CLASSIFICATION AND IDENTIFICATION OF DIABETIC RETINOPATHY SEVERITY STAGES IN THAI PATIENTS USING DEEP LEARNING BASED CONVOLUTION NEURAL NETWORKS

摘要

Diabetic retinopathy (DR) is one of the complications caused by long-term diabetes, have an effect on vision quality of the patient until the level of permanent vision loss. This research has developed an automation method for screening and identifying the severity of diabetic retinopathy in Thai patients by using image processing principle together with deep learning (DL) technique based convolution neural network (CNN). The purposed method has been learned with 9,900 color fundus retinal images and taken to testing with 3,300 sets of DR images. The result showed 99.36% of accuracy level and 99.48% of sensitivity level when compared with the diagnosis of an ophthalmologist.