Diabetic Retinopathy Detection: EfficientNet with SAP BTP

Diabetic Retinopathy (DR) is the most common diabetic disease that damages the retina. It is the leading cause of blindness globally, and early identification can help people save their sight. However, detecting Diabetic Retinopathy early is a tough task that requires the interpretation of fundus images by clinical experts. In this study we will go through the numerous challenges faced in the detection of DR and various approaches used to tackle them. Comparing the performance of CNN, DNN and BNNs. Furthermore, how SAP Business Technology Platform (BTP) may have the potential to enhance diagnostic accuracy and scalability. Google’s EfficientNet yields in better accuracy of about 98.26% on 5 stage DR classification and 96.70% on legion segmentation and integrating it with BTP platform will make DR detection easy, streamlined, computationally efficient, and fast.