Early detection of Diabetic Retinopathy from retinal fundus images using Eigen value analysis

摘要

Diabetic Retinopathy is the frequent cause of blindness. Diabetic retinopathy is a complication of diabetes mellitus. The longer one is suffering with diabetes the more is the probability of diabetic retinopathy(DR). DR leads to several abnormalities like Micro aneurisms(MA); hard exudates (HE), hemorrhages (Hem) and Cotton wool spots(CWS). MA is first symptom of DR. the detection of MA is important for the early detection of DR. We have proposed MA detection based on Eigen value analysis using hessian matrix in retinal images. The objective of the proposed work is, to develop an algorithm for improved MA detection. The algorithm employed Eigen value analysis from Hessian matrix to detect MAs. MA particle analysis is employed to measure the area of MA in retinal image. When the method was evaluated on visible MAs using 89 images from the diaretDB0 database, the true positive rate was 91% with eight false positive images. Lab VIEW software is used to implement the algorithm.