Implanted high-density electrode arrays to connect with the human spinal cord

Neural interface technologies aim to bridge the gap between the nervous system and external environment, offering both clinical applications for individuals with impairments and enhancement capabilities for healthy users. While various interface approaches exist across different neural pathway levels, muscle-level interfaces through electromyography (EMG) present a particularly promising avenue by capturing neural signals as they transform into physical actions. Although surface EMG provides insights into global muscle activity, its effectiveness is limited by filtering of the signal through the conductive tissue. Intramuscular EMG, especially when recorded via multichannel electrodes, offers more direct access to neural signals—yet the full potential of intramuscular multichannel electrodes remains unexplored, with current decoding algorithms not optimized for these advanced recording capabilities. This thesis advances the field through the characterization of a novel intramuscular multichannel micro-electrode array. I collected a novel dataset of multichannel intramuscular EMG recordings from forearm and leg muscles in healthy participants, which supported several key contributions: (1) development of a novel algorithm to extract (decode) the neural information from high-density intramuscular EMG, later adapted for surface EMG applications; (2) systematic analysis of electrode parameters’ impact on decoding performance; and (3) demonstration of a proof-of-concept offline system for gesture control using motor neuron decoding and mapping. These advances lay the groundwork for more sophisticated neural interfaces that could enhance both clinical applications and human-machine interaction.