Researchers have developed an AI-based technique that detects previously hidden signs of multiple sclerosis using routine MRI scans. The method reveals brain lesions that conventional MRI often misses, providing a more complete view of disease-related damage linked to disability and cognitive decline.

Multiple sclerosis can cause damaging lesions in the brain's gray matter, especially the cortex, but these injuries are difficult to detect with routine MRI despite being closely linked to disease progression and disability. Researchers at the University at Buffalo have now developed an AI approach that analyzes existing MRI scans to uncover these hidden lesions, revealing disease patterns that conventional imaging methods often miss.

Figure 1. AI Reveals Hidden MS Cortical Lesions

The new AI technique enables researchers to detect previously invisible cortical lesions in standard MRI scans, marking a significant advance in multiple sclerosis research and patient care. By revealing hidden indicators linked to cognitive decline and disability, the method could improve how clinicians monitor disease progression and assess treatment outcomes. Figure 1 shows AI Reveals Hidden MS Cortical Lesions.

Hidden Lesions Complicated MS Diagnosis

Cortical lesions have been recognized as a feature of multiple sclerosis since the disease was first described, but their role in diagnosis remained limited because routine MRI could not reliably detect them. Although postmortem studies have long confirmed extensive gray matter damage, much of this ongoing disease activity has remained invisible in conventional clinical imaging.

Researchers developed an AI-powered imaging approach that uncovers hidden disease signals by analyzing relationships across multiple MRI images rather than relying on a single scan. Using a new technique called multimodal cortical lesion enhancement (MMCLE), the team successfully identified previously unseen cortical lesions in MRI data from more than 700 participants in the Phase III ORATORIO clinical trial of the MS drug Ocrelizumab.

Researchers Identify More Than 11,000 Hidden Cortical Lesions

By applying AI-guided analysis across multiple MRI image types, researchers uncovered a hidden layer of cortical damage that standard scans could not detect [1]. The technique identified around 15–20 cortical lesions per patient—more than 11,000 across the study—by recognizing subtle tissue changes across images, demonstrating the power of AI to reveal disease signals invisible to conventional MRI analysis.

The researchers believe the discovery of widespread hidden brain damage could transform multiple sclerosis research by enabling scientists to reanalyze past clinical trial data and improve the design and evaluation of future studies.

Reference:

1. https://scitechdaily.com/new-technique-exposes-hidden-multiple-sclerosis-damage-in-routine-mri-scans/

Cite this article:

Janani R (2026), New MRI Technique Reveals Hidden Multiple Sclerosis Damage, AnaTechMaz, pp. 1021