2DPCA在脑磁图棘波信号检测中的应用
|
摘要
棘波是癫痫疾病诊断和癫痫灶评估的重要标志,脑磁图设备能更精确地捕捉到癫痫患者在发作间期的棘波信号。然而,目前临床医生仍依赖于手动方法标记棘波信号,缺少便捷离线的多通道棘波检测方法。提出一种脑磁图的多通道棘波检测方法,针对给定时间宽度的多通道脑磁图信号的时间序列可以看作为一个二维矩阵,利用二维主成分分析(2DPCA)方法提取该矩阵的本征特征,再结合最近邻分类器实现离线的多通道棘波信号检测。通过临床癫痫患者的脑磁图信号验证表明,提出的方法棘波信号检测率高达93.23%,且该方法是有效的。
The epileptic spike plays a key role in the epileptic diagnosis and the epileptogenic zone evaluation, additionally,the Magnetoencephalogram(MEG)equipment can capture the interictal spikes more accurately. However, clinicians still manually mark spikes due to lacking an effective multi-channel spike detection approach. To solve this problem, this paper proposes a spike detection algorithm based on multi-channel MEG signals. The multi-channel MEG signal within a configurable time window is considered as a two-dimensional matrix. Furthermore, the intrinsic features of the matrix are extracted by the Two-Dimensional Principal Component Analysis(2DPCA)and the spikes are detected through the nearest neighbor classifier. Finally, experimental results indicate that the proposed method is effective and the detection rate can be up to 93. 23% on clinical MEG data.
The epileptic spike plays a key role in the epileptic diagnosis and the epileptogenic zone evaluation, additionally,the Magnetoencephalogram(MEG)equipment can capture the interictal spikes more accurately. However, clinicians still manually mark spikes due to lacking an effective multi-channel spike detection approach. To solve this problem, this paper proposes a spike detection algorithm based on multi-channel MEG signals. The multi-channel MEG signal within a configurable time window is considered as a two-dimensional matrix. Furthermore, the intrinsic features of the matrix are extracted by the Two-Dimensional Principal Component Analysis(2DPCA)and the spikes are detected through the nearest neighbor classifier. Finally, experimental results indicate that the proposed method is effective and the detection rate can be up to 93. 23% on clinical MEG data.
引文
[1] Lau M,Yam D,Burneo J G.A systematic review on MEG and its use in the presurgical evaluation of localization-related epilepsy[J].Epilepsy Research,2008,79(23):97-104.
[2] Hu Y,Lin Y,Yang B,et al.Deep source localization with Magnetoencephalography based on sensor array decomposition and beamforming[J].Sensors,2017,17(8):1860.
[3] Hu Y,Lin Y.MEG source imaging algorithm for finding deeper epileptogenic zone[M]//Advanced Computational Methods in Life System Modeling and Simulation.Singapore:Springer,2017:536-544.
[4] Ahmed R,Rutka J T.The role of MEG in pre-surgical evaluation of epilepsy:current use and future diretions[J].Expert Review of Neurotherapeutics,2016,16(7):795-801.
[5] Tzallas A T,Tsipouras M G,Tsalikakis D G,et al.Automated epileptic seizure detection methods:A review study[M]//Epilepsy-Histological,Electroencephalographic and Psychological Aspects.[S.l.]:InTech,2012.
[6] Hwang W J,Wang S H,Hsu Y T.Spike detection based on normalized correlation with automatic template generation[J].Sensors,2014,14(6):11049-11069.
[7]马东华,郑旭媛,王真.基于形态成分分析的癫痫脑电棘波检测[J].生物医学工程学杂志,2013(4):710-713.
[8] Garg H K,Kohli A K.EEG spike detection technique using output correlation method:A Kalman filtering approach[J].Circuits Systems&Signal Processing,2015,34(8):2643-2665.
[9] Garg H K,Kohli A K.Nonstationary-epileptic-spike detection algorithm in EEG signal using SNEO[J].Biomedical Engineering Letters,2013,3(2):80-86.
[10]初孟,邱天爽,鲍海平.基于人工神经网络的癫痫棘波检测方法[J].中国生物医学工程学报,2006,25(4):421-425.
[11]张见.基于视觉组织原则和支持向量机的发作间期癫痫波自动检测方法研究[D].上海:华东理工大学,2013.
[12] Khalid M I,Alotaiby T,Aldosari S A,et al.Epileptic MEG spikes detection using common spatial patterns and linear discriminant analysis[J].IEEE Access,2016,4:4629-4634.
[13]王娜,王汇源.基于环形对称Gabor变换和2DPCA的人脸识别算法[J].计算机工程与应用,2015,51(16):146-150.
[14] Xie H B,Zhou P,Guo T,et al.Multiscale two-directional two-dimensional principal component analysis and its application to high-dimensional biomedical signal classification[J].IEEE Transactions on Biomedical Engineering,2016,63(7):1416-1425.
[15] Yang J,Zhang D,Frangi A F,et al.Two-dimensional PCA:A new approach to appearance-based face representation and recognition[J].IEEE Transactions Pattern Analysis Machine Intelligence,2004,26(1):131-137.
[16] Yang J,Yang J Y.From image vector to matrix:A straightforward image projection technique—IMPCA vs.PCA[J].Pattern Recognition,2002,35(9):1997-1999.
[17] Rozendaal Y J,Van L G,Ossenblok P P.Spatiotemporal mapping of interictal epileptiform discharges in human absence epilepsy:A MEG study[J].Epilepsy Research,2015,119:67-76.
[18] Chowdhury R A,Lina J M,Kobayashi E,et al.MEG source localization of spatially extended generators of epileptic activity:Comparing entropic and hierarchical bayesian approaches.[J].Plos One,2013,8(2):276.
[19] Keshri A K,Singh A,Das B N,et al.LDASpike for recognizing epileptic spikes in EEG[J].International Journal of Information Engineering&Electronic Bu,2013,5(4):41-50.
[20] Lu J,Plataniotis K N,Venetsanopoulos A N.Face recognition using LDA-based algorithms[J].IEEE Transactions on Neural Networks,2003,14(1):195-200.
[2] Hu Y,Lin Y,Yang B,et al.Deep source localization with Magnetoencephalography based on sensor array decomposition and beamforming[J].Sensors,2017,17(8):1860.
[3] Hu Y,Lin Y.MEG source imaging algorithm for finding deeper epileptogenic zone[M]//Advanced Computational Methods in Life System Modeling and Simulation.Singapore:Springer,2017:536-544.
[4] Ahmed R,Rutka J T.The role of MEG in pre-surgical evaluation of epilepsy:current use and future diretions[J].Expert Review of Neurotherapeutics,2016,16(7):795-801.
[5] Tzallas A T,Tsipouras M G,Tsalikakis D G,et al.Automated epileptic seizure detection methods:A review study[M]//Epilepsy-Histological,Electroencephalographic and Psychological Aspects.[S.l.]:InTech,2012.
[6] Hwang W J,Wang S H,Hsu Y T.Spike detection based on normalized correlation with automatic template generation[J].Sensors,2014,14(6):11049-11069.
[7]马东华,郑旭媛,王真.基于形态成分分析的癫痫脑电棘波检测[J].生物医学工程学杂志,2013(4):710-713.
[8] Garg H K,Kohli A K.EEG spike detection technique using output correlation method:A Kalman filtering approach[J].Circuits Systems&Signal Processing,2015,34(8):2643-2665.
[9] Garg H K,Kohli A K.Nonstationary-epileptic-spike detection algorithm in EEG signal using SNEO[J].Biomedical Engineering Letters,2013,3(2):80-86.
[10]初孟,邱天爽,鲍海平.基于人工神经网络的癫痫棘波检测方法[J].中国生物医学工程学报,2006,25(4):421-425.
[11]张见.基于视觉组织原则和支持向量机的发作间期癫痫波自动检测方法研究[D].上海:华东理工大学,2013.
[12] Khalid M I,Alotaiby T,Aldosari S A,et al.Epileptic MEG spikes detection using common spatial patterns and linear discriminant analysis[J].IEEE Access,2016,4:4629-4634.
[13]王娜,王汇源.基于环形对称Gabor变换和2DPCA的人脸识别算法[J].计算机工程与应用,2015,51(16):146-150.
[14] Xie H B,Zhou P,Guo T,et al.Multiscale two-directional two-dimensional principal component analysis and its application to high-dimensional biomedical signal classification[J].IEEE Transactions on Biomedical Engineering,2016,63(7):1416-1425.
[15] Yang J,Zhang D,Frangi A F,et al.Two-dimensional PCA:A new approach to appearance-based face representation and recognition[J].IEEE Transactions Pattern Analysis Machine Intelligence,2004,26(1):131-137.
[16] Yang J,Yang J Y.From image vector to matrix:A straightforward image projection technique—IMPCA vs.PCA[J].Pattern Recognition,2002,35(9):1997-1999.
[17] Rozendaal Y J,Van L G,Ossenblok P P.Spatiotemporal mapping of interictal epileptiform discharges in human absence epilepsy:A MEG study[J].Epilepsy Research,2015,119:67-76.
[18] Chowdhury R A,Lina J M,Kobayashi E,et al.MEG source localization of spatially extended generators of epileptic activity:Comparing entropic and hierarchical bayesian approaches.[J].Plos One,2013,8(2):276.
[19] Keshri A K,Singh A,Das B N,et al.LDASpike for recognizing epileptic spikes in EEG[J].International Journal of Information Engineering&Electronic Bu,2013,5(4):41-50.
[20] Lu J,Plataniotis K N,Venetsanopoulos A N.Face recognition using LDA-based algorithms[J].IEEE Transactions on Neural Networks,2003,14(1):195-200.