Abstract - Electroencephalogram (EEG) is a highly sensitive instrument and is frequently corrupted with eye blinks. Methods based on adaptive noise cancellation (ANC) and discrete wavelet transform (DWT) have been used as a standard technique for removal of eye blink artefacts. However, these methods often require visual inspection and appropriate thresholding for identifying and removing artefactual components from the EEG signal. The proposed work describes an automated windowed method with a window size of 0.45 s that is slid forward and fed to a support vector machine (SVM) classifier for identification of artefacts, after the identification of artefacts, it is fed to an autoencoder for correction of artefacts. The proposed method is evaluated on the data collected from the project entitled ‘Analysis of Brain Waves and Development of Intelligent Model for Silent Speech Recognition’. From the results it is observed that the proposed method performs better in identifying and removing artefactual components from EEG data than existing wavelet and ANC based methods. The proposed method does not require the application of independent component analysis (ICA) before processing and can be applied to multiple channels in parallel.
Abstract - This paper proposes an automatic eyeblink artifacts removal method from corrupted-EEG signals using discrete wavelet transform (DWT) and meta-heuristically optimized threshold. The novel idea of thresholding approximation-coefficients (ACs) instead of detail-coefficients (DCs) of DWT of EEG in a backward manner is proposed for the first time for the removal of eyeblink artifacts. EEG is very sensitive and easily gets affected by eyeblink artifacts. First, the eyeblink corrupted EEG signals are identified using support vector machine (SVM) as a classifier. Then the corrupted EEG signal is decomposed using DWT up to the sixth level. Both the mother wavelet and the level of decomposition are selected using appropriate techniques. Then the ACs are thresholded in backward manner using the optimum threshold values followed by inverse DWT operation to reconstruct the original EEG signal. The AC at level 6 is thresholded and is used in IDWT with DC to get back the AC at level 5. Likewise, the backward thresholding of the ACs followed by IDWT is continued till the artifact free EEG signal is reconstructed at level 1. The optimum values of the thresholds of the ACs at different levels are optimized using two meta-heuristic algorithms, particle swarm optimization (PSO) and grey wolf optimization (GWO) for comparison. The results reveal that the proposed methodology is superior to the recently reported methods in terms of average correlation coefficient (CC) which states that the proposed method is better in terms of the quality of reconstruction in addition to being fully automatic.
Abstract - This study proposes a novel combination of independent component analysis (ICA) in conjunction with support vector machine (SVM) and denoising autoencoder (DA), for the first time, for removal of eyeblink artefacts from the corrupted electroencephalography (EEG). At first the eyeblink corrupted EEG signals are decomposed into independent components (ICs) using ICA, the corrupted-ICs are then identified using SVM as a classifier. From the corrupted-ICs, the artefacted segment is identified with a second SVM classifier and corrected by the pre-trained DA. Finally, inverse-ICA operation is applied on the remaining ICs and the corrected ICs to obtain the artefact-free EEG signal. The proposed methodology modifies only the portion corrupted with artefacts, and does not alter the uncorrupted part, thereby preserving the neural information in the original EEG. The proposed methodology was implemented to remove eyeblinks from the EEG data collected from the publicly available EEGLab data set. The results reveal that the proposed methodology is superior to the other recently reported methods in terms of the mutual information and average correlation coefficient. Further, the proposed method is automatic and does not require any intervention of the operator, whereas the other methods require intervention of the user.
In recent times, emotion recognition is in attention in brain computer interface (BCI) and human computer interaction (HCI) research area to provide a very good communication between brain and computer. The aim is to achieve a good recognition rate, although there are numerous researches have been conducted also there has been created several confusions with the definition of human emotions and the difference between emotions and moods. To detect brain signal, Electroencephalogram (EEG) signal has become biological marker. For its low cost, good time and spatial resolution EEG has been used widely in BCI researches. Extraction of features from EEG signals is one of the vital steps of EEG based emotion recognition. The appropriate feature selection for EEG based automatic emotion recognition is still now a big research topic. In this paper, a comprehensive survey is made on feature extraction methods and their comparative merits and limitations.
Brain computer interfaces (BCI’s) employing electroencephalographic signals are being applied to a wide variety of applications like motor imagery task classification, prosthetics etc. Electroencephalography (EEG) data are inherently non-stationary and noisy, and as such identification of appropriate features for classification is a crucial task. Selection of features based on genetic algorithms (GA) has been applied, but it leads to a redundant set of features. In the present work, grey wolf optimization (GWO) based feature selection method has been applied on EEG data for silent speech classification. The EEG data from the ABISSR (Analysis of Brain Waves and development of intelligent model for silent speech recognition) project was used in the proposed work. An accuracy of 65% was obtained in classifying five imagined vowels /a/, /e/, /i/, /o/ and /u/ from EEG data using support vector machine (SVM). Moreover, it was observed that the GWO outperformed GA in optimization.
Electroencephalography (EEG) data are highly susceptible to noise and are frequently corrupted with eye-blink artifacts. Methods based on independent component analysis (ICA) and discrete wavelet transform (DWT) have been used as a standard for removal of such kinds of artifacts. However, these methods often require visual inspection and appropriate thresholding for identifying and removing artifactual components from the EEG signal. The proposed method presents a windowed method, where an LDA classifier is used for identification of artifacts and RBF neural network is used for correcting artifacts. In the present work, we propose a robust and automated method for identification and removal of artifacts from EEG signals, without the need for any visual inspection or threshold selection. Using test data contaminated with eye-blink artifacts, it is observed that our proposed method performs better in identifying and removing artifactual components from EEG data than the existing thresholding methods and does not require the application of ICA for identification of artifacts and can also be applied to any number of channels.
The proposed work presents a discrete wavelet based feature extraction method in conjunction with a back propagation neural network (BPNN) for the classification of emotions from EEG recordings. The EEG recordings were used from DEAP dataset (A Database for emotion analysis using physiological signals) for evaluation of the method. Russell’s model was used for quantification of emotions in valence and arousal dimension space. The emotions were classified into four classes, namely – high arousal positive valence (HAPV), high arousal negative valence (HANV), low arousal positive valence (LAPV) and low arousal negative valence (LANV). We have used three mother wavelets Sym8, Coif5 and Db8 for extracting four features, namely entropy, mean, standard deviation and variance. Classification of the emotions was done using BPNN. An average classification accuracy of 57.9% was obtained in the classification of the above mentioned four classes using different mother wavelets.
Brain Computer Interface (BCI) enables us to record and process the information generated by the brain and process them. Due to high variability of the Electroencephalogram (EEG) data, multiple trails are recorded for a particular task. The present work aims to improve the accuracy for motor imagery task classification by selecting the most prominent trail from the multiple trails recorded during motor imagery. In this paper, we propose a novel weight optimization algorithm for common spatial filtering (CSP) using evolutionary algorithms (i.e. cuckoo search algorithm (CSA), firefly algorithm (FA) and gravitational search algorithm (GSA)) to select the most prominent trial from the multiple trails recorded for feature extraction. The features extracted from the selected trials were thus used for motor imagery task classification. The performance was evaluated on the extracted features from the selected trials using two classifiers namely linear discriminant analysis (LDA) and support vector machines (SVM). It is observed that FA with band power as a feature gives the best performance in comparison to the earlier reported methods i.e. average, error based and alternating direction method of multipliers (ADMM).