A Closer Look at the Potential of AI Foundation Models for Brain MRI

Could an emerging AI foundation model offer an array of diagnostic and prognostic applications for brain magnetic resonance imaging (MRI)?

In a new study, recently published in Nature Neuroscience, researchers developed and compared the foundation model Brain Imaging Adaptive Core (BrainIAC) with the publicly available in-domain AI foundation models MedicalNet, BrainSegFounder and Scratch for a variety of applications including brain age prediction, isocitrate dehydrogenase (IDH) mutation detection and time-to-stroke prediction. Leveraging self-supervised learning, the BrainIAC model was trained and validated on 48,965 brain MRI scans, according to the study.

In a recent interview with Diagnostic Imaging, study co-author Benjamin Kann, M.D., discussed the impetus behind the research and the potential of AI foundation models to optimize the amount of data obtained from brain MRI scans.

“We’re getting millions of scans on these patients every year, and there's a lot of data that's being left on the table,” noted Dr. Kann, an associate professor of radiation oncology at Harvard Medical School.

“ … We felt that, as we're seeing with large language models rolling out many foundation models, there could really be a role for vision-based foundation models in the major biomedical imaging modalities such as MRI, so you could sort of teach the model to have a baseline foundational knowledge about how a brain MRI structure looks, and then take that knowledge and, with a lot fewer cases of labeled data, be able to develop a powerful model that can adapt to many different situations.”

The study authors found that the brainIAC model had the lowest mean absolute error (MAE) between predicted age and chronological age (6.55 years), the lowest MAE in time-to-stroke prediction (38.87 days) and the highest AUC (79 percent) for IDH mutation prediction in patients with low-grade gliomas.

“ … As far as clinical impact, I think when we talk about things like mutation, prediction of a brain tumor, or the prediction of the development of dementia or stroke, these are really powerful pieces of information that I think we're going to be able to provide patients just from their routine brain MRI, and really uncover a lot of data and a lot of clinical insight from these MRIs that we just could not unlock before. I think that models like BrainIAC and other foundation models are just going to lower the barrier to do that in a variety of scenarios,” maintained Dr. Kann, a faculty member of the Artificial Intelligence in Medicine Program at Mass General Brigham and Harvard Medical School.

(Editor’s note: For related content, see “Updated MRI-Based AI Software Offers Automated Segmentation and Volumetric Reporting of Brain Metastases and Meningiomas,” “MRI-Derived Fat Quantification and Neurologic Impacts: What Emerging Research Reveals” and “FDA Clears MRI-Based AI Software for Assessment of Brain Metastases.”)