东南大学学报（自然科学版）2017，Vol.47Issue(3)：432-437,6.DOI:10.3969/j.issn.1001-0505.2017.03.003
单电极中潜伏期反应的听觉注意特征提取与识别
Feature extraction and recognition of auditory attention in middle latency response from single electrode
摘要
Abstract
The recognition of auditory attention and non-attention states of normal individuals are studied and realized by the extraction of the differences of middle latency response (MLR) of single electrode.First,the MLR signal is preprocessed by wavelet filtering,threshold de-artifact and coherent averaging.Then,the component wave differences of these two states are analyzed.The amplitudes of the Na,Pa,Nb waves and the traditional characteristics such as the energy,the area,the C0 complexity,and the coefficients of the auto regression model(AR) are combined into a new feature vector.Finally,the support vector machine(SVM) and the artificial neural network(ANN) are used to identify the target by using the traditional feature vector and the new one.The experimental results of eight subjects show that the amplitudes of the Na,Pa and Nb waves have significant differences (p ＜ 0.05) under the two different states,while no difference exhibits during the latencies.Using the new feature vector,the mean classification accuracy achieves 85.7％ with the ANN classifier.Therefore,it is effective to use MLR from single electrode to distinguish between the auditory attention state and the non-attention state.关键词
听觉注意/中潜伏期反应/单电极/人工神经网络Key words
auditory attention/middle latency response (MLR)/single electrode/artificial neural network (ANN)分类
医药卫生引用本文复制引用
徐梦圆,邹采荣,梁瑞宇,王力,王青云..单电极中潜伏期反应的听觉注意特征提取与识别[J].东南大学学报（自然科学版）,2017,47(3):432-437,6.基金项目
国家自然科学基金资助项目(61375028,61673108)、江苏省“六大人才高峰”资助项目(2016-DZXX-023)、江苏省博士后科研资助计划资助项目(1601011B)、江苏省“青蓝工程”资助项目、广州大学广东省灯光与声视频工程技术研究中心开放基金资助项目(KF201601,KF201602). （61375028,61673108）
