Modulation of Motor Cortical Activities by Action Observation and Execution in Patients with Stroke: An MEG Study

Abstract

Action observation therapy has recently attracted increasing attention; however, the mechanisms through which action observation and execution (AOE) modulate neural activity in stroke patients remain unclear. This study was aimed at investigating the effects of action observation and two types of AOE on motor cortical activations after stroke using magnetoencephalography. Twenty patients with stroke and 20 healthy controls were recruited for the collection of data on the beta oscillatory activity in the primary motor cortex (M1). All participants performed the conditions of resting, observation only, and video observation combined with execution (video AOE). Stroke patients performed one additional condition of affected hand observation combined with execution (affected hand AOE). The relative change index of beta oscillations was calculated, and nonparametric tests were used to examine the differences in conditions. In stroke patients, the relative change index of M1 beta oscillatory activity under the video AOE condition was significantly lower than that under the observation only and affected hand AOE conditions. Moreover, M1 cortical activity did not significantly differ under the observation only and affected hand AOE conditions. For healthy controls, the relative change index under the video AOE condition was significantly lower than that under the observation only condition. In addition, no significant differences in relative change indices were found under the observation only and video AOE conditions between the 2 groups. This study provides new insight into the neural mechanisms underlying AOE, which supports the use of observing videos of normal movements during action observation therapy in stroke rehabilitation.

Figure 1

Illustration of the conditions during MEG recordings. Note: (v) in some pictures indicates watching video movements. AOE = action observation and execution.

Figure 2

The brain images of stroke lesions and time-frequency analyses of beta oscillatory activities of 2 individual patients. (a) Patient No. 18 had an ischemic stroke in the left-side corona radiata. Left panel: the lesion site of this patient (red arrow) in the diffusion-weighted MRI image. Right panel: time-frequency maps and time courses of beta oscillatory activity under each condition. The powers of beta oscillatory activity of this patient in the resting, observation only, video AOE, and affected hand AOE conditions were 5.05, 4.54, 0.15, and 1.21, respectively. (b) Patient No. 20 had a hemorrhagic stroke in the right-side putamen. Left panel: the lesion site of this patient (red arrow) in the T2-weighted MRI image. Right panel: time-frequency maps and time courses of oscillatory activity under each condition. The powers of beta oscillatory activity of this patient in the resting, observation only, video AOE, and affected hand AOE conditions were 4.71, 2.86, 0.97, and 2.26, respectively. Note: AOE = action observation and execution.

Figure 3

Left panel of (a, b): time-frequency maps of group-averaged electricity-induced beta rebound oscillations (red rectangles) under each condition. Right panel of (a, b): time courses of beta oscillatory activities in the most reactive frequency bands (2 Hz for consecutive bins) in the primary motor cortex (M1). Note: AOE = action observation and execution.

Figure 4

Comparisons of the relative change index of beta oscillatory activity under each condition of the stroke patients and healthy controls. (a) Stroke patients: the relative change index under the video AOE condition decreased significantly more than those under the observation condition and the affected hand AOE condition. (b) Healthy controls: the relative change index under the video AOE condition was significantly lower than that under the observation condition. Data are presented as the median ± IQR values. Note: AOE = action observation and execution.

Figure 5

Comparisons of relative change indices under the conditions of observation only and video AOE between the stroke patients and healthy controls. No significant differences in relative change indices under the observation only and video AOE conditions were found between the 2 groups. Data are presented as the median ± IQR values. Note: AOE = action observation and execution.