Wireless Communication Protocol for backscattered-based Neural Implants

This work presents a novel protocol for bi-directional wireless communication with neural implants that contributes to the growing field of closed-loop brain-computer interfaces (BCIs). BCIs are an emerging technology for studying and treating neurological disorders, such as spinal cord injuries (SCI). Furthermore, BCI heavily rely on neural implants as a crucial element, because they hold the potential to restore functionality of paralyzed limbs. The proposed protocol presents an open configuration to enable neural implants to communicate wirelessly with an external reader. Because computation to extract movement intention is performed externally, computing power is nearly unlimited and the energy consumption of the implant is reduced drastically. To validate the proposed protocol, the downlink (reader to implant) was implemented on a software defined radio (SDR) running GNU-Radio toolkit with custom communication blocks. The uplink (implant to reader) was implemented on an FPGA. Finally, to validate the movement intention decoding, pre-recorded neural data was backscattered from an FPGA-based implant and the decoding was executed successfully.