Comparison of motion correction techniques applied to functional near-infrared spectroscopy data from children
Xiao-Su Hu1, Maria M Arredondo2, Megan Gomba1
1University of Michigan, Center for Human Growth and Development, 300 North Ingalls Street, Ann Arbor, MI 48104.
Insights
Motion artifacts significantly impact pediatric functional near-infrared spectroscopy (fNIRS) data. Moving average and wavelet methods effectively correct these artifacts in children, improving brain imaging analysis quality.
Area of Science:
- Neuroscience
- Biomedical Engineering
- Signal Processing
Background:
- Motion artifacts are a major challenge in pediatric functional near-infrared spectroscopy (fNIRS) data acquisition and analysis.
- Existing motion correction techniques have not been thoroughly evaluated for their effectiveness in noisy pediatric fNIRS data.
- The heterogeneity of artifacts in pediatric fNIRS data complicates correction efforts.
Purpose of the Study:
- To compare the efficacy of six common motion artifact correction techniques for pediatric fNIRS data.
- To identify the most effective methods for improving the quality of fNIRS data from children during a language task.
- To provide guidance for the analysis of pediatric optical brain imaging data.
Main Methods:
- Collected fNIRS data from children during a language acquisition task.
- Applied six prevalent motion artifact correction techniques: wavelet, spline interpolation, principal component analysis, moving average (MA), correlation-based signal improvement, and a wavelet-MA combination.
- Evaluated correction efficacy using five predefined metrics.
Main Results:
- The moving average (MA) and wavelet methods demonstrated the best performance in correcting motion artifacts.
- These methods significantly improved the quality of fNIRS data acquired from pediatric subjects.
- Findings highlight the varied nature of fNIRS artifacts in children and the differential effectiveness of correction algorithms.
Conclusions:
- Moving average and wavelet methods are recommended for motion artifact correction in pediatric fNIRS studies.
- These findings contribute to the theoretical understanding and practical application of optical brain imaging analysis in children.
- Improved artifact correction enhances the reliability of fNIRS data for studying child brain development and function.
Abstract:
Motion artifacts are the most significant sources of noise in the context of pediatric brain imaging designs and data analyses, especially in applications of functional near-infrared spectroscopy (fNIRS), in which it can completely affect the quality of the data acquired. Different methods have been developed to correct motion artifacts in fNIRS data, but the relative effectiveness of these methods for data from child and infant subjects (which is often found to be significantly noisier than adult data) remains largely unexplored. The issue is further complicated by the heterogeneity of fNIRS data artifacts. We compared the efficacy of the six most prevalent motion artifact correction techniques with fNIRS data acquired from children participating in a language acquisition task, including wavelet, spline interpolation, principal component analysis, moving average (MA), correlation-based signal improvement, and combination of wavelet and MA. The evaluation of five predefined metrics suggests that the MA and wavelet methods yield the best outcomes. These findings elucidate the varied nature of fNIRS data artifacts and the efficacy of artifact correction methods with pediatric populations, as well as help inform both the theory and practice of optical brain imaging analysis.
More Related Videos
06:18Qualitative and Comparative Cortical Activity Data Analyses from a Functional Near-Infrared Spectroscopy Experiment Applying Block Design
Published on: December 3, 2020
08:10Neuroimaging Field Methods Using Functional Near Infrared Spectroscopy NIRS Neuroimaging to Study Global Child Development: Rural Sub-Saharan Africa
Published on: February 2, 2018