A novel hybrid decoding neural network for EEG signal representation

Existing research methods for recognizing EEG (Electroencephalogram) signals in motor imagery (MI) often overlook the dynamic changes of brain networks over time, resulting in insufficient classification accuracy for MI tasks. This article addresses the recognition problem of dynamic changes in brain networks during MI tasks and applies an EEG signal classification method based on multi-layer time-varying functional brain networks. This article uses the BCI (Brain-Computer Interface) Competition IV 2a dataset to preprocess the raw EEG signals through bandpass filtering and CSP (Common Spatial Pattern) algorithm. The EEG signals of the MI task are divided into 7 1-second time windows with a step size of 0.5 seconds. Within each time window, Pearson correlation coefficients between EEG channels can be calculated to generate corresponding brain networks, and multiple time-varying functional brain networks can be constructed by stacking the brain networks from multiple time windows. The network topology features, node degree, clustering coefficient, network efficiency, and multi-layer network features of each window can be extracted, including Multiplex Clustering Coefficient (MCC), Multiplex Participation Coefficient (MPC), and inter layer correlation coefficient. By dividing the dataset through 10 fold cross validation, the random forest algorithm can be used to classify and recognize four types of motion imagination tasks. The experimental results show that the average recognition rate of the article’s method in four types of MI tasks reached 89.19%. This method can improve the classification accuracy of MI tasks and enhance a comprehensive understanding of the dynamic changes in brain networks during the process of MI.

Dealing with subjects who are unable to attain a proper level of performance, that is, those with brain–computer interface (BCI) illiteracy or BCI inefficients, is still a major issue in human electroencephalography (EEG) BCI systems. The most suitable approach to address this issue is to analyze the EEG signals of individual subjects independently recorded before the main BCI tasks, to evaluate their performance on these tasks. This study mainly focused on non-linear analyses and deep learning techniques to investigate the significant relationship between the intrinsic characteristics of a prior idle resting state and the subsequent BCI performance. To achieve this main objective, a public EEG motor/movement imagery dataset that constituted two individual EEG signals recorded from an idle resting state and a motor imagery BCI task was used in this study. For the EEG processing in the prior resting state, spectral analysis but also non-linear analyses, such as sample entropy, permutation entropy, and recurrent quantification analyses (RQA), were performed to obtain individual groups of EEG features to represent intrinsic EEG characteristics in the subject. For the EEG signals in the BCI tasks, four individual decoding methods, as a filter-bank common spatial pattern-based classifier and three types of convolution neural network-based classifiers, quantified the subsequent BCI performance in the subject. Statistical linear regression and ANOVA with post hoc analyses verified the significant relationship between non-linear EEG features in the prior resting state and three types of BCI performance as low-, intermediate-, and high-performance groups that were statistically discriminated by the subsequent BCI performance. As a result, we found that the frontal theta rhythm ranging from 4 to 8 Hz during the eyes open condition was highly associated with the subsequent BCI performance. The RQA findings that higher determinism and lower mean recurrent time were mainly observed in higher-performance groups indicate that more regular and stable properties in the EEG signals over the frontal regions during the prior resting state would provide a critical clue to assess an individual BCI ability in the following motor imagery task.

A novel hybrid BCI speller based on RSVP and SSVEP paradigm

Multi-channel Electroencephalograph (EEG) signal is an important source of neural information for motor imagery (MI) limb movement intent decoding. The decoded MI movement intent often serve as potential control input for brain-computer interface (BCI) based rehabilitation robots. However, the presence of multiple dynamic artifacts in EEG signal leads to serious processing challenge that affects the BCI system in practical settings. Hence, this study propose a hybrid approach based on Low-rank spatiotemporal filtering technique for concurrent elimination of multiple EEG artifacts. Afterwards, a convolutional neural network based deep learning model (ConvNet-DL) that extracts neural information from the cleaned EEG signal for MI tasks decoding was built. The proposed method was studied in comparison with existing artifact removal methods using EEG signals of transhumeral amputees who performed five different MI tasks. Remarkably, the proposed method led to significant improvements in MI task decoding accuracy for the ConvNet-DL model in the range of 8.00~13.98%, while up to 14.38% increment was recorded in terms of the MCC: Mathew correlation coefficients at p<0.05. Also, a signal to error ratio of more than 11 dB was recorded by the proposed method.Clinical Relevance- This study showed that a combination of the proposed hybrid EEG artifact removal method and ConvNet-DL can significantly improve the decoding accuracy of MI upper limb movement tasks. Our findings may provide potential control input for BCI rehabilitation robotic systems.

Motor imagery EEG recognition based on conditional optimization empirical mode decomposition and multi-scale convolutional neural network

A novel Morse code-inspired method for multiclass motor imagery brain–computer interface (BCI) design

The brain-computer interface (BCI) work is to provide humans an alternative channel that allows direct transmission of messages from the brain by analyzing the brain’s mental activities [1–7]. The brain activity is recorded by means of multi-electrode electroencephalographic (EEG) signals that are either invasive or noninvasive. Noninvasive recording is convenient and popular in BCI applications so it is commonly used. According to the definition suggested at the first international meeting for BCI technology, the term BCI is reserved for a system that must not depend on the brain’s normal output pathways of peripheral nerves and muscles [2]. It has become popular for BCI systems on motor imagery (MI) EEG signals in the last decade [8]. It reveals that there are special characteristics of event-related desynchronization (ERD) and synchronization (ERS) in mu and beta rhythms over the sensorimotor cortex during MI tasks by discriminating EEG signals between left and right MIs [9, 10]. ERD/ERS is the task-related or event-related change in the amplitude of the oscillatory behavior of specific cortical areas within various frequency bands. An amplitude (or power) increase is defined as event-related synchronization while an amplitude (or power) decrease is defined as event-related desynchronization. As other event-related potentials, ERD/ERS patterns are associated with sensory processing and motor behavior [2]. The principal objective of this study is to propose a BCI system, which combines neuro-fuzzy prediction and multiresolution fractal feature vectors (MFFVs) with support vector machine, for MI classification.

A motor imagery (MI) brain-computer interface (BCI) plays an important role in the neurological rehabilitation training for stroke patients. Electroencephalogram (EEG)-based MI BCI has high temporal resolution, which is convenient for real-time BCI control. Therefore, we focus on EEG-based MI BCI in this paper. The identification of MI EEG signals is always quite challenging. Due to high inter-session/subject variability, each subject should spend long and tedious calibration time in collecting amounts of labeled samples for a subject-specific model. To cope with this problem, we present a supervised selective cross-subject transfer learning (sSCSTL) approach which simultaneously makes use of the labeled samples from target and source subjects based on Riemannian tangent space. Since the covariance matrices representing the multi-channel EEG signals belong to the smooth Riemannian manifold, we perform the Riemannian alignment to make the covariance matrices from different subjects close to each other. Then, all aligned covariance matrices are converted into the Riemannian tangent space features to train a classifier in the Euclidean space. To investigate the role of unlabeled samples, we further propose semi-supervised and unsupervised versions which utilize the total samples and unlabeled samples from target subject, respectively. Sequential forward floating search (SFFS) method is executed for source selection. All our proposed algorithms transfer the labeled samples from most suitable source subjects into the feature space of target subject. Experimental results on two publicly available MI datasets demonstrated that our algorithms outperformed several state-of-the-art algorithms using small number of the labeled samples from target subject, especially for good target subjects.

Motor Imagery (MI)-based Brain Computer Interface (BCI) system is a potential technology for active neurorehabilitation of stroke patients by complementing the conventional passive rehabilitation methods. Research to date mainly focused on classifying left vs. right hand/foot MI of stroke patients. Though a very few studies have reported decoding imagined hand movement directions using electroencephalogram (EEG)-based BCI, the experiments were conducted on healthy subjects. Our work analyzes MI-based brain cortical activity from EEG signals and decodes the imagined hand movement directions in stroke patients. The decoded direction (left vs. right) of hand movement imagination is used to provide control commands to a motorized arm support on which patient's affected (paralyzed) arm is placed. This enables the patient to move his/her stroke-affected hand towards the intended (imagined) direction that aids neuroplasticity in the brain. The synchronization measure called Phase Locking Value (PLV), extracted from EEG, is the neuronal signature used to decode the directional movement of the MI task. Event-related desynchronization/synchronization (ERD/ERS) analysis on Mu and Beta frequency bands of EEG is done to select the time bin corresponding to the MI task. The dissimilarities between the two directions of MI tasks are identified by selecting the most significant channel pairs that provided maximum difference in PLV features. The training protocol has an initial calibration session followed by a feedback session with 50 trials of MI task in each session. The feedback session extracts PLV features corresponding to most significant channel pairs which are identified in the calibration session and is used to predict the direction of MI task in left/right direction. An average MI direction classification accuracy of 74.44% is obtained in performing the training protocol and 68.63% from the prediction protocol during feedback session on 16 stroke patients.

Speller brain-computer interface (BCI) systems can help neuromuscular disorders patients write their thoughts by using the electroencephalogram (EEG) signals by just focusing on the speller tasks. For practical speller-based BCI systems, the P300 event-related brain potential is measured by using the EEG signal. In this paper, we design a robust machine-learning algorithm for P300 target detection. The novel spatial-temporal linear feature learning (STLFL) algorithm is proposed to extract high-level P300 features. The STLFL method is a modified linear discriminant analysis technique focusing on the spatial-temporal aspects of information extraction. A new P300 detection structure is then proposed based on the combination of the novel STLFL feature extraction and discriminative restricted Boltzmann machine (DRBM) for the classification approach (STLFL + DRBM). The effectiveness of the proposed technique is evaluated using two state-of-the-art P300 BCI datasets. Across the two available databases, we show that in terms of average target recognition accuracy and standard deviation values, the proposed STLFL + DRBM method outperforms traditional methods by 33.5, 78.5, 93.5, and 98.5% for 1, 5, 10, and 15 repetitions, respectively, in BCI competition III datasets II and by 71.3, 100, 100, and 100% for 1, 5, 10, and 15 repetitions, respectively, in BCI competition II datasets II and by 67.5 ± 4, 84.2 ± 2.5, 93.5 ± 1, 96.3 ± 1, and 98.4 ± 0.5% for rapid serial visual presentation (RSVP) based dataset in repetitions 1–5. The method has some advantages over the existing variants including its efficiency, robustness with a small number of training samples, and a high ability to create discriminative features between classes.

Research on brain–computer interfaces (BCIs) advances the way scientists understand how the human brain functions. The BCI system, which is based on the use of electroencephalography (EEG) signals to detect motor imagery (MI) tasks, enables opportunities for various applications in stroke rehabilitation, neuroprosthetic devices, and communication tools. BCIs can also be used in emotion recognition (ER) research to depict the sophistication of human emotions by improving mental health monitoring, human–computer interactions, and neuromarketing. To address the low accuracy of MI-BCI, which is a key issue faced by researchers, this study employs a new approach that has been proven to have the potential to enhance motor imagery classification accuracy. The basic idea behind the approach is to apply feature extraction methods from the field of emotion recognition to the field of motor imagery. Six feature sets and four classifiers were explored using four MI classes (left and right hands, both feet, and tongue) from the BCI Competition IV 2a dataset. Statistical, wavelet analysis, Hjorth parameters, higher-order spectra, fractal dimensions (Katz, Higuchi, and Petrosian), and a five-dimensional combination of all five feature sets were implemented. GSVM, CART, LinearSVM, and SVM with polynomial kernel classifiers were considered. Our findings show that 3D fractal dimensions predominantly outperform all other feature sets, specifically during LinearSVM classification, accomplishing nearly 79.1% mean accuracy, superior to the state-of-the-art results obtained from the referenced MI paper, where CSP reached 73.7% and Riemannian methods reached 75.5%. It even performs as well as the latest TWSB method, which also reached approximately 79.1%. These outcomes emphasize that the new hybrid approach in the motor imagery/emotion recognition field improves classification accuracy when applied to motor imagery EEG signals, thus enhancing MI-BCI performance.

Motor imagery is the mental simulation of a motor act that includes preparation for movement, passive observations of action and mental operations of motor representations implicitly or explicitly. Motor imagery as preparation for immediate movement likely involves the motor executive brain regions. Implicit mental operations of motor representations are considered to underlie cognitive functions. Another problem concerning neuro-imaging studies on motor imagery is that the performance of imagination is very difficult to control. The ability of an individual to control its EEG may enable him to communicate without being able to control their voluntary muscles. Communication based on EEG signals does not require neuromuscular control and the individuals who have neuromuscular disorders and who may have no more control over any of their conventional communication abilities may still be able to communicate through a direct brain-computer interface. A brain-computer interface replaces the use of nerves and muscles and the movements they produce with electrophysiological signals and is coupled with the hardware and software that translate those signals into physical actions. One of the most important components of a brain-computer interface is the EEG feature extraction procedure. This paper presents an approach that uses self-organizing fuzzy neural network based time series prediction that performs EEG feature extraction in the time domain only. EEG is recorded from two electrodes placed on the scalp over the motor cortex. EEG signals from each electrode are predicted by a single fuzzy neural network. Features derived from the mean squared error of the predictions and from the mean squared of the predicted signals are extracted from EEG data by means of a sliding window. The architecture of the two auto-organizing fuzzy neural networks is a network with multi inputs and single output.

Motor imagery (MI) is an important brain-computer interface (BCI) paradigm. The traditional MI paradigm (imagining different limbs) limits the intuitive control of the outer devices, while fine MI paradigm (imagining different joint movements from the same limb) can control the mechanical arm without cognitive disconnection. However, the decoding performance of fine MI limits its application. Electroencephalogram (EEG) and functional near-infrared spectroscopy (fNIRS) are widely used in BCI systems because of their portability and easy operation. In this study, a fine MI paradigm including four classes (hand, wrist, shoulder and rest) was designed, and the data of EEG-fNIRS bimodal brain activity was collected from 12 subjects. Event-related desynchronization (ERD) from EEG signals shows a contralateral dominant phenomenon, and there is difference between the ERD of the four classes. For fNIRS signal in the time dimension, the time periods with significant difference can be observed in the activation patterns of four MI tasks. Spatially, the signal peak based brain topographic map also shows difference of these four MI tasks. The EEG signal and fNIRS signal of these four classes are distinguishable. In this study, a bimodal fusion network is proposed to improve the fine MI tasks decoding performance. The features of these two modalities are extracted separately by two feature extractors based on convolutional neural networks (CNN). The recognition performance was significantly improved by the bimodal method proposed in this study, compared with the performance of the single-modal network. The proposed method outperformed all comparison methods, and achieved a four-class accuracy of 58.96%. This paper demonstrates the feasibility of EEG and fNIRS bimodal BCI systems for fine MI, and shows the effectiveness of the proposed bimodal fusion method. This research is supposed to support fine MI-based BCI systems with theories and techniques.

Electroencephalography (EEG) signals are often employed in modern BCI (brain-computer interface) systems. For patients with irreversible nerve injury, motor imagery BCIs enable them to control devices such as artificial limbs simply with brain activities (i.e. imagination) without physical movement. In practice, EEG signals are measured with multiple electrodes and often contaminated with various kinds of artifacts and noise. In this study, a novel EEG motor imagery classification approach is developed based on discrete wavelet transform (DWT) and TSK (Tagaki-Sugeno-Kang) fuzzy inference neural networks. By using energy ratios in the frequency bands that are related with motor imagery activities, fewer features are needed for the classifier, which makes the real-time implementation of BCI systems more feasible. DWT not only provides multi-resolution analysis for non-stationary EEG signals, but also aids for noise and artifact reduction. Furthermore, the TSK fuzzy neural network is able to handle imprecise information with adaptive learning. The proposed algorithm is tested on EEG signals of PhysioNet database and compared with other methods. Satisfactory computer simulation results are obtained.

A brain-computer interface (BCI) based on motor imagery (MI) from the same limb can provide an intuitive control pathway but has received limited attention. It is still a challenge to classify multiple MI tasks from the same limb. The goal of this study is to propose a novel decoding method to classify the MI tasks of four joints of the same upper limb and the resting state. EEG signals were collected from 20 participants. A time-distributed attention network (TD-Atten) was proposed to adaptively assign different weights to different classes and frequency bands of the input multiband Common Spatial Pattern (CSP) features. The long short-term memory (LSTM) and dense layers were then used to learn sequential information from the reweight features and perform the classification. Our proposed method outperformed other baseline and deep learning-based methods and obtained the accuracies of 46.8% in the 5-class scenario and 53.4% in the 4-class scenario. The visualization results of attention weights indicated that the proposed framework can adaptively pay attention to alpha-band related features in MI tasks, which was consistent with the analysis of brain activation patterns. These results demonstrated the feasibility and interpretability of the attention mechanism in MI decoding and the potential of this fine MI paradigm to be applied for the control of a robotic arm or a neural prosthesis.