Block term decomposition for modelling epileptic seizures

Recordings of neural activity, such as EEG, are an inherent mixture of different ongoing brain processes as well as artefacts and are typically characterised by low signal-to-noise ratio. Moreover, EEG datasets are often inherently multidimensional, comprising information in time, along different channels, subjects, trials, etc. Additional information may be conveyed by expanding the signal into even more dimensions, e.g. incorporating spectral features applying wavelet transform. The underlying sources might show differences in each of these modes. Therefore, tensor-based blind source separation techniques which can extract the sources of interest from such multiway arrays, simultaneously exploiting the signal characteristics in all dimensions, have gained increasing interest. Canonical polyadic decomposition (CPD) has been successfully used to extract epileptic seizure activity from wavelet-transformed EEG data (Bioinformatics 23(13):i10–i18, 2007; NeuroImage 37:844–854, 2007), where each source is described by a rank-1 tensor, i.e. by the combination of one particular temporal, spectral and spatial signature. However, in certain scenarios, where the seizure pattern is nonstationary, such a trilinear signal model is insufficient. Here, we present the application of a recently introduced technique, called block term decomposition (BTD) to separate EEG tensors into rank- (Lr,Lr,1) terms, allowing to model more variability in the data than what would be possible with CPD. In a simulation study, we investigate the robustness of BTD against noise and different choices of model parameters. Furthermore, we show various real EEG recordings where BTD outperforms CPD in capturing complex seizure characteristics.

[1]  R. Bro,et al.  A new efficient method for determining the number of components in PARAFAC models , 2003 .

[2]  M. Hallett,et al.  An improved method for localizing electric brain dipoles , 1990, IEEE Transactions on Biomedical Engineering.

[3]  M. De Vos,et al.  Removing muscle and eye artifacts using blind source separation techniques in ictal EEG source imaging , 2009, Clinical Neurophysiology.

[4]  T. Sejnowski,et al.  Removing electroencephalographic artifacts by blind source separation. , 2000, Psychophysiology.

[5]  Fumikazu Miwakeichi,et al.  Decomposing EEG data into space–time–frequency components using Parallel Factor Analysis , 2004, NeuroImage.

[6]  C. W. Hesse,et al.  Tracking and detection of epileptiform activity in multichannel ictal EEG using signal subspace correlation of seizure source scalp topographies , 2006, Medical & Biological Engineering & Computing.

[7]  S. Gómez,et al.  The triangle method for finding the corner of the L-curve , 2002 .

[8]  Laurent Albera,et al.  Muscle Artifact Removal in Ictal Scalp-EEG Based on Blind Source Separation , 2015 .

[9]  P Berg,et al.  Dipole modelling of eye activity and its application to the removal of eye artefacts from the EEG and MEG. , 1991, Clinical physics and physiological measurement : an official journal of the Hospital Physicists' Association, Deutsche Gesellschaft fur Medizinische Physik and the European Federation of Organisations for Medical Physics.

[10]  Panayiotopoulos Cp,et al.  The Epilepsies: Seizures, Syndromes and Management , 2004 .

[11]  K. J. Blinowska,et al.  Linear model of brain electrical activity—EEG as a superposition of damped oscillatory modes , 1985, Biological Cybernetics.

[12]  Lotfi Senhadji,et al.  ICA-based EEG denoising: a comparative analysis of fifteen methods , 2012 .

[13]  K. Arun,et al.  State-space and singular-value decomposition-based approximation methods for the harmonic retrieval problem , 1983 .

[14]  Patrick Dupont,et al.  Canonical decomposition of ictal scalp EEG reliably detects the seizure onset zone , 2007, NeuroImage.

[15]  Kazutomo Yunokuchi,et al.  Tests of EEG localization accuracy using implanted sources in the human brain , 1991, Annals of neurology.

[16]  Henk A L Kiers,et al.  A fast method for choosing the numbers of components in Tucker3 analysis. , 2003, The British journal of mathematical and statistical psychology.

[17]  Abdulhamit Subasi,et al.  EEG signal classification using PCA, ICA, LDA and support vector machines , 2010, Expert Syst. Appl..

[18]  Andrzej Cichocki,et al.  Nonnegative Tensor Factorization for Continuous EEG Classification , 2007, Int. J. Neural Syst..

[19]  F Wendling,et al.  EEG extended source localization: Tensor-based vs. conventional methods , 2014, NeuroImage.

[20]  Liqing Zhang,et al.  Pattern Classification of Visual Evoked Potentials Based on Parallel Factor Analysis , 2008 .

[21]  L. K. Hansen,et al.  Automatic relevance determination for multi‐way models , 2009 .

[22]  Lieven De Lathauwer,et al.  Decompositions of a Higher-Order Tensor in Block Terms - Part I: Lemmas for Partitioned Matrices , 2008, SIAM J. Matrix Anal. Appl..

[23]  Josemir W Sander,et al.  The global burden and stigma of epilepsy , 2008, Epilepsy & Behavior.

[24]  Lars Kai Hansen,et al.  Parallel Factor Analysis as an exploratory tool for wavelet transformed event-related EEG , 2006, NeuroImage.

[25]  J. Chang,et al.  Analysis of individual differences in multidimensional scaling via an n-way generalization of “Eckart-Young” decomposition , 1970 .

[26]  Lieven De Lathauwer,et al.  On the Uniqueness of the Canonical Polyadic Decomposition of Third-Order Tensors - Part I: Basic Results and Uniqueness of One Factor Matrix , 2013, SIAM J. Matrix Anal. Appl..

[27]  Elena Urrestarazu,et al.  Independent Component Analysis Removing Artifacts in Ictal Recordings , 2004, Epilepsia.

[28]  Ahmet Alkan,et al.  Classification of EEG Recordings by Using Fast Independent Component Analysis and Artificial Neural Network , 2008, Journal of Medical Systems.

[29]  Johan Wagemans,et al.  Single trial ERP reading based on parallel factor analysis. , 2013, Psychophysiology.

[30]  Pierre Comon,et al.  Independent component analysis, A new concept? , 1994, Signal Process..

[31]  Rasmus Bro,et al.  Multiway analysis of epilepsy tensors , 2007, ISMB/ECCB.

[32]  Hyunwoo Nam,et al.  Independent Component Analysis of Ictal EEG in Medial Temporal Lobe Epilepsy , 2002, Epilepsia.

[33]  Robert Tibshirani,et al.  Estimating the number of clusters in a data set via the gap statistic , 2000 .

[34]  Wim Van Paesschen,et al.  Canonical Decomposition of Ictal Scalp EEG and Accurate Source Localisation: Principles and Simulation Study , 2007, Comput. Intell. Neurosci..

[35]  Richard W Newton,et al.  The Epilepsies: Seizures, Syndromes and Management , 2006 .

[36]  Thom F. Oostendorp,et al.  The conductivity of the human skull: results of in vivo and in vitro measurements , 2000, IEEE Transactions on Biomedical Engineering.

[37]  Lieven De Lathauwer,et al.  On the Uniqueness of the Canonical Polyadic Decomposition of Third-Order Tensors - Part II: Uniqueness of the Overall Decomposition , 2013, SIAM J. Matrix Anal. Appl..

[38]  Wim Van Paesschen,et al.  Canonical Correlation Analysis Applied to Remove Muscle Artifacts From the Electroencephalogram , 2006, IEEE Transactions on Biomedical Engineering.

[39]  Lieven De Lathauwer,et al.  Blind Separation of Exponential Polynomials and the Decomposition of a Tensor in Rank-(Lr, Lr, 1) Terms , 2011, SIAM J. Matrix Anal. Appl..

[40]  Rasmus Bro,et al.  Tucker core consistency for validation of restricted Tucker3 models. , 2012, Analytica chimica acta.

[41]  J. Kruskal Three-way arrays: rank and uniqueness of trilinear decompositions, with application to arithmetic complexity and statistics , 1977 .

[42]  Sabine Van Huffel,et al.  Classification of Structured EEG Tensors Using Nuclear Norm Regularization: Improving P300 Classification , 2013, 2013 International Workshop on Pattern Recognition in Neuroimaging.

[43]  S. Vanhuffel,et al.  Algorithm for time-domain NMR data fitting based on total least squares , 1994 .

[44]  Enhua Shen,et al.  Advances in Cognitive Neurodynamics ICCN 2007 , 2008 .

[45]  Wim Van Paesschen,et al.  Modeling common dynamics in multichannel signals with applications to artifact and background removal in EEG recordings , 2005, IEEE Transactions on Biomedical Engineering.

[46]  Lieven De Lathauwer,et al.  Optimization-Based Algorithms for Tensor Decompositions: Canonical Polyadic Decomposition, Decomposition in Rank-(Lr, Lr, 1) Terms, and a New Generalization , 2013, SIAM J. Optim..

[47]  Lieven De Lathauwer,et al.  Decompositions of a Higher-Order Tensor in Block Terms - Part II: Definitions and Uniqueness , 2008, SIAM J. Matrix Anal. Appl..

[48]  E. Martínez-Montes,et al.  EEG Pattern Recognition by Multidimensional Space-Time-Frequency Analysis , 2013 .

[49]  H. Kiers,et al.  Three-mode principal components analysis: choosing the numbers of components and sensitivity to local optima. , 2000, The British journal of mathematical and statistical psychology.

[50]  W. van Paesschen,et al.  Improving the Interpretation of Ictal Scalp EEG: BSS–CCA Algorithm for Muscle Artifact Removal , 2007, Epilepsia.

[51]  Sabine Van Huffel,et al.  ResearchArticle Canonical Decomposition of Ictal Scalp EEG and Accurate Source Localisation: Principles and Simulation Study , 2007 .