WANG Yun1,2, WU Qi3, ZHOU Wei-dong1,2, YUAN Sha-sha1,2, YUAN Qi1,2
(1. School of Information Science and Engineering, Shandong University, Ji'nan 250100, China; 2. Suzhou Institute of Shandong University, Suzhou 215123, China; 3. Qilu Hospital, Shandong University, Ji'nan 250100, China)
Abstract： Seizure detection is extremely essential for long-term monitoring of epileptic patients. This paper investigates the detection of epileptic seizures in multi-channel long-term intracranial electroencephalogram (iEEG). The algorithm conducts wavelet decomposition of iEEGs with five scales, and transforms the sum of the three frequency bands into histogram for computing the distance. The proposed method combines a novel feature called EMD-L1, which is an efficient algorithm of earth movers’ distance (EMD), with support vector machine (SVM) for binary classification between seizures and non-seizures. The EMD-L1 used in this method is characterized by low time complexity and high processing speed by exploiting the L1 metric structure. The smoothing and collar technique are applied on the raw outputs of SVM classifier to obtain more accurate results. Several evaluation criteria are recommended to compare our algorithm with other conventional methods using the same dataset from the Freiburg EEG database. Experiment results show that the proposed method achieves a high sensitivity, specificity and low false detection rate, which are 95.73%, 98.45% and 0.33/h, respectively. This algorithm is characterized by its robustness and high accuracy with the possibility of performing real-time analysis of EEG data, and may serve as a seizure detection tool for monitoring long-term EEG.
Key words： electroencephalograph (EEG) signals; earth movers’ distance (EMD); EMD-L1; support vector machine (SVM); wavelet decomposition; seizure detection
CLD number： TN911.7 Document code： A
Article ID： 1674-8042(2014)03-0094-09 doi： 10.3969/j.issn.1674-8042.2014.03.018
References
[1] Shoeb A, Edwards H, Connolly J, et al. Patient-specific seizure onset detection. Epilepsy & Behavior, 2004, 5(4): 483-498.
[2] Aarabi A, Fazel-Rezai R, Aghakhani Y. A fuzzy rule-based system for epileptic seizure in intracranial EEG. Clinical Neurophysiology, 2009, 120(9): 1648-1657.
[3] Adeli H, Zhou Z, Dadmehrc N. Analysis of EEG records in an epileptic patient using wavelet transform. Journal of Neuroscience Methods, 2003, 123(1): 69-87.
[4] Khan Y U, Gotman J. Wavelet based automatic seizure detection in intracerebral electroencephalogram. Clinical Neurophysiology, 2003, 114(5): 898-908.
[5] Temko A, Thomas E, Boylan G, et al. EEG-based neonatal seizure detection with support vector machines. Clinical Neurophysiology, 2011, 122(3): 464-473.
[6] Shoeb A, Guttag J. Application of machine learning to epileptic seizure detection. In: Proceeding of the 27th International Conference on Machine Learning (ICML-10), Haifa, Israel, 2010: 975-982.
[7] Muller-Gerking J, Pfurtscheller G. Designing optimal spatial filers for single-trial EEG classification in a movement task. Clinical Neurophysiology, 1999, 110(5): 787-798.
[8] LIU Yin-xia, ZHOU Wei-dong, YUAN Qi, et al. Automatic seizure detection using wavelet transform and SVM in long-term Intracranial EEG. IEEE Transactions on Neural System and Rehabilitation Engineering, 2012, 20(6): 749-755.
[9] Subasi A. EEG signal classification using wavelet feature extraction and a mixture of expert model. Expert Systems with Applications, 2007, 32(4): 1084-1093.
[10] Murro A, King D, Smith J, et al. Computerized seizure detection of complex partial seizures. Clinical Neurophysiology, 1991, 79: 330-333.
[11] QU Hao, Gotman J. A patient-specific algorithm for the detection of seizure onset in long-term EEG monitoring: possible use as a warning device. IEEE Transactions on Biomedical Engineering, 1997, 44(2): 115-122.
[12] Güler N, beyli E, Güler I. Recurrent neural networks employing Lyapunov exponents for EEG signals classification. Expert Systems with Applications, 2005, 29(3): 506-514.
[13] Kannathal N, Choo M, Acharya U, et al. Entropies for detection of epilepsy in EEG. Computer Methods & Programs in Biomedicine, 2005, 80(3): 187-194.
[14] Rubner Y, Tomasi C, Guibas L J. A metric for distributions with application to image database. In: Proceedings of the 6th International Conference on Computer Vision (ICCV’98), Bombay, India, 1998; 59-67.
[15] Rubner Y, Guibas L J, Tomasi C. The earth mover's distance, multi-dimensional scaling, and color-based image retrieval. DARPA Image Understanding Workshop, 1997: 661-668.
[16] Anthony Y, LIU Wen-yin, DENG Xiao-tie. Detecting phishing web pages with visual similarity assessment based on earth mover's distance (EMD). IEEE Transactions on Dependable and Secure Computing, 2006, 3(4): 301-311.
[17] Assent I, Wenning A, Seidl T. Approximation techniques for indexing the earth mover's distance in multimedia databases. In: Proceedings of the 22nd International Conference on Data Engineering, Atlanta, USA, 2006: 11.
[18] Thorsten J. Transductive inference for text classification using support vector machines. In: Proceedings of the Sixteenth International Conference on Machine Learning, 1999: 200-209.
[19] Huang J, SHAO Xu-hui, Wechsler H. Face pose discrimination using support vector machines (SVM). In: Proceedings of the 14th International Conference on Pattern Recognition, 1998, 1: 154-156.
[20] Parrado-Hernandez E, Mora-Jimenez I, Arenas-Garcia J, et al. Growing support vector classifiers with controlled complexity. Pattern Recognition, 2003, 36(7): 1479-1488.
[21] Chen W H, Shih J Y. A study of Taiwan’s issuer credit rating systems using support vector machines. Expert Systems with Applications, 2006, 30(3): 427-435.
[22] Comak E, Arslan A. A new training method for support vector machines: clustering k-NN support vector machines. Expert Systems with Applications, 2008, 35(3): 564-568.
[23] Hwang J P, Park S. A new weighted approach to imbalanced data classification problem via support vector machine with quadratic cost function. Expert Systems with Applications, 2011, 38(7): 8580-8585.
[24] Lin H J, Yeh J P. Optimal reduction of solutions for support vector machines. Applied Mathematics and Computation, 2009, 214: 329-335.
[25] Tran Q A, LI Xun, DUAN Hai-xin. Efficient performance estimate for one-class support vector machine. Pattern Recognition Letters, 2005, 26(8): 1174-1182.
[26] Rubner Y, Tomasi C, Guibas L J. The earth movers’ distance as a metric for image retrieval. International Journal of Computer Vision, 2000, 40(2): 99-121.
[27] Jones D G, Malik J. Computational framework for determining stereo correspondence from a set of linear spatial filters. Image and Vision Computing, 1992, 10(10): 395-410.
[28] Freeman W, Pasztor E, Carmichael O. Learning low level vision. International Journal of Computer Vision, 2000, 40(1): 24-47.
[29] LING Hai-bin, Okada K. An efficient earth mover's distance algorithm for robust histogram comparison. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2007, 29(5): 840-853.
[30] Chua E C, Patel K, Fitzsimons M, et al. Improved patient specific seizure detection during pre-surgical evaluation. Clinical Neurophysiology, 2011, 122(4): 672-679.
[31] Raghunathan S, Jaitli A, Irazoqui P P. Multistage seizure detection techniques optimized for low-power hardware platforms. Epilepsy & Behavior, 2011, 22(S1): 61-68.
基于推土机距离和支持向量机的脑电癫痫检测算法
王芸1,2， 吴琦3， 周卫东1,2， 袁莎莎1,2， 袁琦1,2
(1. 山东大学 信息科学与工程学院, 山东 济南 250100; 2. 山东大学 苏州研究院， 江苏 苏州 215123； 3. 山东大学 齐鲁医院， 山东 济南 250100)
摘要:对于需要长期脑电图监测的癫痫患者， 癫痫自动检测技术是十分必要的。 本文所提出的基于多导联长程脑电的算法， 首先， 对脑电信号进行五层小波分解， 取其中三层并求和， 然后建立直方图并计算它们之间的距离。 本文使用了一种有效的推土机距离算法， 提出了脑电新特征EMD-L1。 由于EMD-L1利用了L1范式， 有效降低了时间复杂度， 提高了运算速度。 本文采用支持向量机作为分类器， 并对支持向量机的输出做了后处理， 其中包括平滑滤波和“collar”技术， 以获取更准确的检测结果。 本文将所提出的方法与其他使用相同脑电数据库的癫痫检测方法相比较， 实验结果表明本文提出的算法达到了较高的灵敏度95.73%、 特异性98.45%以及较低的误判率0.33/h。 该检测系统不但具有较高的精确度和较强的鲁棒性， 而且可以对脑电数据进行实时分析， 用于长期脑电监测中。
关键词:脑电信号； 推土机距离； EMD-L1； 支持向量机； 小波分解； 癫痫检测
引用格式：WANG Yun, WU Qi, ZHOU Wei-dong, et al. Seizure detection using Earth Movers’ Distance and SVM in Intracranial EEG. Journal of Measurement Science and Instrumentation, 2014, 5(3): 94-102. [doi: 10.3969/j.issn.1674-8042.2014.03.018]
[full text view]
