Participatory AI for skin disease: editable multimodal modeling
Official title: A Model Editing Framework for Participatory Multimodal AI in Dermatology
Funded at University of California Berkeley · NIH-11573186 · Data updated 2026-11-12
A UC Berkeley project will let dermatologists directly guide and correct a multimodal AI system for skin disease using easy natural-language edits.
Quick facts
| Grant type | NIH-funded research |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of California Berkeley NIH-funded |
| Lab location | 1 site (Berkeley, United States) |
| Project start | September 8, 2025 |
| Project end | September 7, 2027 |
| Funding year | 2026 |
| Award (this year) | $958,748 |
| NIH project number | 3OT2OD038079-01S1 |
| Project ID | NIH-11573186 on NIH RePORTER |
Who could benefit from this research
✓Could help
Ideal candidates are individuals with skin conditions evaluated by dermatologists who may benefit from improved diagnostic tools and AI-assisted decision support.
✕Less likely to help
Patients with skin conditions outside the scope of dermatology or those not interacting with AI-assisted diagnosis may not directly benefit.
What this research studies
This project aims to create an editable multimodal AI model for dermatology that allows doctors to intervene and correct the system’s decisions using simple language. It will develop new methods to edit model behavior across images, dermoscopy, pathology, and text data, and will build datasets to support participatory design. The effort emphasizes transparency, clinician involvement, and accountability to reduce black-box failures in real-world practice. The work plans to demonstrate how dermatologist feedback can steer AI recommendations in more accurate and reliable ways.
Why it matters
Potential benefit: If successful, patients could benefit from more accurate computer-aided skin diagnoses and an accountable AI that clinicians can steer in real time.
How similar studies have gone: Prior studies show that clinician-in-the-loop AI can improve performance and trust, though editable MAI in dermatology is an emerging area.
Where this research is happening
-
University of California Berkeley
Berkeley, United States Active
Researchers
- Principal investigator: Alaa, Ahmed — University of California Berkeley
- Study coordinator: Alaa, Ahmed
Related research
Other NIH-funded projects on the same topics.
-
Mobile AI-powered microscope to quickly spot parasites in wastewater
A new AI-enabled mobile microscope will be tested to quickly detect parasite eggs in wastewater from refugee settings to help protect public health.
-
SCAN: Standardized brain imaging for Alzheimer’s and related dementias
A nationwide effort to standardize brain imaging data in people with Alzheimer’s and related dementias so researchers can compare results and learn how to diagnose and track the disease.
-
Cerebellum’s role in thinking and behavior across tasks
This grant explores how the cerebellum helps people think, feel, and behave across different tasks, with a focus on understanding cognitive and affective processing in cerebellar disorders.
-
Genetic clues to how teeth form and replace themselves in sticklebacks
A study in a fish model aims to uncover the genetic factors that control how teeth form and keep renewing, with potential implications for human tooth regeneration.
-
California Abundant Birth Project: does guaranteed income during pregnancy affect birth outcomes and child development?
A government-backed program gives pregnant people in California monthly income to reduce stress, with researchers studying how this helps birth outcomes and early child development.
-
American-Asian Arbovirus Research and Enhanced Surveillance Center
A2CARES is a national program that brings together scientists across disciplines to improve detection, study, and response to arthropod-borne viruses affecting adults in the Americas and Asia.
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.