The innumerable benefits of an improved understanding of the human nervous system motivates this research for developing improved technology for hardware critical to bioelectric signal recording and stimulation.
Much effort has been made in recent history to investigate the human nervous system and to correlate activity in the peripheral nervous system to the behavior of the brain and the regulation of bodily functions. Systematic algorithms to analyze complex neural-signal responses are desired in designing brain-machine interfaces. However, for implantable device applications, the hardware resources such as area and power consumption are limiting factors to many state-of-art approaches on printed circuits. Therefore, the realization of neural signal processing algorithms on implantable devices becomes a challenging yet crucial problem. This project aims to develop algorithms and VLSI architectures of neural signal processors for implantable device applications. The fidelity of neural signals and its robustness to noise prevailing in implant environments can be maintained via digital signal processing (DSP) techniques. Dedicated DSP architectures process incoming neural signals and extract features of interest with the lowest possible computational cost. The neural signal processors not only allow users to study the neural behavior in real time but also contribute to a reduction in data-transmission rate and hence power consumption dominated by the transmitter.