A Causal Brain Network Approach For Driving Fatigue Relief

Driving fatigue is recognized as one of the main factors causing traffic accidents in countries around the world,and its harm to road safety is very great.Therefore,efficient mitigation and accurate detection of driving fatigue is of great importance to improve road safety.This topic analyzes the electroencephalogram(EEG)signals of subjects under the driving fatigue relief experiment based on the causal brain network method,and proposes an efficient driving fatigue relief method by comparing the differences of different driving fatigue relief methods.The main research of this paper is as follows:(1)Driving fatigue relief method and system design.In this paper,a representative acupoint electrical stimulation relief was selected as an external objective stimulus,and a novel subjective cognitive driving subtask relief method(arousal mode)was designed,and finally an integrated stimulation relief method combining objective stimulation and subjective cognition was proposed.Arousal mode mainly consists of customized driving subtask configuration module,task flow adjustment module and keyword reminder device.The customized driving subtask module automatically generates driving subtasks about subjective cognition,and then the keyword reminder device guides the driver to carry out thinking-type cognitive activities to adjust the current cognitive load so as to relieve driving fatigue.In addition,this paper proposes a method for simultaneous electrical stimulation of three acupoints(Láogóng point(劳宫PC8),Nèiguān point(内关PC6),and Hégǔpoint(合谷L14))relative to the traditional single acupoint stimulation method,and designs a new electrical stimulation glove to overcome the disadvantage of inconvenient traditional wearing of electrodes.Finally,a driving fatigue relief system was built and driving fatigue relief experiments were conducted according to the three relief methods.(2)Analytical validation of a driver fatigue relief system based on a causal brain network approach.The causal brain network was constructed by orthogonalized partial directional coherence analysis of the EEG signals of the above experimental subjects,and then the causal brain network of the control group was analyzed based on graph theory,and the results showed that the causal brain network characteristic parameters(clustering coefficients and global efficiency)could accurately reflect the state of driving fatigue.On this basis,the differences in the causal brain network characteristic parameters(clustering coefficient,global efficiency,out degree,and in degree)between the three relief experiments and the control group were compared and analyzed,and the results showed that the integrated stimulation method proposed in this paper was more sustainable and efficient for driving fatigue relief.(3)The brain mechanisms underlying the generation of driving fatigue and its alleviation were investigated.By analyzing the causal information flow in the driving fatigue relief control group experiment,the mechanism of driving fatigue generation based on causal brain network was revealed: as driving time increased,the capacity of frontal brain areas to process executive and decision-making information weakened,as shown by the shift from causal target to causal source in the left prefrontal area and from causal source to causal target in the right prefrontal area,while the capacity of occipital brain areas to process visual information weakened and occipital areas shifted from causal target to causal source.Finally,the causal information flow of the three mitigation experiments was compared,and the results showed significant differences in the effects of external objective stimuli versus subjective cognitive methods of mitigation on the right prefrontal lobe of the brain.