Head Motion in Diffusion Magnetic Resonance Imaging: Quantification, Mitigation, and Structural Associations in...

Area of Science:

• Neuroimaging
• Diffusion Magnetic Resonance Imaging (dMRI)
• Brain Microstructure Analysis

Background:

• Head motion during dMRI scans introduces artifacts and biases in quantitative measurements.
• Comprehensive characterization of motion across diverse cohorts and consortiums is lacking.
• Understanding motion impact is crucial for accurate dMRI data analysis and interpretation.

Purpose of the Study:

• To characterize the magnitude and direction of head motion in a large dMRI dataset.
• To evaluate the effectiveness of state-of-the-art dMRI preprocessing pipelines in mitigating motion-induced biases.
• To investigate potential differences in brain microstructure and structural connectivity between high and low motion participants.

Main Methods:

• Analysis of 16,995 dMRI sessions from 13 cohorts, covering a wide age range and diverse clinical populations.
• Utilized datasets with scan-rescan acquisitions to assess preprocessing pipeline performance.
• Compared quantitative dMRI measures and structural connectivity between participants categorized as 'movers' and 'non-movers' based on motion levels.

Main Results:

• Subjects typically exhibit 1-2 mm/min of motion, primarily translational (anterior-posterior) and rotational (right-left axis).
• Modern dMRI preprocessing pipelines effectively mitigate motion artifacts, rendering biases undetectable with current analytical methods.
• No significant differences in brain microstructure or macrostructural connectivity were observed between high and low motion groups.

Conclusions:

• Head motion in dMRI is characterized by specific patterns, but modern preprocessing pipelines are robust.
• Current analysis techniques cannot detect motion-induced biases after applying advanced preprocessing.
• Motion levels do not appear to influence microstructure or structural connectivity outcomes in dMRI studies.