A Neural Network and Machine Learning-Based Approach for Diabetic Retinopathy Detection and Classification
摘要
Chronic diabetes can cause diabetic retinopathy, resulting in blindness. The increased severity of diabetic retinopathy can be prevented by early detection. Diabetic retinopathy can be detected quickly by an automated system to determine the need for follow-up treatment to prevent further retinal damage. In this study, an effective, robust, and accurate automatic methodology for detecting DR subjects will be developed. It is based on two steps: (1) Reconstruction and enhancement of blood vessels using customised programs; (2) the use of an Artificial Neural Network as a classifier to distinguish people with diabetes with and without diabetic retinopathy (DR) from those with mild to moderate non-proliferative diabetic retinopathy (NPDR).