本研究探討在不同聲音情境下專注與安靜放鬆之腦波狀態。實驗使用腦波儀測量人體頭部表面電訊號，分析受試者於安靜、管弦樂、噪音和混合音等聲音情境下之腦波訊號，以作為後續專注度分析數據。腦波經由訊號處理來濾除雜訊，以減少雜訊干擾，降低分析上的錯誤。接著使用碎形維度(fractal dimension)、近似熵(approximate entropy)、與經過希爾伯特-黃轉換(Hilbert-Huang transform)後所得的alpha與beta頻段功率作為分類的特徵指標，並於後續利用接收者操作特徵曲線(receiver operating characteristic curve)與支持向量機(support vector machine)兩種分析工具進行腦波情境分類，判別受試者在各種情境間的腦波差異狀況及不同情境對專注力影響。;In this study, we discussed the brain wave modes of concentration and relaxation under various sound conditions. Twenty-four participants were recruited to play the attention game in the quiet, orchestra, noise and mixed sound states through human factors experiments. The brain waves were filtered out by signal processing to reduce noise interference and analytical errors. Then, we used the fractal dimension, the approximate entropy and the alpha and beta band power as features in the brainwave analysis. Furthermore, the brainwave classification was performed by using two analysis tools, including receiver operating characteristic curve (ROC curve) and support vector machine (SVM), to identify the difference in brainwave status between different contexts and the effect of different contexts on concentration.
