脑电信号成分稀疏分析范式及其可行证明
|
摘要
稀疏分解技术作为一种可靠的信号处理与传输方法,在包括EEG的多种时变信号分析和处理领域得到了广泛的应用。在EEG信号的成分分析中,现有算法(ICA,EMD)等都存在分解结果与真实成分显著不符的情况,难以对实际成分的波形进行估计。本文在稀疏分解算法基础上,通过对样本稀疏分布情况进行度量,给出了一个经过改良的稀疏性能评价指标(SPI)并以此建立了一个新的成分分析范式和相应的优化函数,经过理论和实际证明,该范式在成分分析领域能比传统方法更有效地使分解结果趋向于真实成分,对EEG信号、乃至其他时变信号的成分解析都具有相当的积极意义。
As a reliable signal processing and transmission method,sparse decomposition technology has been widely applied in many time-varying signal analysis and processing fields including EEG. Existing algorithms such as ICA and EMD often( almost always) bring significant discrepancy between their decomposition results and the real components,so it is difficult to estimate the waveform of the actual components by these algorithms. This paper provides an improved sparse performance index( SPI) by measuring the sparse distribution of the EEG samples,and introduces a new paradigm of component analysis and its optimizing function based on the sparse decomposition algorithm. We also prove by theory and practice that this paradigm can make the decomposition result tend to real components more effectively than traditional methods in the field of compositional analysis,which will provide considerable positive advantages to analysis of EEG and other timevarying signals.
As a reliable signal processing and transmission method,sparse decomposition technology has been widely applied in many time-varying signal analysis and processing fields including EEG. Existing algorithms such as ICA and EMD often( almost always) bring significant discrepancy between their decomposition results and the real components,so it is difficult to estimate the waveform of the actual components by these algorithms. This paper provides an improved sparse performance index( SPI) by measuring the sparse distribution of the EEG samples,and introduces a new paradigm of component analysis and its optimizing function based on the sparse decomposition algorithm. We also prove by theory and practice that this paradigm can make the decomposition result tend to real components more effectively than traditional methods in the field of compositional analysis,which will provide considerable positive advantages to analysis of EEG and other timevarying signals.
引文
[1]Gao B,Dellandrea E,Chen L.Music sparse decomposition onto a MIDI dictionary of musical words and its application to music mood classification[C]∥International Workshop on Content-Based Multimedia Indexing.IEEE,2012:1-6.
[2]Okumura H,Nakashizuka M,Iiguni Y.Single channel speech separation by using a sparse periodic decomposition[J].Technical Report of Ieice Ea,2015,107:7-12.
[3]Minaee S,Abdolrashidi A,Wang Y.Screen content image segmentation using sparse-smooth decomposition[C]∥Asilomar Conference on Signals,Systems and Computers.IEEE,2015:1202-1206.
[4]Gao C,Li H,Ma L.An intrinsic mode function basis dictionary for auditory signal processing[C]∥International Conference on Audio,Language and Image Processing.IEEE,2015:16-21.
[5]高畅,李海峰,马琳.面向内容的语音信号压缩感知研究[J].信号处理,2012,28(6):851-858.Gao C,Li H F,Ma L.Content-based Compressive Sensing for Speech Signal[J].Signal Processing,2012,28(6):851-858.(in Chinese)
[6]Ren J,Liu J,Guo Z.Context-Aware Sparse Decomposition for Image Denoising and Super-Resolution[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2013,22(4):1456-1469.
[7]胡爱军,孙敬敬,向玲.经验模态分解中的模态混叠问题[J].振动、测试与诊断,2011,31(4):429-434.Hu A J,Sun J J,Xiang L.Mode Mixing in Empirical Mode Decomposition[J].Journal of Vibration Measurement&Diagnosis,2011,31(4):429-434.(in Chinese)
[8]白春华,周宣赤,林大超,等.消除EMD端点效应的PSO-SVM方法研究[J].系统工程理论与实践,2013,33(5):1298-1306.Bai C H,Zhou X C,Lin D C,et al.PSO-SVM method based on elimination of end effects in EMD[J].Systems Engineering-Theory&Practice,2013,33(5):1298-1306.(in Chinese)
[9]刘亚新,赵瑞珍,胡绍海,等.用于压缩感知信号重建的正则化自适应匹配追踪算法[J].电子与信息学报,2010,32(11):2713-2717.Liu Y X,Zhao R Z,Hu S H,et al.Regularized Adaptive Matching Pursuit Algorithm for Signal Reconstruction Based on Compressive Sensing[J].Journal of Electronics&Information Technology,2010,32(11):2713-2717.(in Chinese)
[10]张晓阳,柴毅,李华锋.基于K-SVD和残差比的低信噪比图像稀疏表示去噪算法[J].光学技术,2012,38(1):23-29.Zhang Xiaoyang,Chai Yi,Li Huafeng.Low SNR image denoising via sparse and redundant representations over KSVD algorithm and residual ratio iteration termination[J].Optical Technique,2012,38(1):23-29.(in Chinese)
[11]张晗,杜朝辉,方作为,等.基于稀疏分解理论的航空发动机轴承故障诊断[J].机械工程学报,2015,51(1):97-105.Zhang H,Du C H,Fang Z W,et al.Sparse Decomposition Based Aero-engine’s Bearing Fault Diagnosis[J].Journal of Mechanical Engineering,2015,51(1):97-105.(in Chinese)
[12]梁国龙,邱龙皓,邹男.基于稀疏信号重构的阵元位置误差校正方法[J].声学学报,2017(6):677-684.Liang G L,Qiu L H,Zou N.A sparse signal reconstruction perspective for hydrophone array shape calibration[J].Acta Acustica,2017(6):677-684.(in Chinese)
[13]Jiang Z,Lin Z,Davis L S.Label Consistent K-SVD:Learning A Discriminative Dictionary for Recognition[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2013,35(11):2651-2664.
[14]Rubinstein R,Peleg T,Elad M.Analysis K-SVD:A Dictionary-Learning Algorithm for the Analysis Sparse Model[J].IEEE Transactions on Signal Processing,2013,61(3):661-677.
[15]李海峰,徐忠亮,马琳,等.基于稀疏质量优化的语音信号分析方法[C]∥全国人机语音通讯学术会议,2017.Li H F,Xu Z L,Ma L,et al.Speech signal analysis method based on sparse quality optimization[C]∥National Conference on Man-Machine Speech Communication,2017.(in Chinese)
[16]Johannes H,Martijn S,Benjamin B,et al.A Novel 9-Class Auditory ERP Paradigm Driving a Predictive Text Entry System[J].Frontiers in Neuroscience,2011,5:99.
[2]Okumura H,Nakashizuka M,Iiguni Y.Single channel speech separation by using a sparse periodic decomposition[J].Technical Report of Ieice Ea,2015,107:7-12.
[3]Minaee S,Abdolrashidi A,Wang Y.Screen content image segmentation using sparse-smooth decomposition[C]∥Asilomar Conference on Signals,Systems and Computers.IEEE,2015:1202-1206.
[4]Gao C,Li H,Ma L.An intrinsic mode function basis dictionary for auditory signal processing[C]∥International Conference on Audio,Language and Image Processing.IEEE,2015:16-21.
[5]高畅,李海峰,马琳.面向内容的语音信号压缩感知研究[J].信号处理,2012,28(6):851-858.Gao C,Li H F,Ma L.Content-based Compressive Sensing for Speech Signal[J].Signal Processing,2012,28(6):851-858.(in Chinese)
[6]Ren J,Liu J,Guo Z.Context-Aware Sparse Decomposition for Image Denoising and Super-Resolution[J].IEEE Transactions on Image Processing A Publication of the IEEE Signal Processing Society,2013,22(4):1456-1469.
[7]胡爱军,孙敬敬,向玲.经验模态分解中的模态混叠问题[J].振动、测试与诊断,2011,31(4):429-434.Hu A J,Sun J J,Xiang L.Mode Mixing in Empirical Mode Decomposition[J].Journal of Vibration Measurement&Diagnosis,2011,31(4):429-434.(in Chinese)
[8]白春华,周宣赤,林大超,等.消除EMD端点效应的PSO-SVM方法研究[J].系统工程理论与实践,2013,33(5):1298-1306.Bai C H,Zhou X C,Lin D C,et al.PSO-SVM method based on elimination of end effects in EMD[J].Systems Engineering-Theory&Practice,2013,33(5):1298-1306.(in Chinese)
[9]刘亚新,赵瑞珍,胡绍海,等.用于压缩感知信号重建的正则化自适应匹配追踪算法[J].电子与信息学报,2010,32(11):2713-2717.Liu Y X,Zhao R Z,Hu S H,et al.Regularized Adaptive Matching Pursuit Algorithm for Signal Reconstruction Based on Compressive Sensing[J].Journal of Electronics&Information Technology,2010,32(11):2713-2717.(in Chinese)
[10]张晓阳,柴毅,李华锋.基于K-SVD和残差比的低信噪比图像稀疏表示去噪算法[J].光学技术,2012,38(1):23-29.Zhang Xiaoyang,Chai Yi,Li Huafeng.Low SNR image denoising via sparse and redundant representations over KSVD algorithm and residual ratio iteration termination[J].Optical Technique,2012,38(1):23-29.(in Chinese)
[11]张晗,杜朝辉,方作为,等.基于稀疏分解理论的航空发动机轴承故障诊断[J].机械工程学报,2015,51(1):97-105.Zhang H,Du C H,Fang Z W,et al.Sparse Decomposition Based Aero-engine’s Bearing Fault Diagnosis[J].Journal of Mechanical Engineering,2015,51(1):97-105.(in Chinese)
[12]梁国龙,邱龙皓,邹男.基于稀疏信号重构的阵元位置误差校正方法[J].声学学报,2017(6):677-684.Liang G L,Qiu L H,Zou N.A sparse signal reconstruction perspective for hydrophone array shape calibration[J].Acta Acustica,2017(6):677-684.(in Chinese)
[13]Jiang Z,Lin Z,Davis L S.Label Consistent K-SVD:Learning A Discriminative Dictionary for Recognition[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2013,35(11):2651-2664.
[14]Rubinstein R,Peleg T,Elad M.Analysis K-SVD:A Dictionary-Learning Algorithm for the Analysis Sparse Model[J].IEEE Transactions on Signal Processing,2013,61(3):661-677.
[15]李海峰,徐忠亮,马琳,等.基于稀疏质量优化的语音信号分析方法[C]∥全国人机语音通讯学术会议,2017.Li H F,Xu Z L,Ma L,et al.Speech signal analysis method based on sparse quality optimization[C]∥National Conference on Man-Machine Speech Communication,2017.(in Chinese)
[16]Johannes H,Martijn S,Benjamin B,et al.A Novel 9-Class Auditory ERP Paradigm Driving a Predictive Text Entry System[J].Frontiers in Neuroscience,2011,5:99.