A deep learning approach based on stochastic gradient descent and least absolute shrinkage and selection operator for identifying diabetic retinopathy

More than eighty-five to ninety percentage of the diabetic patients are affected with retinopathy (DR) which is an eye disorder that leads blindness. The computational techniques can support detect DR by using retinal images. However, it hard measure raw image. This paper proposes effective method for identification from In this research work, initially Weiner filter used preprocessing Then preprocessed image segmented fuzzy c-mean technique. image, features extracted grey level co-occurrence matrix (GLCM). After extracting fundus feature selection performed stochastic gradient descent, and least absolute shrinkage operator (LASSO) accurate during classification process. inception v3-convolutional neural network (IV3-CNN) model in process classify as or non-DR By applying proposed method, performance IV3-CNN identifying studied. Using identified accuracy about 95%, processed mild DR.