Related Experiment Videos
Movement-related effects in fMRI time-series
K J Friston1, S Williams, R Howard
1Wellcome Department of Cognitive Neurology, Institute of Neurology, United Kingdom.
Magnetic Resonance in Medicine
|March 1, 1996
Summary
This study presents a novel method to remove confounding movement artifacts in functional MRI (fMRI) data. The approach successfully corrects for over 90% of signal contamination caused by subject motion during brain imaging.
Area of Science:
- Neuroimaging
- Biophysics
- Signal Processing
Background:
- Functional MRI (fMRI) is crucial for studying brain activity.
- Subject movement during fMRI scans introduces significant artifacts, confounding activation studies.
- Even after realignment, movement-related effects persist and distort the fMRI signal.
Purpose of the Study:
- To develop and present an approach for modeling and removing movement-related artifacts from fMRI time-series.
- To address confounding spatial and intensity transformations caused by subject motion.
- To improve the accuracy and reliability of fMRI activation studies.
Main Methods:
- Modeling movement-related artifacts using an autoregression-moving average model.
- Accounting for spin excitation history and local saturation differences due to previous displacements.
- Implementing spatial and intensity transformations to correct fMRI data.
Main Results:
- Demonstrated successful removal of movement-related artifacts from fMRI time-series.
- Empirical analyses show that movement can account for over 90% of the fMRI signal in extreme cases.
- The proposed method effectively removes artifactual components, enhancing signal quality.
Conclusions:
- Subject movement is a major source of artifact in fMRI data, impacting signal intensity and spatial information.
- The developed autoregression-moving average model effectively captures and corrects for these movement-related artifacts.
- This approach significantly improves the fidelity of fMRI activation studies by removing substantial motion-induced noise.