Fig 4 - uploaded by Anurag Tiwari
Content may be subject to copyright.
Topographical visualization of 22 EEG channels in dataset 1 after channel setting where the dark boundary electrodes are used as candidate channels in our experiment.
Source publication
Brain-Computer Interface (BCI) provides a direct communicating pathway between the human brain and the external environment. In the BCI systems, electroencephalography (EEG) signals are used to represent different cognitive patterns corresponding to various limb movements or motor imagery (MI) activities. However, EEG signals are multichannel in na...
Contexts in source publication
Context 1
... used in our work consists of a set of 25 EEG channels. Since these channels are Electrooculography (EOG) channels, they are not used for analysis. The data collected from these channels are considered artifacts; therefore, these channels are directly deleted from the dataset, and the remaining 22 channels are used for further processing. In Fig. 4, all 22 EEG channels are topographically shown on the human ...
Context 2
... the true level of performed limb activity. During MI execution, C 3 and C z cover electrical activities generated from the left and right hemisphere, respectively while C 4 collects MI signals from the central part of the human brain. For visualization, the locations of candidate channels in dataset 1 have been shown with dark circles in Fig. ...
Context 3
... used in our work consists of a set of 25 EEG channels. Since these channels are Electrooculography (EOG) channels, they are not used for analysis. The data collected from these channels are considered artifacts; therefore, these channels are directly deleted from the dataset, and the remaining 22 channels are used for further processing. In Fig. 4, all 22 EEG channels are topographically shown on the human ...
Context 4
... MI execution, í µí° ¶ 3 and í µí° ¶ í µí± § cover electrical activities generated from the left and right hemisphere, respectively while í µí° ¶ 4 collects MI signals from the central part of the human brain. For visualization, the locations of candidate channels in dataset 1 have been shown with dark circles in Fig.4. ...
Similar publications
The traditional imagery task for brain–computer interfaces (BCIs) consists of motor imagery (MI) in which subjects are instructed to imagine moving certain parts of their body. This kind of imagery task is difficult for subjects. In this study, we used a less studied yet more easily performed type of mental imagery—visual imagery (VI)—in which subj...
Citations
... The effective EEG channels (DCRCC) are identified in [30] using a dynamic channel relevance (DCR) score. The lowest redundancy maximum relevance paradigm is introduced for choosing the appropriate channels. ...
... Then, the EEG signal is split into four sub-bands using a zero-phase Butterworth bandpass filter. The extracted four sub-bands are Mu (8-13 Hz), low beta (13-22 Hz), high beta (22)(23)(24)(25)(26)(27)(28)(29)(30)(31)(32)(33)(34)(35), and full-band . Figure 4 and 5 shows the sub-bands of channel C3 for subject 'al' and 'a' of the BCI Competition III (IVA) dataset and BCI Competition IV (I) dataset. ...
... In comparison with previous techniques, the proposed method performs better in terms of accuracy across all the subjects. It is compared with the recently developed channel selection algorithm, such as dynamic channel relevance [30], channel selection using a correlation coefficient [25], CSP-R-MF [29], LRFCSP [31], CCS-RCSP [23], and FCCR [32] to understand its superiority. The dynamic channel relevance (DCR) score was utilized to find the effective EEG channels in [30] with a mean classification accuracy of 87.58%. ...
Electroencephalography (EEG) is effectively employed to describe cognitive patterns corresponding to different tasks of motor functions for brain–computer interface (BCI) implementation. Explicit information processing is necessary to reduce the computational complexity of practical BCI systems. This paper presents an entropy-based approach to select effective EEG channels for motor imagery (MI) classification in brain–computer interface (BCI) systems. The method identifies channels with higher entropy scores, which is an indication of greater information content. It discards redundant or noisy channels leading to reduced computational complexity and improved classification accuracy. High entropy means a more disordered pattern, whereas low entropy means a less disordered pattern with less information. The entropy of each channel for individual trials is calculated. The weight of each channel is represented by the mean entropy of the channel over all the trials. A set of channels with higher mean entropy are selected as effective channels for MI classification. A limited number of sub-band signals are created by decomposing the selected channels. To extract the spatial features, the common spatial pattern (CSP) is applied to each sub-band space of EEG signals. The CSP-based features are used to classify the right-hand and right-foot MI tasks using a support vector machine (SVM). The effectiveness of the proposed approach is validated using two publicly available EEG datasets, known as BCI competition III–IV(A) and BCI competition IV–I. The experimental results demonstrate that the proposed approach surpasses cutting-edge techniques.
... For the SVM, we used the Linear Kernel, Polynomial Kernel (degrees 2, 3 and 4), and RBF Kernel with gamma 0.5. Recently variety of SVM and tree-based predictive have been developed to deal with spatiotemporal signals [36][37][38][39][40][41]. The classifiers based on Bayes Decision Theory seek to find a minimum average risk, minimizing the probability of classification error. ...
Music accompanies all phases of our lives, and when we reach old age, music becomes a direct symbol of nostalgia. Autobiographical memories are essential to an individual’s sense of identity, continuity, and meaning. But some pathologies, such as dementia, can interrupt the memory storage process. Music can help recall and evoke memories and can be used in alternative treatments for dementia. This work aims to propose an architecture for a music recommendation system capable of recommending music according to musical genre, with the aim of helping music therapists in therapies addressed to elderly people with dementia in initial states. Here we used data from the public music database Emotify, which is composed of 400 songs labeled by 1595 participants in 7975 sessions. Both channels of the songs were windowed using 10s windows with 5s overlap. The data from these windows were represented by 34 time and frequency features. Then, we assessed and compared the performance of classifiers based on support vector machines, decisions trees and Bayesian network. The most suitable architecture in this experimental study was the Random Forest with 250 trees, with an accuracy of 83.42% ± 1.72%, kappa statistic of 0.78 ± 0.02, AUC-ROC of 0.99 ± 0.00, sensitivity of 0.96 ± 0.02, and specificity of 0.94 ± 0.01. this exploratory study found promising results that indicates the possibility of building recommendation systems to support music therapy based on the automatic classification of songs according to the most appropriate musical genre for the patient.
... Feature engineering is the most popular research topic in machine learning after deep learning [16][17][18][19]. The biggest advantage of feature engineering models is low time complexity [20,21]. ...
The aim of this work is to investigate the use of shadow videos for daily sports activity detection and to contribute to the emerging field of shadow-based classification in machine learning. A novel deep feature engineering model is proposed, and a new shadow video dataset is collected to validate the proposed model. Furthermore, the use of shadow videos ensures privacy protection for individuals. To evaluate the proposed system, a dataset comprising five sports activities (i.e., squat, steady run, standing butterfly, overhead side bend, and knee lift) recorded from 33 participants is used. The proposed model works in the following way: (i) videos are divided into frames and aggregated into non-overlapping blocks of six frames to create images, (ii) deep features are extracted from three fully connected layers of the pre-trained AlexNet, resulting in 4096, 4096, and 1000 features from fc6, fc7, and fc8 layers, respectively. These three feature vectors are then merged to generate a final feature vector with a length of 9192 (= 4096 + 4096 + 1000). (iii) The Chi2 selector is employed to select the most distinctive 1000 features in the feature selection phase, and (iv) the support vector machine (SVM) with leave-one-subject-out (LOSO) validation is used to classify the five sports activities. The proposed deep features coupled with Chi2 based model achieved a classification accuracy of 88.49% using the SVM classifier with LOSO cross-validation (CV) on our collected dataset.
... In recent years, many efficient channel selection algorithms have emerged, such as correlation-based methods, machinelearning-based methods, wrapper-based methods, heuristicsearching-based methods, and so on. Correlation-based method (Park and Chung, 2020;Liu T. et al., 2021;Tiwari and Chaturvedi, 2021) goes through the EEG signals obtained from each channel and uses various information-theoretic concepts to evaluate the correlated channels for each feature and to select these channels. So far, some information theoretic concepts, such as normalized mutual information (NMI) (Yang et al., 2021), entropy (Ghembaza and Djebbari, 2022), correlation coefficient (Jin et al., 2019;Moon et al., 2020), and chi-squared statistics (Baig et al., 2020), have been used to assess the correlation of channels. ...
Background
For non-invasive brain-computer interface systems (BCIs) with multiple electroencephalogram (EEG) channels, the key factor limiting their convenient application in the real world is how to perform reasonable channel selection while ensuring task accuracy, which can be modeled as a multi-objective optimization problem. Therefore, this paper proposed a two-objective problem model for the channel selection problem and introduced a domain knowledge-assisted multi-objective optimization algorithm (DK-MOEA) to solve the aforementioned problem.
Methods
The multi-objective optimization problem model was designed based on the channel connectivity matrix and comprises two objectives: one is the task accuracy and the other one can sensitively indicate the removal status of channels in BCIs. The proposed DK-MOEA adopted a two-space framework, consisting of the population space and the knowledge space. Furthermore, a knowledge-assisted update operator was introduced to enhance the search efficiency of the population space by leveraging the domain knowledge stored in the knowledge space.
Results
The proposed two-objective problem model and DK-MOEA were tested on a fatigue detection task and four state-of-the-art multi-objective evolutionary algorithms were used for comparison. The experimental results indicated that the proposed algorithm achieved the best results among all the comparative algorithms for most cases by the Wilcoxon rank sum test at a significance level of 0.05. DK-MOEA was also compared with a version without the utilization of domain knowledge and the experimental results validated the effectiveness of the knowledge-assisted mutation operator. Moreover, the comparison between DK-MOEA and a traditional classification algorithm using all channels demonstrated that DK-MOEA can strike the balance between task accuracy and the number of selected channels.
Conclusion
The formulated two-objective optimization model enabled the selection of a minimal number of channels without compromising classification accuracy. The utilization of domain knowledge improved the performance of DK-MOEA. By adopting the proposed two-objective problem model and DK-MOEA, a balance can be achieved between the number of the selected channels and the accuracy of the fatigue detection task. The methods proposed in this paper can reduce the complexity of subsequent data processing and enhance the convenience of practical applications.
... Instead of using channels selected based on a priori knowledge, some researchers have proposed channelselection approaches that aimed to find the most effective features from EEG recordings. For example, in [33], the authors proposed a novel approach that used the minimum redundancy maximum relevance paradigm to reduce the number of channels, achieving a 53% reduction rate. This result highlights the importance of considering channel-selection strategies in order to improve the performance of MI-BCI systems. ...
(1) Background: in the field of motor-imagery brain–computer interfaces (MI-BCIs), obtaining discriminative features among multiple MI tasks poses a significant challenge. Typically, features are extracted from single electroencephalography (EEG) channels, neglecting their interconnections, which leads to limited results. To address this limitation, there has been growing interest in leveraging functional brain connectivity (FC) as a feature in MI-BCIs. However, the high inter- and intra-subject variability has so far limited its effectiveness in this domain. (2) Methods: we propose a novel signal processing framework that addresses this challenge. We extracted translation-invariant features (TIFs) obtained from a scattering convolution network (SCN) and brain connectivity features (BCFs). Through a feature fusion approach, we combined features extracted from selected channels and functional connectivity features, capitalizing on the strength of each component. Moreover, we employed a multiclass support vector machine (SVM) model to classify the extracted features. (3) Results: using a public dataset (IIa of the BCI Competition IV), we demonstrated that the feature fusion approach outperformed existing state-of-the-art methods. Notably, we found that the best results were achieved by merging TIFs with BCFs, rather than considering TIFs alone. (4) Conclusions: our proposed framework could be the key for improving the performance of a multiclass MI-BCI system.
... Tiwari and Chaturvedi [21] have developed a BCI system using dynamic channel relevance Liu et al. [23] proposed a recalibration-free cross-device transfer learning framework. ...
This chapter provides essential information about electroencephalography (EEG) signal analysis for the brain-computer interface (BCI). The BCI translates the brain activity signals of the users into commands or messages for carrying out specific tasks. The main objective of the BCI is to aid people with neuromuscular diseases by analyzing and transforming their brain activity patterns into meaningful actions. EEG is a neurophysiological signal that represents the electrical activity of the brain. The EEG data reflects normal and abnormal activities related to the brain and other organs related to the brain. Since EEG is a non-invasive method of extraction of brain-related activity, they are the most often used method. This chapter mainly focuses on the basics of BCI using EEG and provides a brief overview of recent EEG-based BCI approaches.
Keywords: Brain-computer interface (BCI), Electroencephalography (EEG), Machine and Deep Learning schemes.
... M denotes the algorithm's total number of iterations, Cj and Lj stand for predicted and actual class labels, respectively, while N stands for the set of data points in the test dataset. The match function compares two labels and returns one if they are the same and 0 if they aren't [43]. ...
Feature Selection (FS) is choosing a subcategory of features purposed to construct a machine learning model. Among the copious existing FS algorithms, Binary Particle Swarm Optimization Algorithm (BPSO) is prevalent with applications in several domains. However, BPSO suffers from premature convergence that affects exploration, resulting in dilapidation. In this current work, we boost the exploration of BPSO, incorporating the intelligence of crows to hide their food sources from other crows and predators and maintain diversity by implementing a clustering strategy. The clustering technique guarantees that the starting population is evenly distributed over the feature space while including more promising features. Additionally, suppose a crow realizes another crow or a predator is tracking it. In that case, the crow moves randomly to evict the stalker, leading to a better exploration of unexplored regions within the search space. We named the proposed method Hybrid Particle Swarm Optimization and Crow Search Algorithm with clustering initialization strategy (HPSOCSA-CIS). To evaluate the performance of HPSOCSA-CIS, 15 standard UCI datasets are utilized, and the outcomes are compared with recently proposed hybrid and standard optimization algorithms. From observation, HPSOCSA-CIS outperforms the comparing approaches for feature selection challenges on representative datasets that fall in the three-category based on dimensions. The HPSOCSA-CIS improves performance in terms of mean classification accuracy by 8.87%, 17.5%, and 21.90% on Low, medium, and high dimensional datasets, respectively.
... Brain-Computer Interface (BCIcommunicationicating scheme that facilitates the human brain to interact with different intelligent devices such as wheelchair systems (Tiwari et al., 2021;Xu et al., 2020), neuroprosthetics (Ramos-Murguialday et al., 2012) and robots (Waytowich et al., 2010). The BCI aims to translate intrinsic cognitive patterns into computer-based control commands using multiple signal-processing methods. ...
... In the future, our proposed method can be used to improve the performance of various interdisciplinary applications such as design pattern detection [40], channel selection [41], and cognitive imaging [42]. In addition, Rough Set Theory (RST) [43] can be used to investigate positive boundary regions, which may help select more relevant features from high-dimensional datasets. ...
The Butterfly Optimization Algorithm (BOA) is a recently proposed nature-inspired metaheuristic algorithm mimicking the food-foraging behavior of butterflies. Its abilities include simplicity, good convergence rate towards local optima, and avoiding the local optima stagnation problem to some extent. In earlier studies, the performance of Binary BOA (BBOA) is superior to various state-of-the-art methods in different optimization issues, such as search space reduction and solving classical engineering problems. Here, BBOA expands the original search space with all possibilities (Exploration) and seeks to determine the best one from all the produced solutions (Exploitation). Generally, the global performance of BBOA depends on the tradeoff between the Exploration and Exploitation phase and produces quality solutions when a suitable tradeoff is maintained. This study introduces an improved and computationally effective variant of conventional BBOA by improving the local search ability of the Butterfly Optimization Algorithm. Initially, twelve binary variants were produced using three different transfer functions (S, U, V-shaped), and solution quality is evaluated in terms of respective fitness function scores. Next, we explored the local search ability of BOA by another recently developed optimization technique, namely, Adaptive β –Hill Climbing, to compute quality solutions. This optimization process employed two stochastic operators:
N -operator
(Neighborhood operator) and β -operator (Mutation operator) to generate improved offspring compared to parent solutions. This phase is iteratively implemented until the desired level of binary pattern with suitable classification accuracy is obtained. We validated the proposed approach on twenty datasets with eleven state-of-the-art feature selection algorithms. The overall results suggest that the proposed improvements increase the classification accuracy with fewer features on most datasets. In addition, the proposed approach’s time complexity was significantly reduced on eighteenth out of twenty datasets. Moreover, the proposed method balances space exploration and solution exploitation in feature selection problems.
... In this extended intra-subjects comparison, the approaches put forth by Xu B et al [32], Tiwari A et al [38] and Zhang R et al [33] performed effectively achieving respectively mean accuracy of 85.6%, 85.4% and 84% of mean accuracy, although the ARK-ANN achieved the best outcomes. ...
... The ARK-ANN method produced the best results in this cross-subjects comparison with an accuracy of 87.4%, followed by the CNN [32] and DCR [38] methods that are getting almost the same accuracy. Despite achieving a lower level of precision, the DL [33] method is still performing well. ...
... Mean accuracy SNN [31] 81.3% Wavelet T-F Image CNN [32] 85.6% Hybrid DL [33] 84.0% HSS-ELM [34] 79.5% DCR [38] 85.4% ARK-ANN 87.4% ...
More recently, a number of studies show the interest of the use of the Riemannian geometry in EEG classification. The idea is to exploit the EEG covariance matrices, instead of the raw EEG data, and use the Riemannian geometry to directly classify these matrices. This paper presents a novel Artificial Neural Network approach based on an Adaptive Riemannian Kernel, named ARK-ANN, to classify Electroencephalographic (EEG) motor imaging signals in the context of Brain Computer Interface (BCI). A multilayer perceptron is used to classify the covariance matrices of Motor Imagery (MI) signals employing an adaptive optimization of the testing set. The contribution of a geodesic filter is also assessed for the ANN and the original method which use an SVM classifier. The results demonstrate that the ARK-ANN performs better than the other methods and the geodesic filter gives slightly better results in the ARK-SVM, considered here as the reference method, in the case of inter-subject classification (accuracy of 87.4% and 86 % for ARK-ANN and ARK-SVM, respectively). Regarding the cross-subject classification, the proposed method gives an accuracy of 77.3% and increases the precision by 8.2% in comparison to the SVM based method.
