Detection of diabetic severity through dnn algorithm using fundus image

Abstract The diabetic disease is referred to as diabetic retinopathy (DR) which is diagnosed by the ophthalmologist by obtaining the patient’s retinal image through a fundus camera. Diabetic neural network (DNN) may be a prominent technique which is widely utilized in the different biomedical signal analysis because it doesn’t only model the signal but also choose to classify the signal as normal or abnormal. An efficient procedure for automatic detection of diabetic retinopathy has an important impact on further treatment of the patients to avoid vision loss. The proposed research work is aimed to style a classifier system for the diagnosis of diabetic retinopathy using retinal images by a prominent method called Prediction Neural Networks. The classifier system described during this research is developed for the mass screening of patients associated with diabetic retinopathy. This classifier system is developed by using different neural networks like Diabetic Neural Network Stage 1 and Diabetic Neural Network Stage 2.