Comparison Different Vessel Segmentation Methods in Automated Microaneurysms Detection in Retinal Images using Convolutional Neural Networks

TL;DR

This study compares vessel segmentation methods in retinal images to improve microaneurysm detection for diabetic retinopathy, demonstrating high sensitivity across different techniques and datasets.

Contribution

It introduces a comparative analysis of LoG, Canny, and Matched filter segmentation methods combined with CNNs for microaneurysm detection in retinal images.

Findings

LoG achieved over 85% sensitivity on local and public datasets.

Canny segmentation achieved over 80% sensitivity on all images.

Matched filter achieved over 87% sensitivity across datasets.

Abstract

Image processing techniques provide important assistance to physicians and relieve their workload in different tasks. In particular, identifying objects of interest such as lesions and anatomical structures from the image is a challenging and iterative process that can be done by computerized approaches in a successful manner. Microaneurysms (MAs) detection is a crucial step in retinal image analysis algorithms. The goal of MAs detection is to find the progress and at last identification of diabetic retinopathy (DR) in the retinal images. The objective of this study is to apply three retinal vessel segmentation methods, Laplacian-of-Gaussian (LoG), Canny edge detector, and Matched filter to compare results of MAs detection using a combination of unsupervised and supervised learning either in the normal images or in the presence of DR. The steps for the algorithm are as following: 1) Preprocessing and Enhancement, 2) vessel segmentation and masking, 3) MAs detection and Localization using a combination of Matching based approach and Convolutional Neural Networks. To evaluate the accuracy of our proposed method, we compared the output of our method with the ground truth that collected by ophthalmologists. By using the LoG vessel segmentation, our algorithm found a sensitivity of more than 85% in the detection of MAs for 100 color images in a local retinal database and 40 images of a public dataset (DRIVE). For the Canny vessel segmentation, our automated algorithm found a sensitivity of more than 80% in the detection of MAs for all 140 images of two databases. And lastly, using the Matched filter, our algorithm found a sensitivity of more than 87% in the detection of MAs in both local and DRIVE datasets.