Lateralized modulation brought by discrepancy speed ratios of left and right arm movements during human action observation: an EEG study

The left and right movements of a human or humanoid robot as an action observation stimulus will activate the human mirror neuron system (hMNS) to generate contralateral event-related desynchronization (ERD) suppression of the brain. The activation of hMNS response to action observation remains controversial for several reasons. Researches have found the influence of different speed factor for brain’s ERD suppression, but lack of exploring the influence of different speed discrepancy of left and right movements. To explore how is the discrepancy movements of left and right under different speeds influence ERD suppression, this paper invited six healthy subjects to participate action observation experiment under four different speeds (low, moderate, fast, and finalistic). Meanwhile, four different discrepancy speed ratios of movements are applied for such four different speeds to explore the influence of speed discrepancy on the left and right for the hMNS lateralized modulation effect. For the recorded electroencephalography (EEG) signals under different action observation stimulus, this paper selected the occipital and sensorimotor brain regions and used the convolutional neural network to classify EEG signals and measure the ERD suppression. Experimental results have shown that the action observation stimulus with different speed discrepancies improved the lateralized activity during low and moderate speeds, and significantly improved the lateralized activity than non-discrepancy during fast and finalistic speeds. We also analyzed the temporal, spectral, and classification characteristics for speed discrepancy, and discussed that the stimulus material design with speed discrepancy could greatly improve the lateralized modulation of ERD suppression. Action observation material with speed discrepancy of left and right movements could generate significant ERD suppression differences on lateralization of the brain, which could be used to build a more complex and robust brain-computer interface.