Automated Binary Classification of Diabetic Retinopathy by SWIN Transformer
Rasha Ali Dihin, Ebtesam N. AlShemmary, Waleed A. Mahmoud Al‐Jawher
Abstract
Diabetic retinopathy is a medical condition that affects the eyes and is caused by damage to the blood vessels in the retina (the light-sensitive part of the eye) due to high blood sugar levels in individuals with diabetes. This damage can lead to vision loss or even blindness. It is a common complication of diabetes and a leading cause of blindness in working-age adults. In this paper, to automatically classify images of the retina as having either diabetic retinopathy or not. The goal of this classification is to assist medical professionals in diagnosing diabetic retinopathy more accurately and efficiently, potentially improving patient outcomes. In this process, the Swin transformer model is trained on the APTOS dataset of retinal images and then used to automatically classify new images as either positive or negative for diabetic retinopathy. Used CLAHE and Gaussian, to improve the input image, and the model achieved a Test Accuracy of 96%, Sensitivity of 96%, F1 Score of 96% for Swin-T and Test Accuracy of 98% for Swin-B, Sensitivity of 98%, and F1 Score of 98%.