Task-based fMRI in early Multiple Sclerosis: what is the best head motion correction approach?

Abstract

Functional MRI (fMRI) is one of the most common brain imaging modalities used for understanding brain organization and connectivity abnormalities associated with multiple sclerosis (MS). The fMRI signal is highly perturbed by head motion, which degrades data quality and influences all image-derived metrics. Numerous correction approaches have been proposed over the years to overcome the problems induced by head motion, however, despite a few efforts, there are still current and persistent controversies regarding the best correction strategy. The lack of a systematic comparison between different correction approaches motivates the search for optimal correction models, particularly in studies with clinical populations prone to characterize by higher motion. Moreover, motion correction strategies gain more relevance in task-based designs, which are less explored compared to resting-state and may have a crucial role in describing the functioning of the brain and highlighting specific connectivity changes.

We acquired fMRI data from a group of patients with early MS and matched healthy controls (HC) during performance of a visual task, characterized motion in both groups, and compared the most used motion correction methods. We compared task-activation metrics obtained from models without motion correction, models containing 6 or 24 motion parameters (MPs) as nuisance regressors, models containing 6 or 24 MPs and motion outliers detected with FD or DVARS as nuisance regressors (scrubbing) and models with 6 or 24 MPs where motion outliers were corrected through volume interpolation. To our knowledge, volume interpolation is a frequently used approach but was never compared with other existent methods.

Our results showed that there were no differences in motion between groups, suggesting that recently diagnosed MS patients do not present problematic motion. In general, models with 6 MPs present higher Z-scores than models with 24 MPs, suggesting the 6 MPs as the best trade-off between motion correction and preservation of valuable information. However, correction approaches differ between groups, regarding the combination of MPs with correction of motion outliers. Models with 6 MPs and outliers’ volume interpolation or scrubbing with FD presented higher Z-scores in the MS group, while models with 6 MPs and scrubbing with DVARS or volume interpolation were the best combinations for the HC group. Differences between groups in motion correction strategies draw attention to the intrinsic impact of MS on fMRI analyses, which should be carefully addressed.

This work paves the way towards finding an optimal motion correction strategy, which is required to improve the accuracy of fMRI analyses, crucially in clinical studies in MS and other patient populations.