New AI tool predicts brain age, dementia risk, cancer survival — Harvard Gazette
New AI tool predicts brain age, dementia risk, cancer survival — Harvard Gazette
Publish Date: 2026-02-05 05:00:00
Source Domain: news.harvard.edu
-
New AI Model Introduced: A new foundation model named BrainIAC has been developed to extract multiple disease risk signals from routine brain MRI scans.
-
Versatile Application: BrainIAC can estimate brain age, predict dementia risk, detect brain tumor mutations, and predict survival rates from brain cancer.
-
Training and Performance: The model was trained on nearly 49,000 brain MRI scans and outperformed other more task-specific AI models.
-
Self-Supervised Learning: Utilizes a self-supervised learning approach to identify inherent features from unlabeled datasets, enabling it to adapt to various tasks efficiently.
-
Efficiency with Limited Data: BrainIAC particularly excelled in scenarios where training data was limited or task complexity was high.
-
Diverse Validation Tasks: After pretraining on multiple imaging datasets, BrainIAC was validated on 48,965 brain MRI scans for seven different clinical tasks.
-
Potential Impact on Medicine: The model has the potential to accelerate biomarker discovery, enhance diagnostic tools, and integrate AI more broadly into clinical practice.
-
Funding and Support: The study received support from the National Institutes of Health/the National Cancer Institute and Botha-Chan Low Grade Glioma Consortium.
