AI Reveals Previously Invisible Cortical Lesions on Conventional MRI Scans, Transforming MS Research and Clinical Care
Key Insights
A UB-led team developed AI-based image processing methods that detect previously invisible gray matter (cortical) lesions on conventional legacy MRI scans, publishing findings in Communications Medicine.
Applying the new multimodal cortical lesion enhancement (MMCLE) technique to over 700 participants from the phase III ORATORIO trial of Ocrelizumab revealed 15–20 cortical lesions per patient, totaling more than 11,000 across the dataset.
Cortical lesions are key indicators of MS disease progression and cognitive impairment but have been undetectable on standard clinical MRI, frustrating clinicians for decades.
A University at Buffalo-led international research team has harnessed artificial intelligence to solve a decades-old problem in multiple sclerosis (search): the inability to detect gray matter lesions on conventional MRI scans. Published in Communications Medicine, the study demonstrates how AI-based image processing can reveal cortical lesions that have long been known to drive disease progression and cognitive decline but remained invisible to clinicians.
The advance centers on a new technique called multimodal cortical lesion enhancement (MMCLE), which combines multiple image-processing approaches to extract information from the relationships between different MRI contrast images — data that cannot be discerned from any single scan.
"Detecting previously invisible cortical lesions on conventional legacy MRI scans has major implications for MS research and clinical care," said Robert Zivadinov, MD, PhD, senior author on the paper, SUNY Distinguished Professor in the Department of Neurology and director of the Buffalo Neuroimaging Analysis Center (BNAC) in the Jacobs School of Medicine and Biomedical Sciences at UB. "The ability to see for the first time these previously hidden indicators of MS disease progression, including cognitive impairment and disability, is an important advance."
A Long-Standing Blind Spot in MS Imaging
The involvement of cortical lesions in MS has been recognized since the late 19th century, yet they were not included in diagnostic criteria until the 21st century — and even then, their clinical utility was acknowledged to be severely limited by the capabilities of standard MRI. While many new drugs developed over the past decade can significantly slow disease progression, they primarily work by reducing white matter lesions, leaving the gray matter pathology largely unaddressed and unmonitored.
"We have all been very frustrated, knowing that these cortical lesions were there but not being able to see them," said Michael G. Dwyer, PhD, first and corresponding author on the paper, associate professor of neurology and biomedical informatics at the Jacobs School and a researcher with BNAC. "There's a lot of ongoing damage that continues to happen in MS that you won't see with conventional MRI, but that histopathologists have been clearly demonstrating for decades on postmortem tissue."
How AI Uncovers the Invisible
The AI approaches, which build on prior work from co-authors in the Netherlands, were designed to extrapolate vital information from the relationships between multiple images. The researchers applied MMCLE and other techniques to MRI scans from the large, phase III FDA regulatory ORATORIO clinical trial, a study of the MS drug Ocrelizumab that included more than 700 participants.
The results were striking. While individual images of a patient's brain revealed mostly white matter lesions, the AI-based methods uncovered anywhere from 15 to 20 cortical lesions per patient — more than 11,000 across the entire dataset.
"If you look on the original scans, you generally can't see the cortical lesions," Dwyer explained, "but generative AI is very powerful because it can look between the scans and detect tiny differences between them. 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 them, and synthesize what had been missing."
Implications for Past and Future Trials
The international research team included scientists and clinicians from academia and industry, including Genentech, which manufactures Ocrelizumab and provided partial support for the research. Zivadinov emphasized that the collaborative breadth was key to the breakthrough.
"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," Zivadinov said.
Dwyer described the achievement as "a real success story for applying AI in the medical arena," adding: "We now have access to these incredibly useful data on MRI scans that were there but you couldn't see them without using AI to pull them out. The computational methods are finally at the point where we can do this."
In addition to Zivadinov and Dwyer, UB co-authors include Niels P. Bergsland, PhD, assistant professor of neurology; Alexander Bartnik, PhD, postdoctoral researcher; and Dejan Jakimovski, MD, PhD, research adviser at BNAC. Other co-authors are Samantha Noteboom, Menno M. Schoonheim, and Martijn D. Steenwijk of the MS Center Amsterdam, Vrije Universiteit Amsterdam, and Jinglan Pei and David Clayton of Genentech Inc.