AI to make 3D mammogram screening more accurate
Artificial Intelligence for Improved Breast Cancer Screening Accuracy: External Validation, Refinement, and Clinical Translation
This project uses artificial intelligence to help 3D mammograms find breast cancer more reliably and cut down on false alarms.
Quick facts
| Grant type | R37 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of Washington NIH-funded |
| Lab location | 1 site (Seattle, United States) |
| Project ID | NIH-11322059 on NIH RePORTER |
What this research studies
The team is taking a high-performing AI model originally developed for 2D mammograms and adapting it to read 3D (tomosynthesis) screening images. They will train and refine the algorithm using large collections of de-identified mammogram images and linked clinical data from many sites and then externally validate its performance on separate 3D datasets. An academic–industry partnership with DeepHealth supports scaling and translating the tool toward routine clinical use. The work includes testing in real-world settings to measure effects on missed cancers, recall rates, and how the AI could fit into current screening workflows.
Who could benefit from this research
Good fit: People who get routine screening mammograms—especially those offered 3D tomosynthesis—are the ones who could benefit or be asked to provide de-identified images for validation.
Not a fit: People who do not undergo mammography screening, such as those without breast tissue or who are not eligible for routine screening, would not directly benefit from this project.
Why it matters
Potential benefit: If successful, this could detect more cancers earlier and reduce the number of false-positive callbacks and unnecessary follow-up tests from screening.
How similar studies have performed: Previous AI work and the DREAM crowdsourced challenge showed promise for 2D mammogram algorithms, but translating and validating those tools for 3D screening is newer and less tested.
Where this research is happening
Seattle, United States
- University of Washington — Seattle, United States (Active)
Researchers
- Principal investigator: Lee, Christoph I — University of Washington
- Study coordinator: Lee, Christoph I
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.