Preliminary Study On EEG-based Evaluation Of Visually Induced Motion Sickness

In recent years,virtual reality(VR)technology is developing rapidly.It is becoming more and more commercial,marketable,and industrial.VR technology creates the virtual environment through head-mounted display(HMD)or large projection screen.VR can provide users with rich three-dimensional information,which brings people strong sense of stereo and immersion.However,the most prominent problem in VR technology is that the users may have discomfort symptoms,including dizziness,nausea,and vomiting.These symptoms are recognized as visually induced motion sickness(VIMS)or virtual reality motion sickness(VRMS),which seriously threaten people's health and restrict the applications of VR technology in different fields.At present,few studies focus on VIMS in virtual reality.Therefore,it is of great significance to carry out this research.The purpose of this dissertation was to analyse the relationship between the electroencephalogram(EEG)and subjective visually induced motion sickness level(VIMSL)and find the EEG markers for VIMS evaluation,through the VR-based vehicle driving simulator(VDS)experiment.In this dissertation,VR-based vehicle driving simulator was used to induce VIMS symptoms,and the Muse device was used to collect EEG data of subjects throughout the experiment.The subjective and objective methods were combined to evaluate the VIMS.As for the subjective method,this dissertation analysed the VIMSL data of different subjects collected by verbally report.As for the objective method,the EEG data were preprocessed and then divided into non-VIMS and VIMS state according to the VIMSL.Finally,the preprocessed EEG data of different states were analyzed using the power spectrum density(PSD),gravity frequency(GF)and power spectral entropy(PSE).Our studies suggest that:(1)There were individual differences in subjective feeling of VIMS for different subjects.(2)Subjective feelings were different in repeated experiments for the same subject.(3)There were no significant differences in the mean value of PSD and mean value of PSE between non-VIMS state and VIMS state.Thus,they could not be used as EEG markers for VIMS evaluation.(4)For the following frequency bands and channels of EEG,mean values of GF between non-VIMS and VIMS states are significant: theta FP1,beta FP1,alpha TP9,alpha FP2,and alpha TP10.Mean values of GF in non-VIMS state are greater than that of GF in VIMS state.(5)For the following frequency bands and channels of EEG,standard deviations of GF between non-VIMS and VIMS states are significant: alpha TP9,alpha FP1,alpha FP2,alpha TP10,alpha FP2-FP1(the difference between FP2 and FP1).Standard deviations of GF in non-VIMS state are greater than that of GF in VIMS state,which indicates the dispersion decrease of standard deviation of GF.The results above indicated that there were individual differences in subjective feeling of VIMS for different subjects.There is a relationship between EEG and VIMS: there were significant differences in GF of EEG between non-VIMS state and VIMS state,which may be used as the marker for VIMS evaluation.The results of this dissertation can provide the basis for evaluation and reduction of VIMS in virtual reality;hence promote the applications of VR technology.