How Does the Degree of Valence Influence Affective Auditory P300-Based BCIs?
How Does the Degree of Valence Influence Affective Auditory P300-Based BCIs?
复制标题
DOI：
10.3389/fnins.2019.00045
复制
发表时间：
2019-02-19
影响因子：
4.3
通讯作者：
Nakagawa, Seiji
中科院分区：
文献类型：
--
作者：
Onishi, Akinari;Nakagawa, Seiji
脑机接口（BCI）将大脑信号转换为控制设备和通信的命令。BCI使残疾人能够与外界交流。积极和消极的情感声音已经被引入到基于P300的BCI中;然而，效价的程度（例如，非常积极或正面的）影响的BCI尚未被研究。为了进一步研究情感声音在基于P300的脑机接口中的影响，我们将具有五个效价的声音应用于基于P300的脑机接口。声音的效价范围从非常消极的非常积极的，确定由谢弗的方法。通过波形分析评估声音效价对BCI的影响，然后评估离线刺激分类准确性。结果表明，P300的晚期成分对非常积极和非常消极的声音的反应比对其他声音的反应表现出更高的点-双稳态相关系数。从感兴趣区域估计离线刺激分类准确度。分析表明，非常消极的声音达到了最高的准确度，非常积极的声音达到了第二高的准确度，这表明可能需要非常积极的声音和非常消极的声音来提高准确度。
A brain-computer interface (BCI) translates brain signals into commands for the control of devices and for communication. BCIs enable persons with disabilities to communicate externally. Positive and negative affective sounds have been introduced to P300-based BCIs; however, how the degree of valence (e.g., very positive or positive) influences the BCI has not been investigated. To further examine the influence of affective sounds in P300-based BCIs, we applied sounds with five degrees of valence to the P300-based BCI. The sound valence ranged from very negative to very positive, as determined by Scheffe's method. The effect of sound valence on the BCI was evaluated by waveform analyses, followed by the evaluation of offline stimulus-wise classification accuracy. As a result, the late component of P300 showed significantly higher point-biserial correlation coefficients in response to very positive and very negative sounds than in response to the other sounds. The offline stimulus-wise classification accuracy was estimated from a region-of-interest. The analysis showed that the very negative sound achieved the highest accuracy and the very positive sound achieved the second highest accuracy, suggesting that the very positive sound and the very negative sound may be required to improve the accuracy.