25. Physiologic Sensorimotor Neural Signaling With the Composite Regenerative Peripheral Nerve Interface (C-RPNI) in a Rodent Model

PURPOSE: Naturalistic prosthetic function requires both the reliable detection of the user’s motor intention for accurate prosthetic control and the communication of sensory information from the prosthetic to the user. Currently, there is lack of an effective patient-prosthetic interface that can effectively support this bidirectional sensorimotor signaling. We developed the Composite Regenerative Peripheral Nerve Interface (C-RPNI), a transected nerve implanted between a dermal and muscle graft, as a biologic sensorimotor interface. C-RPNI reinnervation supports independent motor and sensory signaling in an anesthetized rat. In this study, we assessed whether stimulation of the C-RPNI dermal component in awake rats can facilitate sensory feedback and whether physiologically accurate electromyography (EMG) signals can be recorded from the C-RPNI muscle component during volitional ambulation. METHODS: Three Fischer rats underwent C-RPNI surgery on transected CP nerve and underwent weekly treadmill ambulation training. After 3 months, the C-RPNI muscle component was instrumented with a recording electrode and the dermal component with a stimulating electrode. Electrodes were attached to a headcap. Ten days post-implantation, sensory thresholding was performed. Electrical pulses were applied with stepwise increasing amplitude. Sensory threshold was reached when the rat first demonstrated reaction to the stimulus (e.g. hindlimb shifting) and discomfort threshold when the rat demonstrated pain response (e.g. hindlimb withdrawal, vocalization). Three days post-thresholding, EMG were recorded from C-RPNI during volitional ambulation. A motion tracking system and neural network was used to assess the rats’ gait. EMG and kinematic data were synchronized to assess physiologic accuracy of EMG (i.e. CP nerve EMG signals during the swing phase of gait cycle). Bidirectional signaling during ambulation was assessed by determining changes in gait immediately post-stimulation. Only descriptive statistics are reported due to the study’s pilot nature. RESULTS: All C-RPNI were viable, revascularized, and reinnervated. With sensory stimulation, the average sensory threshold was 0.36 mA and the average discomfort threshold was 0.57mA with biphasic pulse widths of 12.5ms, frequency of 20Hz, and duration of 500ms. EMG recordings obtained from the C-RPNI muscle component were easily and reliably distinguishable from the baseline electrical noise with an SNR of 11.33. Synchronization of EMG and kinematic data confirmed that the EMG signals occurred during each swing phase of the gait cycle. At their baseline running rate, rats had an average of 2.2 gait cycles/second with an average EMG duration of 0.32s. Immediately post-stimulation with an amplitude 0.1mA above sensory threshold, the rats ran faster at 3.6 gait cycles/second with shorter EMG duration of 0.24ms. CONCLUSION: The C-RPNI EMG signals were physiologically accurate and easily identifiable from background electrical noise, which supports that the C-RPNI can facilitate accurate detection of motor intent needed in accurate prosthetic control. Behavioral and gait changes with stimulation suggests that the C-RPNI can facilitate sensory feedback. Therefore, the C-RPNI has the potential to facilitate physiologic bidirectional sensorimotor signaling needed for naturalistic prosthetic function.