Study Of Driver’s Brain Function Assessment Based On Near-Infrared Spectroscopy And Virtual Reality

Traffic accidents are always a problem ever since transportation vehicles had been developed,causing casualties and financial losses.How to decrease the probability of traffic accident and make transportation safer is a hot research topic.The key to less traffic accident lies in a deeper understanding of the mechanism of traffic accident.It is how much we know about traffic accident mechanism that depends how many accidents we can prevent.Vehicle driver is one of the main factors of the "Driver-Vehicle-Road" system,and decides whether an accident will occur at a great extent.Thus the probability of a driver causing an accident can be predicted if the driving ability of the driver could be evaluated.Driving ability is a direct reflection of brain function.This thesis is a preliminary attempt of drivers’ brain function evaluation using functional near-infrared spectroscope(fNIRS).The content of this article can be divided into three parts.Part one,the developing of an adaptive superficial layer noise canceling algorithm based on short"Source-Detector"(S-D)separation for fNIRS devices.Starts from the measure mechanism of the fNIRS devices and the anatomical structure of human head,summarizing the previous discovery of short S-D prob arrangements,we designed two short S-D prob arrangements for two specific fNIRS devices.Then,aiming at the demand of brain functional connectivity analysis of wavelet phase coherence and that of brain effective connectivity analysis of phase coupling function,we developed an adaptive noise canceler based on continue wavelet transform.Part two,the development of the driving simulator based on Unity 3D.The mechanical structure is developed based on author’s undergraduate graduation project thesis.The present work including sensors’ selection and installation,the development of the single chip machine(SCM)system for sensors,the development of the virtual reality scene for different experiment task designs.Part three,we conducted an experiment of driving task using the driving simulator while recording the brain function of the subjects using fNIRS devices.Then the recorded data was pre-processed using the adaptive noise canceler that we developed.After that,wavelet phase coherence analysis was applied to both pre-processed data and original data.Statistical result suggests that the adaptive noise canceler is effective.Above all,this thesis developed an adaptive superficial layer noise canceling algorithm for fNIRS devices,and conducted an experiment of simulation driving task to confirm the effect of the algorithm.For brain function measurement using fNIRS devices,this algorithm can improve the measurement precision to a further degree,which might be useful for evaluating the drivers’ brain function in a real-time,dynamic and precised way.