Design And Implementation Of A Multi-channel Data Acquisition System With Brain Computer Interface

The extraction of EEG signals related to biological behavior information from the brain is of great significance to the research and understanding of the feedback and function of brain information and the brain-computer interface(BCI).In recent years,neuronal action potential pulses have been collected by using microelectrode array technology,and this action potential is also called a spike.Due to the small amplitude of the spike and the susceptibility to interference,it is necessary to simultaneously record the signals of surrounding groups of neurons through multiple closely arranged channels.However,in the case of multi-channel recording signals,the high-speed data collection and uploading and the real-time storage and display of the upper computer have brought us many challenges.In view of the problems caused by the multi-channel high-speed acquisition of spike signals,this paper studies and implements three aspects:real-time acquisition of spike signals,real-time data acquisition and storage display of PC desktop software,and cloud data upload and storage:1.Real-time acquisition of spike potential signals:connect high-speed ADC sampling modules through multiple SPI interfaces to collect spike potential signals.Each ADC sampling module has eight acquisition channels.After data integration is completed,it is uploaded to the upper computer and supports USB and Ethernet communications.And to ensure the scalability of the system,the reliability of data acquisition can still be ensured after the acquisition channel is expanded.2.PC desktop software real-time data acquisition and storage display:desktop software is connected to the embedded acquisition system via USB,and the best storage solution is found by comparing various storage methods and the size of the operating data block.And through the use of Qcustomplot and OpenGL acceleration and other technologies to complete the real-time display of data.3.Cloud data upload and storage:the host computer software pushes data to the cloud through tcp,the cloud realizes efficient data push through the MPMC asynchronous multi-threaded model,and realizes efficient data storage through technologies such as zero copy.