Automated Computer Aided Detection of Diabetic Retinopathy Using Machine Learning Hybrid Model

Diabetic retinopathy is a potentially fatal condition that affects diabetics worldwide, resulting in blurred vision or total blindness. A technique for identifying diabetic retinopathy using the fundus image obtained from the patient's retina is proposed in this paper. The method entails processing a digital image of the fundus image, which assists the ophthalmologist in examining DR. A neural network was utilized to diagnose a micro-aneurysm, a type of diabetic retinopathy that is the first stage. A comparison was made between the proposed Support Vector Machine and the existing Naive Bayes classifier. For experimental validation, the programed MATLAB/SIMULINK is employed. The preprocess image was used as input data for pattern recognition using a neural network. There has been a significant improvement in terms of sensitivity, specificity, and accuracy when compared to the aforementioned existing techniques.