High-density microelectrode arrays (MEAs) permit the ensemble recording of large populations of neurons which mediate sensory and motor processing, perception, and learning. MEAs have the potential to monitor functional alterations in neuronal circuits within awake and behaving animal models of neurological disorders and serve as brain machine interfaces to provide real-time data to control assistive devices in persons with high level paralysis. A critical issue limiting the utility of wireless MEAs for basic science and clinical applications is the problem of the large bandwidth requirements imposed by measuring and transmitting the activity from many neurons simultaneously.
The objective of this project is to develop a scalable, low-power neural signal processor for extracting biologically relevant information from high-yield neural data in real-time prior to wireless telemtry. The information sought consists of neuronal spike trains and local field potential activity. The challenge in this project is the ability to extract this information within the constraints imposed by implantability requirements such as chip size, power dissipation, real-time operation, and limited telemetry bandwidth. Our approach relies heavily on statistical signal processing and is typically validated through adaptive models of neuronal signaling and benchmarked against wired data acquistion systems commercially available. The intended application of the system under development is geared towards large-scale monitoring of brain activity for basic neuroscience research and clinical Brain Machine Interface applications.