Understanding and trusting AI in clinical settings
Official title: XAI-TRUST: Explainable AI Techniques to Rigorously Understand, Scrutinize, and Trust Clinical AI
Funded at University of Washington · NIH-11063288 · Data updated 2026-11-12
This study is working on making AI tools easier to understand for doctors and researchers, so they can see how things like chest X-ray images influence the AI's predictions, helping everyone trust and use these technologies safely in healthcare.
Quick facts
| Grant type | R01 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of Washington NIH-funded |
| Lab location | 1 site (Seattle, United States) |
| Project start | April 1, 2024 |
| Project end | March 31, 2028 |
| Funding year | 2025 |
| Award (this year) | $445,848 |
| NIH project number | 5R01EB035934-02 |
| Project ID | NIH-11063288 on NIH RePORTER |
Who could benefit from this research
✓Could help
Ideal candidates for this research include patients undergoing chest imaging procedures, particularly those whose diagnoses may be influenced by AI interpretations.
✕Less likely to help
Patients who do not require chest imaging or whose conditions are not assessed using AI tools may not benefit from this research.
What this research studies
This research focuses on developing explainable AI (XAI) techniques to help biomedical researchers and healthcare providers better interpret complex machine learning models used in clinical applications. By analyzing input features, such as chest X-ray images, the project aims to identify which aspects of these images contribute to AI-generated predictions, enhancing transparency and trust in AI systems. The methodology involves addressing current limitations of XAI, such as the complexity of understanding feature attributions and improving the interpretability of AI models. This work is crucial for ensuring that AI tools can be effectively and safely integrated into clinical practice.
Why it matters
Potential benefit: If successful, this research could lead to more reliable and interpretable AI tools in healthcare, improving diagnostic accuracy and patient outcomes.
How similar studies have gone: Other research has shown promise in using explainable AI techniques in clinical settings, indicating that this approach has potential for success.
Where this research is happening
-
University of Washington
Seattle, United States Active
Researchers
- Principal investigator: Lee, Su-in — University of Washington
- Study coordinator: Lee, Su-in
Related research
Other NIH-funded projects on the same topics.
-
Making genetic testing more accessible to the public
This study is all about making it easier for people at risk of certain cancers, like breast, ovarian, and colon cancers, to get genetic testing, so we can find out what helps them access these important services and understand how they can benefit from them.
-
Breast cancer diagnosis in low-resource settings using millifluidics and fast digital imaging in Ghana
A new project will test a fast, low-cost way to diagnose breast cancer at the point of care in Ghana using small-batch fluidics and direct digital imaging.
-
Binaural Hearing and Speech in Noise in Children
This project explores how children hear speech in noisy classrooms using how the two ears work together to see if binaural processing differences explain why some kids struggle more than others.
-
Neural circuits that control pain aversion in the brain
Researchers are exploring how a specific brain area helps people feel and avoid pain, with potential to lead to better pain treatments.
-
United World Antiviral Research Network (UWARN)
A collaborative, international effort is testing new virus surveillance and evolution ideas to help people living with HIV and communities in LMICs stay safer from current and emerging viruses.
-
HF endotypes in heart failure using ethical multi-modal AI and molecular data
A patient-friendly project that uses advanced AI and multiple kinds of health data to identify distinct heart failure subtypes that could guide personalized care.
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.