Harnessing AI for Accurate Diabetic Retinopathy Detection in Retinal Imaging

Iqbal, Muhammad Adeel (2025) Harnessing AI for Accurate Diabetic Retinopathy Detection in Retinal Imaging. Masters thesis, Dublin, National College of Ireland.

Preview	PDF (Master of Science) Download (1MB) | Preview
Preview	PDF (Configuration Manual) Download (1MB) | Preview

Abstract

Diabetic retinopathy (DR) has been one of the major causes of preventable vision loss globally, and early-stage diagnosis plays a fundamental role in effective screening. However, micro-aneurysms as indicators of the pathological state are usually missed in traditional screening systems. This study proposes a unified deep learning framework that combines the DR classification and performs the lesion-level segmentation at the same architecture to enhance the diagnostic accuracy and interpretability. A ResNet34-based encoder-decoder model, trained on the e-ophtha dataset made publicly available, is used to recognize the occurrence of DR and to provide lesion probability maps highlighting the localization of pathological areas simultaneously. The accuracy of the classification component was 0.9868, precision was 1.0000, Recall was 0.9655, F1-score was 0.9825, PR-AUC was 0.9873, and ROC-AUC was also 0.9875. The segmentation module achieved a Dice coefficient and Intersection-over-Union of 0.9173 and 0.8545, which indicate high precision in lesion localization of the segmentation module. The findings show the value of integrating classification and segmentation in early detection of DR, which gives clinicians predictive results as well as visual interpretations.

Item Type:	Thesis (Masters)
Supervisors:	Name Email Shahid, Abdul UNSPECIFIED
Uncontrolled Keywords:	Diabetic Retinopathy; Early Detection; Lesion Segmentation; Deep Learning; Unified Framework
Subjects:	R Medicine > RE Ophthalmology Q Science > QH Natural history > QH301 Biology > Methods of research. Technique. Experimental biology > Data processing. Bioinformatics > Artificial intelligence Q Science > Q Science (General) > Self-organizing systems. Conscious automata > Artificial intelligence R Medicine > Healthcare Industry
Divisions:	School of Computing > Master of Science in Artificial Intelligence
Depositing User:	Ciara O'Brien
Date Deposited:	20 Aug 2026 09:40
Last Modified:	20 Aug 2026 09:40
URI:	https://norma.ncirl.ie/id/eprint/9560

Actions (login required)

View Item