The Osseointegrated Neural Interface (ONI): A Rabbit Model for Chronic Peripheral Nerve Interfacing in Bone with Percutaneous Osseointegrated Connectors.
Aaron M. Dingle, PhD1, Jared P. Ness, MS1, Joseph Novello, MS2, Weifeng Zeng, MD1, Jacqueline S. Israel, MD1, Conner Feldman, BS1, Brett Nemke, BS1, Yan Lu, MD1, Aaron J. Suminski, PhD1, Mark D. Markel, DVM1, Justin C. Williams, PhD1, Sameul O. Poore, MD, PhD1.
1University Wisconsin, Madison, WI, USA, 2University Wisconsin, Madison, AB, USA.
PURPOSE: Peripheral nerve interfaces represent a paradigm shift in the treatment and prevention of amputation neuromas. Rather than simply bury a transected nerve in muscle or bone in an effort to prevent/treat painful neuromas, attention has shifted to exploiting the regenerative capacity of these nerves to carry the control signals needed to animate advanced robotic prostheses. Caveats to current interfacing technology are the motion artifact and long-term stability of these devices in dynamic soft tissue environments.
The Osseointegrated Neural Interface (ONI) represents a novel approach to peripheral nerve interfacing- utilizing the medullary cavity of the amputated long bone to house and protect the amputated nerve/delicate electrode interfaces from motion artifact and damage. The ONI is based on the transposition of nerve in bone to treat amputation neuromas, first described by Edwin Boldrey in 1943 and relies on the premise that bone provides stability and protection from external stimuli that may cause neuropathic pain. These same principals of stability and protection are also key components of any technology seeking to achieve a robust, chronic interface with the peripheral nervous system. Objective: To evaluate the stability and longevity of chronically implanted ONI devices.
METHODS: We have developed a novel dual cuff electrode implant with a percutaneous, osseointegrated connecter that enables chronic electrophysiological evaluation of the ONI. The device consists of two, bipolar cuff electrodes with percutaneous connectors secured to a stainless steel intramedullary rod. Transfemoral amputation was performed in New Zealand white rabbits. The terminal end of the transected sciatic was passed through a proximal corticotomy in the femur, threaded into the medullary cavity and out the end of the bone. The terminal end of the nerve was secured in one cuff electrode and inserted back into the medullary cavity, followed by the intramedullary rod, which was then secured with bone cement. The second cuff electrode was attached to the nerve external to the bone, proximal to the corticotomy. Stimulation (monophasic, cathodal pulses- 30µs duration, 100µA-8mA) evoked afferent and efferent compound nerve action potentials (CNAPs) were recorded 3 and 5 weeks post implant with cortical somatosensory evoked potentials (SSEPs) recorded at 5 weeks. Differences in peak CNAP amplitude were investigated with one-way ANOVA.
RESULTS: Efferent (motor) CNAPs are present at 3 weeks and improve by 5 weeks, as indicated by a significantly greater peak CNAP response from a smaller stimulus (n=6, p= ≤0.05). There were no obvious afferent (sensory) CNAPs recorded at the same time points; however, afferent SSEPs were measured. The ability to record afferent SSEPs demonstrates that there is still a connection between the distal end of the amputated nerve and the cortex, indicating that the ability to transmit sensory information through the ONI is not lost. Future work will investigate if the ability to record afferent signals improves with longer implant duration.
CONCLUSIONS: This is the first evidence that physiological activity of nerves transposed and interfaced within bone can be harnessed chronically, towards future prosthesis control via an ONI.
Back to 2018 Program