Durable and Fatigue-Resistant Soft Peripheral Neuroprosthetics for In Vivo Bidirectional Signaling

Abstract

Soft neuroprosthetics that monitor signals from sensory neurons and deliver motor information can potentially replace damaged nerves. However, achieving long­term stability of devices interfacing peripheral nerves is challenging, since dynamic mechanical deformations in peripheral nerves cause material degradation in devices. Here, a durable and fatigue­resistant soft neuroprosthetic device is reported for bidirectional signaling on peripheral nerves. The neuroprosthetic device is made of a nanocomposite of gold nanoshell (AuNS)­coated silver (Ag) flakes dispersed in a tough, stretchable, and self­healing polymer (SHP). The dynamic self­healing property of the nanocomposite allows the percolation network of AuNS­coated flakes to rebuild after degradation. Therefore, its degraded electrical and mechanical performance by repetitive, irregular, and intense deformations at the device？nerve interface can be spontaneously self­recovered. When the device is implanted on a rat sciatic nerve, stable bidirectional signaling is obtained for over 5 weeks. Neural signals collected from a live walking rat using these neuroprosthetics are analyzed by a deep neural network to predict the joint position precisely. This result demonstrates that durable soft neuroprosthetics can facilitate collection and analysis of large­sized in vivo data for solving challenges in neurological disorders.