Noise removal in electroencephalogram signals using an artificial neural network based on the simultaneous perturbation method

Abstract Electroencephalogram (EEG) recordings often experience interference by different kinds of noise, including white, muscle and baseline, severely limiting its utility. Artificial neural networks (ANNs) are effective and powerful tools for removing interference from EEGs. Several methods have been developed, but ANNs appear to be the most effective for reducing muscle and baseline contamination, especially when the contamination is greater in amplitude than the brain signal. An ANN as a filter for EEG recordings is proposed in this paper, developing a novel framework for investigating and comparing the relative performance of an ANN incorporating real EEG recordings. This method is based on a growing ANN that optimized the number of nodes in the hidden layer and the coefficient matrices, which are optimized by the simultaneous perturbation method. The ANN improves the results obtained with the conventional EEG filtering techniques: wavelet, singular value decomposition, principal component analysis, adaptive filtering and independent components analysis. The system has been evaluated within a wide range of EEG signals. The present study introduces a new method of reducing all EEG interference signals in one step with low EEG distortion and high noise reduction.

[1]  D. Lehmann,et al.  Functional imaging with low-resolution brain electromagnetic tomography (LORETA): a review. , 2002, Methods and findings in experimental and clinical pharmacology.

[2]  Hugo Vélez-Pérez,et al.  Blind source separation, wavelet denoising and discriminant analysis for EEG artefacts and noise cancelling , 2012, Biomed. Signal Process. Control..

[3]  Simon Haykin,et al.  Neural Networks and Learning Machines , 2010 .

[4]  Tricia J. Willink.  Efficient Adaptive SVD Algorithm for MIMO Applications , 2008, IEEE Transactions on Signal Processing.

[5]  Geng Chen,et al.  Dynamic weighting ensemble classifiers based on cross-validation , 2011, Neural Comput. Appl..

[6]  Yingyi Hong,et al.  Passive Filter Planning Using Simultaneous Perturbation Stochastic Approximation , 2010, IEEE Transactions on Power Delivery.

[7]  Lan-Da Van,et al.  Energy-Efficient FastICA Implementation for Biomedical Signal Separation , 2011, IEEE Transactions on Neural Networks.

[8]  A. M. Torres,et al.  Dynamic Fuzzy Neural Network Based Learning Algorithms for Ocular Artefact Reduction in EEG Recordings , 2013, Neural Processing Letters.

[9]  Boualem Boashash,et al.  Wavelet Denoising Based on the MAP Estimation Using the BKF Prior With Application to Images and EEG Signals , 2013, IEEE Transactions on Signal Processing.

[10]  Rik Vullings,et al.  An Adaptive Kalman Filter for ECG Signal Enhancement , 2011, IEEE Transactions on Biomedical Engineering.

[11]  Ki Hwan Eom,et al.  Learning method of the ADALINE using the fuzzy logic system , 2004, Neural Computing & Applications.

[12]  Michel J. A. M. van Putten,et al.  Reduction of TMS Induced Artifacts in EEG Using Principal Component Analysis , 2013 .

[13]  Chen Geng,et al.  Dynamic weighting ensemble classifiers based on cross-validation , 2011, Neural Computing and Applications.

[14]  Pierre Comon,et al.  Robust Independent Component Analysis by Iterative Maximization of the Kurtosis Contrast With Algebraic Optimal Step Size , 2010, IEEE Transactions on Neural Networks.

[15]  Simon Haykin,et al.  Neural networks expand SP's horizons , 1996, IEEE Signal Process. Mag..

[16]  James C. Spall.  Feedback and Weighting Mechanisms for Improving Jacobian Estimates in the Adaptive Simultaneous Perturbation Algorithm , 2009, IEEE Trans. Autom. Control..

[17]  M. Z. U. Rahman,et al.  Efficient and Simplified Adaptive Noise Cancelers for ECG Sensor Based Remote Health Monitoring , 2012, IEEE Sensors Journal.

[18]  Mario Sansone,et al.  Adaptive removal of gradients-induced artefacts on ECG in MRI: a performance analysis of RLS filtering , 2010, Medical & Biological Engineering & Computing.

[19]  James C. Spall,et al.  Introduction to stochastic search and optimization - estimation, simulation, and control , 2003, Wiley-Interscience series in discrete mathematics and optimization.

[20]  Ferat Sahin,et al.  New classification techniques for electroencephalogram (EEG) signals and a real-time EEG control of a robot , 2011, Neural Computing and Applications.

[21]  Daniel E. Rivera,et al.  Plant-Friendly Signal Generation for System Identification Using a Modified Simultaneous Perturbation Stochastic Approximation (SPSA) Methodology , 2011, IEEE Transactions on Control Systems Technology.

[22]  Octavian Fratu,et al.  Imperfect cross-correlation and amplitude balance effects on conventional multiuser decoder with turbo encoding , 2010, Digit. Signal Process..

[23]  Alberto J. Palma,et al.  Efficient wavelet-based ECG processing for single-lead FHR extraction , 2013, Digit. Signal Process..

[24]  Osamu Saotome,et al.  Vehicle inductive signatures recognition using a Madaline neural network , 2009, Neural Computing and Applications.

[25]  Marcello R. Napolitano,et al.  Bounding set calculation for neural network-based output feedback adaptive control systems , 2010, Neural Computing and Applications.

[26]  S. Dandapat,et al.  ECG signal denoising using higher order statistics in Wavelet subbands , 2010, Biomed. Signal Process. Control..

[27]  Michel J. A. M. van Putten,et al.  Reduction of TMS Induced Artifacts in EEG Using Principal Component Analysis , 2013, IEEE Transactions on Neural Systems and Rehabilitation Engineering.

[28]  Rui J. P. de Figueiredo,et al.  Learning rules for neuro-controller via simultaneous perturbation , 1997, IEEE Trans. Neural Networks.

[29]  Daniel Graupe,et al.  Principles of Artificial Neural Networks , 2018, Advanced Series in Circuits and Systems.

[30]  Danwei Wang,et al.  A robust recurrent simultaneous perturbation stochastic approximation training algorithm for recurrent neural networks , 2013, Neural Computing and Applications.

[31]  J. Spall,et al.  Nonlinear adaptive control using neural networks: estimation with a smoothed form of simultaneous perturbation gradient approximation , 1994, Proceedings of 1994 American Control Conference - ACC '94.

[32]  Junshui Ma,et al.  Muscle artifacts in multichannel EEG: Characteristics and reduction , 2012, Clinical Neurophysiology.

[33]  R. Komanduri,et al.  Nonlinear adaptive wavelet analysis of electrocardiogram signals. , 2007, Physical review. E, Statistical, nonlinear, and soft matter physics.

[34]  M. P. S. Chawla,et al.  PCA and ICA processing methods for removal of artifacts and noise in electrocardiograms: A survey and comparison , 2011, Appl. Soft Comput..

[35]  Simona Halunga,et al.  Performance evaluation for conventional and MMSE multiuser detection algorithms in imperfect reception conditions , 2010, Digit. Signal Process..

[36]  Rangaraj M. Rangayyan,et al.  Biomedical Signal Analysis: A Case-Study Approach , 2001 .

[37]  Chong Jin Ong,et al.  Automatic EEG Artifact Removal: A Weighted Support Vector Machine Approach With Error Correction , 2009, IEEE Transactions on Biomedical Engineering.

[38]  James C. Spall,et al.  Introduction to stochastic search and optimization - estimation, simulation, and control , 2003, Wiley-Interscience series in discrete mathematics and optimization.

[39]  Christian Jutten,et al.  A Nonlinear Bayesian Filtering Framework for ECG Denoising , 2007, IEEE Transactions on Biomedical Engineering.

[40]  R. Ward,et al.  EMG and EOG artifacts in brain computer interface systems: A survey , 2007, Clinical Neurophysiology.

[41]  James C. Spall,et al.  Introduction to Stochastic Search and Optimization. Estimation, Simulation, and Control (Spall, J.C. , 2007 .

[42]  Anil Kumar,et al.  EEG/ERP Adaptive Noise Canceller Design with Controlled Search Space (CSS) Approach in Cuckoo and Other Optimization Algorithms , 2013, IEEE/ACM Transactions on Computational Biology and Bioinformatics.

[43]  Springer-Verlag London Limited.  Quantitative methods for detecting cerebral infarction from multiple channel EEG recordings , 2012 .

[44]  Pablo Laguna,et al.  Bioelectrical Signal Processing in Cardiac and Neurological Applications , 2005 .

[45]  Babak Sokouti,et al.  A framework for diagnosing cervical cancer disease based on feedforward MLP neural network and ThinPrep histopathological cell image features , 2012, Neural Computing and Applications.

[46]  Joseph D. Bronzino,et al.  The Biomedical Engineering Handbook , 1995 .

[47]  Chris J. B. Macnab,et al.  Neural-adaptive control using alternate weights , 2011, Neural Computing and Applications.

[48]  Yunfeng Wu,et al.  Filtering electrocardiographic signals using an unbiased and normalized adaptive noise reduction system. , 2009, Medical engineering & physics.

[49]  Yutaka Maeda,et al.  Simultaneous perturbation learning rule for recurrent neural networks and its FPGA implementation , 2005, IEEE Transactions on Neural Networks.

[50]  Nurettin Acir.  Estimation of brainstem auditory evoked potentials using a nonlinear adaptive filtering algorithm , 2012, Neural Computing and Applications.

[51]  Daniel P. Ferris,et al.  Removal of movement artifact from high-density EEG recorded during walking and running. , 2010, Journal of neurophysiology.

[52]  Frank W. Sharbrough,et al.  Use of Principal Component Analysis in the Frequency Domain for Mapping Electroencephalographic Activities: Comparison with Phase‐Encoded Fourier Spectral Analysis , 2004, Brain Topography.

[53]  J. Spall.  Adaptive stochastic approximation by the simultaneous perturbation method , 1998, Proceedings of the 37th IEEE Conference on Decision and Control (Cat. No.98CH36171).

[54]  Christopher J. James,et al.  Employing spatially constrained ICA and wavelet denoising, for automatic removal of artifacts from multichannel EEG data , 2012, Signal Process..

[55]  Bernard Widrow,et al.  30 years of adaptive neural networks: perceptron, Madaline, and backpropagation , 1990, Proc. IEEE.

[56]  Yeng Chai Soh,et al.  Robust neural network tracking controller using simultaneous perturbation stochastic approximation , 2003, 42nd IEEE International Conference on Decision and Control (IEEE Cat. No.03CH37475).

[57]  Laurent Peyrodie,et al.  Improvements of Adaptive Filtering by Optimal Projection to filter different artifact types on long duration EEG recordings , 2012, Comput. Methods Programs Biomed..

[58]  S. Luo,et al.  A review of electrocardiogram filtering. , 2010, Journal of electrocardiology.

[59]  Umit Aydin,et al.  A Kalman filter-based approach to reduce the effects of geometric errors and the measurement noise in the inverse ECG problem , 2011, Medical & Biological Engineering & Computing.

[60]  Yutaka Maeda,et al.  A learning rule of neural networks via simultaneous perturbation and its hardware implementation , 1995, Neural Networks.

[61]  Dominic H. ffytche,et al.  A Novel Method for Reducing the Effect of Tonic Muscle Activity on the Gamma Band of the Scalp EEG , 2012, Brain Topography.

[62]  C Jutten,et al.  Model-based Bayesian filtering of cardiac contaminants from biomedical recordings , 2008, Physiological measurement.

[63]  P. Comon,et al.  Ica: a potential tool for bci systems , 2008, IEEE Signal Processing Magazine.

[64]  M. Ramasubba Reddy,et al.  A transform domain SVD filter for suppression of muscle noise artefacts in exercise ECG's , 2000, IEEE Transactions on Biomedical Engineering.

[65]  Jeff H. Duyn,et al.  Statistical feature extraction for artifact removal from concurrent fMRI-EEG recordings , 2012, NeuroImage.

[66]  M. P. S. Chawla,et al.  A comparative analysis of principal component and independent component techniques for electrocardiograms , 2009, Neural Computing and Applications.