Reducing respiratory effect in motion correction for EPI images with sequential slice acquisition order.
1Imaging Research Facility, Indiana University, Bloomington, IN, United States; Department of Psychological and Brain Sciences, Indiana University, Bloomington, IN, United States.
Journal of Neuroscience Methods
|February 25, 2014
Summary
Respiration causes motion-like artifacts in functional MRI (fMRI) data. A new method improves fMRI analysis by performing motion correction on segments of sequentially acquired echo planar imaging (EPI) data, enhancing temporal signal-to-noise ratio (TSNR).
Area of Science:
- Neuroimaging
- Biomedical Engineering
- Medical Physics
Background:
- Accurate motion correction is essential for functional MRI (fMRI) time-series analysis.
- Subject respiration can induce magnetic field fluctuations, causing slice misalignments in echo planar imaging (EPI) that mimic head motion.
- Acquisition delays between slices exacerbate respiration-induced artifacts, leading to position-dependent shifts.
Purpose of the Study:
- To investigate the impact of respiration on fMRI motion correction.
- To develop and evaluate a novel motion correction strategy for mitigating respiration-induced artifacts in EPI.
- To improve the temporal signal-to-noise ratio (TSNR) in fMRI data acquired with sequential slice acquisition.
Main Methods:
- Acquisition of fast-sampled fMRI data using multi-band EPI.
- Simulation of various acquisition schemes to characterize respiration effects.
- Proposal and implementation of a segmented motion correction approach for sequentially acquired EPI data.
- Comparison of segmented motion correction with whole-volume correction.
Main Results:
- Respiration introduces significant noise into fMRI data even after standard motion correction.
- Increased effective repetition time (TR) amplifies signal variations between volumes post-motion correction.
- Segmented motion correction significantly increased TSNR compared to whole-volume correction for sequential acquisition.
- TSNR gains were more pronounced in superior slices and absent in interleaved acquisition schemes.
Conclusions:
- Respiration-induced artifacts are a critical confound in fMRI analysis, particularly with longer TRs.
- Segmented motion correction is a highly effective strategy for mitigating respiratory noise in sequentially acquired EPI data.
- This approach offers substantial TSNR improvements, especially for superior brain regions, comparable to advanced retrospective correction methods.
