Texture Feature Analysis of Digital Fundus Images for Early Detection of Diabetic Retinopathy

Abstract

Diabetic retinopathy (DR) is a complication where the retina of a diabetic patient is damaged due to fluid leakage from the blood vessels into the retina and the patient may suffer from complete blindness if untreated. Hemorrhages and Microaneurysms (HMAs) are the early signs that appear in retina at the initial stage of DR. Early diagnosis of HMAs is crucial to prevent blindness and fundus image is used for this purpose. We have focused on the analysis of texture micro-patterns of the regions of interest (ROIs), which are suspicious regions in a fundus image, for the detection of HMAs. Texture micro-structures of ROIs are analyzed through Local Binary Pattern (LBP) for their description. Finally Support Vector Machine (SVM) is employed to identify whether an ROI contains HMAs or not.