Spectral Graph Theory-Based Spatio-spectral Filters for Motor Imagery Brain–Computer Interface
01 Jan 2021-pp 247-256
TL;DR: A novel approach that utilizes graph theory-based unsupervised feature selection method to determine a reduced set of non-redundant and relevant frequency bands is proposed and shows improvement in classification performance.
Abstract: Motor imagery brain–computer interfaces are one of the widely adopted techniques for imparting basic communication capability to motor disabled patients The preciseness of a motor imagery BCI task classification is highly dependent on identifying the subject-specific relevant subset of frequency filters This article proposes a novel approach that utilizes graph theory-based unsupervised feature selection method to determine a reduced set of non-redundant and relevant frequency bands The empirical analysis of the proposed method is conducted on publicly available datasets, and the obtained results show improvement in classification performance Further, the performed Friedman statistical test also establishes that the proposed approach surpasses the baseline techniques in classification accuracy
References
More filters
Book•
24 Jan 2012
TL;DR: The Future of BCIs: Meeting the Expectations Jonathan R. Wolpaw and Elizabeth Winter Wolpaws Index finds that BCI Therapeutic Applications for Improving Brain Function and Ethical Issues in BCI Research are becoming more important than ever.
Abstract: Contributors PART I: INTRODUCTION 1. Brain-Computer Interfaces: Something New under the Sun Jonathan R. Wolpaw and Elizabeth Winter Wolpaw PART II: BRAIN SIGNALS FOR BCIs 2. Neuronal Activity in Motor Cortex and Related Areas Lee E. Miller and Nicholas Hatsopoulos 3. Electric and Magnetic Fields Produced by the Brain Paul L. Nunez 4. Signals Reflecting Brain Metabolic Activity Nick F. Ramsey PART III: BCI DESIGN, IMPLEMENTATION, AND OPERATION 5. Acquiring Brain Signals from Within the Brain Kevin Otto, Kip A. Ludwig, Daryl R. Kipke 6. Acquiring Brain Signals from Outside the Brain Ramesh Srinivasan 7. BCI Signal Processing: Feature Extraction Dean J. Krusienski, Dennis J. McFarland, and Jose C. Principe 8. BCI Signal Processing: Feature Translation Dennis J. McFarland and Dean J. Krusienski 9. BCI Hardware and Software J. Adam Wilson, Christoph Guger, and Gerwin Schalk 10. BCI Operating Protocols Steven G. Mason, Brendan Z. Allison, and Jonathan R. Wolpaw 11. BCI Applications Jane E. Huggins and Debra Zeitlin PART IV: EXISTING BCIs 12. BCIs that Use P300 Event-Related Potentials Eric W. Sellers, Yael Arbel, and Emanuel Donchin 13. BCIs that Use Sensorimotor Rhythms Gert Pfurtscheller and Dennis J. McFarland 14. BCIs that Use Steady-State Visual Evoked Potentials or Slow Cortical Potentials Brendan Z. Allison, Josef Faller, and Christa Neuper 15. BCIs that Use Electrocorticographic (ECoG) Activity Gerwin Schalk 16. BCIs that Use Signals Recorded in Motor Cortex John P. Donoghue 17. BCIs that Use Signals Recorded in Parietal or Premotor Cortex Hansjorg Scherberger 18. BCIs that Use Brain Metabolic Signals Ranganatha Sitaram, Sangkyung Lee, and Niels Birbaumer PART V: USING BCIs 19. BCI Users and Their Needs Leigh R. Hochberg and Kim D. Anderson 20. Clinical Evaluation of BCIs Theresa M. Vaughan, Eric W. Sellers, and Jonathan R. Wolpaw 21. Dissemination: Getting BCIs to the People Who Need Them Frances J.R. Richmond and Gerald E. Loeb 22. BCI Therapeutic Applications for Improving Brain Function Janis J. Daly and Ranganatha Sitaram 23. BCI Applications for the General Population Benjamin Blankertz, Michael Tangermann, and Klaus-Robert Mu?ller 24. Ethical Issues in BCI Research Mary-Jane Schneider, Joseph J. Fins, and Jonathan R. Wolpaw PART VI: CONCLUSION 25. The Future of BCIs: Meeting the Expectations Jonathan R. Wolpaw and Elizabeth Winter Wolpaw Index
760 citations
TL;DR: It is shown that a suitably arranged interaction between these concepts can significantly boost BCI performances and derive information-theoretic predictions and demonstrate their relevance in experimental data.
Abstract: Noninvasive electroencephalogram (EEG) recordings provide for easy and safe access to human neocortical processes which can be exploited for a brain-computer interface (BCI). At present, however, the use of BCIs is severely limited by low bit-transfer rates. We systematically analyze and develop two recent concepts, both capable of enhancing the information gain from multichannel scalp EEG recordings: 1) the combination of classifiers, each specifically tailored for different physiological phenomena, e.g., slow cortical potential shifts, such as the premovement Bereitschaftspotential or differences in spatio-spectral distributions of brain activity (i.e., focal event-related desynchronizations) and 2) behavioral paradigms inducing the subjects to generate one out of several brain states (multiclass approach) which all bare a distinctive spatio-temporal signature well discriminable in the standard scalp EEG. We derive information-theoretic predictions and demonstrate their relevance in experimental data. We will show that a suitably arranged interaction between these concepts can significantly boost BCI performances.
614 citations
02 May 2007
TL;DR: This work proposes a new method called sub-band common spatial pattern (SBCSP), which outperforms the other two approaches and achieves similar result as compared to the best one in the literature which was obtained by a time-consuming fine-tuning process.
Abstract: Brain-computer interface (BCI) is a system to translate humans thoughts into commands. For electroencephalography (EEG) based BCI, motor imagery is considered as one of the most effective ways. Different imagery activities can be classified based on the changes in mu and/or beta rhythms and their spatial distributions. However, the change in these rhythmic patterns varies from one subject to another. This causes an unavoidable time-consuming fine-tuning process in building a BCI for every subject. To address this issue, we propose a new method called sub-band common spatial pattern (SBCSP) to solve the problem. First, we decompose the EEG signals into sub-bands using a filter bank. Subsequently, we apply a discriminative analysis to extract SBCSP features. The SBCSP features are then fed into linear discriminant analyzers (LDA) to obtain scores which reflect the classification capability of each frequency band. Finally, the scores are fused to make decision. We evaluate two fusion methods: recursive band elimination (RBE) and meta-classifier (MC). We assess our approaches on a standard database from BCI Competition III. We also compare our method with two other approaches that address the same issue. The results show that our method outperforms the other two approaches and achieves similar result as compared to the best one in the literature which was obtained by a time-consuming fine-tuning process.
280 citations
TL;DR: Experimental results demonstrate that the proposed method significantly outperforms the existing methods in terms of classification error, and two well-known classifiers (LDA and SVM) are investigated.
Abstract: This paper presents a novel algorithm (CVSTSCSP) for determining discriminative features from an optimal combination of temporal, spectral and spatial information for motor imagery brain computer interfaces. The proposed method involves four phases. In the first phase, EEG signal is segmented into overlapping time segments and bandpass filtered through frequency filter bank of variable size subbands. In the next phase, features are extracted from the segmented and filtered data using stationary common spatial pattern technique (SCSP) that can handle the non- stationarity and artifacts of EEG signal. The univariate feature selection method is used to obtain a relevant subset of features in the third phase. In the final phase, the classifier is used to build adecision model. In this paper, four univariate feature selection methods such as Euclidean distance, correlation, mutual information and Fisher discriminant ratio and two well-known classifiers (LDA and SVM) are investigated. The proposed method has been validated using the publicly available BCI competition IV dataset Ia and BCI Competition III dataset IVa. Experimental results demonstrate that the proposed method significantly outperforms the existing methods in terms of classification error. A reduction of 76.98%, 75.65%, 73.90% and 72.21% in classification error over both datasets and both classifiers can be observed using the proposed CVSTSCSP method in comparison to CSP, SBCSP, FBCSP and CVSCSP respectively.
39 citations