Motor Cortical Activity during Observing a Video of Real Hand Movements versus Computer Graphic Hand Movements: An MEG Study

Abstract

Both action observation (AO) and virtual reality (VR) provide visual stimuli to trigger brain activations during the observation of actions. However, the mechanism of observing video movements performed by a person's real hand versus that performed by a computer graphic hand remains uncertain. We aimed to investigate the differences in observing the video of real versus computer graphic hand movements on primary motor cortex (M1) activation by magnetoencephalography. Twenty healthy adults completed 3 experimental conditions: the resting state, the video of real hand movements (VRH), and the video of computer graphic hand movements (CGH) conditions with the intermittent electrical stimuli simultaneously applied to the median nerve by an electrical stimulator. The beta oscillatory activity (~20 Hz) in the M1 was collected, lower values indicating greater activations. To compare the beta oscillatory activities among the 3 conditions, the Friedman test with Bonferroni correction (p-value < 0.017 indicating statistical significance) were used. The beta oscillatory activities of the VRH and CGH conditions were significantly lower than that of the resting state condition. No significant difference in the beta oscillatory activity was found between the VRH and CGH conditions. Observing hand movements in a video performed by a real hand and those by a computer graphic hand evoked comparable M1 activations in healthy adults. This study provides some neuroimaging support for the use of AO and VR in rehabilitation, but no differential activations were found.

Keywords: action observation; magnetoencephalography; primary motor cortex; virtual reality.

Figure 1

Flowchart of experimental procedure. The beta rebound oscillatory power (red rectangle), ranging from 400 to 900 ms after the onset of electrical stimulation, represents the M1 activity. Notes: CGH = the video of computer graphic hand movements; VRH = the video of real hand movements; M1 = primary motor cortex.

Figure 2

The grand-averaged time–frequency maps and corresponding beta rebound oscillations of M1 in the 3 conditions. The left panel presents time–frequency maps of electricity-induced beta rebound oscillations (red rectangles) averaged across participants in the 3 conditions: resting state, CGH, and VRH. The right panel shows the temporal evolution of the beta oscillations, which were identified as the mean strength of the most reactive frequency bands (2 Hz), relative to the baseline power in M1. The mean of beta frequency was 18.6 Hz (~20 Hz), SD was 2.59 Hz, and the range was 13 to 26 Hz. Notes: CGH = the video of computer graphic hand movements; VRH = the video of real hand movements; M1 = primary motor cortex.

Figure 3

The levels of beta oscillation activities of M1 in the 3 conditions. Based on the Wilcoxon signed-rank test with Bonferroni correction, a p-value < 0.017 indicated statistical significance (i.e., 3 comparisons: resting state vs. VRH, resting state vs. CGH, and VRH vs. CGH). An asterisk (*) indicates statistical significance. Notes: CGH = the video of computer graphic hand movements; VRH = the video of real hand movements; M1 = primary motor cortex. Data are presented as median and interquartile range.