Ensemble Diabetic Retinopathy Detection in 2-D Color Fundus Retina Scan

Diabetic Retinopathy (DR) is a burgeoning malady in Asian territories. DR causes 5–7% of the total blindness throughout the region. The main aim of this research is to determine whether a patient is suffering from DR or not by the dint of 2-D color fundus retina scans. In this paper, authors have proposed a GUI-based technique an Ensemble Diabetic Retinopathy Detection (EDRD). This method helps in detecting 2D color fundus scans with an efficient approach for finding DR-affected persons within a few seconds using ensemble techniques of CNN and RLU. The visual geometry group (VGG-19) model and adaptive moment estimation optimizer are used for training and reducing error for the developed technique. A maximum accuracy of 92% was obtained for an 80% training set with a 0.001 learning rate and 25 batch size. The proposed research contribution definitively detects whether the given OCT scan with an efficient approach for finding DR-affected persons within a few seconds.