北京生物醫(yī)學(xué)工程

視覺(jué)刺激下基于“是-否”狀態(tài)的腦電信號(hào)分類(lèi)研究

Classification of electroencephalogrambased on state of ‘Yes-No’ during visual stimuli experiment

作者：李康寧李明鈺李萌杜若瑜

單位：南京郵電大學(xué)地理與生物信息學(xué)院 (南京 210023) 江蘇省智慧健康大數(shù)據(jù)分析與位置服務(wù)工程實(shí)驗(yàn)室(南京 210023)

關(guān)鍵詞：腦電信號(hào)；視覺(jué)刺激；支持向量機(jī)；經(jīng)驗(yàn)?zāi)Ｊ椒纸猓还部臻g模式

分類(lèi)號(hào)：R318.04

出版年·卷·期（頁(yè)碼）：2020·39·3（257-263）

摘要：

目的為了探究腦機(jī)接口中腦電信號(hào)與判斷認(rèn)知心理活動(dòng)之間的識(shí)別問(wèn)題，本文在視覺(jué)刺激誘發(fā)實(shí)驗(yàn)設(shè)計(jì)中采用文字和圖片結(jié)合的方式進(jìn)行腦電分類(lèi)研究以期提高識(shí)別率。方法通過(guò)設(shè)計(jì)視覺(jué)刺激誘發(fā)判斷認(rèn)知腦電實(shí)驗(yàn)采集到15名受試者在“是”或“否”狀態(tài)下的腦電信號(hào)，經(jīng)過(guò)預(yù)處理和事件相關(guān)擾動(dòng)(event-related spectral dynamics ，ERSP)特征分析，運(yùn)用經(jīng)驗(yàn)?zāi)Ｊ椒纸猓╡mpirical mode decomposition ，EMD）優(yōu)化共空間模式(common spatial pattern ，CSP)的特征提取算法進(jìn)行分類(lèi)識(shí)別。首先，利用EMD對(duì)預(yù)處理后的腦電信號(hào)進(jìn)行有效的固有模態(tài)函數(shù)(intrinsic mode function，IMF)頻段篩選；其次，使用CSP濾波器進(jìn)行濾波提取特征向量；最后，使用支持向量機(jī)(support vector machine ，SVM)進(jìn)行分類(lèi)識(shí)別，并對(duì)測(cè)試組進(jìn)行檢驗(yàn)。結(jié)果經(jīng)過(guò)EMD-CSP優(yōu)化濾波后進(jìn)行SVM分類(lèi)正確率可達(dá)88.97%，相比單獨(dú)利用CSP進(jìn)行特征提取下的SVM分類(lèi)結(jié)果提高了約5%。結(jié)論EMD-CSP優(yōu)化濾波方法對(duì)判斷認(rèn)知腦電識(shí)別的可行性和有效性，為進(jìn)一步研究腦機(jī)接口應(yīng)用產(chǎn)品的開(kāi)發(fā)提供認(rèn)知參考依據(jù)。

Objective To explore the relationship of cognitive judgement and brain oscillatory activity in communication cognitive task for brain-computer interface (BCI), this paper uses a combination of text and pictures in the design of visual stimulation-induced experiments to conduct EEG classification research in order to improve the recognition rate. Methods The EEG signals of 15 subjects in "yes" or "no" state are collected through the experiment of visual stimulation induced two basic cognitive judgementstates. After preprocessing and event-related spectral dynamics(ERSP) analysis, the feature extraction algorithm of common spatial pattern (CSP) is optimized by empirical mode decomposition (EMD) for classification and recognition. First of all, EMD is used to filter the frequency band of IMFs. Secondly, CSP filter is used to extract feature vector. Finally, support vector machine (SVM) is used for classification and identification, and the test group is tested. Results After optimized filtering by EMD-CSP, the accuracy of SVM classification can reach 88.97%, which is about 5% higher than that of SVM classification using CSP for feature extraction alone. Conclusions The validity of EMD-CSP optimized filtering method for EEG classification provides a reference for further study of "Yes-No" EEG classification and recognition.

參考文獻(xiàn)：

[1] 賈花萍. 思維狀態(tài)的EEG信號(hào)分類(lèi)方法研究[J]. 電子設(shè)計(jì)工程, 2010

[1] 賈花萍. 思維狀態(tài)的EEG信號(hào)分類(lèi)方法研究[J]. 電子設(shè)計(jì)工程, 2010,18(6):124-126.

Jia HP. Classification for EEG signals of different mental tasks[J]. Electronic Design Engineering, 2010，18(6):124-126.

[2] 李穎潔, 樊飛燕, 陳興時(shí). 腦電分析在認(rèn)知研究中的進(jìn)展[J]. 北京生物醫(yī)學(xué)工程,2006, 25(3):321-324.

Li YJ, Fan FY, Chen XS. Development of EEG analysis in the research of cognitive science[J]. Beijing Biomedical Engineering, 2006, 25(3):321-324.

[3] 白芷嫣,卞毓彩,朱穎穎,杜若瑜.基于腦電信號(hào)的抑郁情緒傾向研究[J].北京生物醫(yī)學(xué)工程,2018，37(1):21-26.

Bai ZY, Bian YC, Zhu YY, et al. Depressive emotion based on electroencephalography [J].Beijing Biomedical Engineering, 2018, 37(1):21-26.

[4] 李文輝. 自我無(wú)關(guān)、相關(guān)的肯否句句末情緒詞加工的行為與腦電研究[D]. 大連：遼寧師范大學(xué), 2011.

Li WH. Behavioral and ERP study of emotional words processing in Self-related／unrelated affirmation and negation sentences[D]. Dalian：Liaoning Normal University, 2011.

[5]陳陽(yáng). “所與”的認(rèn)知實(shí)驗(yàn)及腦電計(jì)算分析[D]. 廈門(mén)：廈門(mén)大學(xué), 2014.

Cheng Y. Cognitive experiment and EEG calculation analysis of "the given"[D]. Xiamen：Xiamen University, 2014.

[6] 李子光, 劉國(guó)忠. 聽(tīng)覺(jué)誘發(fā)的“是”與“非”腦電信號(hào)研究[J]. 電子科技, 2013, 26(4):8-11.

Li ZG, Liu GZ. Research into “Yes” and “No” Responses by Auditory Stimuli in Human EEG[J]. Electronic Science and Technology,2013, 26(4):8-11.

[7] 喬曉艷, 彭佳卉. 視聽(tīng)誘發(fā)腦電特征提取與大腦認(rèn)知機(jī)理分析[J]. 測(cè)試技術(shù)學(xué)報(bào), 2013, 27(6):509-515.

Qiao XY, Peng JH. The analysis on the brain cognitive mechanism and audio-visual evoked EEG feature extraction[J]. Journal of Test and Measurement Technology, 2013, 27(6):509-515.

[8]戚光遠(yuǎn). 漢字和圖片的認(rèn)知腦電測(cè)謊研究[J]. 智富時(shí)代, 2016(增刊2).374-375.

Qi GY. A study of cognitive EEG lie detection of Chinese characters and pictures, The Fortune Times, 2016(Suppl 2).374-375.

[9] Huang NE，Shen SSP. Hilbert–Huang transform and its applications [M]. Singapore: World Scientific Publishing, 2005.

[10]孔祥浩, 馬琳, 薄洪健,等. CNN與CSP相結(jié)合的腦電特征提取與識(shí)別方法研究[J]. 信號(hào)處理, 2018, 34(2):164-173.

Kong XH, Ma L, Bo HJ, et al. Research on EEG feature extraction and recognition method based on CNN and CSP[J].Journal of Signal Processing, 2018, 34(2):164-173.

[11]康莎莎, 周蚌艷, 呂釗,等. 基于CSP的多類(lèi)運(yùn)動(dòng)想象腦電特征自動(dòng)選擇算法[J]. 北京生物醫(yī)學(xué)工程, 2016, 35(4):339-346.

Kang SS, Zhou BY, Lu Z, et al.Automatic selection algorithm for multi-class motor imagery of EEG eigenvalues based on CSP[J].Beijing Biomedical Engineering, 2016, 35(4):339-346.

[12]Mehmood RM,Du RY, Lee HJ. Optimal feature selection and deep learning ensembles method for emotion recognition from human brain EEG sensors[J]. IEEE Access, 2017, 5(99):14797-14806.

[13]沃建中, 林崇德. 大學(xué)生對(duì)句子、圖形的內(nèi)部加工過(guò)程研究[J]. 心理發(fā)展與教育, 1997, 13(4):23-28.

Wo JZ, Lin CD, Research on the internal processing process of sentences and graphics of college student[J]. Psychological Development and Education, 1997, 13(4):23-28.

[14]Sander N,Yeung N,van den Wildenberg W,et al.Electrophysiological correlates of anterior cingulate function in a go/no-go task: Effects of response conflict and trial type frequency[J]. Cognitive Affective & Behavioral Neuroscience, 2003, 3(1):17-26.

[15]Delorme A,Makeig S.EEGLAB: an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis[J]. Journal of Neuroscience Methods, 2004, 134(1):9-21.

[16] Du RY, Lee HJ. Frontal alpha asymmetry during the audio emotional experiment revealed by event-related spectral perturbation[C]// International Conference on Biomedical Engineering and Informatics. IEEE, 2016:531-536.

[17] Gábor C, Gábor S, Sarolta K, et al. Event-related theta synchronization predicts deficit in facial affect recognition in schizophrenia[J]. Journal of Abnormal Psychology, 2014, 123(1):178-189.

[18]VapnikV.Statistical learning theory [M] . New York: John Wiley&Sons Inc.,1998.

[19] Arroyo FV. Negatives in context[J]. Journal of Verbal Learning & Verbal Behavior, 1982, 21(1):118-126.

[20]Steinhauer K , Drury JE . On the early left-anterior negativity (ELAN) in syntax studies[J]. Brain & Language, 2012, 120(2):135-162.

[21] Gouvea AC , Phillips C , Kazanina N , et al. The linguistic processes underlying the P600[J]. Language and Cognitive Processes, 2010, 25(2):149-188.

[22]Berkum JJAV , Hagoort P , Brown CM . Semantic integration in sentences and discourse: evidence from the N400[J]. Journal of Cognitive Neuroscience, 1999, 11(6):657-671.

服務(wù)與反饋：

【文章下載】【加入收藏】

提示：您還未登錄，請(qǐng)登錄！點(diǎn)此登錄

51黑料吃瓜在线观看,51黑料官网|51黑料捷克街头搭讪_51黑料入口最新视频