An automated quantitative report for multiple sclerosis using only 3D T2-fluid-attenuated inversion recovery MRI

Area of Science:

• Neuroimaging
• Radiology
• Biomarker Discovery

Background:

• Automated tools for multiple sclerosis (MS) imaging biomarkers often require specialized MRI sequences and lack comprehensive reference data.
• This limits the routine clinical application and interpretation of quantitative imaging metrics in MS patient care.

Purpose of the Study:

• To develop an open-source quantitative report (QReport) integrating validated 3D T2-FLAIR quantification methods with multi-center MS and healthy reference models.
• To present quantitative imaging biomarkers in a structured graphical report for improved contextualized interpretation.

Main Methods:

• Retrospective collection of 2516 3D T2-FLAIR scans from 1723 healthy controls (HC) and 793 people with MS (pwMS) across 14 centers.
• Application of validated T2-FLAIR-based algorithms to quantify lesion count (LC), lesion volume (LV), brain volume (BV), and BrainAGE.
• Development of a QReport to display single-subject measures against normative distributions (5th, 50th, 95th percentiles) for pwMS and HC.

Main Results:

• The QReport successfully contextualized quantitative imaging biomarkers, including BrainAGE, against reference distributions.
• In evaluations by four neuroradiologists, QReport outputs showed Moderate-to-Complete agreement with visual assessment in 94% of cases.
• The QReport was rated as useful in 82% of the evaluated cases, demonstrating its clinical utility.

Conclusions:

• An MS-specific QReport was developed, requiring only 3D T2-FLAIR MRI.
• The QReport integrates validated quantification algorithms and BrainAGE within a clinically interpretable framework.
• This tool supports the contextualized interpretation of MS imaging biomarkers using readily available MRI sequences.