New AI Eye Scans Can Detect Diabetes Without Blood Tests

New AI Eye Scan Developed by India and US Researchers Detects Diabetes Without

Using Blood Tests

A groundbreaking collaboration between researchers in India and the United States has led to the development of an artificial intelligence system capable of detecting diabetes through simple eye scans. By analyzing high-resolution images of the retina, this AI technology offers a non-invasive alternative to traditional blood tests and fasting requirements.

Revolutionizing Diabetes Screening

The study, published in Diabetes Technology and Therapeutics, highlights how machine learning can identify microscopic changes in retinal blood vessels. These subtle patterns, often invisible to the human eye, allow the AI to distinguish between healthy individuals and those with diabetes or prediabetes.

The research team—including experts from Chennai, Emory University, and Yenepoya University—analyzed over 200 vessel features from 139 participants. The results were impressive: the AI achieved a 95% sensitivity rate in identifying the disease.

Why This Matters

For countries like India, which is home to over 100 million people living with diabetes, early detection is a major challenge. Many individuals remain undiagnosed until complications arise. Dr. V. Mohan, a lead researcher on the project, noted that this tool could revolutionize real-time screening because it removes the need for needles, fasting, or expensive lab equipment.

While the technology shows immense potential for rural and underserved areas, researchers emphasize that larger, more diverse clinical trials are needed before a global rollout. If successful, this "instant" eye-based checkup could become a standard part of preventative healthcare, catching diabetes in its earliest, most reversible stages.

Key Takeaways

• Non-Invasive Detection: The AI uses retinal photographs instead of blood samples, eliminating the need for needles or fasting.

• High Accuracy: The system identifies diabetes with 95% sensitivity by spotting vessel changes invisible to doctors.

• Early Intervention: The tool can detect prediabetes, allowing patients to make lifestyle changes before the condition worsens.

Adapted From

India New England