A 32-Channel Low-Power Neural Recording System with Continuously Monitoring for ECoG Signal Detection

Abstract

In this paper, a 32-ch low power neural recording system is developed to analyze neural activities. As neural activity recording system plays important role in neuroscience, as well as in the neuroprosthetics devices, treating of neurological disease and recovery of disabilities, research has been developed over several decades. For example, by monitoring neural signals, we can predict the behavior of paralyzed patients or explain the causality between behavior and neural activity. A conventional recording system with an integrated circuit has several limitations in terms of noise and power consumption. Here, we propose a 32-channel low power fully implantable neural signal recording system. Low noise amplifier, low pass filter, analog MUX, and shift register are designed for full system. Experiment results show low noise and frequency response of the system. Also, in the experiment that measured ECG signal, up to 100uV small signal enable to be detected clearly. For the 32-ch neural recording system, only 1.2V supply voltage is needed with a total power consumption of 50uW. The total gain is 54dB with input referred noise of 3uVrms. Bandwidth is 2Hz~500Hz for targeting ECoG signal. The system is designed with a standard 0.18um CMOS technology, to measure the neural signal with very low power consumption.