Active haptic training improves somatosensory-motor feedback control in the motor skills of experts

Abstract

We perceive the external world through both externally generated and self-generated sensory stimuli (i.e., passive and active perception). While the performance in passive perception is improved by training or repetition through the functional and structural reorganization of the central nervous system (i.e., perceptual learning), the mechanisms by which perceptual learning occurs in active perception remain unclear. Here, we sought to explore the mechanisms underlying the improvement of active somatosensory perception and sensorimotor skills through active perceptual learning. Because we previously found that active perceptual learning depends on the expertise of the motor task to be performed, this study focused primarily on trained individuals. To this end, active haptic training (AHT) that targets the active somatosensory perception during the piano keystroke was used as a means of inducing active perceptual learning. We found four main results. First, participants actively modulated the muscle coordination patterns to optimize the movements required by the task through AHT without any explicit instruction on it by the experimenter, suggesting the involvement of active exploration in active perceptual learning. Second, AHT increased the relative reliance on afferent sensory information relative to the predicted one during the piano keystroke. Third, perceptual sensitivity of externally generated keystroke motions remained unchanged through AHT. Finally, AHT improved feedback control of repetitive keystroke movements in expert pianists. These results suggest that active perceptual learning involves changes in both the predictive integration process and active exploration and that the improved feedback control of fine movements benefits from the improvement in active perception. Significant statementOur ability to perceive both externally generated and self-generated sensory stimuli can be enhanced through training, known as passive and active perceptual learning. Here, we sought to explore the mechanisms underlying active perceptual learning by using active haptic training (AHT), which has been demonstrated to enhance the somatosensory perception of a finger in a trained motor skill. First, AHT reorganized the muscular coordination during the piano keystroke. Second, AHT increased the relative reliance on afferent sensory information relative to predicted one, in contrast to no increment of overall perceptual sensitivity. Finally, AHT improved feedback movement control of keystrokes. These results suggest that active perceptual learning involves active exploration and adaptation of predictive sensory integration, which underlies the co-enhancement of active perception and feedback control of movements.