“…Importantly, these parameters could be easily and accurately estimated from pool or cellular spiking in humans (Farina et al, 2017;Afsharipour et al, 2020;Hassan et al, 2020). While decoding MN spiking activity for estimating the synaptic input (as a measure of motor intent) or for estimating grip force have been recently examined in animals (Thompson et al, 2018) and humans (Farina et al, 2017;Twardowski et al, 2019;Afsharipour et al, 2020;Hassan et al, 2020), only one decoder in the literature has shown real-time performance for prosthetic control (decoding time < 10 ms, Twardowski et al, 2019). When compared to conventional amplitudebased myoelectric decoders, MN firing activity-based decoders provide more responsive, smooth, and proportional control, supporting the high performance in Twardowski et al (2019) and even higher performance observed in our results (see Pearson's CC, RMSE, and NRMSE, next section).…”