns: Speech pattern analysis is a rich domain
Below, we present the most critical features identified by within machine learning applications for diagnosing
the studies for the diagnosis of chronic diseases. neurological and psychiatric conditions [184]. Subtle
• Genetic markers: Genetic markers are specific sequences changes in speech, such as pauses, intonation, and word
in the genome indicating an increased likelihood of selection, can be early indicators of Alzheimer’s disease
developing certain diseases [178]. They are critical as cognitive decline begins to affect language centers
for conditions with a hereditary component, such as in the brain. In depression and schizophrenia, speech
Alzheimer’s and Parkinson’s disease, as well as many can become monotone or disjointed, reflecting the
heart conditions. Identifying these markers can lead underlying psychological disturbances. Machine learn-
to preemptive measures and personalized treatment ing algorithms excel at detecting these subtle changes by
plans [179], [180]. For instance, in Alzheimer’s, genetic analyzing large datasets of speech recordings, which can
markers like the APOE ε4 allele have been associated be particularly helpful for tracking disease progression
with a higher risk, and their detection can warrant closer or response to treatment [185].
monitoring and early intervention strategies. In Parkin- • Movement data: Regarding diseases like Parkinson’s
son’s, mutations in genes such as LRRK2 and PARK7 and other motor disorders, movement data is critical to
are insightful for risk assessment [181]. In cardiology, diagnosis and monitoring [186], [187]. Gait patterns, for

122576 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

example, can reveal much about the severity and type of adjusting medications, and developing neurofeedback
a motor disorder. Patients with Parkinson’s disease often therapies to help patients manage their conditions more
have a shuffling gait with reduced arm swing and some- effectively [197].
times freezing of gait, where they temporarily are unable • Cognitive test results: Cognitive test results are a rich
to move. Tremor analysis is another essential feature, data source for machine learning models that identify
as resting tremors are a hallmark of Parkinson’s disease, and track the progression of cognitive impairments and
while intention tremors are more associated with cere- mental health issues [198], [199]. These tests often
bellar diseases [188], [189]. Advanced machine learning assess various aspects of cognition, including memory,
algorithms can analyze these movement patterns, often attention, problem-solving skills, and language abilities.
captured by motion sensors or video recordings, to dis- For conditions like Alzheimer’s disease or mild cog-
cern subtle differences in motor function that may not nitive impairment, declining scores in memory-related
be noticeable to the naked eye. This data can be used tasks can indicate the progression of the disease. In men-
to track the disease’s progression over time, evaluate the tal health, cognitive tests can help diagnose conditions
efficacy of treatments, and adjust medications or ther- such as depression or schizophrenia, where patients may
apy as needed. Moreover, movement data can provide show difficulties in concentration, memory, or execu-
insight into the risk of falls and help develop preventive tive function. Machine learning can analyze longitudinal
strategies, which are crucial for maintaining the quality test data to predict the trajectory of a disease, monitor
of life in patients with motor impairments [190]. changes over time, and evaluate the effectiveness of
• Blood test results: Blood tests provide a wealth of interventions. The nuanced analysis can also help dif-
information through various biomarkers that reflect the ferentiate between similar cognitive conditions, aiding
functioning of different body systems [191], [192]. in more accurate diagnoses and personalized treatment
In the case of inflammatory diseases like rheumatoid strategies [200].
arthritis, markers such as C-reactive protein (CRP) and • Ocular measurements: Ocular measurements, like
erythrocyte sedimentation rate (ESR) can indicate the intraocular pressure (IOP) and optic nerve health,
presence and intensity of inflammation [193], [194]. provide quantifiable data crucial for diagnosing and
For Hepatitis C virus infection, liver enzymes such as managing eye diseases [201]. In glaucoma, elevated
AST, ALT, and viral load measurements are critical. IOP can lead to optic nerve damage and vision loss;
Anemia, elevated inflammatory markers, and nutrient therefore, monitoring IOP is essential for early detection
deficiencies may indicate inflammatory bowel diseases and prevention of the disease’s progression. Similarly,
like Crohn’s disease and ulcerative colitis. Blood mark- in diabetic retinopathy, changes in retinal blood ves-
ers can also help monitor the effectiveness of treatments, sels can be quantified and monitored [202]. Machine
such as reducing inflammation in response to medica- learning algorithms can analyze these ocular parameters
tion. Machine learning models can use these markers to identify patterns that may indicate disease before
to identify patterns that may predict disease flares or significant damage occurs. They can also track the
remissions. Additionally, by analyzing changes in blood effectiveness of treatments like pressure-lowering med-
markers over time, machine learning can help predict ications in glaucoma or anti-VEGF injections [203] in
the course of a disease and tailor personalized treatment diabetic retinopathy. Furthermore, advanced imaging
plans, which can be vital for chronic conditions that techniques like optical coherence tomography provide
require long-term management. detailed images of the retina, enabling the detection of
• Electrophysiological signals: Electrophysiological sig- subtle changes that may not be visible during a standard
nals, especially those derived from electroencephalogra- eye exam [204].
phy (EEG), are critical in diagnosing and understanding • Histopathological data: Histopathological data gathered
neurological conditions like epilepsy, characterized by from tissue samples under a microscope includes cel-
abnormal brain activity that can lead to seizures and lular morphology, tissue architecture, and the presence
other severe symptoms [195]. EEG signals capture the of specific markers [205], [206]. This data is pivotal
brain’s electrical activity and reveal telltale patterns in diagnosing conditions such as cirrhosis, where liver
associated with seizure activity. Additionally, EEG fea- tissue undergoes fibrosis and loses its typical architec-
tures are studied in autism spectrum disorders (ASD), ture, or cancer, where abnormal cell growth patterns are
where they may indicate neural connectivity differences, evident. Machine learning models trained on histopatho-
and in depression, where they can reflect altered brain logical images can assist pathologists in detecting
function [196]. By analyzing these signals, machine disease characteristics that might be too subtle or com-
learning algorithms can learn to recognize the unique plex for the human eye to discern consistently [207].
electrical patterns associated with these conditions, aid- They can also help quantify the extent of disease,
ing in early detection and intervention. Moreover, EEG which is critical for staging cancer or assessing liver
data can be valuable in monitoring treatment response, damage. Furthermore, these algorithms can uncover

VOLUME 12, 2024 122577

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

correlations between histopathological features and automated chronic disease detection until 2019. Post-2019,
patient outcomes, aiding in prognosis and the tailoring DL has surged in popularity, recognized for its ability to man-
of therapy to individual patient profiles [208]. age large datasets and achieve enhanced outcomes. DL differs
• Patient history: Patient history encompasses a wealth of from ML’s minimal need for manual feature engineering and
information vital for personalized healthcare [209]. This selection, offering a more automated process. This automa-
includes past medical diagnoses, treatment regimens, tion streamlines the data processing and supports the flexible
medication adherence, and lifestyle choices such as diet, use of various algorithms, making DL a powerful and increas-
exercise, and smoking status. This historical context can ingly preferred tool in chronic disease diagnosis.
help machine learning systems predict disease progres- The reliability and success of AI applications in diagnosing
sion and potential complications for chronic diseases. chronic diseases are deeply dependent on the integrity of
Patient history can inform risk stratification and preven- the underlying data. AI systems require a robust dataset—
tive heart disease or asthma strategies [210]. For mental complete, consistent, and accurate—to make precise predic-
health conditions, historical data on symptom onset, tions and diagnoses. The data collection process, which often
duration, and response to previous treatments can guide falls under the purview of nursing staff, is a critical step;
therapeutic decisions. Machine learning can analyze any discrepancies, omissions, or inaccuracies at this stage
these complex, multidimensional data to identify pat- can significantly compromise the effectiveness of AI diag-
terns that inform risk prediction models, enhance under- nostic tools. Nurses play a vital role in ensuring the quality
standing of disease etiologies, and optimize treatment of patient data. They are on the front lines, providing care
approaches tailored to individual patient histories [211]. and meticulously documenting various patient metrics and
Figure 7 shows the distribution of critical features consid- clinical findings. Their contributions to the medical records
ered crucial for diagnosing chronic diseases as identified by form the foundation upon which AI algorithms are trained
various studies. The largest segment, at 28.6%, is devoted and tested. Therefore, any variation in data quality, whether
to imaging data, emphasizing its pivotal role in diagnos- due to human error or systemic issues, can lead to AI systems
ing conditions through MRI, CT scans, and X-rays. Genetic making unreliable predictions. Moreover, the volume of data
markers constitute 14.3%, reflecting the growing recognition is just as important as its quality. AI models, especially those
of genetics in the development and progression of chronic dis- employing deep learning techniques, require large datasets to
eases. Blood test results, capturing 11.4% of the pie, are also learn the complex patterns associated with chronic diseases.
significant, likely due to their role in revealing physiological Inadequate data can hinder the model’s generalization of
abnormalities. Movement data and cognitive test results each its findings to broader patient populations. Training AI sys-
account for 10% and 8.6%, respectively, highlighting their tems with comprehensive and accurately annotated datasets
importance in diagnosing diseases that affect motor skills and allows for more sophisticated and nuanced understandings
cognitive functions. Electrophysiological signals, which can of chronic conditions. This, in turn, can lead to earlier and
indicate various conditions, from heart rhythm disorders to more accurate identifications of such diseases, potentially
neurological diseases, make up 5.7% of the chart. Speech improving patient outcomes. Consequently, the healthcare
patterns are considered in 4.3% of the cases, which could industry must emphasize the critical role of data collection
be critical in conditions like Parkinson’s disease or after a and encourage rigorous training and protocols to minimize
stroke. Patient history and ocular measurements each have errors in data entry and maintenance. This will ensure that
7.1%, signifying the value of comprehensive medical histo- AI-assisted diagnostic tools can reach their full potential in
ries and eye-related metrics in chronic disease management. the fight against chronic diseases.
Histopathological data, the analysis of tissue changes caused The AI models scrutinized in our review reveal specific
by disease, accounts for 2.9%, showing its specific but vital data-related constraints. The complexity lies in the intricate
role in diagnosing certain conditions. interplay between the available data and the data processing
methods used to extract relevant insights. A notable data
V. DISCUSSION challenge is the prevalence imbalance between classes, where
This study highlights the application of cutting-edge technol- data for non-disordered instances outweighs that for disor-
ogy in the automated diagnosis of chronic diseases. Through dered cases. For AI models to function without bias, it is
a thorough analysis of 76 scholarly articles, we have explored crucial to establish a balanced dataset. Data augmentation
various automated methods for chronic disease detection strategies are frequently employed to rectify this discrepancy,
spanning the last decade, from 2014 to 2024. Our investiga- primarily to bolster the representation of the less prevalent,
tion has tracked the evolution and development of automated disordered class. This step is vital to counter the issues that
detection systems tailored to chronic illnesses, focusing on arise from such imbalances, ensuring that the AI models can
ML and DL strategies as documented in references. Our learn to identify disorders with equal precision across varied
review found that ML was the primary method used in instances.
65 studies for the automatic detection of insomnia, with only Automated methods for chronic disease detection and
11 studies utilizing DL models, as illustrated in Figure 4. intense learning models are often described as black boxes
The trend indicates that ML was the dominant approach in because their decision-making processes are not easily

122578 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

FIGURE 7. Feature types used in chronic disease research.

interpretable by humans. Nurses may need help trusting tools can lead to more user-friendly solutions that are better
or acting on AI recommendations if they understand their aligned with clinical workflows and the practical needs of
rationale. This lack of transparency can hinder the effective care providers. Collaborative efforts between AI developers,
integration of AI tools into clinical decision-making pro- healthcare professionals, and patients are essential to create
cesses. The inability to interpret AI decisions affects trust AI systems that enhance rather than complicate healthcare
and limits the nurse’s ability to explain treatment choices to delivery.
patients, potentially impacting patient satisfaction and adher- Chronic diseases, characterized by their long-lasting and
ence to treatment plans. Moreover, the black-box nature of AI persistent nature, frequently exhibit a wide range of symp-
models raises significant challenges regarding accountability toms that can vary significantly among individuals, particu-
and error tracing. In cases where an AI-driven diagnosis leads larly in the context of aging. These conditions, such as heart
to an adverse outcome, it can be challenging to analyze the disease, diabetes, arthritis, and Alzheimer’s disease, often
decision-making process and identify where errors occurred. involve multiple factors, including genetic predisposition,
This ambiguity complicates efforts to improve AI systems lifestyle choices, and environmental exposures, making their
and can lead to legal and ethical dilemmas about responsibil- diagnosis and management a challenging task. The progres-
ity for medical decisions. To address these issues, there is a sion and manifestation of symptoms in chronic diseases are
growing call for developing more interpretable AI models and not linear and can be influenced by the intricate interplay
incorporating explainable AI (XAI) principles into health- between these factors. For AI models to effectively capture
care applications. XAI aims to make AI decision-making the complexity of chronic diseases, especially in elderly pop-
processes more transparent and understandable to human ulations, they must be designed to account for the nuanced
users, thereby increasing trust and facilitating more informed ways these diseases present and progress. Physiological
decision-making by healthcare professionals. In addition to changes that come with aging, such as decreased organ func-
improving AI interpretability, enhancing nurse education tion, altered metabolism, and reduced physical resilience, can
and training regarding AI technologies is crucial. Providing significantly affect the manifestation of diseases. For exam-
nurses with a better understanding of how AI tools function, ple, an elderly individual might show atypical symptoms
their potential benefits, and limitations can empower them of a heart attack, such as fatigue and shortness of breath,
to use these technologies more effectively and confidently. rather than the classic chest pain. Similarly, the response to
Training should focus not only on the technical aspects of treatment in older adults can differ from younger individuals
AI but also on ethical considerations, data privacy, and the due to changes in drug metabolism and increased suscepti-
importance of integrating AI assistance with human judgment bility to side effects. Incorporating these complexities into AI
and patient-centered care. Furthermore, involving nurses and models requires a multifaceted approach that includes diverse
other healthcare professionals in designing and developing AI datasets encompassing various ages, ethnicities, and comor-

VOLUME 12, 2024 122579

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

bidities. Advanced algorithms capable of processing and • Ethical and Privacy Considerations in AI: As AI
learning from these heterogeneous data sources are essential. technologies become more integrated into healthcare,
Moreover, AI systems need to be dynamic and capable of addressing ethical concerns and ensuring patient pri-
adapting to new information and evolving patterns of disease vacy will be crucial. Future studies should explore
manifestation. Interdisciplinary collaboration among clini- frameworks and guidelines for ethical AI use in clin-
cians, gerontologists, data scientists, and AI experts is crucial ical settings, ensuring that these technologies benefit
to ensure that the models are clinically relevant and capable patients without compromising their rights or autonomy.
of handling the intricacies of chronic diseases in the elderly. • Cross-disciplinary Collaborations: Encouraging collab-
Furthermore, ethical considerations and biases in AI devel- orations between computer scientists, healthcare profes-
opment must be addressed to ensure equitable and effective sionals, and ethicists could lead to more effective and
healthcare outcomes. Elderly populations are often underrep- ethically sound AI solutions. These collaborations can
resented in clinical trials and datasets, leading to potential ensure that AI technologies are developed with a deep
biases in AI models. Ensuring the inclusion of robust and understanding of clinical needs and ethical standards.
representative data from older adults in developing AI sys- • Longitudinal Studies on AI Implementation: Conduct-
tems is essential for creating models that accurately capture ing longitudinal studies to assess the long-term effects
the complexity of chronic diseases in this demographic. of AI integration in geriatric care can provide insights
Future works in chronic disease detection in geriatric care, into its efficacy, patient outcomes, and potential chal-
leveraging AI technologies, could consider several avenues lenges. These studies could help refine AI applications
to address the identified gaps and enhance the precision for chronic disease management in elderly populations.
and reliability of current methods. These potential directions
include: VI. CONCLUSION
• Developing Interpretable AI Models: Future research This systematic literature review emphasizes the transforma-
could focus on creating more interpretable AI models tive impact of AI on geriatric clinical care. Adopting ML
that allow healthcare professionals, particularly nurses, and DL technologies, including Random Forest, Logistic
to understand the rationale behind the AI’s decisions. Regression, and CNNs, has significantly improved real-time
This transparency could improve trust in AI-assisted health monitoring and early disease detection. These AI
diagnostics and enable nurses to make more informed tools play a vital role in addressing the complex health
decisions regarding patient care. requirements of older adults, who typically contend with
• Expanding Dataset Diversity: To improve the robustness various chronic conditions that demand comprehensive care
of AI models, future studies should aim to incorporate strategies. AI’s integration into geriatric care enhances the
a broader range of data sources, including electronic precision and efficiency of diagnosing and managing neu-
health records, imaging data, and wearable device data. rological, mental health, and physical disorders, which
Diverse datasets can help train more comprehensive are common among elderly populations. Advanced algo-
models capable of detecting a wider array of chronic rithms that process extensive and intricate datasets boost
conditions in elderly patients. decision-making effectiveness, leading to prompt and precise
• Integrating Multimodal Data Analysis: Combining vari- interventions. Furthermore, AI supports ongoing monitoring,
ous data types (e.g., clinical notes, lab results, patient- enabling the early identification of potential health declines
reported outcomes) through multimodal AI models to prevent severe complications and enhance overall patient
could provide a more holistic view of the patient’s health outcomes. This capability is precious in environments like
status. This approach can enhance the early detection of nursing homes or home care, where continuous professional
chronic diseases by capturing subtle signs that might be supervision is challenging. Although the benefits of AI in
overlooked when analyzing data sources in isolation. geriatric care are clear, continued research and technological
• Personalized AI Models: Future works could explore the improvement are essential to maximize its effectiveness and
development of personalized AI models that take into ensure dependable integration into daily clinical settings.
account the unique characteristics of each patient, such To enhance the credibility of our findings and guide future
as genetic information, lifestyle factors, and comorbidi- research, it is essential to address the limitations identified
ties. Personalized models could lead to more accurate in our review. A primary concern is the generalizability of
predictions and tailored interventions, improving patient our results, drawn mainly from high-impact journals. This
outcomes. selection may overlook pioneering studies from lesser-known
• Real-time Monitoring and Prediction Systems: Imple- or nascent platforms. Future research should include a wider
menting AI systems capable of real-time monitoring and variety of sources, thereby enriching the diversity and inclu-
predictive analytics could enable more timely interven- sivity of the research examined. Additionally, while our
tions. Future research could focus on creating systems study categorizes articles into three major disorder groups,
that detect existing conditions and predict the risk of it needs to adequately investigate the interactions among
developing new chronic diseases or exacerbations of these disorders or assess the impact of multimorbidity on the
existing ones. effectiveness of AI technologies. A more detailed exploration

122580 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

of these dynamics would deepen our understanding of how [5] S. Mancin, M. Sguanci, D. Cattani, F. Soekeland, G. Axiak, B. Mazzoleni,
AI can manage complex health scenarios in older adults. M. G. De Marinis, and M. Piredda, ‘‘Nutritional knowledge of nursing
students: A systematic literature review,’’ Nurse Educ. Today, vol. 126,
Moreover, integrating AI with fields like bioinformatics, Jul. 2023, Art. no. 105826.
pharmacology, and gerontology could spawn more innova- [6] H. Karabuga Yakar, S. Oguz, B. Bulut, C. Kapusuz, P. Abi, and R. Yavuz,
tive solutions specifically designed for the aging population. ‘‘Compassion fatigue in nurses caring for chronic diseases,’’ Int. J. Occu-
pational Saf. Ergonom., vol. 29, no. 1, pp. 109–114, Jan. 2023.
Addressing ethical and privacy concerns in the deployment
[7] E. Akleman, ‘‘Deep learning,’’ Computer, vol. 53, no. 9, pp. 17–17,
of AI in healthcare is also critical. Developing robust frame- Sep. 2020.
works to safeguard patient data while maximizing the benefits [8] N. Mavaddat, K. Michailidou, J. Dennis, M. Lush, L. Fachal, and A. Lee,
of AI is imperative, especially in sensitive areas like geri- ‘‘Polygenic risk scores for prediction of breast cancer and breast cancer
subtypes,’’ Amer. J. Human Genet., vol. 104, no. 1, pp. 21–34, 2019.
atric care. Taking these steps will allow future studies to [9] L. Hong, M. H. Modirrousta, M. H. Nasirpour, M. M. Chargari,
significantly enhance our comprehension of AI’s role in F. Mohammadi, S. V. Moravvej, L. Rezvanishad, M. Rezvanishad,
geriatric care, ensuring that technological advancements lead I. Bakhshayeshi, R. Alizadehsani, I. Razzak, H. Alinejad-Rokny,
and S. Nahavandi, ‘‘GAN-LSTM-3D: An efficient method for lung
to practical improvements in the management of elderly tumour 3D reconstruction enhanced by attention-based LSTM,’’ CAAI
care. Trans. Intell. Technol., vol. 1, pp. 1–24, May 2023.
[10] A. Saeed, A. AlShafea, A. Bin Saeed, M. Nasser, and R. Ali, ‘‘Robotics
and artificial intelligence and their impact on the diagnosis and treatment
ABBREVIATIONS of cardiovascular diseases,’’ Cureus, vol. 15, no. 7, pp. 1–26, Jul. 2023.
The following abbreviations are used in this study: [11] O. Arigbede, T. Amusa, and S. G. Buxbaum, ‘‘Exploring the use of arti-
ficial intelligence and robotics in prostate cancer management,’’ Cureus,
NN Neural network. vol. 15, no. 9, pp. 1–32, Sep. 2023.
SVM Support vector machine. [12] C. Jailin, S. Mohamed, R. Iordache, P. Milioni De Carvalho, S. Y. Ahmed,
ANN Artificial neural network. E. A. Abdel Sattar, A. F. I. Moustafa, M. M. Gomaa, R. M. Kamal, and
L. Vancamberg, ‘‘AI-based cancer detection model for contrast-enhanced
AUC Area Under the Curve. mammography,’’ Bioengineering, vol. 10, no. 8, p. 974, Aug. 2023.
MRI Magnetic Resonance Imaging. [13] H. E. C. D. Silva, G. N. M. Santos, A. F. Leite, C. R. M. Mesquita,
COPD Chronic Obstructive Pulmonary Disease. P. T. D. S. Figueiredo, C. M. Stefani, and N. S. de Melo, ‘‘The use of
LSN Siamese neural-network. artificial intelligence tools in cancer detection compared to the traditional
diagnostic imaging methods: An overview of the systematic reviews,’’
XGBoost Extreme gradient boosting. PLoS One, vol. 18, no. 10, Oct. 2023, Art. no. e0292063.
DBN Deep belief network. [14] S. K. Bhatt and S. Srinivasan, ‘‘Lung cancer detection using AI and
KNN K-nearest neighbors. different techniques of machine learning,’’ Int. J. Intell. Syst. Appl. Eng.,
vol. 12, no. 8s, pp. 630–638, 2024.
MLP Multilayer neural networks. [15] H. Chen, E. Cohen, D. Wilson, and M. Alfred, ‘‘A machine learn-
ML Machine learning. ing approach with human-AI collaboration for automated classification
DNN Deep neural networks. of patient safety event reports: Algorithm development and validation
study,’’ JMIR Hum. Factors, vol. 11, Jan. 2024, Art. no. e53378.
CNN Convolution neural network.
[16] E. A. Johnson, K. M. Dudding, and J. M. Carrington, ‘‘When to ERR is
LSTM Long Short-Term Memory. inhuman: An examination of the influence of artificial intelligence-driven
RF Random forest. nursing care on patient safety,’’ Nursing Inquiry, vol. 31, no. 1, Jan. 2024,
LR Logistic regression. Art. no. e12583.
[17] M. Bekbolatova, J. Mayer, C. W. Ong, and M. Toma, ‘‘Transformative
DT Decision tree. potential of AI in healthcare: Definitions, applications, and navigating
DL Deep learning. the ethical landscape and public perspectives,’’ Healthcare, vol. 12, no. 2,
GBT Gradient boosting. p. 125, Jan. 2024.
[18] E. Mosqueira-Rey, E. Hernández-Pereira, D. Alonso-Ríos, J. Bobes-
GBM Gradient boosting machine. Bascarán, and Á. Fernández-Leal, ‘‘Human-in-the-loop machine learn-
BN Bayesian network. ing: A state of the art,’’ Artif. Intell. Rev., vol. 56, no. 4, pp. 3005–3054,
PCA Principal component analysis. Apr. 2023.
NLP Natural language processing. [19] M. Khalifa and M. Albadawy, ‘‘AI in diagnostic imaging: Revolutionising
accuracy and efficiency,’’ Comput. Methods Programs Biomed. Update,
vol. 5, Jun. 2024, Art. no. 100146.
[20] S. Nyholm, ‘‘Artificial intelligence and human enhancement: Can
REFERENCES AI technologies make us more (artificially) intelligent?’’ Cambridge
[1] G. K. Vincent, The Next Four Decades: The Older Population in the Quart. Healthcare Ethics, vol. 33, no. 1, pp. 76–88, Jan. 2024.
United States: 2010 To 2050, document 1138, U.S. Dept. Commerce, [21] A. Choudhury, E. Renjilian, and O. Asan, ‘‘Use of machine learning
Econ. Statist. Admin., 2010. in geriatric clinical care for chronic diseases: A systematic literature
[2] K. E. Miller, R. G. Zylstra, and J. B. Standridge, ‘‘The geriatric patient: review,’’ JAMIA Open, vol. 3, no. 3, pp. 459–471, Oct. 2020.
A systematic approach to maintaining health,’’ Amer. Family Physician, [22] A. Das and P. Dhillon, ‘‘Application of machine learning in measurement
vol. 61, no. 4, pp. 1089–1104, 2000. of ageing and geriatric diseases: A systematic review,’’ BMC Geriatrics,
[3] C. Hawk, M. Schneider, P. Dougherty, B. J. Gleberzon, and vol. 23, no. 1, p. 841, Dec. 2023.
L. Z. Killinger, ‘‘Best practices recommendations for chiropractic [23] A. Husnain, H. K. Hussain, H. M. Shahroz, M. Ali, and Y. Hayat, ‘‘A
care for older adults: Results of a consensus process,’’ J. Manipulative precision health initiative for chronic conditions: Design and cohort study
Physiological Therapeutics, vol. 33, no. 6, pp. 464–473, Jul. 2010. utilizing wearable technology, machine learning, and deep learning,’’
[4] J. Barrio-Cortes, A. Castaño-Reguillo, M. T. Beca-Martínez, Int. J. Adv. Eng. Technol. Innov., vol. 1, no. 2, pp. 118–139, 2024.
M. Bandeira-de Oliveira, C. López-Rodríguez, and M. Á. Jaime- [24] M. N. A. Tawhid, S. Siuly, K. Wang, and H. Wang, ‘‘Automatic and
Sisó, ‘‘Chronic diseases in the geriatric population: Morbidity and use of efficient framework for identifying multiple neurological disorders from
primary care services according to risk level,’’ BMC Geriatrics, vol. 21, EEG signals,’’ IEEE Trans. Technol. Soc., vol. 4, no. 1, pp. 76–86,
no. 1, pp. 1–11, Dec. 2021. Mar. 2023.

VOLUME 12, 2024 122581

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

[25] M. Mujahid, A. Rehman, T. Alam, F. S. Alamri, S. M. Fati, and T. Saba, [45] M. L. Alarcon, U. Roy, A. Kambhampati, N. Nguyen, M. Attari, R. Surya,
‘‘An efficient ensemble approach for Alzheimer’s disease detection using F. Bunyak, M. Maschmann, K. Palaniappan, and P. Calyam, ‘‘Learning-
an adaptive synthetic technique and deep learning,’’ Diagnostics, vol. 13, based image analytics in user-AI agent interactions for cyber-enabled
no. 15, p. 2489, Jul. 2023. manufacturing,’’ in Proc. IEEE 4th Int. Conf. Hum.-Mach. Syst. (ICHMS),
[26] R. J. Woodman and A. A. Mangoni, ‘‘A comprehensive review of machine May 2024, pp. 1–7.
learning algorithms and their application in geriatric medicine: Present [46] T. A. Assegie, T. Suresh, R. Purushothaman, S. Ganesan, and
and future,’’ Aging Clin. Experim. Res., vol. 35, no. 11, pp. 2363–2397, N. K. Kumar, ‘‘Early prediction of gestational diabetes with parameter-
Sep. 2023. tuned K-nearest neighbor classifier,’’ J. Robot. Control (JRC), vol. 4,
[27] M. Hamaker, C. Lund, M. T. Molder, P. Soubeyran, H. Wildiers, and no. 4, pp. 452–457, Jul. 2023.
L. van Huis, ‘‘Geriatric assessment in the management of older patients [47] M. Kateri, ‘‘Contingency table analysis,’’ Statist. Ind. Technol., vol. 525,
with cancer-a systematic review (update),’’ J. geriatric Oncol., vol. 13, pp. 1–24, Sep. 2014.
no. 6, pp. 761–777, 2022. [48] P. K. Misra, N. Kumar, A. Misra, and A. Khang, ‘‘Heart disease prediction
[28] H. Cai, Y. Jin, R. Liu, Q. Zhang, Z. Su, G. S. Ungvari, Y.-L. Tang, using logistic regression and random forest classifier,’’ in Data-Centric AI
C. H. Ng, X.-H. Li, and Y.-T. Xiang, ‘‘Global prevalence of depression in Solutions and Emerging Technologies in the Healthcare Ecosystem. Boca
older adults: A systematic review and meta-analysis of epidemiological Raton, FL, USA: CRC Press, 2023, pp. 83–112.
surveys,’’ Asian J. Psychiatry, vol. 80, Feb. 2023, Art. no. 103417. [49] A. Das, ‘‘Logistic regression,’’ in Encyclopedia of Quality of Life and
[29] J. C. Correia, A. Waqas, J.-P. Assal, M. J. Davies, F. Somers, A. Golay, and Well-Being Research: Springer. Cham, Switzerland: Springer, 2024,
Z. Pataky, ‘‘Effectiveness of therapeutic patient education interventions pp. 3985–3986.
for chronic diseases: A systematic review and meta-analyses of random- [50] R. Bhardwaj and S. Vatta, ‘‘Implementation of ID3 algorithm,’’
ized controlled trials,’’ Frontiers Med., vol. 9, Jan. 2023, Art. no. 996528. Int. J. Adv. Res. Comput. Sci. Softw. Eng., vol. 3, no. 6, pp. 1–24, 2013.
[30] D. Collado-Mateo, A. M. Lavín-Pérez, C. Peñacoba, J. Del Coso,
[51] E. O. Paul, ‘‘Hybrid decision tree-based machine learning models for
M. Leyton-Román, A. Luque-Casado, P. Gasque, M. Á. Fernández-del-
diabetes prediction,’’ SCIREA J. Inf. Sci. Syst. Sci., vol. 8, pp. 1–30,
Olmo, and D. Amado-Alonso, ‘‘Key factors associated with adherence
Jan. 2024.
to physical exercise in patients with chronic diseases and older adults:
An umbrella review,’’ Int. J. Environ. Res. Public Health, vol. 18, no. 4, [52] M. Attari, N. P. Nguyen, K. Palaniappan, and F. Bunyak, ‘‘Multi-loss
p. 2023, Feb. 2021. topology-aware deep learning network for segmentation of vessels in
microscopy images,’’ in Proc. IEEE Appl. Imag. Pattern Recognit. Work-
[31] Y. Kumar, A. Koul, R. Singla, and M. F. Ijaz, ‘‘Artificial intelligence in
shop (AIPR), Sep. 2023, pp. 1–7.
disease diagnosis: A systematic literature review, synthesizing framework
and future research agenda,’’ J. Ambient Intell. Humanized Comput., [53] M. V. N, A. N, and K. S. K, ‘‘Decision tree-based explainable AI for
vol. 14, no. 7, pp. 8459–8486, Jul. 2023. diagnosis of chronic kidney disease,’’ in Proc. 5th Int. Conf. Inventive
[32] A. Bin Sawad, B. Narayan, A. Alnefaie, A. Maqbool, I. Mckie, J. Smith, Res. Comput. Appl. (ICIRCA), Aug. 2023, pp. 947–952.
B. Yuksel, D. Puthal, M. Prasad, and A. B. Kocaballi, ‘‘A system- [54] L. Jiang, H. Zhang, and Z. Cai, ‘‘A novel Bayes model: Hidden naive
atic review on healthcare artificial intelligent conversational agents for Bayes,’’ IEEE Trans. Knowl. Data Eng., vol. 21, no. 10, pp. 1361–1371,
chronic conditions,’’ Sensors, vol. 22, no. 7, p. 2625, Mar. 2022. Oct. 2009.
[33] A. Merabet, A. Saighi, Z. Laboudi, and M. A. Ferradji, ‘‘Multiple diseases [55] S. Halder, S. Kumar, D. P. Acharjya, and S. Dutta, ‘‘Forecast of health
forecast through AI and IoMT techniques: Systematic literature review,’’ risk for chronic kidney disease: A comparison between Naïve Bayes (NB)
in Proc. Int. Conf. Intell. Syst. Pattern Recognit., 2024, pp. 189–206. and support vector machine (SVM) models,’’ in Proc. Comput. Vis. AI-
[34] T. P. Theodore Armand, K. A. Nfor, J.-I. Kim, and H.-C. Kim, ‘‘Appli- Integr. IoT Technol. Med. Ecosystem, 2023, pp. 267–287.
cations of artificial intelligence, machine learning, and deep learning [56] H. Gharagozlou, J. Mohammadzadeh, A. Bastanfard, and S. S. Ghidary,
in nutrition: A systematic review,’’ Nutrients, vol. 16, no. 7, p. 1073, ‘‘RLAS-BIABC: A reinforcement learning-based answer selection
Apr. 2024. using the BERT model boosted by an improved ABC algorithm,’’
[35] C. Cortes and V. Vapnik, ‘‘Support-vector networks,’’ Mach. Learn., Comput. Intell. Neurosci., vol. 2022, pp. 1–21, May 2022, doi:
vol. 20, pp. 273–297, Jul. 1995. 10.1155/2022/7839840.
[36] V. C. R, Asha, V, A. Prasad, S. Das, S. Kumar, and S. S. P, ‘‘Support vector [57] S. Reddy, N. Sethi, R. Rajender, and G. Mahesh, ‘‘Forecasting dia-
machine (SVM) and artificial neural networks (ANN) based chronic kid- betes correlated non-alcoholic fatty liver disease by exploiting Naïve
ney disease prediction,’’ in Proc. 7th Int. Conf. Comput. Methodologies Bayes tree,’’ ICST Trans. Scalable Inf. Syst., vol. 10, Jul. 2018,
Commun. (ICCMC), Feb. 2023, pp. 469–474. Art. no. 173975.
[37] P. Ebrahimi, H. Dustmohammadloo, H. Kabiri, P. Bouzari, and M. Fekete- [58] T. Chen and C. Guestrin, ‘‘XGBoost: A scalable tree boosting system,’’
Farkas, ‘‘Transformational entrepreneurship and digital platforms: A in Proc. 22nd ACM SIGKDD Int. Conf. Knowl. Discovery Data Mining,
combination of ISM-MICMAC and unsupervised machine learning algo- Aug. 2016, pp. 785–794.
rithms,’’ Big Data Cognit. Comput., vol. 7, no. 2, p. 118, Jun. 2023. [59] H. A. Al-Jamimi, ‘‘Synergistic feature engineering and ensemble learn-
[38] A. S. Shanthakumari and R. Jayakarthik, ‘‘Utilizing support vec- ing for early chronic disease prediction,’’ IEEE Access, vol. 12,
tor machines for predictive analytics in chronic kidney diseases,’’ pp. 62215–62233, 2024.
Mater. Today, Proc., vol. 81, pp. 951–956, May 2023. [60] S. Danaei, A. Bostani, S. V. Moravvej, F. Mohammadi, R. Alizadehsani,
[39] L. Breiman, ‘‘Bagging predictors,’’ Mach. Learn., vol. 24, no. 2, A. Shoeibi, H. Alinejad-Rokny, and S. Nahavandi, ‘‘Myocarditis diag-
pp. 123–140, Aug. 1996. nosis: A method using mutual learning-based ABC and reinforcement
[40] H. Dustmohammadloo, A. Nikeghbalzadeh, F. F. Karkaragh, and S. Souri, learning,’’ in Proc. IEEE 22nd Int. Symp. Comput. Intell. Informat. 8th
‘‘Knowledge sharing as a moderator between organizational learning and IEEE Int. Conf. Recent Achievements Mechatronics, Autom., Com-
error management culture in academic staff,’’ Int. J. Multicultural Educ., put. Sci. Robot. (CINTI-MACRo), Nov. 2022, pp. 000265–000270.
vol. 25, no. 3, pp. 1–24, 2023. [61] H. Zareiamand, A. Darroudi, I. Mohammadi, S. V. Moravvej, S. Danaei,
[41] Z. Ghasemi Kooktapeh, H. Dustmohammadloo, H. Mehrdoost, and and R. Alizadehsani, ‘‘Cardiac magnetic resonance imaging (CMRI)
F. Fatehi, ‘‘In the line of fire: A systematic review and meta-analysis of applications in patients with chest pain in the emergency depart-
job burnout among nurses,’’ 2023, arXiv:2312.14853. ment: A narrative review,’’ Diagnostics, vol. 13, no. 16, p. 2667,
[42] G. Airlangga, ‘‘Comparative analysis of machine learning models for Aug. 2023.
chronic disease indicator classification using U.S. chronic disease indi- [62] S. Bhaskar and T. T. M., ‘‘LSTM model for visual speech recognition
cators dataset,’’ MALCOM, Indonesian J. Mach. Learn. Comput. Sci., through facial expressions,’’ Multimedia Tools Appl., vol. 82, no. 4,
vol. 4, no. 3, pp. 1034–1042, 2024. pp. 5455–5472, Feb. 2023.
[43] N. Rankovic, D. Rankovic, I. Lukic, N. Savic, and V. Jovanovic, ‘‘Unveil- [63] S. V. Moravvej, R. Alizadehsani, S. Khanam, Z. Sobhaninia, A. Shoeibi,
ing the comorbidities of chronic diseases in Serbia using ML algorithms F. Khozeimeh, Z. A. Sani, R.-S. Tan, A. Khosravi, S. Nahavandi,
and Kohonen self-organizing maps for personalized healthcare frame- N. A. Kadri, M. M. Azizan, N. Arunkumar, and U. R. Acharya, ‘‘RLMD-
works,’’ J. Personalized Med., vol. 13, no. 7, p. 1032, Jun. 2023. PA: A reinforcement learning-based myocarditis diagnosis combined
[44] T. Cover and P. Hart, ‘‘Nearest neighbor pattern classification,’’ IEEE with a population-based algorithm for pretraining weights,’’ Contrast
Trans. Inf. Theory, vols. IT13, no. 1, pp. 21–27, 1967. Media Mol. Imag., vol. 2022, no. 1, pp. 1–18, Jan. 2022.

122582 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

[64] M. Roshanzamir, R. Alizadehsani, S. V. Moravvej, J. H. Joloudari, [83] M. A. Al-Jarrah and H. Shatnawi, ‘‘Non-proliferative diabetic retinopa-
H. Alinejad-Rokny, and J. M. Gorriz, ‘‘Enhancing interpretability in thy symptoms detection and classification using neural network,’’
machine learning: A focus on genetic network programming, its variants, J. Med. Eng. Technol., vol. 41, no. 6, pp. 498–505, Aug. 2017.
and applications,’’ in Proc. Int. Work-Conf. Interplay Between Natural [84] DIARETDB1. Accessed: Sep. 2, 2024. [Online]. Available: https://
Artif. Comput., 2024, pp. 98–107. www.kaggle.com/datasets/nguyenhung1903/diaretdb1-standard-
[65] M. Esmaeili, S. H. Goki, B. H. K. Masjidi, M. Sameh, H. Gharago- diabetic-retinopathy-database
zlou, and A. S. Mohammed, ‘‘ML-DDoSnet: IoT intrusion detection [85] I. Beheshti, N. Maikusa, M. Daneshmand, H. Matsuda, H. Demirel, and
based on denial-of-service attacks using machine learning methods and G. Anbarjafari, ‘‘Classification of Alzheimer’s disease and prediction of
NSL-KDD,’’ Wireless Commun. Mobile Comput., vol. 2022, pp. 1–16, mild cognitive impairment conversion using histogram-based analysis of
Aug. 2022. patient-specific anatomical brain connectivity networks,’’ J. Alzheimer’s
[66] A. Mirzaee Moghaddam Kasmaee, A. Ataei, S. V. Moravvej, Disease, vol. 60, no. 1, pp. 295–304, Aug. 2017.
R. Alizadehsani, J. M. Gorriz, Y.-D. Zhang, R.-S. Tan, and U. R. Acharya, [86] Japanese Alzheimer’s Disease Neuroimaging Initiative. Accessed:
‘‘ELRL-MD: A deep learning approach for myocarditis diagnosis using Aug. 28, 2024. [Online]. Available: https://humandbs.biosciencedbc.
cardiac magnetic resonance images with ensemble and reinforcement jp/en/hum0043-v1
learning integration,’’ Physiological Meas., vol. 45, no. 5, May 2024, [87] M. Hubenthal, G. Hemmrich-Stanisak, F. Degenhardt, S. Szymczak,
Art. no. 055011. Z. Du, A. Elsharawy, A. Keller, S. Schreiber, and A. Franke, ‘‘Sparse
[67] S. V. Moravvej, M. J. M. Kahaki, M. S. Sartakhti, and A. Mirzaei, modeling reveals MiRNA signatures for diagnostics (don’t short) of
‘‘A method based on attention mechanism using bidirectional long-short inflammatory bowel disease,’’ PLoS One, vol. 10, no. 10, Oct. 2015,
term memory(BLSTM) for question answering,’’ in Proc. 29th Iranian Art. no. e0140155.
Conf. Electr. Eng. (ICEE), May 2021, pp. 460–464. [88] U.K. Biobank. Accessed: Sep. 3, 2024. [Online]. Available: https://
[68] S. Vahid Moravvej, A. Mirzaei, and M. Safayani, ‘‘Biomedical text sum- www.ukbiobank.ac.uk
marization using conditional generative adversarial network(CGAN),’’ [89] J. A. Andrews, R. F. Harrison, L. J. E. Brown, L. M. MacLean, F. Hwang,
2021, arXiv:2110.11870. T. Smith, E. A. Williams, C. Timon, T. Adlam, H. Khadra, and A. J. Astell,
[69] C. Hema and F. P. Garcia Marquez, ‘‘Emotional speech recognition using ‘‘Using the NANA toolkit at home to predict older adults’ future depres-
CNN and deep learning techniques,’’ Appl. Acoust., vol. 211, Aug. 2023, sion,’’ J. Affect. Disorders, vol. 213, pp. 187–190, Apr. 2017.
Art. no. 109492. [90] J. Choi, J. Choi, and W. J. Choi, ‘‘Predicting depression among commu-
[70] M. Rayhan Ahmed, S. Islam, A. K. M. Muzahidul Islam, and S. Shatabda, nity residing older adults: A use of machine learning approch,’’ in Nursing
‘‘An ensemble 1D-CNN-LSTM-GRU model with data augmentation for Informatics 2018. Amsterdam, The Netherlands: IOS Press, 2018,
speech emotion recognition,’’ Expert Syst. Appl., vol. 218, May 2023, p. 265.
Art. no. 119633. [91] National Social Life, Health, and Aging Project. Accessed:
[71] K. Modi, I. Singh, and Y. Kumar, ‘‘A comprehensive analysis of artifi- Aug. 29, 2024. [Online]. Available: https://www.icpsr.umich.edu/
cial intelligence techniques for the prediction and prognosis of lifestyle web/ICPSR/studies/20541
diseases,’’ Arch. Comput. Methods Eng., vol. 30, no. 8, pp. 4733–4756, [92] C. M. Hatton, L. W. Paton, D. Mcmillan, J. Cussens, S. Gilbody,
Nov. 2023. and P. A. Tiffin, ‘‘Predicting persistent depressive symptoms in older
[72] P. R. Kotha, M. Attari, M. Maschmann, and F. Bunyak, ‘‘Deep style trans- adults: A machine learning approach to personalised mental healthcare,’’
fer for generation of photo-realistic synthetic images of CNT forests,’’ in J. Affect. Disorders, vol. 246, pp. 857–860, Mar. 2019.
Proc. IEEE Appl. Imag. Pattern Recognit. Workshop (AIPR), Sep. 2023, [93] N. Bhagwat, J. D. Viviano, A. N. Voineskos, M. M. Chakravart, and
pp. 1–7. A. D. Initiative, ‘‘Modeling and prediction of clinical symptom trajecto-
[73] K. Sinha, Z. Uddin, H. I. Kawsar, S. Islam, M. J. Deen, and ries in Alzheimer’s disease using longitudinal data,’’ PLoS Comput. Biol.,
M. M. R. Howlader, ‘‘Analyzing chronic disease biomarkers using vol. 14, no. 9, 2018, Art. no. e1006376.
electrochemical sensors and artificial neural networks,’’ TrAC Trends [94] Alzheimer’s Disease Neuroimaging Initiative. Accessed: Sep. 2, 2024.
Anal. Chem., vol. 158, Jan. 2023, Art. no. 116861. [Online]. Available: https://adni.loni.usc.edu
[74] S. Taherinavid, S. V. Moravvej, Y.-L. Chen, J. Yang, C. S. Ku, [95] S. L. James, M. Romero, D. Ramírez-Villalobos, S. Gómez, K. Pierce,
and P. L. Yee, ‘‘Automatic transportation mode classification using a A. Flaxman, P. Serina, A. Stewart, C. J. Murray, E. Gakidou,
deep reinforcement learning approach with smartphone sensors,’’ IEEE R. Lozano, and B. Hernandez, ‘‘Validating estimates of prevalence of non-
Access, vol. 12, pp. 514–533, 2024. communicable diseases based on household surveys: The symptomatic
[75] G. Kim, H. Lim, Y. Kim, O. Kwon, and J.-H. Choi, ‘‘Intra-person multi- diagnosis study,’’ BMC Med., vol. 13, no. 1, pp. 1–12, Dec. 2015.
task learning method for chronic-disease prediction,’’ Sci. Rep., vol. 13, [96] Population Health Metrics Research Consortium Study. Accessed:
no. 1, p. 1069, Jan. 2023. Aug. 20, 2024. [Online]. Available: https://www.georgeinstitute.org.
in/projects/population-health-metrics-research-consortium-phmrc-
[76] A. Chunduru, A. R. Kishore, B. K. Sasapu, and K. Seepana, ‘‘Multi
project
chronic disease prediction system using CNN and random forest,’’ Social
Netw. Comput. Sci., vol. 5, no. 1, p. 157, Jan. 2024. [97] G. Lee, K. Nho, B. Kang, K. A. Sohn, and D. Kim, ‘‘Predict-
ing Alzheimer’s disease progression using multi-modal deep learning
[77] A. A. Malibari, ‘‘An efficient IoT-artificial intelligence-based disease
approach,’’ Sci. Rep., vol. 9, no. 1, p. 1952, 2019.
prediction using lightweight CNN in healthcare system,’’ Meas., Sensors,
vol. 26, Apr. 2023, Art. no. 100695. [98] H. Bogunovic, A. Montuoro, M. Baratsits, M. G. Karantonis,
S. M. Waldstein, F. Schlanitz, and U. Schmidt-Erfurth, ‘‘Machine
[78] S. Chitra and V. Jayalakshmi, ‘‘Heart disease and chronic kidney dis- learning of the progression of intermediate age-related macular
ease prediction based on Internet of Things using FRNN algorithm,’’ degeneration based on OCT imaging,’’ Investigative Opthalmology
in Proc. 6th Int. Conf. Comput. Methodologies Commun. (ICCMC), Vis. Sci., vol. 58, no. 6, Jun. 2017, Art. no. BIO141.
Mar. 2022, pp. 467–479.
[99] B. Cheng, M. Liu, H.-I. Suk, D. Shen, and D. Zhang, ‘‘Multimodal
[79] L. Gorzelniak, K. Schultz, M. Wittmann, J. Rudnik, R. Jörres, A. Horsch, manifold-regularized transfer learning for MCI conversion
and A. Dias, ‘‘Classification of exacerbation episodes in chronic obstruc- prediction,’’ Brain Imag. Behav., vol. 9, no. 4, pp. 913–926,
tive pulmonary disease patients,’’ Methods Inf. Med., vol. 53, no. 2, Dec. 2015.
pp. 108–114, 2014.
[100] K. M. Kuo, P. C. Talley, C. H. Huang, and L. C. Cheng, ‘‘Predicting
[80] Y. Du, R. Almajalid, J. Shan, and M. Zhang, ‘‘A novel method to predict hospital-acquired pneumonia among schizophrenic patients: A machine
knee osteoarthritis progression on MRI using machine learning methods,’’ learning approach,’’ BMC Med. Informat. Decis. Making, vol. 19, no. 1,
IEEE Trans. Nanobiosci., vol. 17, no. 3, pp. 228–236, Jul. 2018. pp. 1–8, Dec. 2019.
[81] Osteoarthritis Initiative Database. Accessed: Aug. 30, 2024. [101] A. J. C. C. Lins, M. T. C. Muniz, A. N. M. Garcia, A. V. Gomes,
[Online]. Available: https://nda.nih.gov/oai R. M. Cabral, and C. J. A. Bastos-Filho, ‘‘Using artificial neural networks
[82] A. Akl, B. Taati, and A. Mihailidis, ‘‘Autonomous unobtrusive detection to select the parameters for the prognostic of mild cognitive impair-
of mild cognitive impairment in older adults,’’ IEEE Trans. Biomed. Eng., ment and dementia in elderly individuals,’’ Comput. Methods Programs
vol. 62, no. 5, pp. 1383–1394, May 2015. Biomed., vol. 152, pp. 93–104, Dec. 2017.

VOLUME 12, 2024 122583

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

[102] A. Ortiz, J. Munilla, J. M. Górriz, and J. Ramírez, ‘‘Ensembles of deep [122] NCT02484690. Accessed: Sep. 2, 2024. [Online]. Available: https://
learning architectures for the early diagnosis of the Alzheimer’s disease,’’ clinicaltrials.gov/study/NCT02484690
Int. J. Neural Syst., vol. 26, no. 7, Nov. 2016, Art. no. 1650025. [123] J. Lu, Y. Cheng, F. E. Hiya, M. Shen, G. Herrera, Q. Zhang, G. Gregori,
[103] Ashraf A and Taati B, ‘‘Automated video analysis of handwashing behav- P. J. Rosenfeld, and R. K. Wang, ‘‘Deep-learning-based automated
ior as a potential marker of cognitive health in older adults,’’ IEEE measurement of outer retinal layer thickness for use in the assessment
J. Biomed. Health Informat., vol. 20, no. 2, pp. 682–690, Jul. 2015. of age-related macular degeneration, applicable to both swept-source
[104] M. Christopher, A. Belghith, R. N. Weinreb, C. Bowd, M. H. Goldbaum, and spectral-domain OCT imaging,’’ Biomed. Opt. Exp., vol. 15, no. 1,
L. J. Saunders, F. A. Medeiros, and L. M. Zangwill, ‘‘Retinal nerve fiber p. 413, 2024.
layer features identified by unsupervised machine learning on optical [124] M. Elsharkawy, A. Sharafeldeen, F. Khalifa, A. Soliman, A. Elnakib,
coherence tomography scans predict glaucoma progression,’’ Investiga- M. Ghazal, A. Sewelam, A. Thanos, H. S. Sandhu, and A. El-Baz,
tive Opthalmology Vis. Sci., vol. 59, no. 7, p. 2748, Jun. 2018. ‘‘A clinically explainable AI-based grading system for age-related
[105] Diagnostic Innovations in Glaucoma. Accessed: Jul. 18, 2024. macular degeneration using optical coherence tomography,’’ IEEE
[Online]. Available: https://grantome.com/grant/NIH/R01-EY027510-05 J. Biomed. Health Informat., vol. 28, no. 4, pp. 2079–2090, Apr. 2024.
[106] A. König, N. Linz, R. Zeghari, X. Klinge, J. Tröger, J. Alexandersson, [125] DRIONS-DB. Accessed: Aug. 31, 2024. [Online]. Available: https://
and P. Robert, ‘‘Detecting apathy in older adults with cognitive disorders www.idiap.ch/software/bob/docs/bob/bob.db.drionsdb/master/index.html
using automatic speech analysis,’’ J. Alzheimer’s Disease, vol. 69, no. 4, [126] A. Khan and S. Zubair, ‘‘An improved multi-modal based machine learn-
pp. 1183–1193, Jun. 2019. ing approach for the prognosis of Alzheimer’s disease,’’ J. King Saud
[107] A. Weakley, J. A. Williams, M. Schmitter-Edgecombe, and D. J. Cook, Univ.-Comput. Inf. Sci., vol. 34, no. 6, pp. 2688–2706, Jun. 2022.
‘‘Neuropsychological test selection for cognitive impairment classifica- [127] G. Muni Nagamani and E. Rayachoti, ‘‘Deep learning network (DL-
tion: A machine learning approach,’’ J. Clin. Experim. Neuropsychology, Net) based classification and segmentation of multi-class retinal diseases
vol. 37, no. 9, pp. 899–916, Oct. 2015. using OCT scans,’’ Biomed. Signal Process. Control, vol. 88, Feb. 2024,
[108] M. Treder, J. L. Lauermann, and N. Eter, ‘‘Deep learning-based detec- Art. no. 105619.
tion and classification of geographic atrophy using a deep convolutional [128] I. M. Saied, T. Arslan, and S. Chandran, ‘‘Classification of Alzheimer’s
neural network classifier,’’ Graefe’s Arch. Clin. Experim. Ophthalmology, disease using RF signals and machine learning,’’ IEEE J. Electromagn.,
vol. 256, no. 11, pp. 2053–2060, Nov. 2018. RF Microw. Med. Biol., vol. 6, no. 1, pp. 77–85, Mar. 2022.
[109] S. M. Naushad, T. Hussain, B. Indumathi, K. Samreen, S. A. Alrokayan, [129] M. H. Wang, J. Zhou, C. Huang, Z. Tang, X. Yu, G. Hou, J. Yang,
and V. K. Kutala, ‘‘Machine learning algorithm-based risk prediction Q. Yuan, K. K. Chong, and L. Huang, ‘‘Fusion learning methods for
model of coronary artery disease,’’ Mol. Biol. Rep., vol. 45, no. 5, the age-related macular degeneration diagnosis based on multiple sources
pp. 901–910, Oct. 2018. of ophthalmic digital images,’’ in Proc. 2nd Int. Conf. Electr., Electron.,
[110] C. Cabral, P. M. Morgado, D. Campos Costa, and M. Silveira, ‘‘Pre- Inf. Eng., Jan. 2024, pp. 470–492.
dicting conversion from MCI to AD with FDG-PET brain images at
[130] C.-H. Liu, C.-H. Peng, L.-Y. Huang, F.-Y. Chen, C.-H. Kuo, C.-Z. Wu,
different prodromal stages,’’ Comput. Biol. Med., vol. 58, pp. 101–109,
and Y.-F. Cheng, ‘‘Comparison of multiple linear regression and machine
Mar. 2015.
learning methods in predicting cognitive function in older Chinese type 2
[111] K.-H. Thung, C.-Y. Wee, P.-T. Yap, and D. Shen, ‘‘Identification of
diabetes patients,’’ BMC Neurol., vol. 24, no. 1, p. 11, Jan. 2024.
progressive mild cognitive impairment patients using incomplete longitu-
[131] G. Battineni, M. A. Hossain, N. Chintalapudi, E. Traini, V. R. Dhulipalla,
dinal MRI scans,’’ Brain Struct. Function, vol. 221, no. 8, pp. 3979–3995,
M. Ramasamy, and F. Amenta, ‘‘Improved Alzheimer’s disease detection
Nov. 2016.
by MRI using multimodal machine learning algorithms,’’ Diagnostics,
[112] Y. Zhan, K. Chen, X. Wu, D. Zhang, J. Zhang, L. Yao, and X. Guo, ‘‘Iden-
vol. 11, no. 11, p. 2103, Nov. 2021.
tification of conversion from normal elderly cognition to Alzheimer’s
disease using multimodal support vector machine,’’ J. Alzheimer’s Dis- [132] S. Hong, B. Lu, S. Wang, and Y. Jiang, ‘‘Comparison of logistic
ease, vol. 47, no. 4, pp. 1057–1067, Aug. 2015. regression and machine learning methods for predicting depression risks
[113] J. Lara, F. X. López-Labrador, F. González-Candelas, M. Berenguer, and among disabled elderly individuals: Results from the China health and
Y. E. Khudyakov, ‘‘Computational models of liver fibrosis progression for retirement longitudinal study,’’ Hubei Univ. Med., China, Tech. Rep.
hepatitis c virus chronic infection,’’ BMC Bioinf., vol. 15, no. S8, pp. 1–9, 2021WNLOKF020, 2024.
Jul. 2014. [133] P. Kishore, C. U. Kumari, M. Kumar, and T. Pavani, ‘‘Detection and anal-
[114] GenBank. Accessed: Sep. 3, 2024. [Online]. Available: https://www. ysis of Alzheimer’s disease using various machine learning algorithms,’’
ncbi.nlm.nih.gov/genbank/ Mater. Today, Proc., vol. 45, pp. 1502–1508, Jun. 2021.
[115] U. Schmidt-Erfurth, S. M. Waldstein, S. Klimscha, A. Sadeghipour, [134] C. Kavitha, V. Mani, S. R. Srividhya, O. I. Khalaf, and C. A. Tavera
X. Hu, B. S. Gerendas, A. Osborne, and H. Bogunovic, ‘‘Prediction of Romero, ‘‘Early-stage Alzheimer’s disease prediction using machine
individual disease conversion in early AMD using artificial intelligence,’’ learning models,’’ Frontiers Public Health, vol. 10, Mar. 2022,
Investigative Opthalmology Vis. Sci., vol. 59, no. 8, p. 3199, Jul. 2018. Art. no. 853294.
[116] HARBOR. Accessed: Sep. 3, 2024. [Online]. Available: https:// [135] A. Singh and R. Kumar, ‘‘Brain MRI image analysis for Alzheimer’s dis-
clinicaltrials.gov/study/NCT04910685 ease (AD) prediction using deep learning approaches,’’ Social Netw. Com-
[117] F. C. Morabito, M. Campolo, N. Mammone, M. Versaci, S. Franceschetti, put. Sci., vol. 5, no. 1, p. 160, Jan. 2024.
F. Tagliavini, V. Sofia, D. Fatuzzo, A. Gambardella, A. Labate, [136] C. Montorsi, A. Fusco, P. Van Kerm, and S. P. A. Bordas, ‘‘Predicting
L. Mumoli, G. G. Tripodi, S. Gasparini, V. Cianci, C. Sueri, E. Ferlazzo, depression in old age: Combining life course data with machine learning,’’
and U. Aguglia, ‘‘Deep learning representation from electroencephalog- Econ. Hum. Biol., vol. 52, Jan. 2024, Art. no. 101331.
raphy of early-stage creutzfeldt-jakob disease and features for differen- [137] H. Park, C. Kang, A.-C. Team, W. Lee, I. Song, D. Kwon, J. Oh,
tiation from rapidly progressive dementia,’’ Int. J. Neural Syst., vol. 27, J. Moon, J. Park, J. Min, E. Kim, H. Jang, and H. Kim, ‘‘Particulate
no. 2, Mar. 2017, Art. no. 1650039. matters (PM2.5 , PM10 ) and the risk of depression among middle-aged
[118] V. Pedoia, J. Haefeli, K. Morioka, H. Teng, L. Nardo, R. B. Souza, and older population: Analysis of the Korean longitudinal study of aging
A. R. Ferguson, and S. Majumdar, ‘‘MRI and biomechanics multidimen- (KLoSA), 2016–2020 in South Korea,’’ Environ. Health, vol. 23, no. 1,
sional data analysis reveals R2 -R1 ρ as an early predictor of cartilage p. 4, Jan. 2024.
lesion progression in knee osteoarthritis,’’ J. Magn. Reson. Imag., vol. 47, [138] Z. Qu, Y. Wang, D. Guo, G. He, C. Sui, Y. Duan, X. Zhang,
no. 1, pp. 78–90, Jan. 2018. L. Lan, H. Meng, Y. Wang, and X. Liu, ‘‘Identifying depres-
[119] J. L. Avila-Jimenez, V. Canton-Habas, M. del Pilar Carrera-Gonzslez, sion in the United States veterans using deep learning algorithms,
M. Rich-Ruiz, and S. Ventura, ‘‘A deep learning model for Alzheimer’s NHANES 2005–2018,’’ BMC Psychiatry, vol. 23, no. 1, p. 620,
disease diagnosis based on patient clinical records,’’ Comput. Biol. Med., Aug. 2023.
vol. 169, Sep. 2024, Art. no. 107814. [139] National Health and Nutrition Examination Survey. Accessed: Sep. 3,
[120] National Alzheimer’s Coordinating Center. Accessed: Sep. 2, 2024. 2024. [Online]. Available: https://www.cdc.gov/nchs/nhanes/index.htm
[121] H. Yao, Z. Wu, S. S. Gao, R. H. Guymer, V. Steffen, H. Chen, M. Hejrati, [140] S. R. B. S. Gomes, M. von Schantz, and M. Leocadio-Miguel, ‘‘Predicting
and M. Zhang, ‘‘Deep learning approaches for detecting of nascent geo- depressive symptoms in middle-aged and elderly adults using sleep data
graphic atrophy in age-related macular degeneration,’’ Ophthalmology and clinical health markers: A machine learning approach,’’ Sleep Med.,
Sci., vol. 4, no. 3, May 2024, Art. no. 100428. vol. 102, pp. 123–131, Feb. 2023.

122584 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

[141] Sleep Health Index. Accessed: Aug. 29, 2024. [Online]. Available: [161] Md. A. Awal, Md. S. Hossain, K. Debjit, N. Ahmed, R. D. Nath,
https://www.thensf.org/sleep-health-index/ G. M. M. Habib, Md. S. Khan, Md. A. Islam, and M. A. P. Mahmud,
[142] S. Lin, Y. Wu, L. He, and Y. Fang, ‘‘Prediction of depressive symp- ‘‘An early detection of asthma using BOMLA detector,’’ IEEE Access,
toms onset and long-term trajectories in home-based older adults using vol. 9, pp. 58403–58420, 2021.
machine learning techniques,’’ Aging Mental Health, vol. 27, no. 1, [162] S. Sarkar, S. Bhattacherjee, P. Bhattacharyya, M. Mitra, and S. Pal,
pp. 8–17, Jan. 2023. ‘‘Automatic identification of asthma from ECG derived respiration using
[143] Health and Retirement Study. Accessed: Aug. 30, 2024. [Online]. Avail- complete ensemble empirical mode decomposition with adaptive noise
able: https://hrs.isr.umich.edu/ and principal component analysis,’’ Biomed. Signal Process. Control,
[144] Y. Lin, B. N. Liyanage, Y. Sun, T. Lu, Z. Zhu, Y. Liao, Q. Wang, vol. 77, Aug. 2022, Art. no. 103716.
C. Shi, and W. Yue, ‘‘A deep learning-based model for detecting depres- [163] O. P. Singh, R. Palaniappan, and M. Malarvili, ‘‘Automatic quantitative
sion in senior population,’’ Frontiers Psychiatry, vol. 13, Nov. 2022, analysis of human respired carbon dioxide waveform for asthma and non-
Art. no. 1016676. asthma classification using support vector machine,’’ IEEE Access, vol. 6,
[145] H. V. Nguyen and H. Byeon, ‘‘Explainable deep-learning-based depres- pp. 55245–55256, 2018.
sion modeling of elderly community after COVID-19 pandemic,’’ [164] J. L. M. Amaral, A. J. Lopes, J. Veiga, A. C. D. Faria, and
Mathematics, vol. 10, no. 23, p. 4408, Nov. 2022. P. L. Melo, ‘‘High-accuracy detection of airway obstruction in asthma
[146] R. Kaur, R. W. Motl, R. Sowers, and M. E. Hernandez, ‘‘A vision- using machine learning algorithms and forced oscillation measurements,’’
based framework for predicting multiple sclerosis and Parkinson’s disease Comput. Methods Programs Biomed., vol. 144, pp. 113–125, Jun. 2017.
gait dsfunctions—A deep learning approach,’’ IEEE J. Biomed. Health [165] R. Haba, G. Singer, S. Naftali, M. R. Kramer, and A. Ratnovsky, ‘‘A
Informat., vol. 27, no. 1, pp. 190–201, Jan. 2023. remote and personalised novel approach for monitoring asthma sever-
[147] CMU MoCap Gait Database. Accessed: Aug. 23, 2024. ity levels from EEG signals utilizing classification algorithms,’’ Expert
[Online]. Available: https://www.kaggle.com/datasets/dasmehdixtr/cmu- Syst. Appl., vol. 223, Aug. 2023, Art. no. 119799.
motion-capture-walking-database [166] D. Spathis and P. Vlamos, ‘‘Diagnosing asthma and chronic obstructive
[148] A. Tousignant, Prediction of Future Multiple Sclerosis Disease Progres- pulmonary disease with machine learning,’’ Health Informat. J., vol. 25,
sion Using Deep Learning Analysis of MRI Data. New York, NY, USA: no. 3, pp. 811–827, Sep. 2019.
McGraw-Hill, 2022. [167] S. Susanty, H. Sufriyana, E. C.-Y. Su, and Y.-H. Chuang, ‘‘Questionnaire-
[149] W. Zhu, C. Chen, L. Zhang, T. Hoyt, E. Walker, S. Venkatesh, F. Zhang, free machine-learning method to predict depressive symptoms among
F. Qureshi, J. F. Foley, and Z. Xia, ‘‘Association between serum multi- community-dwelling older adults,’’ PLoS One, vol. 18, no. 1, Jan. 2023,
protein biomarker profile and real-world disability in multiple sclerosis,’’ Art. no. e0280330.
Brain Commun., vol. 6, no. 1, Dec. 2023, Art. no. fcad300. [168] S. D. Pande, S. K. Hasane Ahammad, M. N. Gurav, O. S. Faragallah,
[150] B. M. Meyer, L. J. Tulipani, R. D. Gurchiek, D. A. Allen, L. Adamowicz, M. M. A. Eid, and A. N. Z. Rashed, ‘‘Depression detection based on social
D. Larie, A. J. Solomon, N. Cheney, and R. S. McGinnis, ‘‘Wearables networking sites using data mining,’’ Multimedia Tools Appl., vol. 83,
and deep learning classify fall risk from gait in multiple sclerosis,’’ no. 9, pp. 25951–25967, Aug. 2023.
IEEE J. Biomed. Health Informat., vol. 25, no. 5, pp. 1824–1831, [169] M. Du, M. Li, X. Yu, S. Wang, Y. Wang, W. Yan, Q. Liu, M. Liu,
May 2021. and J. Liu, ‘‘Development and validation of prediction models for poor
[151] R. Kaur, Z. Chen, R. Motl, M. E. Hernandez, and R. Sowers, ‘‘Predicting sleep quality among older adults in the post-COVID-19 pandemic era,’’
multiple sclerosis from gait dynamics using an instrumented treadmill: A Ann. Med., vol. 55, no. 2, Dec. 2023, Art. no. 2285910.
machine learning approach,’’ IEEE Trans. Biomed. Eng., vol. 68, no. 9, [170] Z. Sabouri, N. Gherabi, M. Nasri, M. Amnai, H. E. Massari, and
pp. 2666–2677, Sep. 2021. I. Moustati, ‘‘Prediction of depression via supervised learning models:
[152] B. G. Rajagopal and M. Arock, ‘‘Application of machine Performance comparison and analysis,’’ Int. J. Online Biomed. Eng.,
learning techniques for study of drug interactions using clinical vol. 19, no. 9, pp. 93–107, Jul. 2023.
parameters for creutzfeldt-jakob disease,’’ in Proc. IEEE-EMBS [171] V. Adarsh, P. Arun Kumar, V. Lavanya, and G. R. Gangadharan, ‘‘Fair and
Conf. Biomed. Eng. Sci. (IECBES), Mar. 2021, pp. 415–420. explainable depression detection in social media,’’ Inf. Process. Manage.,
[153] G. Gimhani, A. C. Aponso, and N. Krishnarajah, ‘‘Critical evaluation of vol. 60, no. 1, Jan. 2023, Art. no. 103168.
different biomarkers and machine-learning-based approaches to identify [172] Y. Zhou, X. Yao, W. Han, Y. Li, J. Xue, and Z. Li, ‘‘Measurement
dementia disease in early stages,’’ in Proc. 4th Int. Congr. Inf. Com- of neuropsychiatric symptoms in the older adults with mild cognitive
mun. Technol., vol. 1, 2019, pp. 353–364. impairment based on speech and facial expressions: A cross-sectional
[154] A. Y. Bayahya, W. Alhalabi, and S. H. Alamri, ‘‘Older adults get lost in observational study,’’ Aging Mental Health, vol. 28, no. 5, pp. 828–837,
virtual reality: Visuospatial disorder detection in dementia using a voting May 2024.
approach based on machine learning algorithms,’’ Mathematics, vol. 10, [173] A. Ibrahim, ‘‘Leveraging deep learning models for early detection of
no. 12, p. 1953, Jun. 2022. chronic diseases in healthcare,’’ in Oral Surgery Innovations: Advancing
[155] C. Hercus and A.-R. Hudaib, ‘‘Delirium misdiagnosis risk in psychia- Dental Procedures for Optim. Outcomes. China: Inkbound Publishers,
try: A machine learning-logistic regression predictive algorithm,’’ BMC 2023, p. 143.
Health Services Res., vol. 20, no. 1, pp. 1–7, Dec. 2020. [174] C. Kim, Y. Son, and S. Youm, ‘‘Chronic disease prediction using
[156] N. Fazakis, O. Kocsis, E. Dritsas, S. Alexiou, N. Fakotakis, and K. Mous- character-recurrent neural network in the presence of missing informa-
takas, ‘‘Machine learning tools for long-term type 2 diabetes risk tion,’’ Appl. Sci., vol. 9, no. 10, p. 2170, May 2019.
prediction,’’ IEEE Access, vol. 9, pp. 103737–103757, 2021. [175] S. V. Moravvej, M. J. M. Kahaki, M. S. Sartakhti, and M. Joodaki, ‘‘Effi-
[157] English Longitudinal Study of Ageing. Accessed: Sep. 3, 2024. cient GAN-based method for extractive summarization,’’ J. Electr. Com-
[Online]. Available: https://www.elsa-project.ac.uk/ put. Eng. Innov. (JECEI), vol. 10, no. 2, pp. 287–298, 2022.
[158] C. Kupper, S. Stroth, N. Wolff, F. Hauck, N. Kliewer, [176] M. Zeinolabedini, M. Pazoki, and P. Saeid, ‘‘Optimization of parameters
T. Schad-Hansjosten, I. Kamp-Becker, L. Poustka, V. Roessner, that affect biogas production of anaerobic digestion using a particle
K. Schultebraucks, and S. Roepke, ‘‘Identifying predictive features of swarm algorithm,’’ Int. J. Energy Water Resour., vol. 7, no. 1, pp. 29–41,
autism spectrum disorders in a clinical sample of adolescents and adults Mar. 2023.
using machine learning,’’ Sci. Rep., vol. 10, no. 1, p. 4805, Mar. 2020. [177] P. Saeid, M. Pazoki, and M. Zeinolabedini, ‘‘Optimization of biomass
[159] L. Bishop-Fitzpatrick, A. Movaghar, J. S. Greenberg, D. Page, production from sugar bagasse in anaerobic digestion using genetic
L. S. DaWalt, M. H. Brilliant, and M. R. Mailick, ‘‘Using machine algorithm,’’ Model. Earth Syst. Environ., vol. 9, no. 2, pp. 2183–2198,
learning to identify patterns of lifetime health problems in decedents with Jun. 2023, doi: 10.1007/S40808-022-01598-X.
autism spectrum disorder,’’ Autism Res., vol. 11, no. 8, pp. 1120–1128, [178] H. Shukla, J. L. Mason, and A. Sabyah, ‘‘Identifying genetic mark-
Aug. 2018. ers associated with susceptibility to cardiovascular diseases,’’ in
[160] M. B. Usta, K. Karabekiroglu, B. Sahin, M. Aydin, A. Bozkurt, T. Karaos- Proc. Adv. Med. Biochemistry, Genomics, Physiol., Pathol., 2021,
man, A. Aral, C. Cobanoglu, A. D. Kurt, N. Kesim, I. Sahin, and E. Ürer, pp. 177–200.
‘‘Use of machine learning methods in prediction of short-term outcome [179] N. Peć ina-Šlaus, A. Kafka, I. Salamon, and A. Bukovac, ‘‘Mismatch
in autism spectrum disorders,’’ Psychiatry Clin. Psychopharmacology, repair pathway, genome stability and cancer,’’ Frontiers Mol. Biosciences,
vol. 29, no. 3, pp. 320–325, Jul. 2019. vol. 7, p. 122, Jun. 2020.

VOLUME 12, 2024 122585

 M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

[180] D. J. M. Crouch and W. F. Bodmer, ‘‘Polygenic inheritance, GWAS, [200] F. de Arriba-Pérez, S. García-Méndez, F. J. González-Castaño, and
polygenic risk scores, and the search for functional variants,’’ E. Costa-Montenegro, ‘‘Automatic detection of cognitive impairment
Proc. Nat. Acad. Sci. USA, vol. 117, no. 32, pp. 18924–18933, in elderly people using an entertainment chatbot with natural language
Aug. 2020. processing capabilities,’’ J. Ambient Intell. Humanized Comput., vol. 14,
[181] C. Blauwendraat, M. A. Nalls, and A. B. Singleton AB, ‘‘The genetic no. 12, pp. 16283–16298, Dec. 2023.
architecture of Parkinson’s disease,’’ Lancet Neurol., vol. 19, no. 2, [201] R. H. Somogye, C. J. Roberts, E. Spoerl, K. R. Pillunat, L. E. Pillunat, and
pp. 170–178, 2020. R. H. Small, ‘‘Estimating pulsatile ocular blood volume from intraocular
[182] J. Younger, A. Lo, L. McCormack, J. McGaughran, S. Prasad, and pressure, ocular pulse amplitude, and axial length,’’ PLoS One, vol. 18,
J. J. Atherton, ‘‘Hypertrophic cardiomyopathy: Challenging the status no. 3, Mar. 2023, Art. no. e0283387.
quo?’’ Heart, Lung Circulat., vol. 29, no. 4, pp. 556–565, Apr. 2020. [202] Y. Yang, X. Zhang, Z. Chen, Y. Wei, Q. Ye, Y. Fan, N. Nathwani,
[183] D. Pfeiffer, F. Pfeiffer, and E. Rummeny, ‘‘Advanced X-ray imaging G. Gazzard, and M. Yu, ‘‘Intraocular pressure and diurnal fluctuation of
technology,’’ Mol. Imag. Oncol., vol. 1, pp. 3–30, Aug. 2020. open-angle glaucoma and ocular hypertension: A baseline report from
[184] K. Dikaios, S. Rempel, S. H. Dumpala, S. Oore, M. Kiefte, and R. Uher, the LiGHT China trial cohort,’’ Brit. J. Ophthalmology, vol. 107, no. 6,
‘‘Applications of speech analysis in psychiatry,’’ Harvard Rev. Psychiatry, pp. 823–827, Jun. 2023.
vol. 31, no. 1, pp. 1–13, 2023. [203] Q. Daka, N. Špegel, M. Atanasovska Velkovska, T. Steblovnik, M. Kolko,
[185] D. M. Low, K. H. Bentley, and S. S. Ghosh, ‘‘Automated assessment of B. Neziri, and B. Cvenkel, ‘‘Exploring the relationship between anti-
psychiatric disorders using speech: A systematic review,’’ Laryngoscope VEGF therapy and glaucoma: Implications for management strategies,’’
Investigative Otolaryngology, vol. 5, no. 1, pp. 96–116, Feb. 2020. J. Clin. Med., vol. 12, no. 14, p. 4674, Jul. 2023.
[186] A. K. Schalkamp, K. J. Peall, N. A. Harrison, and C. Sandor, ‘‘Wear- [204] J. J. O. N. van den Bosch, V. Pennisi, K. Mansouri, R. N. Weinreb,
able movement-tracking data identify Parkinson’s disease years before H. Thieme, M. B. Hoffmann, and L. Choritz, ‘‘Effect of eyelid muscle
clinical diagnosis,’’ Nature Medicine, vol. 29, no. 8, pp. 2048–2056, action and rubbing on telemetrically obtained intraocular pressure in
2023. patients with glaucoma with an IOP sensor implant,’’ Brit. J. Ophthal-
[187] Z. Ayaz, S. Naz, N. H. Khan, I. Razzak, and M. Imran, ‘‘Automated mology, vol. 107, no. 10, pp. 1425–1431, Oct. 2023.
methods for diagnosis of Parkinson’s disease and predicting sever- [205] M. Moscalu, R. Moscalu, C. G. Dascau, V. Trca, E. Cojocaru, I. M. Costin,
ity level,’’ Neural Comput. Appl., vol. 35, no. 20, pp. 14499–14534, E. Trca, and I. L. serban, ‘‘Histopathological images analysis and pre-
Jul. 2023. dictive modeling implemented in digital pthology—Current affairs and
[188] S. Farashi, A. Sarihi, M. Ramezani, S. Shahidi, and M. Mazdeh, ‘‘Parkin- perspectives,’’ Diagnostics, vol. 13, no. 14, p. 2379, Jul. 2023.
son’s disease tremor prediction using EEG data analysis—A preliminary [206] A. Palladino, A. Salerno, A. Crasto, C. Lucini, L. Maruccio, L. D’Angelo,
and feasibility study,’’ BMC Neurol., vol. 23, no. 1, p. 420, Nov. 2023. P. A. Netti, P. de Girolamo, A. Cacchioli, C. Attanasio, and F. Ravanetti,
[189] Z. Chen, C. He, P. Zhang, X. Cai, W. Huang, X. Chen, M. Xu, L. Wang, ‘‘Integration of micro-CT and histology data for vasculature morpho-
and Y. Zhang, ‘‘Abnormal cerebellum connectivity patterns related to functional analysis in tissue regeneration,’’ Ann. Anatomy-Anatomischer
motor subtypes of Parkinson’s disease,’’ J. Neural Transmiss., vol. 130, Anzeiger, vol. 245, Jan. 2023, Art. no. 152019.
no. 4, pp. 549–560, Apr. 2023. [207] D. A. Naik, R. M. Mohana, G. Ramu, Y. S. Lalitha, M. SureshKumar,
[190] C. H. F. Camargo, S. A. Ferreira-Peruzzo, D. I. R Ribas, G. L. Franklin, and K. V. Raghavender, ‘‘Analyzing histopathological images by using
and H. A. G. Teive, ‘‘Imbalance and gait impairment in Parkinson’s machine learning techniques,’’ Appl. Nanoscience, vol. 13, no. 3,
disease: Discussing postural instability and ataxia,’’ Neurological Sci., pp. 2507–2513, Mar. 2023.
vol. 1, pp. 1–12, Apr. 2023. [208] M. Cooper, Z. Ji, and R. G. Krishnan, ‘‘Machine learning in com-
[191] M. M. Berger, D. Talwar, and A. Shenkin, ‘‘Pitfalls in the interpretation putational histopathology: Challenges and opportunities,’’ in Genes,
of blood tests used to assess and monitor micronutrient nutrition status,’’ Chromosomes and Cancer. Hoboken, NJ, USA: Wiley, 2023.
Nutrition Clin. Pract., vol. 38, no. 1, pp. 56–69, Feb. 2023. [209] J.-P. Calbimonte, O. Aidonopoulos, F. Dubosson, B. Pocklington,
[192] C. Schulz, B. Kalali, A. Link, M. Gerhard, and P. Malfertheiner, ‘‘New I. Kebets, P.-M. Legris, and M. Schumacher, ‘‘Decentralized seman-
rapid Helicobacter Pylori blood test based on dual detection of FliD and tic provision of personal health streams,’’ J. Web Semantics, vol. 76,
CagA antibodies for on-site testing,’’ Clin. Gastroenterology Hepatology, Apr. 2023, Art. no. 100774.
vol. 21, no. 1, pp. 229–231, Jan. 2023. [210] A. Carnevale, E. A. Tangari, A. Iannone, and E. Sartini, ‘‘Will big data
[193] H. Kaplan, G. Cengiz, S. Sas, and Y. Ö. Eldemir, ‘‘Is the C-reactive and personalized medicine do the gender dimension justice?’’ AI Soc.,
protein-to-albumin ratio the most remarkable simple inflammatory vol. 38, no. 2, pp. 829–841, Apr. 2023.
marker showing active disease in patients with axial spondyloarthri- [211] A. E. W. Johnson, L. Bulgarelli, L. Shen, A. Gayles, A. Shammout,
tis, psoriatic arthritis, and rheumatoid arthritis?’’ Clin. Rheumatology, S. Horng, T. J. Pollard, S. Hao, B. Moody, B. Gow, L.-W.-H. Lehman,
vol. 42, no. 11, pp. 2959–2969, Nov. 2023. L. A. Celi, and R. G. Mark, ‘‘MIMIC-IV, a freely accessible electronic
[194] W. Shi, Y. Jiang, H. Tian, Y. Wang, Y. Zhang, T. Yu, and T. Li, ‘‘C-reactive health record dataset,’’ Sci. Data, vol. 10, no. 1, p. 1, Jan. 2023.
protein-to-albumin ratio (CAR) and C-reactive protein-to-lymphocyte
ratio (CLR) are valuable inflammatory biomarker combination for the
accurate prediction of periprosthetic joint infection,’’ Infection Drug
Resistance, vol. 1, pp. 477–486, Jan. 2023.
[195] H. Altaheri, G. Muhammad, M. Alsulaiman, S. U. Amin, G. A. Altuwaijri,
W. Abdul, M. A. Bencherif, and M. Faisal, ‘‘Deep learning techniques for
classification of electroencephalogram (EEG) motor imagery (MI) sig-
nals: A review,’’ Neural Comput. Appl., vol. 35, no. 20, pp. 14681–14722,
Jul. 2023.
[196] I. Lambert and L. Peter-Derex, ‘‘Spotlight on sleep stage classification
based on EEG,’’ Nature Sci. Sleep, vol. 1, pp. 479–490, Jun. 2023.
[197] J. Dong, Y. Peng, Y. Zhang, Y. Chai, J. Long, Y. Zhang, Y. Zhao, Y. Huang,
and T. Liu, ‘‘Superelastic radiative cooling metafabric for comfortable
epidermal electrophysiological monitoring,’’ Nano-Micro Lett., vol. 15, MAHDIEH POODINEH MOGHADAM is cur-
no. 1, p. 181, Dec. 2023. rently pursuing the Ph.D. degree in nursing with
[198] S. S. Mukerji, K. J. Petersen, K. M. Pohl, R. M. Dastgheyb, H. S. Fox, Zabol University of Medical Sciences, extensive
R. M. Bilder, M.-J. Brouillette, A. L. Gross, L. A. J. Scott-Sheldon, expertise in computer and nursing. She is an
R. H. Paul, and D. Gabuzda, ‘‘Machine learning approaches to under- Esteemed Scholar with Zabol University of Medi-
stand cognitive phenotypes in people with HIV,’’ J. Infectious Diseases, cal Sciences. Her research enriched by her passion
vol. 227, no. 1, pp. S48–S57, Mar. 2023. for artificial intelligence, is focused on ground-
[199] A. K. Bowe, G. Lightbody, A. Staines, and D. M. Murray, ‘‘Big data, breaking applications in nursing.
machine learning, and population health: Predicting cognitive outcomes
in childhood,’’ Pediatric Res., vol. 93, no. 2, pp. 300–307, Jan. 2023.

122586 VOLUME 12, 2024

M. P. Moghadam et al.: Impact of Artificial Intelligence in Nursing for Geriatric Clinical Care

ZABIH ALLAH MOGHADAM is currently pur- PAWEŁ PŁAWIAK was born in Ostrowiec,
suing the Ph.D. degree in computer software Poland, in 1984. He received the B.Eng. and M.Sc.
engineering with Birjand Branch, Islamic Azad degrees in electronics and telecommunications and
University, Birjand, Iran, known for his extensive the Ph.D. degree (Hons.) in biocybernetics and
expertise in computer. He is an Esteemed Scholar biomedical engineering from the AGH University
with Birjand Branch, Islamic Azad University. of Science and Technology, Kraków, Poland, in
He is deeply involved in cutting-edge research, 2012 and 2016, respectively, and the D.Sc. degree
particularly in the areas of machine learning, deep in technical computer science and telecommunica-
learning, and computer sciences. tions from the Silesian University of Technology,
Gliwice, Poland, in 2020. He is currently the Dean
of the Faculty of Computer Science and Telecommunications and an Asso-
ciate Professor with Cracow University of Technology, Kraków, the Deputy
Director of Research with the National Institute of Telecommunications,
Warsaw, and an Associate Professor with the Institute of Theoretical and
Applied Informatics, Polish Academy of Sciences, Gliwice. He has pub-
lished more than 90 articles in refereed international SCI-IF journals. His
research interests include machine learning and computational intelligence
MOHAMMAD REZA CHALAK QAZANI (e.g., artificial neural networks, genetic algorithms, fuzzy systems, sup-
received the B.Eng. degree in manufacturing and port vector machines, k-nearest neighbors, and hybrid systems), ensemble
production from the University of Tabriz, Tabriz, learning, deep learning, evolutionary computation, classification, pattern
Iran, in 2010, the master’s degree in robotic and recognition, signal processing and analysis, data analysis and data mining,
mechanical engineering from Tarbiat Modares sensor techniques, medicine, biocybernetics, biomedical engineering, and
University, Tehran, Iran, in 2013, and the Ph.D. telecommunications. He is an academic editor and a reviewer of many
degree in modeling and simulation of a motion prestigious and reputed journals.
cueing algorithm using prediction and computa-
tional intelligence techniques from the Institute
for Intelligent Systems Research and Innovation ROOHALLAH ALIZADEHSANI (Member, IEEE)
(IISRI), Deakin University, Australia, in 2021. He was an Alfred Deakin received the Bachelor of Science and Master of
Postdoctoral Research Fellow with IISRI for two years working in the areas Science degrees in computer engineering-software
of model predictive control, motion cueing algorithms, and soft comput- from Sharif University of Technology. He is
ing controllers. He is currently an Assistant Professor with the Faculty currently a Research Fellow with Deakin Univer-
of Computing and Information Technology (FoCIT), Sohar University, sity, Australia. His research interests include data
Sohar, Oman. His teaching and research interests include data structure mining, machine learning, bioinformatics, heart
and algorithms, enterprise resource planning modeling and implementation, disease, skin disease, diabetes disease, hepatitis
modeling and visualization, computer architecture, introduction to artificial disease, and cancer disease.
intelligence, and advanced machine learning.

VOLUME 12, 2024 122587