AI-based screening for diabetic eye diseases launched

Metro India News | AMARAVATI

In a significant step towards integrating artificial intelligence (AI) into public healthcare, the Andhra Pradesh government has decided to pilot AI-powered screening for diabetic retinopathy, a major vision-threatening complication among diabetes patients. The initiative will be implemented in government hospitals at Guntur, Kurnool and Visakhapatnam, Health Minister Satyakumar Yadav announced on Wednesday.

The pilot project aims to facilitate early detection of retinal damage among diabetic patients by using AI to analyze images captured through fundus cameras. The technology can identify the presence and severity of diabetic retinopathy and indicate whether a patient requires immediate specialist consultation, helping bridge the gap caused by a shortage of retina specialists.

Currently, diagnosis depends on the availability of both fundus cameras and ophthalmologists. Under the new system, trained optometrists can upload retinal images to an AI platform, which generates a preliminary assessment and helps determine the urgency of treatment. Officials believe this will enable earlier intervention and prevent avoidable vision loss.

According to the Health Department, nearly 20 percent of diabetes patients develop eye-related complications such as blurred vision, retinal damage and, in severe cases, retinal detachment. At Guntur Government General Hospital alone, around 3,000 diabetic patients visit every month, with 15-20 percent showing signs of retinopathy.

The AI solution, named “Madhunetra,” has been developed by the Wadhwani Foundation in collaboration with AIIMS institutions under the Union Health Ministry. The system has already been trained using data from nearly 13,000 diabetic retinopathy patients screened at AIIMS Delhi and AIIMS Rishikesh and is currently being used in several government medical institutions across India.

Screening began at Guntur GGH on Wednesday and will be launched in Kurnool and Visakhapatnam later this month. The three-month pilot aims to screen around 9,000 patients.