"A new AI foundation model has been developed that can accurately extract multiple disease risk signals from routine brain MRIs, including: estimating a person's "brain age"; predicting dementia risk; detecting brain tumor mutations; and predicting survival from brain cancer, according to investigators from Harvard-affiliated Mass General Brigham. The model, a brain imaging adaptive core called BrainIAC, was trained on nearly 49,000 brain MRI scans. The tool outperformed other, more task-specific AI models, and was especially efficient when limited training data were available."
"According to researchers, despite recent advances in medical AI approaches, there is a lack of publicly available models that focus on broad, brain MRI analysis. Most conventional frameworks perform specific tasks and require extensive training with large, annotated datasets that can be hard to obtain. Furthermore, brain MRI images from different institutions can vary in appearance and based on their intended applications (such as in neurology versus oncology care), making it challenging for AI frameworks to learn similar information from them."
BrainIAC is a brain imaging adaptive core trained on nearly 49,000 brain MRI scans to extract multiple disease risk signals, including brain age estimation, dementia risk prediction, tumor mutation detection, and brain cancer survival prediction. The foundation model outperformed task-specific AI models and remained especially efficient when training data were limited. BrainIAC employs self-supervised learning to discover inherent features from unlabeled datasets and adapts those features across applications. Validation used 48,965 diverse brain MRI scans across seven tasks of varying clinical complexity. The model generalized across healthy and abnormal images and addressed variability in MRI appearance across institutions and clinical use cases. Results were published in Nature Neuroscience.
Read at Harvard Gazette
Unable to calculate read time
Collection
[
|
...
]