Comparing different motion correction approaches for resting-state functional connectivity analysis with functional...

Area of Science:

• Neuroimaging
• Biomedical Engineering
• Signal Processing

Background:

• Motion artifacts pose a significant challenge in functional near-infrared spectroscopy (fNIRS) data acquisition.
• Effective strategies for mitigating motion artifacts in resting-state fNIRS data remain underexplored.

Purpose of the Study:

• To evaluate the impact of various motion artifact correction techniques on functional connectivity analysis in resting-state fNIRS.
• To determine the optimal correction approach based on the type and extent of motion contamination.

Main Methods:

• Utilized semi-simulated resting-state fNIRS datasets with controlled spike-like and baseline-shift motion artifacts.
• Applied fifteen distinct preprocessing pipelines with varying motion correction strategies.
• Assessed pipeline performance using three quantitative metrics on group-level functional connectivity.

Main Results:

• Multiple correction approaches proved effective at low contamination levels.
• Pipeline reliability decreased significantly with increased motion artifact contamination.
• Discarding frames post-preprocessing was optimal for datasets with minimal baseline shifts, while pre-processing frame discarding was superior when both artifact types were present.

Conclusions:

• The efficacy of motion artifact correction in fNIRS is highly dependent on the specific characteristics of the artifacts.
• A one-size-fits-all approach is insufficient; customized strategies are necessary for robust resting-state functional connectivity analysis.