平滑伪Wigner-Ville分布在脑电信号提取中的应用

运用平滑伪Wigner-Ville分布提取大脑皮层的稳态视觉诱发电(steady state visual evoked potential,SSVEP)信号,并作为输入信号应用于脑机接口系统.研究结果表明,该方法不需叠加平均信号就可以提高信噪比,提取出的稳态视觉诱发电位信号能够准确反映使用者的控制意图,为脑机接口选取特征提供了依据.     

闪光视觉诱发电位在脑机接口中的应用研究

运用叠加平均与快速傅立叶变化相结合的方法,提取大脑皮层的稳态视觉诱发电位SSVEP（steady state visual evoked potential）信号,并作为输入信号应用于脑控机器人系统。通过实验确定了测试电极、视觉刺激颜色与背景亮度等模式参数,并且研究了不同使用者在不同视觉刺激频率下的SSVEP幅值变化趋势。研究结果表明,该方法能更好地提高信噪比,提取出的稳态视觉诱发电位信号能够准确反映使用者的控制意图,应用于脑控机器人系统,取得了很好的效果。     

一种基于双特征的联合脑-机接口系统设计

与传统基于单一脑电信号的脑-机接口相比,基于多种特征信号的联合脑-机接口能有效提高脑-机接口性能.在基于稳态视觉诱发电位和P300诱发电位的联合使用的可行性基础上,提出了新的刺激编码方式,构建了一种基于两种特征的联合脑-机接口系统.通过设计3×3字符刺激矩阵,矩阵中纵列按各自设定频率闪烁诱发稳态视觉诱发电位,横行随机出现蓝色框诱发P300.实验表明,当受试者注视并关注目标字符,两种特征脑电信号能够被同时诱发,且对脑电信号中两种特征进行识别能够检测出受试者选取的字符.与传统基于P300的字符脑-机接口相比,刺激诱发时间减少了一半,从根本上提高了脑-机接口的速度.在以后工作中,系统可以扩展到更大矩阵(如6×6),构建更为实用的联合脑-机接口系统.     

基于稳态视觉诱发电位的脑机接口系统的设计与研究

目的设计与研究将稳态视觉诱发电位作为输入信号的脑机接口系统.系统以显示器图形闪烁模块作为稳态视觉诱发电位的刺激源,经过滤波、放大等信号预处理后,对采集到的脑电信号中的稳态视觉诱发电位成分采用一种基于滑动窗的迭代式逐点频谱监测方法进行诱发电位的特征提取和识别,并将其转换为相应的控制命令以实现对伺服机械手臂6种运动方向的实...     

基于虚拟仪器LabVIEW的脑—机接口系统

给出了基于瞬态视觉诱发电位的脑-机接口系统在LabVIEW环境下的实现方案。该方案的关键部分是视觉刺激器的设计和脑电信号的提取两部分。不同的刺激模块代表了多种可能的选择,受试者注视屏幕上其中一个目标,分析诱发电位可判别试者所注视的目标。采用累加平均和滤波的方法提高信噪比,用于提取微弱的脑电信号。该方案能有效地诱发出可识别的具有特征性的视觉诱发电位,并且通过离线的信号处理能够提取出所诱发的视觉诱发电位。     

脑机接口的国产ADC芯片模组及其应用

首先，介绍了脑机接口的基本概念，分析了低功耗高分辨率模/数转换芯片在脑机接口系统中的必要性。然后，讨论了国产ADC芯片BMF21A1的性能，以脑电模拟仪测试验证该芯片可应用于脑电微弱信号的采集。最后，以该芯片构成的ADC模组为例，介绍了眼电检测、专注度及冥想度检测以及稳态视觉诱发电位3种典型的脑机接口应用。     

基于FPGA的脑机接口实时系统

给出了以FPGA为核心,实现基于瞬态视觉诱发电位的脑机接口实时系统的方案。该方案包括脑电采集电路、基于FPGA的VGA视觉刺激器和FPGA开发板三部分。用FPGA取代计算机,作为脑机接口的控制和信息处理器。利用VHDL编程,在FPGA中实时处理采集的脑电信号,提取并识别瞬态视觉诱发电位信号,转换为控制命令,反馈给视觉刺激器。实验结果表明,本方案可以有效地实现脑机接口实时系统,并达到较高的正确率和通信速度。     

基于稳态视觉诱发电位的脑机接口设计

为了设计出更适合的脑机接口，本文对基于稳态视觉诱发电位的概念进行了阐释，着重介绍了SSVEP信号原理与基础。对传统的三种刺激方法进行了简述，并最后选用LCD显示器实现简单闪烁刺激进行实验，采用正弦编码的方法对目标方块的灰度值进行调制，最终完成在线SSVEP-BCI系统设计。     

基于Emotiv的脑机接口研究

针对脑机接口中存在的信号容易受到干扰、操作复杂的问题,利用经济便携式脑电采集设备Emotiv EPOC+搭载了一套基于稳态视觉刺激的脑机交互系统。该脑机接口系统首次将功率谱密度分析、典型相关性分析等方法按照不同权重相结合,使得目标识别的准确率高达98.6%,且该系统具有很高的抗噪能力和可拓展性。     

基于瞬态视觉诱发电位的识别算法研究

基于瞬态视觉诱发电位的研究是脑机接口研究中的一种方法,其核心在于瞬态视觉诱发电位的识别算法研究。采用累加平均和小波分解滤波从强噪声背景下提取微弱的视觉诱发电位,采用主成分分析提取诱发电位的特征,用K近邻算法对得到的特征信号进行模式识别。采集三名受试者的脑电数据作为处理对象,识别准确率可以达到95%。实验结果表明:该方法可以比较准确地识别瞬态视觉诱发电位。     

新颖的稳态视觉诱发电位脑机接口系统

传统的稳态视觉诱发电位（SSVEP）脑机接口系统无法与现实世界进行交互,长时间单调的光闪烁刺激容易导致视觉疲劳,影响识别精度。为了增强人与机器的交互以及对环境的感知,设计了增强现实（AR）和SSVEP结合的范式,在真实环境下对物体进行识别与追踪,并将闪烁块对物体进行标记,采用滤波器组典型相关分析（FBCCA）方法对脑电信号进行处理。结果表明,系统的控制信号传输速度达到50.69 bit/min,FBCCA的识别正确率为99.68%,能够在1 s内对4个目标中的意图目标进行有效区分。研究表明,基于SSVEP和增强现实的脑机接口系统更适合于复杂的现实环境。     

基于高频稳态视觉诱发电位的仿人机器人导航

针对中低频稳态视觉诱发电位（steady-state visual evoked potential,SSVEP）易于导致视觉疲劳的缺点以及现有高频范式对视觉激励载体要求较高的不足,通过应用相位编码方法,基于普通液晶显示器搭建了面向仿人机器人导航的高频SSVEP用户界面;同时针对高频SSVEP较难识别的特点,提出一种新的模糊分类方法来提高脑电信号解码效率．仿人机器人导航实验表明,基于中频SSVEP范式的脑—机器人导航系统的准确率、碰撞次数、操作频率分别为92.44%、2.14次、11.23次/min,而且极易使受试感到不适．而高频SSVEP的应用缓解了视觉疲劳,将导航系统的准确率、碰撞次数、操作频率分别提升至93.31%、1.89次与12.05次/min．     

RSVP与SSVEP混合脑电信号刺激与多类事件检测

提出一种新的基于快速序列视觉呈现（Rapid Serial Visual Presentation,RSVP）与稳态视觉诱发电位（Steady-State Visually Evoked Potential,SSVEP）组合范式的脑电信号（Electroencephalogram,EEG）刺激与多类事件检测方法。对诱发的原始脑电信号通过电位重参考、基线去除、空间滤波等预处理操作去除数据的伪迹和噪声,通过自举聚合决策树（Bagging Tree,BT）和支持向量机（Supported Vector Machine,SVM）等机器学习算法,对14名受试者双重刺激诱发的脑电信号进行目标与频率相结合的多类事件检测,通过实验验证了该组合范式诱发的脑电信号具有良好的多类可分性,为开发基于RSVP和SSVEP两种范式的混合型脑-机接口应用提供了一种新的有效途径。同时,实验结果还表明,基于机器学习的BT和SVM模型对RSVP和SSVEP组合范式诱发的EEG信号进行多类识别的性能明显优于传统的典型关联分析（Canonical Correlation Analysis,CCA）算法的性能。     

Constructing an exactly periodic subspace for enhancing SSVEP based BCI

A novel exactly periodic spatial filtering (EPSD) approach, that provides a robust detection performance, is introduced and evaluated in this study. The proposed method exploits the temporal properties of the steady-state visual evoked potential (SSVEP) response to construct an orthogonal and exactly periodic mapping that enhances the signal to noise ratio (SNR) of the SSVEP embedded in the electroencephalogram (EEG) data. The subspace of interest is constructed via the elimination of the signals spaces that does not constitute the exact period of the target frequency. The EPSD is evaluated on a 35 subject benchmark dataset collected using a 40 target SSVEP BCI system. The results reveal that the proposed EPSD spatial filter significantly enhances the performance of target detection. Further statistical tests also confirm that the EPSD is a potential alternative to the existing SSVEP spatial filters for realizing an efficient BCI system.     

A time-frequency convolutional neural network for the offline classification of steady-state visual evoked potential responses

A new convolutional neural network architecture is presented. It includes the fast Fourier transform between two hidden layers to switch the signal analysis from the time domain to the frequency domain inside the network. This technique allows the signal classification without any special pre-processing and uses knowledge from the problem in the network topology. The first step allows the creation of different spatial and time filters. The second step is dedicated to the signal transformation in the frequency domain. The last step is the classification. The system is tested offline on the classification of EEG signals that contain steady-state visual evoked potential (SSVEP) responses. The mean recognition rate of the classification of five different types of SSVEP response is 95.61% on a time segment length of 1 s. The proposed strategy outperforms other classical neural network architecures.     

A high performance hybrid SSVEP based BCI speller system

The existing EEG based keyboard/speller systems have a tradeoff between the target detection time and classification accuracy. This study focuses on increasing the accuracy and probability of target classification rates in the SSVEP based speller system. We proposed two different types of hybrid SSVEP system by combining SSVEP with vision based eye gaze tracker (VET) and electro-oculogram (EOG). Thirty six targets were randomly chosen for this study and their corresponding visual stimulus was presented with unique frequencies. The visual stimuli were segregated into three groups and each group were arranged into different regions (left/middle/right) of the keyboard/speller layout for improving the probability of target detection rate. The VET/ EOG data were utilized to identify the regions that belong to the selected target. The region/group determination decreases the issue of misclassification of SSVEP frequencies. The averaged spelling accuracies of SSVEP-VET and SSVEP-EOG system for all the subjects is 91.2% and 91.39% respectively. Later, a visual feedback was added to the SSVEP-EOG system (SSVEP-EOG-VF) for improving the target detection rate. In this case, an average classification accuracy of 98.33% was obtained with the information transfer rate (ITR) of 69.21 bits/min for all the subjects. An accuracy of 100% was obtained for five subjects with the ITR of 74.1 bits/min in this system.     

基于主动干电极传感器的无线可穿戴脑电采集系统

为了减少脑电信号(EEG)采集系统的复杂性、实验前长时间繁琐的准备过程,以及改善采集系统的抗干扰性、可穿戴性和舒适性,设计了一种带有8通道主动干电极传感器的无线可穿戴脑电采集系统,具有24位ADC,109 dB信噪比,±500 mV/10 Hz的交流输入范围.实验结果表明:该系统和商业系统记录的脑电图波形之间的相关系数为0.89,增加了可穿戴无线采集系统的可靠性,改善穿戴的舒适性.采用该系统可实现脑电的静态记录和基于稳态视觉诱发(SSVEP)的脑-机接口(BCI)应用.     

基于受限玻尔兹曼机的疲劳脑电特性分析

为了准确提取和分类视觉疲劳所引起的脑电特征，以此提醒过度用眼的工作人员及时休息，提出了多通道受限玻尔兹曼机算法和卷积神经网络(CNN)算法结合的深度学习混合模型，利用该模型对枕叶区10个通道的脑电信号进行自动提取内在特征和分类。在基于SSVEP的视觉疲劳脑电数据集上进行评估，深度学习混合模型的平均准确率达到88.63%,比传统的特征提取和分类方法高10%。实验结果证明了深度学习混合模型取得的分类效果较好，并且克服了传统手动提取特征方法不全面的不足，对疲劳脑电的研究具有现实的意义。     

一种基于人体脑电信号的机械手臂控制系统研究

文章提出了一种基于脑电信号的机械手臂控制系统的设计思路。该系统主要由电极、脑电采集电路、在线检测算法、外设等部分组成。系统采用闪烁刺激使操作者产生基于稳态视觉的诱发电位信号,通过采集电路将信号送入计算机中,由软件对其进一步处理和分析,转换成相应的控制命令控制机械手臂操作。检测算法中解决了脑电信号基线漂移和能量波动问题的困扰。实验显示,系统具有很高的检测实日寸性和准确率。     

Statistical Characterization of Steady-State Visual Evoked Potentials and Their Use in Brain–Computer Interfaces

Steady-state visual evoked potential (SSVEP)-based brain–computer interfaces (BCIs) use the spectral power of the potentials for classification as they can be voluntarily enhanced or diminished by the subject by means of selective attention. The features traditionally extracted from the EEG and used for BCIs have been characterized as a normal distribution, although some studies have shown recently that this normal distribution is not the most appropriate for SSVEPs. In this paper we attempt to characterize the power of SSVEPs as a random variable that follows Rayleigh and exponential distributions when the stimulus is attended and ignored, respectively. BCIs based on SSVEPs can improve the transfer-bit and successful-classification rates if this new model is used instead of the traditional one based on the normal distribution.