Research On Wireless EEG Acquisition Technology And Intelligent Terminal Application Algorithm

Scalp surface electroencephalogram(EEG)is the sum of postsynaptic potentials formed by the synchronous discharge of cortical pyramidal neurons.Traditional EEG acquisition generally adopts wired transmission,which is cumbersome in operation and poor in wearing comfort,which is not conducive to the current more and more flexible research scenes and applications.Therefore,EEG acquisition technology based on wireless transmission is very important.In the application of EEG acquisition technology,in recent years,the research of noninvasive EEG and f MRI fusion analysis has been widely carried out,which realizes the high-precision observation of the brain in time and space,which is conducive to solving the problems of EEG traceability and positioning.At the same time,it also puts forward the requirements of magnetic compatibility for the acquisition equipment,which should not only meet the safety of subjects,but also ensure the quality of data;In addition,the current drug abuse phenomenon is becoming more and more serious.More and more studies show that drug abuse seriously damages the brain nerve function of smokers,and the drug abuse detection technology based on EEG will help to improve the problems of long time limit and cross infection risk of traditional biochemical detection methods.Therefore,the research of wireless EEG acquisition technology and its application in EEG f MRI and drug abuse detection have extensive practical value.In view of the above problems,In this study,a EEG acquisition system based on wireless transmission is designed and developed,which has the characteristics of high density and high expansion.The availability of this set of wireless EEG acquisition system is verified by experiments.In addition,this paper does magnetic compatibility processing for this set of wireless EEG acquisition system,and collects the in vivo animal experimental data in the nuclear magnetic environment for verification and analysis.This paper also explores the feasibility of using resting EEG data for drug abuse detection and analysis.The main content and innovations of this article are as follows:1.Research and development of wireless EEG acquisition system: This paper implements a 32 channel EEG acquisition and wireless transmission system.The single channel sampling rate can reach up to 16 KHz,and can be expanded to 1024 channels through daisy chain cascade,and the single channel sampling rate can reach up to 500 Hz.At the same time,the system is equipped with active preamplifier module and independent external reference voltage module to improve the signal-to-noise ratio.In addition,the main control program based on state machine and special wireless communication protocol based on CRC verification are designed to ensure the high availability of wireless transmission system and the integrity of signal transmission.The performance of this set of wireless EEG acquisition system is tested and verified ? Wave acquisition experiments verify its effectiveness.2.Development of magnetic compatibility processing and gradient artifact removal algorithm for wireless EEG acquisition system.In response to the needs of EEG-f MRI multi-modal research,this article has done magnetic compatibility processing for this wireless EEG acquisition system,and expanded the MRI scan sequence synchronization signal acquisition module.Aiming at the gradient artifact signal coupled in the EEG signal during multimodal synchronous acquisition,this paper develops an artifact removal algorithm based on autoencoder and semi-classical signal analysis.3.Application of EEG based drug abuse detection algorithm: This paper explores the application of machine learning algorithm based on resting EEG in drug abuse detection scene,designs the corresponding drug abuse detection algorithm model,and expands the complexity measurement,nonlinear dynamics(recursive quantitative analysis)and wavelet transform coefficient features on the basis of time-domain and frequency-domain features,including 57 feature inputs.The empirical method and tree structure are combined to select the feature vector of Engineering simplification.The model was verified by the modeling experiment of injecting drugs in rats,and the results achieved more than 90% classification accuracy.Compared with the traditional time-consuming biochemical analysis and detection methods,this method can provide a more reliable and convenient recognition paradigm for drug abuse detection.The complexity of experimental design is less,the detection cost is lower,but the accuracy is equivalent.