Comparing resting state fMRI de-noising approaches using multi- and single-echo acquisitions.

Area of Science:

• Neuroimaging
• Functional Magnetic Resonance Imaging (fMRI)
• Signal Processing

Background:

• Head motion is a significant challenge in resting-state fMRI (rfMRI), compromising data reliability and reproducibility.
• Existing single-echo denoising methods like MWC, FIX, and ICA-AROMA have limitations in artifact removal.

Purpose of the Study:

• To compare the effectiveness of multi-echo ICA (ME-ICA) against popular single-echo denoising techniques for artifact removal in rfMRI.
• To assess the impact of denoising methods on data quality, motion artifact reduction, and default mode network (DMN) functional connectivity.

Main Methods:

• Comparison of ME-ICA with regression of motion parameters, white matter, and cerebrospinal fluid signals (MWC), FMRIB's ICA-based X-noiseifier (FIX), and ICA-based Automatic Removal Of Motion Artifacts (ICA-AROMA).
• Data acquired using a clinical scanner from healthy participants and ADHD patients with varying degrees of head motion.
• Evaluation metrics included post-cleaning data quality, BOLD signal-motion uncoupling, and preservation of DMN functional connectivity.

Main Results:

• Most tested cleaning methods improved data quality.
• ME-ICA demonstrated superior robustness, effectively minimizing motion's impact on functional connectivity, particularly in high-motion participants.
• ME-ICA preserved the functional connectivity structure of the DMN.

Conclusions:

• ME-ICA is a highly effective method for artifact removal in rfMRI, outperforming traditional single-echo techniques.
• The multi-echo approach enables reliable rfMRI data acquisition in clinical settings, even with significant head motion.