摘要
视觉注意一直是心理学、神经生理学的重要研究领域,近年来随着脑功能探测技术的发展,对于视觉注意力的研究已经从早期的定性描述发展到利用各种神经生理学信号对大脑不同状态下的注意力水平进行定量化分析的阶段。目前,视觉注意力水平的定量化研究尚在选取表征注意力的特征参数与量化的分级方面存在一定的缺陷。寻找新的能客观准确地反映视觉注意力水平的特征参数,建立能稳定、可靠地监测并可动态调控操作者视觉注意力水平的生物反馈系统,将不仅有利于揭示人脑中视觉注意力的心理学、神经生理学方面的作用机制,还将有益于与注意力相关脑神经疾病的诊断和治疗,促进脑-机接口等神经工程新技术及在康复医学中的应用。
本文设计了两种用于测试多级视觉注意力水平的脑电反馈实验,通过要求受试者注视屏幕动画并按照提示进行不同强度的想象动作过程并记录多个注意集中程度下的脑电数据。分别采用脑电信号的近似熵、样本熵、多尺度熵、δ-样本熵等基于序列复杂性测度的非线性动力学参数来表征受试者的视觉注意集中程度。经14名受试者实验及其脑电信号熵参数统计分析结果表明,在绝大多数额叶导联(F3、F4、F7、F8、Fz)和部分颞叶导联(T3、T4等)处的多个注意集中程度之间存在显著性差异,并且随受试者注意集中程度的降低,其熵值呈下降趋势;发现样本熵在区分多个注意集中程度时的敏感度最高。为实现多级注意力水平的量化分级,本文优化设计了支持向量机分类器,对注意力水平进行了分级识别,识别率达到85.24%,可以将多种注意力集中程度区分开。最后对受试者注意力集中程度与其想象动作电位特征之间的相关性进行了初步研究。以上基于脑电熵参数的视觉注意力分级研究方法和结果可为脑与认知神经科学基础研究、脑电生物反馈训练以及在线脑-机接口系统设计提供参考依据与技术支持。
Research of visual attention is one of the important domains of psychology and neurophysiology. In recent years, with the development of brain function detecting technology, the research of visual attention has improved from the stage of qualitative analysis to that of quantitative analysis under different functional conditions by utilizing variety of neurophysiologic signals. However, there are still certain defects on the feature parameters selection and attention levels classification in this quantitative research. To seek new parameters which can reflect the levels of attention accurately and objectively, and to establish a stable and reliable biofeedback system to monitor and control the operator’s visual attention levels dynamically will not only be helpful in revealing the mechanism of visual attention in the psychology and neurophysiology field, but also could be useful to diagnose and treat the brain diseases related to the attention. In this way, it can promote new technology, such as brain-computer interface (BCI) etc, and application in neural and rehabilitation engineering.
In this thesis, there were two electroencephalogram (EEG) feedback experiments to measure the different levels of visual attention. In these experiments, following flash stimulus displayed on the screen, all subjects were required to finish different imaginary motor tasks corresponding to multi attention levels and EEG was recorded for further research. In order to assess different visual attention levels, the next step was to process EEG data with nonlinear dynamics parameters based on sequence complexity, such as approximate entropy, sample entropy, multiscale entropy,δ- sample entropy. According to the statistics analysis of entropy parameters of EEG signals of 14 subjects, there are significant differences in attention intensity in most of the electrodes in the frontal regions, such as F3、F4、F7、F8、Fz, and some of the electrodes in temporal regions, such as T3, T4. The values of entropy show a declining tendency with the level of attention declining and sample entropy has the highest sensitivity when discriminating different levels of attention. To realize attention levels quantification, a classifier based on support vector machine was optimized and was used to recognize the levels of attention with recognition ratio of 85.24%. Thus, it can distinguish the different levels of attention. The last segment of this thesis is the analysis of relevance between attention levels and motor imaginary potentials. With the mentioned research methods and results above, it can supply the reference and technical support for the basic research in brain and cognition neuroscience, EEG biofeedback training system and BCI online system design.
引文
[1]王玲,深度运动问题的脑机制研究:[博士学位论文],成都;电子科技大学,2008
[2]罗跃嘉主编,认知神经科学教程,北京:北京大学出版社,2006
[3] Zhou XL, Luo YJ. Cognitive neuroscience in China. Intern J Psychol, 2003, 38(4): 50~61
[4] Chen L. Topological structure in visual perception. Science, 1982, 218: 699~700
[5] Tang SM, Guo AK. Choice behavior of Drosophila facing contradictory visual cues. Science, 2001, 294:1543~1547
[6] Yao HS, Li CY. Clustered organization of neurons with similar extra-receptive field property in the primary visual cortex. Neuron, 2002, 35: 547~553
[7] Luo YJ, Wei JH. Cross-modal selective attention to visual auditory stimuli modulates endogenous ERP components. Brain Research, 1999, 842: 30~38
[8] Luo YJ, Greenwood PM, Parasuraman R. Dynamics the spatial scale of visual attention revealed by brain Event-related Potentials. Cognitive Brain Research, 2001, 12: 371~381
[9] Han S, He X. Modulation of neural activities by enhanced local selection in the processing of compound stimuli. Human Brain Mapping, 2003, 19(4): 273~281
[10] Mai XQ, Luo J, Wu JH, Luo YJ.”Aha!”Effects in Guessing Riddle Task: An ERP Study. Human Brain Mapping, 2004, 22(4): 261~270
[11] Peng DL, et al. Automatic phonological activation of Chinese characters: An fMRI study. Cognitive Brain Research, 2004, 20(2): 156~164
[12]周仲兴,复合下肢想象动作电位的特征识别新技术研究:[博士学位论文],天津;天津大学,2009
[13]陈国田,曹起龙主编,简明神经电生理诊断学,北京:科学技术文献出版社,1998
[14]陈滨津,基于想象动作电位的BCI系统设计与实验研究:[硕士学位论文],天津;天津大学,2005
[15] Paul S, Wolfgang K. What does phase information of oscillatory brain activity tell us about cognitive process. Neuroscience and Biobehavioral Reviews, 2008, 32: 1001~1013
[16]寿天德主编,神经生物学第2版,北京:高等教育出版社,2006
[17]王甦,汪安圣,认知心理学,北京:北京大学出版社,2003
[18]陈桂芳,张开发,脑电生物反馈治疗注意缺陷多动障碍患儿的疗效评估现状,中国组织工程研究与临床康复,2008,30(12):5969～5972
[19]张凤华,张劲松,金星明,脑电生物反馈改善注意缺陷多动障碍患儿行为问题的评价,中国临床康复,2006,10(10):74～76
[20]李红辉,黎宁真,谢朝德,莫少玲,常静,脑电生物反馈治疗注意缺陷多动障碍儿童的临床效果分析,中国康复理论与实践,2006,3(12):188～189
[21] S.Shioiri, T. lnoue, K. Matsumura ,H. Yaguchi. Movement of Visual Attention, Systems, Man, and Cybernetics. IEEE SMC’99 Conference Proceedings, 1999 IEEE International Conference on 1999, (2):5~9
[22] Li CL, Takanori K., Wu JL., Chen DH., Takaaki T., Kunihiko O.. Attention systems and neural responses to visual and auditory stimuli: an fMRI study. 2007 IEEE/ICME International Conference on Complex Medical Engineering, 1494~1498
[23] http://www.playattention.com
[24]傅娟,脑电生物反馈系统及其研究:[硕士学位论文],南京;南京大学,2001
[25]燕楠,王珏,魏娜,宗良,基于样本熵的注意力相关脑电特征信息提取与分类,2007,10(41):1237～1241
[26]吴正华,稳态视觉诱发电位在脑机接口及认知过程中的应用研究:[博士学位论文],成都;电子科技大学,2008
[27] The International 10-20 System Electrode Placement, http://www.telemedx.com/protocol/section7.2.html
[28] Lehnertz K, et al. Non-linear time series analysis of intracranial EEG recording in patients with epilepsy. International Journal of Psychophysiology, 1999, 34: 45~52
[29] Pezard L, Martinerie J, Varela FJ, et al. Entropy maps characterize drug effects on brain dynamics in Alzheimer’s disease. Neuroscience Letters, 1998, 253: 5~8
[30] Jeong J, Kim DJ, et al. Nonlinear analysis of the EEG of schizophrenics with optimal embedding dimension. Medical Engineering & Physics, 1998, 20: 669~676
[31]洪波,唐庆生,杨福生,陈天祥,近似熵、互近似熵的性质、快速算法及其在脑电与认知研究中的初步应用,信号处理,1999,15(2):100～107
[32] Li M, Vitanyi P. An Introduction to Kolmogorov Complexity and Its Applications. Second Edition, Springer Verlag, 1997
[33] Gao JB. Detecting nonstationarity and state transitions in a time series. Physical Review E, 2001, 63: 066~202
[34] Lempel A, Ziv J. On the complexity of finite sequences. IEEE Trans. Inf. Theory, 1976, 22: 75~81
[35] Kaspar F, Schuster HG. Easily calculable measure for the complexity of spatiotemporal patterns. Phys. Rev. A, 1987, 36: 842~848
[36] Pezard L, Martinerie J, Varela FJ, et al. Entropy maps characterize drug effects on brain dynamics in Alzheimer’s disease. Neuroscience Letters, 1998, 253: 5~8
[37] Tong QY, Kong J, Xu JH. A Note on Analyzing Schizophrenic EEG with Complexity Measure, Chaos, Solitons & Fractals, 1996, 7: 371~375
[38] Pincus SM., Goldberger AL. Physiological time-series analysis: what does regularity quantify? American Journal of Physiology, 1994, 266(4): 1643~1656
[39] Pincus S, Singer BH. Randomness and degrees of irregularity. Proc. Natl. Acad. Sci. USA, Mathematics, 1996, 93: 2083~2088
[40]杨福生,廖旺才,近似熵:一种适用于短数据的复杂性度量,中国医疗器械杂志,1997,21(5):283～286
[41] Richman JS, Randall Moorman J. Physiological time-series analysis using approximate entropy and sample entropy. Am J Physiol Heart Circ Physiol, 2000, 278: H2039~H2049
[42]刘慧,和卫星,陈晓平,生物时间序列的近似熵和样本熵方法比较,仪器仪表学报,2004,25(4):806～812
[43] Costa M, Goldberger A.L., Peng C.-K.. Multiscale entropy analysis of biological signals. Physical Review E, 2005, 021906: 1~18
[44]蔡瑞,卞春华,宁新宝,多尺度熵在心率变异信号复杂性分析中的应用,北京生物医学工程,2007,26(5):543～547
[45]葛家怡,睡眠分期及低频磁场睡眠诱导的研究:[博士学位论文],天津;天津大学,2008
[46] Govindan R.B., Wilson J.D., Eswaran H., Lowery C.L., Preiβl H. Revisiting sample entropy analysis. Physica A, 2007, 376: 158~164
[47] Steven M, Pincus and L.Goldberger. Physiological time-series analysis: what does regularity quantify? Am J. Physiol., 1994, 266: H1643~H1656
[48] Joachims T. Text categorization with support vector machines: Learning with many relevant features. Proceeding of the European Conference on Machine Learning,Berlin:Springer, 1998: 137~142
[49] Cortes C, Vapnik. Support vector network. Machines Learning, 1995, 20: 237~297
[50] LeCun Y, et al. Comparison of learning algorithms for handwritten digit recognition. Proceedings ICANN’95 International Conference on Artificial Neural Networks, volumeⅡ. EC2, 1995: 53~60
[51] Jaakkola TS, Haussler D. Exploiting generative models in discrimination classifiers. Advances in Neural Information Processing System,11. MIT Press, 1998
[52] Blanz V, Scholkopf B, Bolthoff H et al. Comparison of view based object recognition algorithms using realistic 3D models. Aetificial Neural Networks ICANN’96 Spingers Lecture Notes in Computer Science-Berlin, 1996, 1112: 251~256
[53] Brown M, Lewis HG, Gunn SR. Linear spectral mixture models and support vector machines for remote sensing. IEEE Trans on Geosciences and Remote Sensing, 2000, 38(5): 2346~2360
[54] Mike Fugate, James, R Gattikers. Computer intrusion detection with claasification and anomaly detection using SVMs. International Journal of Pattern Recognition and Arificial Intelligence, 2003, 17(3): 441~458
[55] Burges C. A tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery, 1998, 2(2): 121~167
[56] Vapnik VN. Estimation of Dependencies Based on Empirical Data. Berlin: Springer-Verlag, 1982
[57] Vapnik VN. The Nature of Statistical Learning Theory. NY: Springer-Verlag, 1995
[58] Cherkassky V, Mulier F. Learning from Data: Concepts, Theory and Methods. NY: JohnViley & Sons, 1997
[59] Vapnik VN, Levin E, LeCun Y. Measuring the VC-dimension of a learning machine, Neural Computation, 1994, 6: 851~876
[60] Burges CJC. A tutorial on support vector machines for pattern recognition. Data Mining and Knowledge Discovery, 1998, 2(2)
[61] Boser B, Guyon I, Vapnik VN. A training algorithm for optimal margin classifiers. Fifth Annual Workshop on Computational Learning Theory. Pittsburgh: ACM Press, 1992
[62]张学工,关于统计学习理论与支持向量机,自动化学报,2000,26(1):32～42
[63] Cortes C, Vapnik VN. Support vector networks. Machine Learning, 1995, 20: 273~297
[64] Schoblkopf B, Burges C, Vapnik VN. Extracting support data for a given task. In: FayyadUM, Uthurusamy R (eds.). Proc. of First Intl. Conf. on Knowledge Discovery & DataMining, AAA I Press, 1995, 262~267
[65]綦宏志,脑—机接口中上肢想象动作侧向性识别的关键技术研究:[博士学位论文],天津;天津大学,2008
[66] Pfurtscheller G., Lopes da Silva F.H.. Event-related EEG/MEG synchronization and desychronization: basic principles. Clin Neurophysiol 1999, 110: 1842~1857
[67] Steven J. Luck: An Introduction to the Event-Related Potential Technique. Cambridge, Mass.: The MIT Press, 2005
[68]黄远桂,吴声伶,临床脑电图学,北京:人民卫生出版社,1998,1～9
[69] Sayers B, Mc A, Beayley HA, et al. The mechanism of auditory evoked EEG responses. Nature 1974, 247: 481~483
[70] van Winsum W, Sergeant J, Gueze R. The functional significance of eventrelated desynchronization of alpha rhythms in attentional and activating tasks. Electroenceph clin Neurophysiol, 1984, 58: 519~524
[71] Pfurtscheller G, Neuper C, Schl?gl A, Lugger K. Separability of EEG signals recorded during right and left motor imagery using adaptive autoregressive parameters. IEEE Trans Rehabil Eng., 1998, 6(3): 316~325
[72] Lopes da Silva FH. Neural mechanisms underlying brain waves: from neural membranes to networks. Electroenceph clin Neurophysiol, 1991, 79: 81~93
[73]周颖,王涛,冯焕清,江朝晖,运动想象脑电的ERD/ERS分析,北京生物医学工程,2004,4(23):263～268