Time domain Feature extraction and classification of EEG data for Brain Computer Interface
29 May 2012-pp 1136-1139
TL;DR: The classification accuracy with TD features is found to be considerably increasedBesides reduction in the memory space and processing time of the classifier used in BCI applications, the classification accuracies are obtained for pair-wise classification.
Abstract: In the recent past Brain Computer Interface (BCI) has become popular in the field of rehabilitation engineering for physically challenged people to improve their day-to-day activities independently. A proper BCI can possibly be achieved by proper classification and feature extraction techniques from the Electroencephalogram (EEG) data acquired from the brain. In this paper time domain (TD) features, like Mean Absolute Value (MAV), Zero Crossings (ZC), Slope Sign Changes (SSC) and Waveform Length (WL) is considered for classification of six channels of EEG data with time window of size 1-sec containing 250 data with an overlap of 125 data. A pair-wise combination of five different mental tasks has been considered for classification using Linear Discriminate Analysis (LDA) for seven subjects. Classification accuracies ranging from 67%–100% is obtained for pair-wise classification. The classification accuracy with TD features is found to be considerably increased besides reduction in the memory space and processing time of the classifier used in BCI applications.
Citations
More filters
TL;DR: The feasibility of fusing sEMG and EEG signals towards improving motion classification accuracy for above-elbow amputees is demonstrated, which might enhance the control performances of multifunctional myoelectric prostheses in clinical application.
Abstract: Most of the modern motorized prostheses are controlled with the surface electromyography (sEMG) recorded on the residual muscles of amputated limbs. However, the residual muscles are usually limited, especially after above-elbow amputations, which would not provide enough sEMG for the control of prostheses with multiple degrees of freedom. Signal fusion is a possible approach to solve the problem of insufficient control commands, where some non-EMG signals are combined with sEMG signals to provide sufficient information for motion intension decoding. In this study, a motion-classification method that combines sEMG and electroencephalography (EEG) signals were proposed and investigated, in order to improve the control performance of upper-limb prostheses. Four transhumeral amputees without any form of neurological disease were recruited in the experiments. Five motion classes including hand-open, hand-close, wrist-pronation, wrist-supination, and no-movement were specified. During the motion performances, sEMG and EEG signals were simultaneously acquired from the skin surface and scalp of the amputees, respectively. The two types of signals were independently preprocessed and then combined as a parallel control input. Four time-domain features were extracted and fed into a classifier trained by the Linear Discriminant Analysis (LDA) algorithm for motion recognition. In addition, channel selections were performed by using the Sequential Forward Selection (SFS) algorithm to optimize the performance of the proposed method. The classification performance achieved by the fusion of sEMG and EEG signals was significantly better than that obtained by single signal source of either sEMG or EEG. An increment of more than 14% in classification accuracy was achieved when using a combination of 32-channel sEMG and 64-channel EEG. Furthermore, based on the SFS algorithm, two optimized electrode arrangements (10-channel sEMG + 10-channel EEG, 10-channel sEMG + 20-channel EEG) were obtained with classification accuracies of 84.2 and 87.0%, respectively, which were about 7.2 and 10% higher than the accuracy by using only 32-channel sEMG input. This study demonstrated the feasibility of fusing sEMG and EEG signals towards improving motion classification accuracy for above-elbow amputees, which might enhance the control performances of multifunctional myoelectric prostheses in clinical application. The study was approved by the ethics committee of Institutional Review Board of Shenzhen Institutes of Advanced Technology, and the reference number is SIAT-IRB-150515-H0077.
145 citations
TL;DR: Experimental results show that the combination strategy of sample entropy and extreme learning machine has shown great performance, which obtains good classification accuracy and low training time.
Abstract: EEG signals play an important role in both the diagnosis of neurological diseases and understanding the psychophysiological processes. Classification of EEG signals includes feature extraction and feature classification. This paper uses approximate entropy and sample entropy based on wavelet package decomposition as the feature exaction methods and employs support vector machine and extreme learning machine as the classifiers. Experiments are performed in epileptic EEG data and five mental tasks, respectively. Experimental results show that the combination strategy of sample entropy and extreme learning machine has shown great performance, which obtains good classification accuracy and low training time.
57 citations
TL;DR: This study proposes a method to improve the accuracy of a 3-class BCI using electroencephalographic (EEG) signals and extracts time-domain features and classifies them using a support vector machine (SVM) with a radial basis kernel function (RBF).
Abstract: Brain-computer interface (BCI) allows collaboration between humans and machines. It translates the electrical activity of the brain to understandable commands to operate a machine or a device. In this study, we propose a method to improve the accuracy of a 3-class BCI using electroencephalographic (EEG) signals. This BCI discriminates rest against imaginary grasps and elbow movements of the same limb. This classification task is challenging because imaginary movements within the same limb have close spatial representations on the motor cortex area. The proposed method extracts time-domain features and classifies them using a support vector machine (SVM) with a radial basis kernel function (RBF). An average accuracy of 74.2% was obtained when using the proposed method on a dataset collected, prior to this study, from 12 healthy individuals. This accuracy was higher than that obtained when other widely used methods, such as common spatial patterns (CSP), filter bank CSP (FBCSP), and band power methods, were used on the same dataset. These results are encouraging and the proposed method could potentially be used in future applications including BCI-driven robotic devices, such as a portable exoskeleton for the arm, to assist individuals with impaired upper extremity functions in performing daily tasks.
30 citations
TL;DR: Modified WAMP and modified SSC prove that for BMI applications, in which a few errors can be tolerated, these combinations of feature-classifier are suitable.
Abstract: The tradeoff between computational complexity and speed, in addition to growing demands for real-time BMI (brain–machine interface) systems, expose the necessity of applying methods with least possible complexity. Willison amplitude (WAMP) and slope sign change (SSC) are two promising time–domain features only if the right threshold value is defined for them. To overcome the drawback of going through trial and error for the determination of a suitable threshold value, modified WAMP and modified SSC are proposed in this paper. Besides, a comprehensive assessment of statistical time–domain features in which their effectiveness is evaluated with a support vector machine (SVM) is presented. To ensure the accuracy of the results obtained by the SVM, the performance of each feature is reassessed with supervised fuzzy C-means. The general assessment shows that every subject had at least one of his performances near or greater than 80%. The obtained results prove that for BMI applications, in which a few errors can be tolerated, these combinations of feature–classifier are suitable. Moreover, features that could perform satisfactorily were selected for feature combination. Combinations of the selected features are evaluated with the SVM, and they could significantly improve the results, in some cases, up to full accuracy.
28 citations
TL;DR: This work aims to verify whether the choice of an epileptic EEG segment as reference can affect the performance of classifiers built from data, and proposes a CC with artificial reference (CCAR) method in order to reduce possible consequences of the random selection of a signal as reference.
Abstract: Several neurological disorders, such as epilepsy, can be diagnosed by electroencephalogram (EEG). Data mining supported by machine learning (ML) techniques can be used to find patterns and to build classifiers for the data. In order to make it possible, data should be represented in an appropriate format, e.g. attribute-value table, which can be built by feature extraction approaches, such as the cross-correlation (CC) method, which uses one signal as reference and correlates it with other signals. However, the reference is commonly selected randomly and, to the best of our knowledge, no studies have been conducted to evaluate whether this choice can affect the ML method performance. Thereby, this work aims to verify whether the choice of an epileptic EEG segment as reference can affect the performance of classifiers built from data. Also, a CC with artificial reference (CCAR) method is proposed in order to reduce possible consequences of the random selection of a signal as reference. Two experimental evaluations were conducted in a set of 200 EEG segments to induce classifiers using ML algorithms, such as J48, 1NN, naive Bayes, BP-MLP, and SMO. In the first study, each epileptic EEG segment was selected as reference to apply CC and ML methods. The evaluation found extremely significant difference, evidencing that the choice of an EEG segment as reference can influence the performance of ML methods. In the second study, the CCAR method was performed, in which statistical tests, only in comparisons involving the SMO classifier, showed not-so-good results.
16 citations
Cites background from "Time domain Feature extraction and ..."
...[26], time domain features, such as mean absolute value (MAV) [51, 68], zero crossings (ZC) [8], slope sign changes (SSC) [33], and waveform length (WL) [29] were considered in classification of EEG collected from seven subjects in order to label motor imagery (MI) tasks....
[...]
References
More filters
01 Jun 2000
TL;DR: The data indicate that a P300-based BCI is feasible and practical, however, these conclusions are based on tests using healthy individuals, which indicates that an off line version of the system can communicate at the rate of 7.8 characters a minute and achieve 80% accuracy.
Abstract: Describes a study designed to assess a brain-computer interface (BCI), originally described by Farwell and Donchin in 1988. The system utilizes the fact that the rare events in the oddball paradigm elicit the P300 component of the event-related potential (ERP). The BCI presents the user with a matrix of 6 by 6 cells, each containing one letter of the alphabet. The user focuses attention on the cell containing the letter to be communicated while the rows and the columns of the matrix are intensified. Each intensification is an event in the oddball sequence, the row and the column containing the attended cell are "rare" items and, therefore, only these events elicit a P300. The computer thus detects the transmitted character by determining which row and which column elicited the P300. The authors report an assessment, using a bootstrapping approach, which indicates that an off line version of the system can communicate at the rate of 7.8 characters a minute and achieve 80% accuracy. The system's performance in real time was also assessed. The authors' data indicate that a P300-based BCI is feasible and practical. However, these conclusions are based on tests using healthy individuals.
1,233 citations
"Time domain Feature extraction and ..." refers methods in this paper
...The second method uses single-trial visual evoked potential (VEP) signals, where the subject is asked to observe a screen with alphabets or instructions [2]....
[...]
01 Jun 2000
TL;DR: This paper describes a research approach to develop a brain-computer interface (BCI) based on recognition of subject-specific EEG patterns using EEG signals recorded from sensorimotor areas during mental imagination of specific movements.
Abstract: Describes a research approach to develop a brain-computer interface (BCI) based on recognition of subject-specific EEG patterns. EEG signals recorded from sensorimotor areas during mental imagination of specific movements are classified on-line and used e.g. for cursor control. In a number of on-line experiments, various methods for EEG feature extraction and classification have been evaluated.
533 citations
TL;DR: The feasibility of establishing an alternative mode of communication between man and his surroundings using only the subject's brain waves was studied, indicating that it is possible to accurately distinguish between any two of the five tasks investigated.
Abstract: The feasibility of establishing an alternative mode of communication between man and his surroundings was studied. The form of communication proposed uses only the subject's brain waves, with no overt physical action required. The subject's electroencephalograms (EEG) were recorded while various mental tasks designed to elicit hemispheric responses were performed. Features formed from the EEG recording were then used as inputs into a Bayes quadratic classifier to test classification accuracy between the various tasks. The results obtained indicate that it is possible to accurately distinguish between any two of the five tasks investigated. A comparison between three different methods for creating the feature sets is also presented. >
466 citations
01 Jan 2007
TL;DR: Cluster analysis of the resulting neural networks’ hidden-unit weight vectors identifies which EEG channels are most relevant to this discrimination problem.
Abstract: Neural networks are trained to classify half-second segments of six-channel, EEG data into one of five classes corresponding to five cognitive tasks performed by four subjects. Two and three-layer feedforward neural networks are trained using 10-fold cross-validation and early stopping to control over-fitting. EEG signals were represented as autoregressive (AR) models. The average percentage of test segments correctly classified ranged from 71% for one subject to 38% for another subject. Cluster analysis of the resulting neural networks’ hidden-unit weight vectors identifies which EEG channels are most relevant to this discrimination problem.
156 citations
"Time domain Feature extraction and ..." refers result in this paper
...The use of time domain feature extraction and LDA classifier is very promising for the field of BCI as higher classification accuracies can be achieved in comparison to other methods, [10-12]....
[...]
18 Sep 2006
TL;DR: The results indicated that the classification performance and training time of the BCI design were improved through the use of additional gamma band features and classification performances were nearly invariant to the number of ENN hidden units or feature extraction method.
Abstract: A common method for designing brain-computer Interface (BCI) is to use electroencephalogram (EEG) signals extracted during mental tasks. In these BCI designs, features from EEG such as power and asymmetry ratios from delta, theta, alpha, and beta bands have been used in classifying different mental tasks. In this paper, the performance of the mental task based BCI design is improved by using spectral power and asymmetry ratios from gamma (24-37 Hz) band in addition to the lower frequency bands. In the experimental study, EEG signals extracted during five mental tasks from four subjects were used. Elman neural network (ENN) trained by the resilient backpropagation algorithm was used to classify the power and asymmetry ratios from EEG into different combinations of two mental tasks. The results indicated that 1) the classification performance and training time of the BCI design were improved through the use of additional gamma band features; 2) classification performances were nearly invariant to the number of ENN hidden units or feature extraction method
114 citations
"Time domain Feature extraction and ..." refers methods in this paper
...Palaniappan [6] used 4 subjects data for classification of mental tasks using Elman neural network (ENN) trained by the resilient backpropagation (BP) algorithm and obtained an accuracy of 86%....
[...]