MUMBAI, India, Oct. 5 -- Intellectual Property India has published a patent application (202641114706 A) filed by Jayaraju Gaddala on September 24, 2026, for 1. Title Of The Invention Edge-Computing System And Method For Explainable Grading Of Diabetic Retinopathy From Fundus Images.
Inventors include Dr P Indira Priyadarsini; Dr. G. Joel Sunny Deol; Dr. Satyakeerthigorripati; Dr. Ch. Sita Kumari; Ch. Samsonu; and Ramya Asalatha Busi.
The application for the patent was published on October 02, 2026, under issue no. 40/2026.
Abstract: The invention relates to computer-aided grading of fundus images. A fundus camera (100) captures a retinal image and sends it to a single-board edge-computing device (120), such as a Raspberry Pi 5, which checks image quality (110), preprocesses the image (130) and classifies it with a lightweight convolutional neural network (140), for example EfficientNet-Lite0, into one of five diabetic retinopathy grades. A grading module (150) selects the predicted grade and a confidence-output module (160) reports its probability. A Grad-CAM explainability module (170) generates a heatmap of the image regions that contributed to the prediction and overlays it on the fundus image. The grade, confidence value and heatmap are shown on a display (180). Because inference runs on the device, images need not be sent to a remote server, which suits screening sites with limited connectivity.
Disclaimer: Curated by HT Syndication.
