Discriminative Power Feature Selection Method for Motor Imagery EEG Classification in Brain Computer Interface Systems

Yu, Xinyang; Park, Seung Min; Ko, Kwang-Eun; Sim, Kwee-Bo

doi:10.5391/ijfis.2013.13.1.12

Cited by 3 publications

(5 citation statements)

References 13 publications

(10 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The EEG acquisition system is a g.tec Guger Technologies based one [27]. The active electrodes were mounted at positions C3, Cz, C4, P3, Pz, P3, CP3, CP4, according to [10][11][12][13][14][15][16][17][18][19][20] International System of electrode placement. The mentioned channels are considered significant to highlight real or imagined motor activity [11,28].…”

Section: Databasesmentioning

confidence: 99%

“…Sensorimotor rhythms (SMR) represent oscillations recorded in the motor cortex. The brain oscillations are classified according to the following specific frequency bands: Delta (0.1 -4 Hz), Theta (4 -8 Hz), Alpha (8 -12 Hz), Beta (12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23)(24)(25)(26)(27)(28)(29)(30) and Gamma . The Alpha rhythm activity recorded in the sensorimotor area is also called Mu rhythm.…”

Section: Introductionmentioning

confidence: 99%

“…The most known feature extraction methods implemented for discriminating the motor tasks are: the principal component analysis (PCA) [13][14][15] the independent component analysis (ICA) [16][17][18], the autoregressive spectral estimation [19], the fractal spectra [20], the phase synchronization [18,[21][22][23] and the wavelet transform [24][25][26].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Amplitude Modulation Index as Feature in a Brain Computer Interface

Eva¹,

Lazar²

2019

View full text Add to dashboard Cite

The goal of this research involving a motor imagery brain-computer interface paradigm is to assess the possibility of enhancing the classification rate handling a feature vector based on the modulation of electrophysiological brain activity in specific bands. A new amplitude modulation energy index of the cerebral rhythms is proposed as feature vector concept. The method is proven on a public database and on a set of electroencephalographic data recorded in our own laboratory. In both cases, only eight electrodes are used in order to reach high performance classifying rates. The discrimination of motor tasks (imagination of right and left hand movements) is analyzed by means of five classifiers: support vector machine, k nearest neighbor, linear discriminant analysis, quadratic discriminant analysis and Mahalanobis distance based classifier. For our database, the medians of the classification rates for two of classifiers are very high (94.62 % -97.76 %) when some rhythms are modulated in theta and alpha bands. Significantly higher classification rates reported herein (greater than 90 % for both of the databases) compared with classifiers trained on the other features prove that the index may be very useful for highlighting the modulation found in certain bands of the EEG rhythms.

show abstract

Section: Databasesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Amplitude Modulation Index as Feature in a Brain Computer Interface

Eva¹,

Lazar²

2019

View full text Add to dashboard Cite

show abstract

“…LS-SVM achieved higher success rates than the two logistic regression classifiers (first dataset: 95.72% ± 4.35% versus 89.54% ± 8.61% and 93.38% ± 6.76%; second dataset: 97.89% ± 2.96% versus 95.31% ± 5.88% and 94.87% ± 6.98%). In [90] a PCA-based feature selection and RBF-SVM allowed discriminating data from two different datasets, including right-hand and foot motor imagery and lefthand and foot motor imagery, with accuracies ranging from 61.36% to 90.63%. In [91] movement-related independent components and RBF-SVM were used for the classification of left-hand, right-hand and foot motor imagery, with accuracy of 65%.…”

Section: Motor Tasksmentioning

confidence: 99%

Support vector machines to detect physiological patterns for EEG and EMG-based human–computer interaction: a review

et al. 2017

View full text Add to dashboard Cite

Support vector machines (SVMs) are widely used classifiers for detecting physiological patterns in human-computer interaction (HCI). Their success is due to their versatility, robustness and large availability of free dedicated toolboxes. Frequently in the literature, insufficient details about the SVM implementation and/or parameters selection are reported, making it impossible to reproduce study analysis and results. In order to perform an optimized classification and report a proper description of the results, it is necessary to have a comprehensive critical overview of the applications of SVM. The aim of this paper is to provide a review of the usage of SVM in the determination of brain and muscle patterns for HCI, by focusing on electroencephalography (EEG) and electromyography (EMG) techniques. In particular, an overview of the basic principles of SVM theory is outlined, together with a description of several relevant literature implementations. Furthermore, details concerning reviewed papers are listed in tables and statistics of SVM use in the literature are presented. Suitability of SVM for HCI is discussed and critical comparisons with other classifiers are reported.

show abstract

“…In recent decades, an entropy based qEEG has been long of interest in characterizing dynamics of EEG recording because EEG signals have nonstationarity [5,6]. As a primary tool for quantifying EEG signals, an entropy measures a degree of regularity of the underlying EEG signals [7,8].…”

Section: Introductionmentioning

confidence: 99%

Improved Multiscale Permutation Entropy Measure for Analysis of Brain Waves

Choi

2017

IJFIS

View full text Add to dashboard Cite

This paper presents a novel multiscale entropy measure for analyzing brain rhythms, called electroencephalograms (EEGs), with an aim of unveiling the underlying dynamics of EEG over multiple time scales. This work is inspired by the known fact that neurological signals such as EEG has distinct dynamics over different temporal scales. To reflect the nonlinear and nonstationary nature of EEGs, the recently developed empirical mode decomposition is incorporated, allowing an EEG to be decomposed into its inherent multiple temporal scale elements, referred to as intrinsic mode functions (IMFs). By computing the permutation entropy of each IMF in a time-dependent manner and averaging them over multiple scales, it yields a data-adaptive multiscale permutation entropy (DAMPE) measure for analyzing the brain waves. Simulation and experimental results show that the proposed DAMPE is efficient to reveal the dynamical changes over multiple scales.

show abstract

Discriminative Power Feature Selection Method for Motor Imagery EEG Classification in Brain Computer Interface Systems

Cited by 3 publications

References 13 publications

Amplitude Modulation Index as Feature in a Brain Computer Interface

Amplitude Modulation Index as Feature in a Brain Computer Interface

Support vector machines to detect physiological patterns for EEG and EMG-based human–computer interaction: a review

Improved Multiscale Permutation Entropy Measure for Analysis of Brain Waves

Contact Info

Product

Resources

About