DRDr: Automatic Masking of Exudates and Microaneurysms Caused By Diabetic Retinopathy Using Mask R-CNN and Transfer Learning

TL;DR

This paper presents an automated method using Mask R-CNN and transfer learning to accurately segment and detect exudates and microaneurysms in fundus images for diabetic retinopathy diagnosis.

Contribution

It introduces a tailored Mask R-CNN approach with data augmentation and transfer learning to improve lesion detection in small datasets of fundus images.

Findings

Achieved a test mAP of 0.45 for lesion detection.

Effectively used transfer learning with ResNet101.

Enhanced small lesion detection through data augmentation.

Abstract

This paper addresses the problem of identifying two main types of lesions - Exudates and Microaneurysms - caused by Diabetic Retinopathy (DR) in the eyes of diabetic patients. We make use of Convolutional Neural Networks (CNNs) and Transfer Learning to locate and generate high-quality segmentation mask for each instance of the lesion that can be found in the patients' fundus images. We create our normalized database out of e-ophtha EX and e-ophtha MA and tweak Mask R-CNN to detect small lesions. Moreover, we employ data augmentation and the pre-trained weights of ResNet101 to compensate for our small dataset. Our model achieves promising test mAP of 0.45, altogether showing that it can aid clinicians and ophthalmologist in the process of detecting and treating the infamous DR.