Abstract
The stabilizing role of sensory feedback in relation to movement dynamics remains to be poorly understood in realistic three-dimensional movements of limbs. The objective of this experimental and computational study was to classify the contribution of sensory feedback from muscle spindles to the control of assistive and resistive limb dynamics during human pointing movements. We integrated a human upper-limb musculoskeletal model with a model of Ia primary afferent discharge to analyze motion and muscle activation patterns during reaching movements in virtual reality (VR). The reaching target locations in VR were selected to define movements with varying roles of gravity and interaction torques that created diverse dynamical contexts. Nine healthy human subjects performed the VR task by pointing to the reaching targets with visual feedback of their arm location. Motion capture and electromyography (EMG) were recorded, and joint torques and Ia primary afferent discharge were estimated using the integrated model. The experimental and simulated data were analyzed with hierarchal clustering (Gritsenko et al., 2016). The clustering analysis of EMG and predicted proprioceptive signals showed a divergent relationship between muscle activation and sensory feedback patterns. Even though the Ia models had nonlinear dynamical components, their output was still most related to the anatomical grouping of muscles and less so to the dynamical contexts of each movement, reflected in muscle activations. Altogether, these results suggest that sensory feedback is nonlinearly related to muscle activation profiles and that it may contribute information necessary for coupling between proximal and distal muscle groups.