Research And Application Of High-resolution Wearable Near-infrared Spectroscopy Equipment

Near infrared spectroscopy(NIRS)is one kind of non-invasive brain neuroimaging technology,which uses the optical window of human body to detect the concentration of hemoglobin.Compared with the traditional brain nerve detection technology,NIRS has its own advantages.NIRS is non-invasive,low cost,highly wearable and has strong anti-interference ability.With the improvement of its reliability and accuracy,it has attracted more and more attention in the field of neurology research and clinical practice,providing a new way for the development of long-term brain detection and brain computer interface under natural circumstances.This paper aims to design a set of continuous-wave near infrared spectroscopy hardware system which can support 16 channel light sources and 16 channel detectors at most.First of all,we study and optimize the probe layout,collects the activity data of prefrontal brain area through NIRSIT system and analyzes the variance of each channel.Based on this analysis,we study the brain activity of prefrontal area and optimize the probe layout structure And in order to improve the practicability and expansibility,the system adopts modular design.The whole system consists of six modules: light source module,detector module,probe fixing module,motherboard,Wi-Fi module and controller module.The controller adopts the FPGA platform based on Zynq7000 to control the whole system.Combining the data parallel processing ability of FPGA and the software programming ability of MCU,the system can realize the 43.8Hz data acquisition frequency of 16*16 channels.Compared with similar equipment at home and abroad,it can realize miniaturized size and provide higher spatial and temporal resolution.In addition,a multi-channel signal acquisition mode of two-level light intensity excitation is proposed,which effectively improves the signal quality and the number of channels.On the basis of optimized layout and two-level light intensity excitation mode,the in vivo test and classification tasks are carried out.Compared with the more advanced commercial equipment NIRSIT,the classification accuracy of four kinds of frontal area cognitive tasks is 89.6% under the premise of only using 10% channels.The results show that the optimized layout and two-level light intensity excitation mode are effective.