Using AI across clinics to read eye scans for diabetic eye disease
Official title: Distributed foundational models for multi-task learning in diabetic retinopathy
Funded at University of North Carolina Charlotte · NIH-11208613 · Data updated 2026-11-12
This project trains AI on eye scans held at different hospitals to help doctors detect, predict, and track diabetic retinopathy in people with diabetes.
Quick facts
| Grant type | R01 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of North Carolina Charlotte NIH-funded |
| Lab location | 1 site (Charlotte, United States) |
| Project start | May 1, 2026 |
| Project end | March 31, 2030 |
| Funding year | 2026 |
| Award (this year) | $617,244 |
| NIH project number | 1R01EY037828-01 |
| Project ID | NIH-11208613 on NIH RePORTER |
Who could benefit from this research
✓Could help
People with diabetes who receive OCT/OCTA imaging at one of the participating eye clinics, particularly those at risk for or showing signs of diabetic retinopathy, are the ideal candidates.
✕Less likely to help
Patients who do not get OCT/OCTA imaging, who have other unrelated eye conditions, or who receive care outside participating centers may not see direct benefit from this project.
What this research studies
The team will train large AI models using OCT, OCTA, and enface images collected at multiple clinics while keeping your raw data on site through privacy-preserving federated learning. The models are designed to learn shared features that support several tasks at once—classifying disease stage, predicting future worsening, and estimating treatment outcomes. Participating sites send model updates rather than patient images, which helps protect privacy and builds a more diverse dataset. The goal is AI that works for people from many backgrounds and can be applied to different clinical questions about diabetic eye disease.
Why it matters
Potential benefit: If successful, this could lead to earlier detection, better prediction of who will worsen, and more informed treatment decisions to help prevent vision loss from diabetic retinopathy.
How similar studies have gone: AI using retinal images has shown promise for diabetic retinopathy detection, but federated, multi-task models using OCT/OCTA data are relatively new and less tested in routine clinical care.
Where this research is happening
-
University of North Carolina Charlotte
Charlotte, United States Active
Researchers
- Principal investigator: Alam, Minhaj Nur — University of North Carolina Charlotte
- Study coordinator: Alam, Minhaj Nur
Related research
Other NIH-funded projects on the same topics.
-
Integrative multi-omics to understand diabetic retinopathy
A doctor-scientist team will study blood markers and genetic signals to see why diabetic retinopathy happens and how it affects Hispanic/Latino people.
-
Smartphone retinal imaging for diabetic eye disease in Egypt
This project uses a low-cost smartphone camera and AI to find signs of diabetic eye damage in adults with diabetes in Egypt.
-
Retinal Cell Loss in Diabetic Retinopathy
This project looks at how mitochondrial problems and a specific kind of cell death called ferroptosis may cause retinal cells to die in people with diabetes-related eye disease.
-
Redox-triggered REDD1 blockers for diabetic retinopathy
A research project tests new ways to stop a stress protein (REDD1) from causing vision problems in people with diabetic retinopathy.
-
Ultrafast eye imaging to detect and track diabetic retinopathy
A doctor-science team is developing ultra-fast eye scans to better see tiny blood vessels in people with diabetic retinopathy, helping catch changes early and guide treatment.
-
Practical device to automatically track abnormal retinal blood vessel growth in diabetic eye disease
A new, easier-to-use imaging device aims to automatically detect and monitor abnormal retinal blood vessel growth in people with diabetic retinopathy.
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page.