Abstract
Understanding human motor control requires integrating cortical and spinal activity, muscle mechanics, and electrophysiology, which are typically studied in isolation. We present MyoGen, an open-source framework that unifies spinal circuitry, proprioceptive feedback, musculotendon dynamics, and multimodal electromyography generation. Using data-driven optimization benchmarked against human motor unit recordings, MyoGen produces motor unit population activity matching experimental discharge-rate distributions and variability across muscles. The framework generates decomposable surface and intramuscular electromyography, reproduces beta-band modulation of descending drive and its transformation into force, and implements key sensorimotor loop components. Dimensionality reduction of simulated agonist-antagonist electromyography reveals low-dimensional control manifolds consistent with recordings from able-bodied individuals and those with spinal cord injury. Here, we show that MyoGen provides a physiologically grounded simulation substrate, enabling systematic investigation of motor control principles, benchmarking of signal-processing algorithms, and exploration of experimentally inaccessible sensorimotor interactions.

This publication has 59 references indexed in Scilit: