多通道实时神经信号采集与峰电位检测系统

A multi-channel system for real-time Acquisition and Detection of neural spikes

研制了一种可应用于在体神经检测的新型多通道实时神经信号采集与峰电位检测系统。该系统主要包括自主研发的植入式硅基微电极、集成多通道神经信号放大电路以及模数转换器(ADC)和数字信号处理器(DSP)等部分商用器件。该系统采用局部非线性能量算子(NEO)对采集到的32通道神经信号进行同步峰电位检测,并在终端屏幕上显示检测结果。局部NEO峰电位检测算法硬件实现简单,能够完全自主地对神经信号峰电位进行实时检测。在生理盐水模拟的在体实验环境中的测试表明,整个系统的信噪比(SNR)为27.4dB,在此信噪比的条件下,局部NEO峰电位检测算法的检测精度可达95%。

The research on multichannel, realtime acquisition and detection of neural spikes was conducted, and a no vel multichannel system for realtime acquisition and detection of neural spikes was built up by using the selfde signed micromachined silicon recording probes and integrated multichannel biomedical amplifier circuit, as well as some commercial devices such as analogtodigital converters and digital signal processor （DSP）. The local non linear energy operator （NEO） algorithm was applied to the 32channel system for detection of neural spike signals. The detection results were displayed on the PC screen. ＂Local NEO＂ algorithm can work autonomously and is sim ply implemented by hardware. The SNR of the system tested in a saline experiment environment was 27.4dB, while the accuracy of the spike detection algorithm was 95 %.