Fast-LIFE Neural Interfaces Provide Naturalistic Sensation and Individual Digit Control During Robotic Hand Use
Jonathan Cheng, MD1, Edward Keefer, PhD2.
1University of Texas Southwestern Medical Center, Dallas, TX, USA, 2Nerves Incorporated, Dallas, TX, USA.
PURPOSE: 18-25,000 upper limb amputations are estimated in the US annually. We previously described using Fast-LIFE interfaces to target fascicular groups within the residual forearm nerves of upper limb amputees. We have developed a suite of tools to provide these patients with naturalistic sensorimotor control of robotic hands.
METHODS: Case study of our 5th clinical subject, a transradial amputee, who represents a culmination of our efforts in the DARPA HAPTIX direct-to-translation research program. He was implanted with 4 Fast-LIFE + cuff electrodes in the residual median and ulnar nerves of the forearm.
RESULTS: History, screening, surgery, and electrodes will be discussed. Sensory restoration data over the 6+ month implantation will be presented, including receptive fields, body-centered proprioceptive maps, and perception thresholds. Our group is the first ever to develop independent motor control of the individual digits of a robotic hand using signals recorded from the peripheral nerve. This is achieved using a custom microchip for acquiring neural signals with hardware-based artifact rejection. Motor intention is decoded using an artificial intelligence (AI) process, with individual decode algorithms persisting for >1 month. Insights on the role of compression neuropathy in phantom limb pain will be explored.
CONCLUSION: Fast-LIFE interfacing of the forearm nerves can be combined with patient-specific sensory stimulation strategies, custom-fabricated neural recording hardware, and AI decoding of recorded neural signals, to provide a transradial amputee with closed-loop naturalistic sensorimotor control of a robotic hand prosthesis. There may be a role for peripheral nerve decompression in alleviating phantom pain.
Back to 2019 Abstracts