A computational framework with voltage-dependent synaptic function explains LTP-dominant plasticity during functional electrical stimulation therapy

Abstract

Functional Electrical Stimulation (FES) therapy is a widely used neurorehabilitation technique that restores motor function by delivering electrical stimulation to target muscles during voluntary contraction. Despite its clinical effectiveness, the mechanisms by which FES therapy induces neuroplasticity remain poorly understood. Previous work has proposed that positive plasticity arises from Hebbian interactions at corticospinal-motoneuronal synapses when voluntary descending motor commands coincide with antidromic firing elicited by FES therapy. However, if spike-timing-dependent plasticity (STDP) is assumed to underlie this Hebbian mechanism, an unresolved question remains: why does FES therapy produce long-term potentiation (LTP) reliably, rather than the mixture of LTP and LTD predicted from classical STDP rules? Here, we test the hypothesis that interactions between voluntary descending spikes and stimulation-evoked antidromic spikes generate multi-spike patterns that bias plasticity toward potentiation. To investigate this mechanism, we developed a computational framework implementing a voltage-dependent plasticity rule that incorporates postsynaptic membrane dynamics and higher-order spike interactions. This framework enables simulation of synaptic plasticity during FES therapy while systematically varying stimulation frequency, input heterogeneity, and spike timing structure. Our simulations show that voltage-dependent dynamics strongly bias synaptic changes toward LTP during FES therapy-like conditions. In particular, physiological interspike interval variability promotes potentiation, whereas highly regular inputs bias synapses toward depression. These results indicate that postsynaptic voltage dynamics and spike-interaction structure, rather than pairwise spike timing alone, govern plasticity outcomes during FES therapy. Our findings provide a mechanistic explanation for why FES therapy reliably induces LTP-dominant plasticity and offer a computational framework for optimizing neuromodulation therapies.