Real-Time Neural Signal Sensing and Spike Sorting System Using a Modified Zero-Crossing Feature with Highly Efficient Data Computation and Transmission

Abstract

Highly efficient data computation and transmission are necessary to implement the high-performance analysis of real-time neural signals. In a previous study, a zero-crossing feature (ZCF) was proposed as a computationally efficient spike feature that enabled enhanced data transmission efficiency. In this paper, a real-time neural signal sensing and spike sorting system using a modified ZCF is presented. In the system developed, the ZCF extraction algorithm is modified to improve spike sorting accuracy. The data transmission efficiency is highly enhanced by the use of a minimized data set size. The size of the data set is reduced by 90.7% as compared with the raw data transmission. In an in vivo experiment using spinal nerves of a Sprague Dawley rat, neural signals were successfully detected and amplified with a maximum gain of 85.1 dB and a signal-to-noise ratio (SNR) of 19.7 dB. This experimental result shows that the neural spikes were more accurately detected and classified with the modified ZCF extraction algorithm as compared with the original algorithm. From these results, it is expected that this system can be effectively used for high-performance real-time neural signal analysis.