Task-dependent modulation of oscillatory neural activity during movements: Abstract

Neural oscillations in different frequency bands have been observed in a range of sensorimotor tasks and have been linked to coupling of spatially distinct neurons. The goal of this study was to detect a general motor network that is activated during phasic and tonic movements and to study the task-dependent modulation of frequency coupling within this network. To this end we recorded 122-multichannel EEG in 13 healthy subjects while they performed three simple motor tasks. EEG data source modeling using individual MR images was carried out with a multiple source beamformer approach. A bilateral motor network connecting frontal, cerebellar and central motor regions, was consistently activated throughout the motor tasks. Quantification of observed spectral responses using dynamic causal modeling revealed strong coupling in the c-band (30–48 Hz) between frontal and central motor regions when a slow finger movement had to be adjusted to an external trigger. During a self-paced fast finger tapping (presumably sensory) coupling was strongest in the h-band (4–7 Hz), while b-band (13–30 Hz) coupling was dominant during an isometric contraction of the forearm. During these two highly automatic movements effective connectivity was strongest between central and cerebellar regions. Our results show that neural coupling within motor networks is modulated in distinct frequency bands depending on the motor task. They provide evidence that dynamic causal modeling in combination with EEG source analysis is a valuable tool for inferring on architecture and coupling parameters of neural networks.