利用整体经验模态分解和随机森林的脑电信号分类研究
|
摘要
癫痫脑电信号的自动监测与分类在临床医学上具有重要意义。针对脑电信号的非平稳特点,提出一种基于整体经验模态分解和随机森林相结合的脑电信号分类方法。选取波恩大学脑电信号数据集中癫痫发作间期和发作期的200个单通道信号,共819 400个数据作为样本。首先利用整体模态分解将癫痫脑电信号分解成多个固有模态函数,然后对各阶固有模态函数提取有效特征,最后分别用随机森林和最小二乘支持向量机对脑电信号的特征进行分类。将随机森林与最小二乘支持向量机分类正确识别率对比,结果表明,随机森林分类方法对发作期和发作间期的癫痫脑电信号的分类效果比较理想,识别精度为99.60%,高于最小二乘支持向量机的准确性。该方法的提出能有效提高临床癫痫脑电信号分析的效率。
It is of great importance to automaticaly monitor and classify epileptic EEG in clinical medicine. In view of the non-stationary characteristics of EEG signals, a new method for feature extraction and recognition of EEG based on ensemble empirical mode decomposition(EEMD) and random forest(RF) was proposed in this paper. In this study 200 single-channel signals of epileptic ictal and interictal EEG were selected from EEG data of Bonn University, and 819400 data were used as samples. Firstly, the EEG signals were decomposed into several intrinsic mode functions(IMF) by EEMD, and then the effective features were extracted from each IMF component. Finally, the features of each IMF component were classified by RF and least squares support vector machine(LSSVM). We compared the classification results of RF and LSSVM. The results showed that the classification effect of RF algorithm on epileptic EEG signals in ictal and interictal periods was effective. The recognition accuracy was 99.60%, which was higher than the accuracy of LSSVM. The proposed method could effectively improve the efficiency of clinical EEG signal analysis.
It is of great importance to automaticaly monitor and classify epileptic EEG in clinical medicine. In view of the non-stationary characteristics of EEG signals, a new method for feature extraction and recognition of EEG based on ensemble empirical mode decomposition(EEMD) and random forest(RF) was proposed in this paper. In this study 200 single-channel signals of epileptic ictal and interictal EEG were selected from EEG data of Bonn University, and 819400 data were used as samples. Firstly, the EEG signals were decomposed into several intrinsic mode functions(IMF) by EEMD, and then the effective features were extracted from each IMF component. Finally, the features of each IMF component were classified by RF and least squares support vector machine(LSSVM). We compared the classification results of RF and LSSVM. The results showed that the classification effect of RF algorithm on epileptic EEG signals in ictal and interictal periods was effective. The recognition accuracy was 99.60%, which was higher than the accuracy of LSSVM. The proposed method could effectively improve the efficiency of clinical EEG signal analysis.
引文
[1] Kiloh L, Osselton G. Clinical electroencephalography [J]. Journal of Nervous & Mental Disease, 1963, 136(3): 307-308.
[2] Berger H. über das elektrenkephalogramm des menschen [J]. European Archives of Psychiatry and Clinical Neuroscience, 1938, 23(3): 121-124.
[3] Vefik B A. Fundamentals of Electroencephalography [M]. New York: Harper & Row, 1971.
[4] Jones-Gotman M, Bouwer MS, Gotman J. EEG slow waves and memory performance during the intracarotid amobarbital test [J]. Epilepsia, 1994, 35(1): 61-69.
[5] Saab R, Mckeown MJ, Myers LJ, et al. A wavelet based approach for the detection of coupling in EEG signals [C]//International IEEE EMBS Conference on Neural Engineering. Arlington: IEEE, 2005: 616-620.
[6] Gabor AJ, Seyal M. Automated Interictal EEG spike detection using artificial neural networks [J]. Electroencephalography & Clinical Neurophysiology, 1992, 83(5): 271-280.
[7] Ocak H. Automatic detection of epileptic seizures in EEG using discrete wavelet transform and approximate entropy [J]. Expert Systems with Applications, 2009, 36(2): 2027-2036.
[8] Oweis RJ, Abdulhay EW. Seizure classification in EEG signals utilizing Hilbert-Huang transform [J]. BioMedical Engineering OnLine, 2011, 10(1): 1-15.
[9] 李淑芳, 周卫东, 蔡冬梅,等. EMD和SVM结合的脑电信号分类方法 [J]. 生物医学工程学杂志, 2011(5): 891-894.
[10] 韩清鹏. 利用EEG信号的小波包变换与非线性分析实现精神疲劳状态的判定 [J]. 振动与冲击, 2013, 32(2): 182-188.
[11] 罗志增, 周瑛, 高云园. 基于双密度小波邻域相关阈值处理的脑电信号消噪方法 [J]. 模式识别与人工智能, 2014, 27(5): 403-409.
[12] Celka P, Boashash B, Colditz P. Preprocessing and time-frequency analysis of newborn EEG seizures [J]. IEEE Engineering in Medicine & Biology Magazine, 2001, 20(5): 30-39.
[13] Anderson CW, Stolz EA, Shamsunder S. Multivariate autoregressive models for classification of spontaneous electroencephalographic signals during mental tasks [J]. IEEE Transactions on Bio-medical Engineering, 1998, 45(3): 277-286.
[14] Jandó G, Siegel R M, Horváth Z, et al. Pattern recognition of the electroencephalogram by artificial neural networks [J]. Electroencephalography & Clinical Neurophysiology, 1993, 86(2): 100-09.
[15] Fisch BJ. Fisch & Spehlmann′s EEG primer: Basic principles of digital and analong EEG[M]. 3rd ed. New York: Elsevier, 1999:199-206.
[16] 李营, 吕兆承. 基于EEMD和LS-SVM的癫痫脑电信号识别 [J]. 淮南师范学院学报, 2014, 16(3): 8-11.
[17] 袁玲, 杨帮华, 马世伟. 基于HHT和SVM的运动想象脑电识别 [J]. 仪器仪表学报, 2010, 31(3): 649-654.
[18] Breiman L. Random forest [J]. Machine Learning, 2001, 45: 5-32.
[19] 李贤杰. 基于随机森林的失真图像分类 [D]. 大连:大连海事大学, 2013.
[20] 张学清, 梁军. 基于EEMD-近似熵和储备池的风电功率混沌时间序列预测模型 [J]. 物理学报, 2013, 62(5): 68-77.
[21] 张超, 陈建军, 郭迅. 基于EEMD能量熵和支持向量机的齿轮故障诊断方法 [J]. 中南大学学报(自然科学版), 2012, 43(3):216-220.
[2] Berger H. über das elektrenkephalogramm des menschen [J]. European Archives of Psychiatry and Clinical Neuroscience, 1938, 23(3): 121-124.
[3] Vefik B A. Fundamentals of Electroencephalography [M]. New York: Harper & Row, 1971.
[4] Jones-Gotman M, Bouwer MS, Gotman J. EEG slow waves and memory performance during the intracarotid amobarbital test [J]. Epilepsia, 1994, 35(1): 61-69.
[5] Saab R, Mckeown MJ, Myers LJ, et al. A wavelet based approach for the detection of coupling in EEG signals [C]//International IEEE EMBS Conference on Neural Engineering. Arlington: IEEE, 2005: 616-620.
[6] Gabor AJ, Seyal M. Automated Interictal EEG spike detection using artificial neural networks [J]. Electroencephalography & Clinical Neurophysiology, 1992, 83(5): 271-280.
[7] Ocak H. Automatic detection of epileptic seizures in EEG using discrete wavelet transform and approximate entropy [J]. Expert Systems with Applications, 2009, 36(2): 2027-2036.
[8] Oweis RJ, Abdulhay EW. Seizure classification in EEG signals utilizing Hilbert-Huang transform [J]. BioMedical Engineering OnLine, 2011, 10(1): 1-15.
[9] 李淑芳, 周卫东, 蔡冬梅,等. EMD和SVM结合的脑电信号分类方法 [J]. 生物医学工程学杂志, 2011(5): 891-894.
[10] 韩清鹏. 利用EEG信号的小波包变换与非线性分析实现精神疲劳状态的判定 [J]. 振动与冲击, 2013, 32(2): 182-188.
[11] 罗志增, 周瑛, 高云园. 基于双密度小波邻域相关阈值处理的脑电信号消噪方法 [J]. 模式识别与人工智能, 2014, 27(5): 403-409.
[12] Celka P, Boashash B, Colditz P. Preprocessing and time-frequency analysis of newborn EEG seizures [J]. IEEE Engineering in Medicine & Biology Magazine, 2001, 20(5): 30-39.
[13] Anderson CW, Stolz EA, Shamsunder S. Multivariate autoregressive models for classification of spontaneous electroencephalographic signals during mental tasks [J]. IEEE Transactions on Bio-medical Engineering, 1998, 45(3): 277-286.
[14] Jandó G, Siegel R M, Horváth Z, et al. Pattern recognition of the electroencephalogram by artificial neural networks [J]. Electroencephalography & Clinical Neurophysiology, 1993, 86(2): 100-09.
[15] Fisch BJ. Fisch & Spehlmann′s EEG primer: Basic principles of digital and analong EEG[M]. 3rd ed. New York: Elsevier, 1999:199-206.
[16] 李营, 吕兆承. 基于EEMD和LS-SVM的癫痫脑电信号识别 [J]. 淮南师范学院学报, 2014, 16(3): 8-11.
[17] 袁玲, 杨帮华, 马世伟. 基于HHT和SVM的运动想象脑电识别 [J]. 仪器仪表学报, 2010, 31(3): 649-654.
[18] Breiman L. Random forest [J]. Machine Learning, 2001, 45: 5-32.
[19] 李贤杰. 基于随机森林的失真图像分类 [D]. 大连:大连海事大学, 2013.
[20] 张学清, 梁军. 基于EEMD-近似熵和储备池的风电功率混沌时间序列预测模型 [J]. 物理学报, 2013, 62(5): 68-77.
[21] 张超, 陈建军, 郭迅. 基于EEMD能量熵和支持向量机的齿轮故障诊断方法 [J]. 中南大学学报(自然科学版), 2012, 43(3):216-220.