Evaluation of adaptive parafac alogorithms for tracking of simulated moving brain sources

In this paper, we proposed an online 2D localization method for tracking of dynamic moving brain sources. For this purpose, we used an adaptive version of PARAllel FACtor (PARAFAC) analysis for factorization of electroencephalographic (EEG) signals. We utilized Boundary Element Method (BEM) with four layers to solve the forward problem for the simulated EEG signals caused by two moving dipoles within the brain. Then, we created an appropriate tensor built by second order statistics of EEG signals. We adopted an online method to brain source localization called the Recursive Least Squares Tracking (RLST) as an adaptive PARAFAC algorithm with two windowing schemes. Finally, we evaluated the performance of the method applied to EEG signals.

[1]  R. Van de Walle,et al.  Muscle and eye movement artifact removal prior to EEG source localization , 2006, 2006 International Conference of the IEEE Engineering in Medicine and Biology Society.

[2]  Diego Clonda,et al.  Bayesian spatio-temporal approach for EEG source reconstruction: conciliating ECD and distributed models , 2006, IEEE Transactions on Biomedical Engineering.

[3]  Lan-Da Van,et al.  FPGA implementation of 4-channel ICA for on-line EEG signal separation , 2008, 2008 IEEE Biomedical Circuits and Systems Conference.

[4]  Saeid Sanei,et al.  Simultaneous localization and separation of biomedical signals by tensor factorization , 2009, 2009 IEEE/SP 15th Workshop on Statistical Signal Processing.

[5]  Nikos D. Sidiropoulos,et al.  Adaptive Algorithms to Track the PARAFAC Decomposition of a Third-Order Tensor , 2009, IEEE Transactions on Signal Processing.

[6]  A. Cichocki,et al.  Steady-state visually evoked potentials: Focus on essential paradigms and future perspectives , 2010, Progress in Neurobiology.

[7]  Scott Makeig,et al.  Neuroelectromagnetic Forward Head Modeling Toolbox , 2010, Journal of Neuroscience Methods.

[8]  Ganesh R. Naik,et al.  An Overview of Independent Component Analysis and Its Applications , 2011, Informatica.

[9]  S. Sanei,et al.  Factorization based blind identification and separation of nonstationary seizure signals , 2012, The 16th CSI International Symposium on Artificial Intelligence and Signal Processing (AISP 2012).