Related Experiment Video
Updated: May 29, 2026
Cerebral Blood Flow-Based Resting State Functional Connectivity of the Human Brain using Optical Diffuse Correlation Spectroscopy
Published on: May 27, 2020
Anticorrelations in resting state networks without global signal regression
Xiaoqian J Chai1, Alfonso Nieto Castañón, Dost Ongür
1Department of Brain and Cognitive Sciences, McGovern Institute for Brain Research and Poitras Center for Affective Disorders Research, Massachusetts Institute of Technology, Cambridge, MA 02139, USA. xiaoqian@mit.edu
Anticorrelations in resting-state fMRI are likely biological, not artifacts of global signal regression. The component-based noise reduction method (CompCor) effectively reveals these default mode and task-positive network relationships.
Area of Science:
- Neuroscience
- Cognitive Neuroscience
- Functional Neuroimaging
Background:
- Resting-state functional magnetic resonance imaging (fMRI) reveals spontaneous signal fluctuations.
- Anticorrelated networks, such as the default mode network (DMN) and task-positive network (TPN), have been proposed.
- The validity of observed anticorrelations, particularly concerning global signal regression (GSR), is debated.
Purpose of the Study:
- To investigate whether anticorrelations in resting-state fMRI are genuine or artifacts of preprocessing.
- To compare the CompCor noise reduction method with GSR in revealing anticorrelations.
Main Methods:
- Resting-state fMRI data preprocessing using Component-based Noise Reduction (CompCor) and Global Signal Regression (GSR).
- Analysis of positive and negative correlations in resting-state functional connectivity.
- Seed-based correlation analysis using a default mode network region in the medial prefrontal cortex.
Main Results:
- Robust anticorrelations between the DMN and TPN were observed using both CompCor and GSR.
- The specificity of anticorrelations was comparable between the two preprocessing methods.
- CompCor demonstrated higher specificity and sensitivity for positive correlations compared to GSR.
Conclusions:
- Resting-state anticorrelations are likely not artifacts of GSR and may have biological origins.
- The CompCor method is suitable for examining valid resting-state anticorrelations.
- Findings support the intrinsic organization of anticorrelated brain networks during rest.
