Motion Artifact Correction in Functional Magnetic Resonance Imaging

Motion artifacts remain one of the most pervasive challenges in functional magnetic resonance imaging, undermining the fidelity of blood oxygenation level dependent (BOLD) signals and biasing connectivity and activation measures. Even submillimetre head movements can introduce spurious correlations, distort topological metrics of functional networks and reduce statistical power. A broad array of strategies has been developed to mitigate these effects. Prospective motion correction seeks to track head position in real time and update imaging gradients to maintain alignment of successive volumes. Retrospective approaches include regression of motion parameters, censoring or “scrubbing” of high‐motion time points, and nuisance regression techniques such as global signal regression. More sophisticated denoising pipelines employ independent component analysis (ICA) or multi‐echo acquisitions to disentangle motion‐related variance from neural signals. Recent advances have also explored the integration of continuous evaluation frameworks to benchmark and refine preprocessing software. Taken together, these methods have converged towards hybrid pipelines that combine real‐time tracking, robust regression of residual motion parameters and data‐driven component classification, delivering improved temporal signal‐to‐noise ratios and more reliable functional connectivity estimates. Practical applications extend from paediatric imaging, where compliance is variable, to clinical populations with movement disorders, and underpin large‐scale initiatives seeking reproducible biomarkers of brain function.

Research from Nature Portfolio

No recent Nature Portfolio content available.

Motion Artifact Correction in Functional Magnetic Resonance Imaging publication trend

The graph below shows the total number of articles in motion artifact correction in functional magnetic resonance imaging across all publications each year (not limited to Nature Index journals).

Technical terms

Motion artifact: Spurious signal fluctuations in fMRI data arising from subject movement during image acquisition.

BOLD signal: Blood oxygenation level dependent contrast, reflecting changes in deoxyhaemoglobin concentration related to neural activity.

Prospective motion correction: Real-time tracking of head position to update imaging parameters and maintain volume alignment.

Retrospective motion correction: Post-acquisition methods, including regression of motion parameters, scrubbing and nuisance signal removal.

Independent component analysis (ICA): Data-driven decomposition technique separating signal sources, used to identify and remove motion components.

Global signal regression: Nuisance regression approach that removes the mean brain signal from each voxel time course to reduce widespread noise.

Scrubbing: Censoring or exclusion of time points with excessive motion to prevent contaminated volumes from influencing analyses.

References

1. Continuous evaluation of denoising strategies in resting-state fMRI connectivity using fMRIPrep and Nilearn. PLOS Computational Biology (2024).
2. Denoising task-correlated head motion from motor-task fMRI data with multi-echo ICA. Imaging Neuroscience (2024).
3. Prospective motion correction of 3D echo-planar imaging data for functional MRI using optical tracking. NeuroImage (2015).
4. Addressing head motion dependencies for small-world topologies in functional connectomics. Frontiers in Human Neuroscience (2013).