https://github.com/mariam-zaidi/monkey-pox_detection-thesis

M-Pox detection using deep learning models and eXplainable AI
https://github.com/mariam-zaidi/monkey-pox_detection-thesis

cnn deep-learning explainable-ai layerwiserelevancepropogation medical-imaging multi-class-classification smote transfer-learning

Last synced: 24 days ago
JSON representation

M-Pox detection using deep learning models and eXplainable AI

• Host: GitHub
• URL: https://github.com/mariam-zaidi/monkey-pox_detection-thesis
• Owner: Mariam-Zaidi
• Created: 2024-09-16T03:50:09.000Z (5 months ago)
• Default Branch: main
• Last Pushed: 2024-09-16T06:35:30.000Z (5 months ago)
• Last Synced: 2024-11-19T10:12:51.162Z (3 months ago)
• Topics: cnn, deep-learning, explainable-ai, layerwiserelevancepropogation, medical-imaging, multi-class-classification, smote, transfer-learning
• Language: Jupyter Notebook
• Homepage:
• Size: 14.2 MB
• Stars: 0
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# Monkey-Pox_detection-Thesis

M-Pox detection using deep learning models and eXplainable AI

The rapid escalation of the Monkeypox Virus (MPV) has raised concerns about its potential to 

become a global pandemic, especially in the wake of the COVID-19 pandemic. Therefore, to have appropriate protocols and countermeasures

in place, to prevent another international outbreak is of paramount importance. Since Deep 

learning techniques have shown promising results in detecting and predicting COVID-19 cases, 

expecting to show similar results with Monkeypox diagnosis. This study investigates the use of 

Deep Neural Networks (DNNs) to diagnose monkeypox from skin lesion images, leveraging 

Transfer Learning to overcome data scarcity. The performance of the developed models was evaluated using four metrics. Our results 

consistently showed that DenseNet121 outperformed the other models in terms of all metrics 

when pre-trained with ImageNet dataset showed 88.9% accuracy and with HAM10000 dataset showed 75.9% accuracy on test data. While different studies may 

have outperformed others in specific evaluation metrics, our modified DenseNet201 exhibited 

superior performance in terms of recall and F1-Score for both source datasets.

Additionally, the application of explainable AI 

techniques, such as LRP, will be explored to be able to explain the model's predictions.

Explainable AI (XAI) plays a significant role in healthcare by improving the transparency, 

interpretability, and trustworthiness of AI models, leading to better decision-making and patient 

diagnosis.

Link to video presentation : https://drive.google.com/drive/folders/1__gGoFGY7VhYL_IGG8xfmvZEhUADRxCW?usp=sharing