A Microscale Neural Implant ASIC for Scalable Multichannel Brain-Machine Interface

Abstract

The Brain-Machine Interface (BMI) is a pivotal tool for the study and repair of the nervous system, facilitating connections between the brain and external devices. The implementation of a flexible and adaptable multichannel system is essential for effective communication and analysis of a large number of neurons in the anatomically and physiologically complex human nervous system. This paper presents the development of a wireless stimulation and recording ASIC, fundamental for creating a modular BMI system with multiple modules. Simulations reveal the on-chip antenna's peak efficiency at the target frequency within a streamlined design. The inclusive wireless power system, featuring the resonant tank, rectifier, and LDOs, delivers a 1.2 V supply voltage, allocating 22.56 μW to the recording subsystem, while providing 9.7 μW and 2.1 mW to resting and active stimulators, respectively. The spike detection frontend, utilizing an energy operator, conveys data via backscatter communication, and the stimulation backend executes 8-channel biphasic current stimulation, marking a notable stride in the evolution of flexible and adaptive BMI technology.