Study And Design Of Low Noise And High Precision EEG Acquisition System

EEG signal is a kind of weak signal,which is susceptible to external noise.The control of signal to noise ratio(SNR)is the key to improve the accuracy of EEG signal.On the one hand,we should reduce the noise source as much as possible.On the other hand,we need to improve the amplitude of the effective signals and the accuracy of the sampling.In order to meet the requirement of high signal to noise ratio(SNR)of EEG acquisition,a low-noise and high-precision EEG acquisition system is designed in this dissertation.The system consists of three parts: active EEG electrodes,EEG acquisition board and PC monitor.Active EEG electrodes achieve acquisition and amplification of EEG signals.EEG acquisition board achieves signal filtering and analog to digital conversion of EEG signals,and sends the EEG data to PC monitor through serial port.PC monitor achieves the display of EEG signals.To solve the problem in current brain motor that weak EEG signals are susceptible to interference in the process of EEG signals’ transmission through the long wire from the dry electrode to the brain motor,active EEG electrodes,which consist of dry electrodes and amplifiers,are used to collect and amplify the EEG signals before EEG signals are transmitted to the EEG acquisition board to get higher SNR of the EEG signals under the same external noise conditions.To solve the problem that traditional shield transmission line’s parasitic capacitance and parasitic impedance are easy to generate leakage current which brings interference.This dissertation uses the double shield wires,whose outer metal shields are connected to the ground reference potential and inner metal shields are connected to EEG acquisition board’s low impedance PCB loop and the bias potential.These are used to reduce the interference from external noise to EEG signals transmitting on the line and improve SNR,and in the meantime ensuring shielding effect.In order to reduce the requirement of high gain amplifier circuits and improve the accuracy of system,EEG acquisition board uses 24-bit and high-efficiency-bit ADCs.Through the comparative test,there is the result that16-bit ADC can achieve analog-to-digital conversion of signals with peak-to-peak values as low as 61 mVpp,and the 24-bit ADC used in this dissertation can achieve analog-to-digital conversion of the signals with peak-to-peak values as low as 5mVpp,down to about 1/12 of 16-bit ADC ’s.So,the EEG signals,whichare amplified and filtered by active EEG electrodes and filter-regulator circuits,are sent to ADC circuits directly,and don’t need extra amplify circuits to amplify before the ADC circuits.In order to increase the expansion of the system,EEG acquisition board’s four hardware channels use 96 kHz sample-rate ADCs,and each hardware channel can be extended to multiple channels by using a external multiplex-selector switch to meet the application of extraction and analysis of brain’s multiple sites.FPGA and memory have a certain margin in design for enriching function of the system in future.This dissertation completes the design of hardware and software of low-noise and high-precision EEG acquisition system.The system has four hardware channels and the parameters of each hardware channel are as follow: bandwidth of 0.5-100 Hz,analog-to-digital conversion accuracy of 24 bit and maximum sampling rate of 96 kHz.The function of low-noise and high-precision EEG acquisition system designed in this dissertation is verified by test.Compared with the system without using the transmission shielding technology designed in this dissertation,when the EEG signals are amplified to the maximum peak-to-peak values of 15 mV,the system designed in this dissertation can reduce the peak-to-peak noise of about 0.8mV.