Cortical Thin Patch Fraction Reflects Disease Burden in MS: The Mosaic Approach

Area of Science:

• Neuroimaging
• Biomarker Development
• Multiple Sclerosis Research

Background:

• Gray matter (GM) pathology significantly impacts multiple sclerosis (MS) disability progression.
• Neuroradiologic biomarkers are crucial for understanding the diverse cortical disease burden in MS.
• Cortical atrophy assessment requires methods to capture localized pathology.

Purpose of the Study:

• To validate the "mosaic approach" for identifying GM pathology through patch-wise cortical thickness analysis.
• To assess the utility of the mosaic approach's output, the thin patch fraction, as a biomarker.
• To evaluate the mosaic approach's performance in discriminating MS patients from controls and classifying MS phenotypes.

Main Methods:

• Investigated the mosaic approach in 501 MS patients and 651 controls (internal and external).
• Calculated the thin patch fraction as a measure of overall cortical disease burden.
• Employed general linear modeling to assess discrimination, classification, and association with established MS biomarkers.

Main Results:

• The thin patch fraction significantly differed between MS patients and healthy controls.
• The thin patch fraction successfully discriminated among different MS phenotypes.
• Thin patch fraction correlated with disease severity metrics like Expanded Disability Status Scale, cognitive/fatigue scores, and lesion volume.

Conclusions:

• The mosaic approach is a valid neuroradiologic biomarker for assessing and localizing cortical GM pathology in MS.
• The thin patch fraction serves as a candidate biomarker for personalized MS assessment.
• The mosaic approach's validity is confirmed across diverse imaging datasets, supporting AI-driven biomarker development.