AI Sees Invisible Brain Lesions in MS Patients

Artificial intelligence is poised to revolutionize the diagnosis and tracking of multiple sclerosis. A novel AI-driven process, developed by researchers at the University at Buffalo, has demonstrated the ability to detect previously unseen brain lesions within the gray matter. These lesions are critical indicators of MS progression but are undetectable by conventional MRI scans.

The new AI technology functions by meticulously comparing individual patient MRI scans. It identifies subtle differences, revealing minute discrepancies that signify tissue damage. This generative AI approach has already identified an average of 15 to 20 previously hidden lesions per patient in a recent study.

This advancement holds significant implications for both MS research and clinical care. Senior researcher Dr. Robert Zivadinov highlighted that the ability to see these hidden indicators of disease progression, including cognitive impairment and disability, is a major step forward. For the first time, researchers can now assess MS drug effectiveness based on their impact on gray matter lesions, not just white matter lesions.

The research, supported in part by Genentech, utilized MRI scans from an MS drug trial involving over 700 patients. The findings suggest a profound impact on reviewing past clinical trial data and guiding future research, potentially accelerating the development of more effective treatments for multiple sclerosis.