AI Model Offers Earlier Dementia Warnings & Brain Cancer Detection

The model, called Brain Imaging Adaptive Core (BrainIAC), has been trained on nearly 49,000 brain MRIs and is designed to extract clinically relevant information that is not typically captured in standard imaging assessments.

Researchers at Mass General Brigham have developed a new AI model that analyzes brain MRI scans to predict dementia risk, assess brain ageing, and improve detection of brain cancer mutations, potentially enabling earlier intervention and more personalized treatment decisions.

The model, called Brain Imaging Adaptive Core (BrainIAC), has been trained on nearly 49,000 brain MRIs and is designed to extract clinically relevant information that is not typically captured in standard imaging assessments.

Brain MRI is one of the most commonly used tools in neurology and oncology, yet its clinical interpretation is usually limited to answering a specific question, such as identifying a tumour or visible structural abnormality. Large volumes of data embedded in these scans remain underutilized.

Advances in AI and computational imaging are now making it possible to analyze these datasets at scale, uncovering patterns related to disease risk, progression, and outcomes.

BrainIAC was developed by researchers at Mass General Brigham as a foundation model trained using self-supervised learning, meaning it learned from large quantities of unlabeled MRI data.

“BrainIAC is an AI foundation model that is trained on tens of thousands of brain MRI scans to understand how the brain is structured,” said Benjamin Kann, MD, faculty member of the Artificial Intelligence in Medicine (AIM) Program at Mass General Brigham, and corresponding author of the study. “Using this core baseline knowledge, the tool can then be adapted to identify various brain diseases, determine their severity, and predict future risks from these diseases.”

In the study, published in Nature Neuroscience, BrainIAC was shown to estimate brain age, predict dementia risk, detect tumour-related genetic mutations, and forecast survival outcomes in brain cancer patients.

“Identification of these problems will inform clinicians and patients what type of treatment or preventive measures should be taken to reduce future risk, ultimately improving quality of life and survival,” Kann explained.

The model also demonstrated strong performance when limited labelled data was available, outperforming more narrowly trained AI systems. This is significant in healthcare, where well-annotated imaging datasets are often scarce.

BrainIAC has been released as open-source software for research purposes, allowing institutions to adapt it for local use without requiring thousands of labelled scans.

Researchers plan to further refine BrainIAC and expand its applications to additional brain diseases, reinforcing the growing role of AI in dementia research, brain cancer care, and digital health more broadly.

Stay tuned for more such updates on Digital Health News