Estimating cerebral atrophy in multiple sclerosis patients from various MR pulse sequences

Area of Science:

• Neuroimaging
• Medical Physics
• Neurology

Background:

• Cerebral atrophy (CA) is a key indicator of disease progression in multiple sclerosis (MS).
• Accurate quantification of CA is essential for monitoring treatment efficacy in clinical trials.
• Standardization of magnetic resonance imaging (MRI) acquisition and analysis techniques is needed for reliable CA assessment.

Purpose of the Study:

• To investigate the impact of different pulse sequences (PS) and segmentation algorithms (SA) on cerebral atrophy (CA) measurements in MS patients and healthy controls (HCs).
• To evaluate the variability in brain fractional volume (BFV) changes due to varying MRI parameters.

Main Methods:

• MRI scans from 10 relapsing-remitting MS (RRMS) patients and 5 HCs were analyzed over two years.
• T1-weighted, FLAIR, and PD/T2-weighted sequences were used.
• Brain volume segmentation was performed using histogram SA, adaptive fuzzy c-means (AFCM), and adaptive Bayesian with K-means clustering.

Main Results:

• Significant differences in percent change of brain fractional volume (BFV) were observed in MS patients based on combinations of SA and PS.
• BFV changes in MS patients ranged from +2.05% to -1.6% in year one and +0.79% to -3.11% in year two.
• Healthy controls showed smaller BFV fluctuations, ranging from +0.26% to -0.29%.

Conclusions:

• MRI-based CA estimates are demonstrably dependent on the chosen PS and SA.
• Consistency in SA and PS selection is vital for longitudinal MS studies and treatment trials.
• CA progression should be considered a secondary or tertiary outcome measure in MS trials until analysis techniques are better understood.