This book will describe ultra low-power, integrated circuits and systems designed for the emerging field of neural signal recording and processing, and wireless communication. Since neural interfaces are typically implanted, their operation is highly energy-constrained. This book introduces concepts and theory that allow circuit operation approaching the fundamental limits. Design examples and measurements of real systems are provided. The book will describe circuit designs for all of the critical components of a neural recording system, including: Amplifiers which utilize new techniques to improve the trade-off between good noise performance and low power consumption. Analog and mixed-signal circuits which implement signal processing tasks specific to the neural recording application: Detection of neural spikes Extraction of features that describe the spikes Clustering, a machine learning technique for sorting spikes Weak-inversion operation of analog-domain transistors, allowing processing circuits that reduce the requirements for analog-digital conversion and allow low system-level power consumption. Highly-integrated, sub-mW wireless transmitter designed for the Medical Implant Communications Service (MICS) and ISM bands.Spike detections from the circuit were compared with labels from the generating software to determine the sensitivity and selectivity. For comparison, we also applied a threshold-based software spike detector to the same signal. The softwareanbsp;...
|Title||:||Ultra Low-Power Integrated Circuit Design for Wireless Neural Interfaces|
|Author||:||Jeremy Holleman, Fan Zhang, Brian Otis|
|Publisher||:||Springer Science & Business Media - 2010-10-29|