Automatic lesion detection of diabetic retinopathy from color retinal photographs

Automatic detection of microaneurysms in digital color fundus images is still an open issue in medical image processing. In this paper, it propose a method to improve microaneurysm detection. Unlike the well-known approach of considering the output of multiple classifiers, it propose a combination of internal components of microaneurysm detectors such as preprocessing methods and candidate extractors. Since microaneurysm detection is decisive in diabetic retinopathy (DR) grading, also tested the proposed method for this task on the publicly available Messidor database, where a promising result is achieved in a “DR/non-DR”-type classification based on the presence or absence of the microaneurysms.