Related Experiment Video
Updated: Jun 29, 2026
08:19
Simultaneous Data Collection of fMRI and fNIRS Measurements Using a Whole-Head Optode Array and Short-Distance Channels
Published on: October 20, 2023
Evaluation and optimization of fMRI single-subject processing pipelines with NPAIRS and second-level CVA
Jing Zhang1, Jon R Anderson, Lichen Liang
1Health Informatics Graduate Program, University of Minnesota, Minneapolis, MN 55455, USA. jzhang0000@gmail.com
Magnetic Resonance Imaging
|October 14, 2008
Summary
Optimizing functional magnetic resonance imaging (fMRI) preprocessing, particularly spatial smoothing and temporal filtering, significantly enhances multivariate analysis performance and reproducibility. Choosing between univariate and multivariate models is critical for accurate brain activation patterns.
Area of Science:
- Neuroimaging
- Cognitive Neuroscience
- Biostatistics
Background:
- Univariate General Linear Model (GLM) is the standard for brain activation detection in functional magnetic resonance imaging (fMRI).
- Multivariate approaches offer potential for revealing neural networks and functional connectivity.
- Understanding the impact of preprocessing on multivariate fMRI analysis is crucial.
Purpose of the Study:
- To investigate the effect of preprocessing steps on multivariate model-based fMRI processing pipelines.
- To optimize single-subject Canonical Variate Analysis (CVA)-based pipelines using NPAIRS metrics.
- To compare univariate GLM with multivariate CVA pipelines across different software packages.
Main Methods:
- Investigated impact of fMRI preprocessing steps (e.g., spatial smoothing, temporal detrending, motion correction) on multivariate CVA pipelines.
- Optimized pipelines using Nonparametric Prediction, Activation, Influence, and Reproducibility Resampling (NPAIRS) metrics.
- Compared single-subject statistical parametric images (SPIs) from univariate GLM and multivariate CVA using second-level CVA.
Main Results:
- Spatial smoothing, temporal detrending/filtering, and motion correction significantly improved fMRI pipeline performance.
- Combined optimization of spatial smoothing, temporal detrending, and CVA parameters enhanced between-subject reproducibility.
- Key pipeline choices impacting activation patterns include statistical model (univariate vs. multivariate) and spatial smoothing.
Conclusions:
- Preprocessing choices, especially spatial smoothing and temporal filtering, critically influence multivariate fMRI analysis outcomes.
- Multivariate approaches, when optimized, can improve reproducibility and reveal complex brain activity patterns.
- Moving beyond fixed GLM and spatial filtering is important for accurate BOLD fMRI activation detection.
