Research And Development Of High-Throughput Neural Data Acquisition System Based On FPGA

Brain Computer Interface technology is of great significance in the field of Brain Science Research.There are multi-level,multi-scale,and complex neural network connections with dynamic changes over time between the structure and function of the brain.For more detailed and higher-speed recording of massive neuron activity at multiple brain regions or whole brain scales,the signal recording system needs to have more recording channels,higher sampling frequency,larger parallel data processing capacity,and stronger multi-channel spike detection capability.To this end,based on the neural signal recording hardware platform,this thesis has completed the research and developement of a high-throughput neural signal recording system FPGA.The system can be configured through instructions,and supports simultaneous 512 channels neurophysiological signals acquisition at a sampling frequency of up to 30KS/s,digital filtering and neural signal data formatting;to achieve data cache by controlling DDR SDRAM storage interface,and PCI-Express x4 data bus implements instruction interaction and data transmission with the embedded GPU Tegra K1.At the same time,this paper implements manual or automatic online real-time spike detection,respectively using the threshold detection method of amplitude to achieve the manual detection function,and the automatic detection function by the template matching window method.Finally,FPGA data test and simulation experiment of the system prototype were carried out.The results of simulation tests show that the timing of each module of the high-throughput neural signal recording system is accurate,and the data flow processing and instruction configuration are in line with design expectations.The results of simulation experiments show that the neural recording system can collect,process,store and the data can be uploaded.The spike signal can be successfully detected.The performance index meets the design requirements.