Accuracy Enhancement of Diabetic Retinopathy Detection Using Naive Bayes Algorithm

Abstract

Diabetic retinopathy (DR) is a disease of eye which is caused by diabetes. Sometime the DR leads the diabetic patients to complete vision loss. In this scenario, early identification of DR is more essential to protect the eyesight and provide help for timely treatment. The detection of DR can be done manually by ophthalmologists and can also be done by an automated system. An ophthalmologist is required to analyze and explain retinal fundus images in the manual system, which is a time consuming and very expensive task. While, In the automated system, artificial intelligence is used to perform an significant role in the area of ophthalmology and specifically in the early detection of DR over the traditional detection approaches. Recently, numerous advanced studies related to the identification of DR have been reported, But still research for accurate detection of DR is going on. In this paper, a new diabetic retinopathy monitoring model is proposed by using the Naive Bayes method to improve the accuracy of detection of DR. The model is trained on mixture of two datasets Messidor and Kaggle, and evaluated on the Messidor dataset. By using proposed method detection accuracy is found to be higher than existing methods.