Detection of Retinal Microaneurysms using Fractal Analysis and Feature Extraction Technique

In this paper a novel method for the improvement in the candidature detection of microaneurysm (MA) in fundus images has been proposed. In automated screening of diabetic retinopathy it becomes necessary to detect the MAs, which appear as small red dots on retinal fundus images to give the earliest possible sign of the disease. The proposed algorithm consists of two stages. The first stage comprises of image preprocessing and fractal analysis of retinal vascular structure. The effectiveness of the automated screening program gets increased as fractal analysis differentiates normal retina image form the abnormal one. Second stage aims at detection of a typical shape of MAs as the abnormal retinal image goes through canny edge detection and morphological reconstruction. True micro-aneurysms are discriminated from other retinal features based on the analysis of binary object parameters on each segmented candidate. The proposed algorithm has been applied on a set of 89 color fundus images from a published database. The implemented algorithm has achieved a best operating sensitivity of 89.5% and a specificity of 82.1% which makes it feasible for diabetic retinopathy screening.