NEURADERM :A SPECTRUM

OF APPROACHES FOR SKIN

DISEASE DETECTION AND
CLASSIFICATION
 CONTENTS
❏ INTRODUCTION

❏ PROBLEM DEFINITION

❏ OBJECTIVE

❏ LITERATURE SURVEY

❏ PROPOSED SYSTEM

❏ PROJECT SCHEDULE AND TASK ALLOCATION

❏ REFERENCES
 INTRODUCTION

• Skin diseases represent a significant global health burden, affecting millions of people worldwide.

• The World Health Organization (WHO) estimates that skin diseases affect up to 20-30% of the global
population at any given time.

• Skin conditions range from minor irritations to severe and potentially life-threatening diseases.

• Timely detection of skin diseases is crucial for effective treatment and management.

• Early intervention can prevent complications, improve patient outcomes, and reduce healthcare costs.

• Skin cancer, particularly melanoma, is a prime example where early detection can be life-saving.
PROBLEM DEFINITION

• The existing healthcare system faces barriers such as:

⮚Limited access to dermatologists

⮚Delayed appointments
⮚Difficulty of monitoring disease progression

• Individuals encounter difficulties in accessing dermatological expertise, leading to delayed diagnoses

and suboptimal self-care practices.
OBJECTIVE

Integration with Real-Time Skin Disease

Healthcare Professionals Identification
Scalability Comprehensive Disease
Recognition

Data Security and

Personalized Medication Privacy
Recommendations Research and
Progress Tracking Continuous
Improvement
 Literature Survey
TITLE METHOD ADVANTAGES DISADVANTAGES

A Deep Learning Approach Based on Explainable ResNet-18,LIME Framework Powerful deep neural networks, Complexity, susceptibility to overfiiting, and
Artificial Intelligence for Skin Lesion Classification individual prediction limited interpretability
Natasha Nigar, et al

Skin cancer diagnosis using convolutional neural networks Alex-ne,MobileNet- high performance , ability to Large model size, highly dependent on data set
for smartphone images V2 ,Resnet50 ,Data learn complex features quality
Sara Medhat, et al. Augmentation

Detection and Classification of Skin Cancer by Parallel CNN Detection rate is quite more Only 9 types of skin cancer can be classified
Using a Parallel CNN Model compared tom other models
Noortaz Rezaona,et al.

A Comprehensive Joint Learning System to Detect Skin CNN ,LBP Prevents misclassification, CNN not be able to detect other types of skin
Cancer + LBP gives high accuracy lesions,merging of CNN & LBP needs
Lubna Riaz, et al. dimensionality reduction algorithm

Melanoma Skin Cancer Detection Using Support Vector SVM, CNN Accurate Detection Non- Limited Dataset, Limited Hardware
Machines and Convolutional Neural Networks invasive Detection, Dataset and Specification, Lack of Real-Time
Christopher Ebuka Ojukwu Hardware Standardisation Implementation

Skin cancer classification using Convolutional neural Image preprocessing, Image Use of Convolutional Neural Local Optimum,Dataset Limitations,
networks augmentation, Convolution Networks(CNN), High Computational Requirements, Lack of
R Raja Subramanian, et al. Accuracy, Visualization, Data Generalizability Assessment
Preprocessing
PROPOSED SYSTEM
ARCHITECTURE
System Architecture
Data Set
(Images)

Train Dataset Test Dataset

Trained
Model

Model Training
(CNN)

Prediction

OUTPUT
 MODULE IDENTIFICATION

User Module Admin Module Doctor Module

• Login
• Login/Sign Up • Login/Sign Up
• User Management
• Upload Image • Online Consultation
• Data Security
• View Result • Add Doctor • Doctor Location

• Products Recommendations • Machine Learning Model Management • Patient History

• Data Preprocessing
• Skin Care Tips
• Model Training

• Model Testing

• Prediction
 USER MODULE

1. Login/Sign Up
• This component allows users to create an account or log in if they already have one.
• Registration typically involves providing basic information such as name, email, and password.
• Once logged in, users can access the various features and functionalities of the application.
2. Upload Image
• The uploaded images serve as input data for the skin disease detection system.
• Users may be prompted to provide additional information about the skin condition or symptoms
associated with the uploaded image.
 USER MODULE

3. View Result
• After uploading an image, users can view the results of the skin disease detection analysis.
• The system may provide information about the detected skin condition, including potential
diagnoses or recommendations.
4. Skin Care Tips
• The application may feature a section dedicated to providing users with helpful skincare tips and
advice.
• Tips may cover various aspects of skincare, including cleansing routines, moisturizing
techniques, sun protection, and lifestyle habits.
• Users can access informative articles, videos, or tutorials aimed at promoting healthy skin
practices and addressing common skincare concerns.
 ADMIN MODULE

1. Login
• The Admin Module includes a login functionality that allows administrators to securely access the
system.
• Authentication mechanisms ensure that only authorized individuals can log in to the admin
dashboard.
2. User Management
• Admins can manage user accounts, including creating, updating, or deleting user profiles.
• This component allows admins to oversee user activity, track usage metrics, and address any user-
related issues or inquiries.
3. Data Security
• Ensuring the security and integrity of user data is paramount. Admins are responsible for
implementing robust data security measures.
 ADMIN MODULE

1. Add Doctor
• Admins have the authority to add new doctors or healthcare professionals to the system.
• This involves creating doctor profiles, assigning permissions, and providing access to relevant system
functionalities.
2. Machine Learning Model Management
• Admins oversee the management of machine learning models used for skin disease detection.
• This includes deploying, monitoring, updating, and optimizing ML models to ensure their effectiveness and
reliability.
• These are the following steps taken under ML modelling:
1. Data Preprocessing:Prior to model training, data preprocessing is performed to clean, normalize, and transform
input data.
2. Model Training:This involves selecting appropriate algorithms, tuning hyperparameters, and optimizing model
performance.
3. Model Testing:After training, admins conduct rigorous testing to evaluate the performance of ML models.
Dataset
The Dataset is from kaggle
It consist of 25,780 photographs
Class available in the datasets are:
1. Actinic keratosis
2. Basal cell carcinoma
3. Dermatofibroma
4. Melanoma
5. Nevus
6. Pigmented benign keratosis
7. Seborrheic keratosis
8. Squamous cell carcinoma
9. Vascular lesion
 DOCTOR MODULE

1.Login/Sign up
• Doctors can log in to their accounts using secure authentication mechanisms.
• New doctors may sign up for an account by providing necessary credentials and verifying their identity.
2.Online Consultation
• This component enables doctors to conduct online consultations with patients remotely.
• Doctors can interact with patients through video calls, audio calls, or text-based messaging to discuss
skin concerns, provide advice, and recommend treatments.
3.Doctor Location
• Doctors can update information about their current location or view the location of their clinic/hospital.
4.Patient history
• Doctors have access to patients' medical history and previous consultations.
• This component allows doctors to review past diagnoses, treatments, medications, and any relevant
information to inform current consultations and decision-making.
 CONCLUSION

● Meticulously incorporated advanced image preprocessing, feature extraction, and deep learning

techniques to elevate the accuracy of skin disease predictions.

● The deployment of deep neural networks resulted in notably improved precision in identifying skin

diseases and accurately forecasting disease conditions.

● This research underscores the transformative potential of AI in effectively distinguishing diseases.

 REFERENCES

• [1] C.Karimkhani, R. P.Dellavalle, L. E. Coffeng, C. Flohr, R. J. Hay, S. M. Langan, E. O. Nsoesie, A. J. Ferrari, H. E. Erskine, J. I.
Silverberg et al., “Global skin disease morbidity and mortality: an update from the global burden of disease study 2013,” JAMA
dermatology, vol. 153, no. 5, pp. 406–412, 2017.
• [2] T. Gansler, P. A. Ganz, M. Grant, F. L. Greene, P. Johnstone, M. Mahoney, L. A. Newman, W. K. Oh, C. R. Thomas Jr, M. J.
Thun et al., “Sixty years of ca: a cancer journal for clinicians,” CA: a cancer journal for clinicians, vol. 60, no. 6, pp. 345–350, 2010.
• [3] M. Akter, M. S. Hossain, T. Uddin Ahmed, and K. Andersson, “Mosquito classication using convolutional neural network with
data augmentation,” in 3rd International Conference on Intelligent Computing & Optimization 2020, ICO 2020, 2020.
• [4] G. Kasinathan, S. Jayakumar, A. H. Gandomi, M. Ramachandran, S. J. Fong, and R. Patan, “Automated 3-d lung tumor detection
and classification by an active contour model and cnn classifier,” Expert Systems with Applications, vol. 134, pp. 112–119, 2019.
• [5] Z. Gao, L. Wang, L. Zhou, and J. Zhang, “Hep-2 cell image classification with deep convolutional neural networks,” IEEE
journal of biomedical and health informatics, vol. 21, no. 2, pp. 416–428, 2016.
• [6] P. Wang, L. Wang, Y. Li, Q. Song, S. Lv, and X. Hu, “Automatic cell nuclei segmentation and classification of cervical pap
smear images,” Biomedical Signal Processing and Control, vol. 48, pp. 93–103, 2019.
• [7] K. M. Hosny, M. A. Kassem, and M. M. Foaud, “Classification of skin lesions using transfer learning and augmentation with
alex-net,” PloS one, vol. 14, no. 5, p. e0217293, 2019.
• [8] X. He, Z. Yu, T. Wang, B. Lei, and Y. Shi, “Dense deconvolution net: Multi path fusion and dense deconvolution for high
resolution skin lesion segmentation,” Technology and Health Care, vol. 26, no. S1, pp. 307– 316, 2018.
THANK YOU