View Proposal


Proposer
Medhdhar Salem Ali Muthanna Al- Gaashani
Title
Medical Image Classification for Kidney Stone (CT) and Retinal Disease (OCT) Detection
Goal
Description
This project develops deep learning models for classifying medical images using public Kidney Stone CT and Retinal OCT datasets. Students preprocess and augment the images, train and compare transfer-learning models such as MobileNetV2, EfficientNet-B0, or ResNet, and evaluate them using accuracy, precision, recall, F1-score, MCC, and confusion matrices. Model efficiency can also be compared using parameter count and FLOPs. Grad- CAM may be added for explainability, and a simple Streamlit or Flask demo can display predictions and heatmaps.
Resources
Kidney Stone CT dataset (Yildirim et al.): https://github.com/yildirimozal/Kidney_stone_detection — Retinal OCT dataset (Kermany et al.): https://www.kaggle.com/datasets/paultimothymooney/kermany2018
Background
Url
Difficulty Level
Easy
Ethical Approval
None
Number Of Students
1
Supervisor
Medhdhar Salem Ali Muthanna Al- Gaashani
Keywords
Degrees
Bachelor of Science in Computational Sciences and Software Engineering