The NeuroProcessor [electronic resource] : An Integrated Interface to Biological Neural Networks / by Yevgeny Perelman, Ran Ginosar.

Recording From Biological Neural Networks -- The Neuroprocessor -- Integrated Front-End for Neuronal Recording -- NPR03: Mixed-Signal Integrated Front-End for Neuronal Recording -- Algorithms for Neuroprocessor Spike Sorting -- MEA on Chip: In-Vitro Neuronal Interfaces -- Conclusions.

Neuronal electronic interfaces carry significant potential for scientific research and medical applications. Neuroprosthetics may help to restore damaged sensory and motor brain functionality. Neuronal interfaces are evolving into complex micro-fabricated arrays of hundreds or thousands of sensors, and require tighter integration, advanced embedded computation, and wireless communication. At the very least, the electronic circuit of the implanted neuronal interface must acquire the data and transmit it outside. However, the huge data rates produced by large-scale neuronal interfaces exceed the communication bandwidth provided by low-power wireless channels. Hence, extensive embedded computations must be integrated into the interface in order to reduce the amount of transmitted data. This book presents the Neuroprocessor, a novel computational neuronal interface device implemented in VLSI technology. In addition to neuronal signals acquisition, it can process the data, generate stimuli and transmit the data over wireless channels, while using minimum electric energy. The NeuroProcessor opens with a brief background on neuronal communication and microelectrode recording. It introduces three generations of the Neuroprocessor and presents their architecture, circuits and algorithms. Applications to a miniature head-stage for in-vivo experiments and multi-electrode arrays for in-vitro studies are described.