探討動態刺激對不同的認知狀態下

Abstract

在長時間或單調的駕車環境裡，駕駛很容易減低他們的警覺心或注意力。昏睡的駕駛沒辦法專心地開車，會導致一些錯誤的車輛操縱。他們處理訊息的速度與記憶的能力都變差，而開車技巧隨著警覺心的降低開始變糟。之前的研究大多是在靜態的開車環境裡，利用行為上的狀態或生理訊息去預測駕駛的昏睡程度。然而有一些研究發現動態的刺激會影響腦電波(electroencephalogram, EEG)α頻帶(8~12 Hz)能量的變化，並且當成警覺心的指標。在真實的駕駛裡，動態刺激對利用神經活動偵測昏睡程度準確性影響的程度仍是未知。因此我們研究的目的在於有系統的描述動態刺激對不同認知程度的大腦活動影響，特別是在昏睡的部份。 我們利用虛擬環繞場景結合六軸動態平台，獨立成份分析(Independent Component Analysis, ICA)和時頻分析研究從清醒到昏睡時的腦電波活動，並比較平台動與不動的差異。本實驗結果顯示，當受測者昏睡程度增加，使其駕車的能力下降，發現此時大腦枕葉區(occipital)在偏移事件發生前之腦波的α頻帶能量會增加。相似昏睡程度也使偏移事件發生後之腦波的α頻帶能量下降的時間點延後，並增加持續下降的時間。在相同的行為反應下去觀察平台動時腦電波從清醒到昏睡的變化比平台不動時更明顯。本研究的結果第一次證明了動態刺激對虛擬駕車環境的重要性，更進一步指出腦電波的變化比行為狀態更能靈敏地反應出駕駛的昏睡狀態。

It has been found that drivers easily to reduce their vigilance or attention during the prolonged or monotonous driving. The drowsy driver can’t focus on their driving task and tend to commit on manipulating errors. Their information processing speed and working memory capacities are decreased and drastic changes on their task performance occur along with the reduction of the vigilance. Most previous studies that tried to figure out the useful features from behavioral performances or physiological signals for predicting driver’s drowsiness level were done in a static driving environment. However, some studies already showed that the kinesthetic stimulus had influences on fluctuations of brain dynamics especially near the alpha band power, which already used as an index of the vigilance. To what extent the kinesthetic stimulation would affect the accuracy on the predicting drowsiness level from neural activities in real driving is still unclear. Therefore, the aim of this study is to systemically characterized effects of kinesthetic stimulation on the brain activities under different cognitive state, particularly under the drowsiness condition. We used the 3 dimensional surrounded virtual reality scene combined with the six degree motion platform, the independent component analysis (ICA) and time-spectral analysis to explore the fluctuations in spectral dynamics of maximally independent EEG activities from alter to drowsy with or without the enabling of the motion platform. Results showed that subjects’ drowsiness level was increased with the deteriorated of the driving performance which reflected on the tonic increases of the power spectral baselines near the alpha band in the occipital components. The similar drowsy effects also revealed on the changes of the phasic alpha suppressions including the delaying its onset and increases its mean prevalence. With the same behavioral performances, changes on EEG dynamics from alert to drowsiness were further enhanced when the motion platform was enabled. Results of this study first demonstrated the importance of the kinesthetic stimulation in the simulated driving studies. Furthermore, this study also first revealed that the EEG dynamics is more sensitive than the behavioral performance for correctly detecting driver’s drowsiness level.