DETECTION OF MICROANEURYSMS TO INDICATE THE DIABETIC RETINOPATHY DISEASE

Early detection of diabetic retinopathy is essential to avoid blindness. By detecting the early signs of the disease which is the microaneurysms, the progress of the disease can be better controlled. However, there is lack of ophthalmologists to screen hundreds of retina images. Hence, a computer-aided system is developed to help in detecting the microaneurysms in the fundus images. The objectives of this project is to develop a MATLAB coding to detect the microaneurysms which are the early signs of diabetic retinopathy disease and to evaluate between the retina images having microaneurysms and those of not having microaneurysms. The fundus images are retrieved from the online public database that supplies the fundus images for research purpose. This project will focus on detecting only microaneurysms. Thresholding, shade correction, morphological operation and rule-base classifier have been developed in this project. The experiment results demonstrated that microaneuryms could be detected in the fundus images. However, not all microaneurysms are detected.