Home▸AI & Machine Learning▸Inside AI-Assisted Diabetic Retinopathy Screening

👁️ Inside AI-Assisted Diabetic Retinopathy Screening

An interactive 3D fundus camera and retina model that shows how a CNN scans a retinal photograph, highlights lesions with a Grad-CAM heatmap, and outputs an ICDR severity grade.

AI & Machine Learning3DAdvanced60 FPS

A 3D fundus camera captures a retinal photograph, a Grad-CAM heatmap highlights the lesions driving the model's decision, and a monitor panel reports the predicted ICDR severity grade and referral pathway.

🔬 What It Demonstrates

How lesion density (microaneurysms, haemorrhages, exudates, neovascularisation) on a retinal photograph maps onto ICDR severity grades, and how Grad-CAM explains which regions a CNN attended to when grading.

🎮 How to Use

Pick a severity grade, fade the Grad-CAM overlay, toggle CLAHE preprocessing, and click "Run AI screening" to watch a simulated inference pass sweep the retina and update the grading panel.

💡 Did You Know?

Diabetic retinopathy is the leading cause of preventable blindness in working-age adults in the UK, which is why the NHS Diabetic Eye Screening Programme photographs every registered diabetic patient annually.

⚙ Under the hood

An interactive 3D fundus camera and retina model that shows how a CNN scans a retinal photograph, highlights lesions with a Grad-CAM heatmap, and outputs an ICDR severity grade.

artificial intelligencedeep learningmedical imagingretinacnnicdrdiagnosticsThree.js

3D · Three.js / WebGL renderer · 60 FPS target · runs fully client-side, no install