DETECTING DIABETIC RETINOPATHY USING HWT AND CONVOLUTIONAL LENET

Abstract

A common side effect of diabetes mellitus is diabetic retinopathy (DR), which results in retinal lesions that impair vision. Blindness may result if it is not identified in time. Unfortunately, treatment for DR only maintains vision; it is not a reversible process. The risk of vision loss can be considerably decreased with early detection and treatment of DR. In contrast to computer-aided diagnosis systems, ophthalmologists' manual diagnosis of DR retina fundus images is time-consuming, labour-intensive, expensive, and prone to misdiagnosis. Deep learning has recently emerged as one of the most popular methods that has improved performance in many fields, particularly in the analysis and classification of medical images. In medical image analysis, convolutional neural networks are increasingly utilised as a deep learning technique and are extremely