State-dependent differences between functional and effective connectivity of the human cortical motor system

• Published in: NeuroImage (Orlando, Fla.) Vol. 67; pp. 237 - 246
• Main Authors: Rehme, Anne K., Eickhoff, Simon B., Grefkes, Christian
• Format: Journal Article
• Language: English
• Published: Amsterdam Elsevier Inc 15.02.2013; Elsevier; Elsevier Limited
• Subjects: Adult; Biological and medical sciences; Brain mapping; Connectome - methods; Dynamic causal modeling; Estimates; Evoked Potentials, Motor - physiology; Female; fMRI; Fundamental and applied biological sciences. Psychology; Humans; Intrinsic connectivity; Magnetic Resonance Imaging - methods; Male; Mathematical models; Methods; Motor Cortex - anatomy & histology; Motor Cortex - physiology; Movement - physiology; Nerve Net - anatomy & histology; Nerve Net - physiology; NMR; Nuclear magnetic resonance; Parameter estimation; Resting-state; Studies; Vertebrates: nervous system and sense organs; fMRI; Resting-state; Intrinsic connectivity; Dynamic causal modeling; Human; Nuclear magnetic resonance imaging; Functional imaging
• ISSN: 1053-8119; 1095-9572; 1095-9572
• DOI: 10.1016/j.neuroimage.2012.11.027

Neural processing is based on interactions between functionally specialized areas that can be described in terms of functional or effective connectivity. Functional connectivity is often assessed by task-free, resting-state functional magnetic resonance imaging (fMRI), whereas effective connectivity is usually estimated from task-based fMRI time-series. To investigate whether different connectivity approaches assess similar network topologies in the same subjects, we scanned 36 right-handed volunteers with resting-state fMRI followed by active-state fMRI involving a hand movement task. Time-series information was extracted from identical locations defined from individual activation maxima derived from the motor task. Dynamic causal modeling (DCM) was applied to the motor task time-series to estimate endogenous and context-dependent effective connectivity. In addition, functional connectivity was computed for both the rest and the motor task condition by means of inter-regional time-series correlations. At the group-level, we found strong interactions between the motor areas of interest in all three connectivity analyses. However, although the sample size warranted 90% power to detect correlations of medium effect size, resting-state functional connectivity was only weakly correlated with both task-based functional and task-based effective connectivity estimates for corresponding region-pairs. By contrast, task-based functional connectivity showed strong positive correlations with DCM effective connectivity parameters. In conclusion, resting-state and task-based connectivity reflect different components of functional integration that particularly depend on the functional state in which the subject is being scanned. Therefore, resting-state fMRI and DCM should be used as complementary measures when assessing functional brain networks. ► Weak correlations between resting-state functional and task-based effective connectivity ► Robust correlation between task-based functional and effective connectivity ► Connectivity depends on the experimental setting, not on the connectivity method. ► Different connectivity methods represent complementary measures of neural networks.