Back to Annual Meeting Program
Prototype Sensory Regenerative Peripheral Nerve Interface for Artificial Limb Somatosensory Feedback
John V. Larson, BS1, Melanie G. Urbanchek, PhD1, Jana D. Moon, BS1, Daniel A. Hunter, PhD2, Piyaraj Newton, BS2, Philip J. Johnson, PhD2, Matthew D. Wood, PhD2, Theodore A. Kung, MD1, Paul S. Cederna, MD1, Nicholas B. Langhals, PhD1.
1University of Michigan, Ann Arbor, MI, USA, 2Washington University School of Medicine, Saint Louis, MO, USA.
PURPOSE: Major advances in both robotic and neural interface technologies have improved functional control of advanced prosthetic limbs. However, despite considerable success in improving motor function, interface technology supporting long-term somatosensory feedback remains unavailable. For this purpose, our laboratory has developed a prototype sensory regenerative peripheral nerve interface (sRPNI) to receive signals gathered from tactile sensors on an advanced prosthesis and seamlessly transfer this signal to the residual nerves of an amputated limb. As the first major investigation of sensory RPNI technology, this study examined sRPNI viability, reinnervation, and signal fidelity.
METHODS: Twelve rats underwent sRPNI fabrication (Figure 1). To demonstrate reinnervation, after a convalescent period of 3 - 4 months, in vivo sensory recordings were obtained from the sRPNI while electrically stimulating the sural nerve. sRPNI muscle, sRPNI sural nerve, and control EDL muscle and sural nerve from the contralateral limb was then harvested for mass comparisons, and evaluated through the application of both histomorphometric and immunohistochemical techniques.
RESULTS: Upon gross examination, sRPNIs appeared well vascularized, healthy, and viable. The average mass of sRPNI muscle was 75.2% that of control EDL muscle (SD 0.127). Neural-evoked responses from sRPNIs were successfully recorded, confirming sRPNI viability and signal transfer (Figure 2). During electrophysiological testing, the average stimulation threshold required for eliciting a compound action potential was 143.8 μA in 3-month sRPNIs and 99.6 μA in 4-month sRPNIs, indicating continued regeneration and reinnervation. Furthermore, 3-month sRPNIs demonstrated an average compound action potential peak-to-peak amplitude of 0.68 mV, compared with 2.27 mV in 4-month RPNIs. As expected, both groups exhibited similar muscle latencies. Histological analysis of sRPNI muscle revealed normal architecture with the fiber sizes and distributions comparable with control muscle (Figure 3). sRPNI sural nerve also demonstrated healthy nerve fiber architecture consisting of large and small myelinated fibers nearly equivalent to control sural nerve. Immunohistochemical evaluation revealed minimal cellular inflammatory response and numerous axons within sRPNI muscle tissue.
CONCLUSION: Comparisons of electrophysiological responses, muscle mass, and histology indicated that sRPNI muscle and nerve fiber recovery approached equivalence. These findings demonstrate that freely transferred muscle becomes reinnervated and viable when neurotized by a transected sensory nerve. Furthermore, electrophysiological signal is successfully transmitted between the sRPNI and residual nerve. Through electrical stimulation of this functional sRPNI, there is enormous potential for enhancing the recovery and quality of life of thousands of amputees by restoring the sense of touch.
Acknowledgement: This work was sponsored by the Defense Advanced Research Projects Agency (DARPA) MTO, Pacific Grant/Contract No. N66001-11-C-4190.
Back to Annual Meeting Program