Model-based deconvolution of a force signal to estimate motor unit twitch parameters under low-force isometric contractions

Abstract

AbstractMuscle force generation is key in everyday life, yet the mechanisms regulating force generation are not fully resolved. Over the last half-century, attempts have been made to elucidate the mechanisms regulating force generation in vivo, primarily focusing on the motor unit (MU) activity. Although different methods exist to extract MU twitch parameters from the force signal, no method can accurately identify asingleMU twitch given asingleMU spike train. We addressed this problem by developing a model-based deconvolution method. We evaluated the method using extensive simulations mimicking low-force isometric contractions, which resulted in four main findings. First, given an accurate spike train, the method can accurately estimate twitch parameter estimates in a very weak force scenario. Second, the more missed or false spikes were used to estimate twitch parameters, the larger the underestimation is. Third, the higher the ground truth twitch amplitude, the less variance between the true and estimated twitch parameters. Fourth, the method accurately estimates a representative twitch when the force signal comprises unequal successive MU twitch profiles within each MU. To conclude, we anticipate this method enables studying mechanisms regulating force generation during electrically stimulated or voluntarily generated forces.