Spike Timing Depended Plasticity Produces Unsupervised Learning of Synergistic Muscle Feedback in a Synthetical Neural Network

This study investigates how type Ia feedback from muscle spindles can be organized into groups representing agonistic muscle pairs through Spike Timing Dependent Plasticity (STDP). A single degree of freedom joint is actuated with four biologically modeled muscles forming two agonistic pairs. In order to emulate the sensory dynamics of biological muscle spindles, sensors in the model record the active length and velocity states of each muscle, the two primary factors eliciting type Ia afferent responses. In biological networks, synapses from Ia sensory neurons frequently activate interneurons representing agonistic muscle sources. This research investigates whether this organization can emerge in an initially unsorted network through synaptic modulation via STDP. The network of interconnected Ia sensory neurons and interneurons initiates with random conductance values, without knowledge of the desired organization structure. Under semi-randomized muscle activation, STDP in the proprioceptive network demonstrates the ability to organize Ia sensory neuron signals into groups according to their agonistic sources, mirroring known architecture. STDP provides a biologically plausible, unsupervised learning mechanism by which the known connections in vivo may form. With continuing work, STDP may prove itself capable of organizing proprioceptive networks for larger musculoskeletal systems, possibly on the scale of biological creatures. Future investigations will explore how the application of STDP to the other known synaptic connections in the Ia afferent network may assist in additional organization of the network architecture.