Spotting MS Lesions That MRIs Miss

One of the unfortunate truths about multiple sclerosis is that the part of the brain likely to reveal the most about a patient’s disease progression and cognitive impairment has been largely invisible to clinicians.

It’s long been known, from analysis of postmortem brain tissue, that gray matter lesions (also called cortical lesions) play a key role in MS. However, because magnetic resonance imaging has only been able to detect lesions in white matter, there was no way to monitor cortical lesions in a living patient. And while many new drugs developed in the past decade can slow disease progression significantly, they primarily work on reducing white matter lesions.

Now, in a paper published in Communications Medicine, a UB-led team reports that it can use AI to review existing scans in a way that reveals the full picture.

The AI approaches used by the researchers extrapolated vital information from the relationships between multiple images. When the researchers applied these techniques to MRI scans from the ORATORIO clinical trial, a study of the MS drug ocrelizumab that included more than 700 participants, they were able to see anywhere from 15 to 20 cortical lesions for each patient, amounting to more than 11,000 for the whole dataset.

“Generative AI is very powerful because it can look between the scans and detect tiny differences between them,” explains Michael Dwyer, associate professor of neurology and biomedical informatics at the University at Buffalo, and first and corresponding author on the paper. “Because it sees those minor discrepancies, AI can reveal that there’s something going wrong there, that the tissue is not behaving like healthy tissue. The trained models can view multiple MRI images together and synthesize what had been missing.”

The significance of this development, say researchers, cannot be overstated. In the words of Robert Zivadinov, SUNY Distinguished Professor in the Department of Neurology and senior author on the paper, “This work, which has revealed that there is so much invisible pathology in the brain, will have tremendous impact for reviewing data from past clinical trials and also for those going forward.”