Fair combined medical-data models for autoimmune diseases
Unified and fair multimodal representation learning for autoimmune diseases
A new computer model will learn from medical records, tests, scans, genetics, and wearable data to help spot autoimmune diseases earlier in people with symptoms or at higher risk.
Quick facts
| Grant type | NIH-funded research |
|---|---|
| Study type | NIH-funded research |
| Funding institution | Univ of North Carolina Chapel Hill NIH-funded |
| Lab location | 1 site (Chapel Hill, United States) |
| Project ID | NIH-11401314 on NIH RePORTER |
What this research studies
This project will build a single model that brings together many kinds of patient information—doctors' notes, lab tests, imaging, genomics, and activity-tracker data—so care can consider the whole person. The model is designed to accept new types of data over time and to handle missing information by predicting unavailable data from what is present. Developers will add fairness controls so the tool works better for groups often under-represented in medical data, such as Black and Hispanic women who face higher lupus risk. The model will aim to give clear, patient-specific predictions and recommendations about which tests or data would be most useful next.
Who could benefit from this research
Good fit: People with symptoms suggestive of autoimmune disease, those already diagnosed with an autoimmune condition, or patients willing to share medical records, test results, imaging, or wearable data are the best fit.
Not a fit: Those without accessible digital health records, relevant clinical tests, imaging, or wearable data—or people with conditions unrelated to autoimmunity—are unlikely to benefit directly in the short term.
Why it matters
Potential benefit: Could help people get diagnosed sooner and guide more personalized testing and follow-up, potentially reducing long-term organ damage from delayed diagnosis.
How similar studies have performed: Tools using single data types like EHRs or imaging have shown promise for earlier detection, but fully unified multimodal models with built-in fairness and cross-modal generation are relatively new and less tested.
Where this research is happening
Chapel Hill, United States
- Univ of North Carolina Chapel Hill — Chapel Hill, United States (Active)
Researchers
- Principal investigator: Sheikh, Saira Z — Univ of North Carolina Chapel Hill
- Study coordinator: Sheikh, Saira Z
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 below.