Precise modeling of task-related sensorimotor activation based on simultaneous surface electromyography

Abstract Objective To evaluate central and peripheral correlates of motor control using functional magnetic resonance imaging (fMRI) and surface electromyography (EMG), with a focus on the added value of EMG-informed analysis during movement of the lower limb in healthy controls. Methods Twenty participants performed dorsi-/plantarflexion of the ankle (Ankle) and gait imagery (GI) in a block design during fMRI. Accelerometry (Acc) and surface EMG from tibialis anterior (TA) activity were recorded and included as regressors in five analysis models, either with or without temporal derivative (TD) to account for time shift in the task regressor. Voxel-wise analyses complemented by post-hoc region-of-interest (ROI) analyses were performed to compare the amount of variability explained by the models. Results Inclusion of either Acc or EMG on top of the task regressor explained robustly fMRI signal variability in the primary sensorimotor cortices. On top of Acc, EMG additionally explained activation variability mainly in the contralateral thalamus and the secondary somatosensory cortex (S2). This effect was, however, mainly driven by spontaneous signal fluctuations at rest and during imagery. Comparisons between models with and without TD revealed consistent differences in the cerebellum and thalamus across tested models, suggesting that subcortical structures may involve transient signal changes when switching between movement and rest. Conclusion Including EMG in fMRI analysis enhances specificity in detecting motor-related brain activity and enables differentiation of spontaneous or unpredicted motor behavior. TD improved signal detection in the primary sensorimotor cortices, but may have a detrimental effect on signal detection in other, mostly subcortical regions, likely reflecting their different temporal signal dynamics.