The Effect And Evaluation Method Of Virtual Reality Games On Attention

Attention has a significant impact on advanced cognitive abilities such as rational thinking,learning,logical reasoning and decision-making.With the development of virtual reality technology,the research of virtual reality technology combined with gamification in medical assistance and online education is a trend.The study of the impact of virtual reality games on attention is of great significance for the treatment and education of attention deficits.The current research lacks quantitative analysis of the impact of virtual games on attention.In the medical and educational fields,most of researches focused on how to improve attention.However,the research on attention evaluation methods has been neglected.In response to these issues,this thesis studies the attention evaluation method and the influence of virtual reality games on the attention of normal mental adults.With the development of biomedicine,EEG signals have been confirmed to be closely related to attention.This thesis uses feature extraction,machine learning and other related theoretical methods to classify attention state by attention-related EEG features to realize the research of attention evaluation methods.Furthermore,the attention state of virtual reality games is quantified and evaluated in combination with statistical correlation theory.The influence of virtual reality games on normal mental adults is objectively evaluated.The work done in this paper includes:(1)Two parts of the experiment were designed.The first part is the attention test experiment.In the experiment process,multiple channel EEG signals are collected through the brain electric cap.These parts of the experiment data are used as standard data sets to study the attention evaluation method.The second part is a comparison experiment between virtual reality games and 3D games.This paper uses the classification model built in the first part of the data to classify the attention-related EEG signals collected in virtual reality and 3D game environment.(2)In this thesis,combing EEMD with improved wavelet threshold to remove noise for original signal so that the pure EEG signal will be obtained.The method is compared with the reference method.The performance of noise reduction can be evaluated by SNR and RMSE.Compared with the similar methods,the proposed method has larger signal-to-noise ratio and smaller root mean square.Therefore,the algorithm denoising effect of this paper is better than similar algorithms.This method is also applicable to other EEG signal related studies.(3)In terms of attention-related EEG feature extraction,power spectrum analysis,Hilbert-Huang transform and sample entropy methods are used in this thesis.SVM and LSTM algorithm are used to establish the classification model according to the EEG characteristics.The accuracy of the two algorithms is 90.625% and 96.3%.The classification effect of long and short memory networks is better than that of support vector machines,so this model is chosen as the classifier.(4)Comparing the impact of the same game on attention in 3D and VR environments,from the perspective of overall data,in the virtual reality game participants' game performance data is superior to ordinary 3D games.Participants' game performance data is significantly correlated with the game environment.LSTM is used to classify the EEG of 3D and VR.The result shows the degree of attention has a significant correlation with the game environment.Due to individual differences,the time to reach the peak of attention is not significantly different from the game environment.But the time when the participants reached the peak of attention was a significant difference with the gender.The time for the male to reach the peak of attention in the VR environment was shorter than that of the female.The results of this experiment can prove that virtual reality games have a positive effect on the attention of normal adults.The results provide a basis for using virtual reality games as attention training and testing in medicine and education.