Abstract

Diabetic Retinopathy (DR) is progressive dysfunction of the retinal blood vessels caused by chronic hyperglycemia which can be a complication of diabetes type 1 or diabetes type 2. Initially, DR is asymptomatic, if not treated though it can cause low vision and blindness. Diabetic retinopathy is responsible for 1.8 million of the 37 million cases of blindness throughout the world. So the early detection of Diabetic retinopathy through proper screening is essential. The paper presents a Diabetic Retinopathy Screening System which can be used as a primary diagnosis tool by ophthalmologists in the screening process to detect symptoms of Diabetic Retinopathy. The system uses the anatomical structures such as blood vessels, exudates and microaneurysms in retinal images. The retinal images are segmented and classified as normal or DR affected images by extracting features from segmented images and the Gray Level Cooccurrence Matrix (GLCM). The classifier used is Support Vector Machine (SVM) which gives a better accuracy. The system is implemented and tested in MATLAB and LabView for the standard database and need to be optimized for real time screening of images. LabView creates distributable .EXE files and .DLL files which can be downloaded into the FPGA/DSP processor. Hardware implementation on LabView FPGA presents a small learning curve which drastically reduces development time and eliminates the need for custom hardware design.

Authors and Affiliations