听觉通道语音冲突大脑皮层电位的听觉认知控制特征提取方法

认知心理学发现,视觉、听觉接收到信息有冲突时,大脑皮层电位会发生扰动,由此可探索认知冲突控制的“刺激-反应”机制。视觉认知冲突实验较多,成果丰硕,而相应的听觉实验很少,并且得到不一样的结论。本研究利用冲突和非冲突的语音信号刺激,分析研究脑电信号,提出基于三阶段听觉认知控制的时域特征模型。研究人脑听觉通道在出现语音认知冲突时的认知控制的规律下的单次试验脑电数据特征提取方法。根据得到的认知规律,单次试验脑电样本被分成3个部分。被分割的每个阶段使用时域上的平均幅值和Lempel-Ziv复杂度(LZC)进行计算,从而联合3个阶段的特征作为听觉认知脑电样本的特征。结果表明:(1)先发现的认知冲突相关的混合脑电成分“N1-P2&N2&Late-SW”分别体现了听觉认知控制的3个阶段;(2)一个更完整的听觉认知控制过程应包括3个阶段的时域特征:感知阶段:110~140 ms,识别阶段:260~320 ms,解决阶段:500~700 ms;(3)提出针对单次听觉认知控制脑电样本的特征提取方法,联合使用平均幅度和LZC可以获得最好的识别率(99.33%)。实验结果证明了提出的方法能够有效地检测听觉认知控制脑电数据,进而提供人脑认知控制能力评价的声学方法。

The method of auditory cognitive control feature extraction of cerebral cortex potential for auditory modality speech conficting

Cognitive psychology found that the cerebral cortex potential will modulate as the visual or auditory morality receive conflicting information, which can be used to explore the "stimulus-response" mechanism for the cognitive conflict control. The experiments on visual cognitive conflict is more fruitful, while little work has been conducted in the corresponding auditory ones, and inconsistent conclusions have been reported. In this study, we proposed a time feature domain of 3-stage auditory cognitive control by utilizing conflicting and non-conflicting speech signals as stimuli and analyzing and studying the corresponding electroencephalogram (EEG) signals. The paper study the feature extraction for single trial EEG sample based on cognitive control mechanism of speech conflicting. Based on the cognitive mechanism, each EEG sample for a single trial was divided into three parts. Average amplitude in the time-domain and Lempel-Ziv Complexity (LZC) was computed for the divided part of each EEG sample and the union of the feature of the threestages was used as the features of each auditory cognitive EEG sample. The study results are as follows. (1) The cognitive conflicting related complex EEG component "N1-P2&N2&Late Slow Wave (Late-SW)", which was first found, respectively represents each stage of the auditory cognitive control. (2) A more completed auditory cognitive control process should three stages time domain feature: perceptual stage: 110-140 ms, identification stage: 260N320 ms, resolution stage: 500~700 ms. (3) The feature extraction method for a single trial auditory cognitive control EEG sample was proposed, the combination usage of Average amplitude and LZC can achieved the best accuracy (99:33%). The experimental results showed that the proposed method can effectively detect the EEG data of auditory cognitive control. Therefore, we offer an acoustic technique evaluating the ability of our brain auditory cognitive control.