Postmortem validation of MRI cortical volume measurements in MS

Area of Science:

• Neuroimaging
• Neuropathology
• Medical image analysis

Background:

• Grey matter (GM) atrophy is a key indicator of multiple sclerosis (MS) pathology.
• Accurate GM segmentation is crucial for measuring atrophy in MS studies.
• Validating in vivo GM measurement techniques is challenging.

Purpose of the Study:

• To evaluate the correlation between postmortem MRI measurements of cortical volume and thickness with histological measurements.
• To determine the validity of different postprocessing techniques for GM atrophy assessment in MS.
• To assess the impact of manual editing on automated segmentation accuracy.

Main Methods:

• Postmortem MRI (3DT1-weighted) was performed on 16 MS brains.
• Regional cortical volume was measured using FSL-SIENAX, FreeSurfer, and SPM.
• Regional cortical thickness was measured using FreeSurfer and validated against histological measurements.
• Linear regression analyses were used to compare MRI and histological data.

Main Results:

• Significant correlations were found between histological cortical thickness and FSL-SIENAX cortical volume with manual editing (P=0.005).
• Significant correlations were also observed for FreeSurfer cortical volume (P=0.003) and cortical thickness (P=0.003) with manual editing.
• Automated segmentation methods showed no significant correlation, highlighting the importance of manual editing.

Conclusions:

• FSL-SIENAX and FreeSurfer, with manual editing, are supported for assessing GM atrophy in severe MS.
• Manual editing significantly improves the accuracy of automated GM segmentation techniques.
• This study provides a method for validating in vivo MRI measurements using postmortem data.