Related Experiment Video
Updated: May 7, 2026
05:55
Modeling the Functional Network for Spatial Navigation in the Human Brain
Published on: October 13, 2023
Test-retest reliability of fMRI-based graph theoretical properties during working memory, emotion processing, and
Hengyi Cao1, Michael M Plichta, Axel Schäfer
1Central Institute of Mental Health, Department of Psychiatry and Psychotherapy, University of Heidelberg, Medical Faculty Mannheim, Mannheim, Germany.
Neuroimage
|September 24, 2013
Summary
This study assessed the reliability of brain network analysis using functional magnetic resonance imaging (fMRI). Task-regression methods and functional atlases showed superior reliability for brain graph properties in active tasks.
Area of Science:
- Neuroscience
- Cognitive Science
- Data Science
Background:
- Functional magnetic resonance imaging (fMRI) and graph theory are increasingly used to study brain connectivity.
- The robustness of brain graph properties, especially from active fMRI tasks, remains poorly understood.
Purpose of the Study:
- To evaluate the test-retest reliability of brain graphs derived from active and resting-state fMRI tasks.
- To compare different data processing strategies and brain parcellation schemes for reliability.
Main Methods:
- Analyzed fMRI data from 26 healthy participants across n-back, face-matching, and resting-state tasks.
- Utilized AAL and Power et al. functional atlases for node definition.
- Compared intra-class correlation coefficients (ICCs) across five processing strategies.
Main Results:
- Task-regression methods with condition-specific regressors demonstrated superior reliability.
- Resting-state fMRI yielded higher ICCs than active tasks; n-back task was more reliable than face-matching.
- Functional parcellations showed higher reliability than the AAL atlas for global and local network properties.
Conclusions:
- Findings guide the selection of processing strategies, atlases, and outcome measures for fMRI studies employing graph theory.
- Highlights the importance of functional parcellations and task-regression for reliable brain network analysis in active tasks.
