AIIMS Unveils MadhuNETrAI, India’s 1st AI App for Diabetic Retinopathy

Following regulatory approval, AIIMS plans to roll out the app nationwide, prioritizing primary and district-level health facilities.

The All India Institute of Medical Sciences (AIIMS), in collaboration with the Union Health Ministry’s e-health division and Wadhwani AI, has launched MadhuNETrAI, India’s first fully validated artificial intelligence-based mobile application to detect diabetic retinopathy (DR).

With this initiative, AIIMS aims to supply these devices to rural and underserved centers to improve access to timely diabetic eye screening and help reduce preventable vision loss in India.

The app leverages an easy-to-operate AI screening model that analyzes retinal images and identifies early indicators of diabetic retinopathy within seconds.

This will enable primary and district-level healthcare centers to screen high-risk patients even without an ophthalmologist on-site.

Commenting on the significance of the tool, Dr. Rohan Chawla, retina specialist and professor at AIIMS’s RP Centre, said, “While many AI systems for diabetic retinopathy are being built, the key issue is validation. Most are tested on just 90 or 120 images, which isn't enough for clinical confidence. Our model was tested on over 3,000-4,000 retinal images. We now have over 95% detection accuracy.”

The app uses images taken via a fundus camera to categorize scans into normal, mild, and moderate-to-severe DR.

“Patients with mild changes don’t need referral, while moderate or severe cases are flagged for specialist attention,” Dr. Chawla added.

As per reports, the team is working on making MadhuNETrAI compatible with cost-effective fundus cameras and even smartphones, with functionality expected without internet connectivity.

Reportedly, the AI tool is still under evaluation for CDSCO approval. Following regulatory approval, AIIMS plans to roll out the app nationwide, prioritizing primary and district-level health facilities.

While MadhuNETrAI will be free to use, the primary expense lies in acquiring a compatible fundus camera, which costs approximately ₹3 lakh

Stay tuned for more such updates on Digital Health News