Labeled EEG Data Research Articles

Concept-based AI interpretability in physiological time-series data: Example of abnormality detection in electroencephalography

Background and ObjectiveDespite recent performance advancements, deep learning models are not yet adopted in clinical practice on a wide scale. The intrinsic intransparency of such systems is commonly cited as one major reason for this reluctance. This has motivated methods that aim to provide explanations of model functioning. Known limitations of feature-based explanations have led to an increased interest in concept-based interpretability. Testing with Concept Activation Vectors (TCAV) employs human-understandable, abstract concepts to explain model behavior. The method has previously been applied to the medical domain in the context of electronic health records, retinal fundus images and magnetic resonance imaging. MethodsWe explore the usage of TCAV for building interpretable models on physiological time series, using an example of abnormality detection in electroencephalography (EEG). For this purpose, we adopt the XceptionTime model, which is suitable for multi-channel physiological data of variable sizes. The model provides state-of-the-art performance on raw EEG data and is publically available. We propose and test several ideas regarding concept definition through metadata mining, using additional labeled EEG data and extracting interpretable signal characteristics in the form of frequencies. By including our own hospital data with analog labeling, we further evaluate the robustness of our approach. ResultsThe tested concepts show a TCAV score distribution that is in line with the clinical expectations, i.e. concepts known to have strong links with EEG pathologies (such as epileptiform discharges) received higher scores than the neutral concepts (e.g. sex). The scores were consistent across the applied concept generation strategies. ConclusionsTCAV has the potential to improve interpretability of deep learning applied to multi-channel signals as well as to detect possible biases in the data. Still, further work on developing the strategies for concept definition and validation on clinical physiological time series is needed to better understand how to extract clinically relevant information from the concept sensitivity scores.

Advancing passive BCIs: a feasibility study of two temporal derivative features and effect size-based feature selection in continuous online EEG-based machine error detection.

The emerging integration of Brain-Computer Interfaces (BCIs) in human-robot collaboration holds promise for dynamic adaptive interaction. The use of electroencephalogram (EEG)-measured error-related potentials (ErrPs) for online error detection in assistive devices offers a practical method for improving the reliability of such devices. However, continuous online error detection faces challenges such as developing efficient and lightweight classification techniques for quick predictions, reducing false alarms from artifacts, and dealing with the non-stationarity of EEG signals. Further research is essential to address the complexities of continuous classification in online sessions. With this study, we demonstrated a comprehensive approach for continuous online EEG-based machine error detection, which emerged as the winner of a competition at the 32nd International Joint Conference on Artificial Intelligence. The competition consisted of two stages: an offline stage for model development using pre-recorded, labeled EEG data, and an online stage 3 months after the offline stage, where these models were tested live on continuously streamed EEG data to detect errors in orthosis movements in real time. Our approach incorporates two temporal-derivative features with an effect size-based feature selection technique for model training, together with a lightweight noise filtering method for online sessions without recalibration of the model. The model trained in the offline stage not only resulted in a high average cross-validation accuracy of 89.9% across all participants, but also demonstrated remarkable performance during the online session 3 months after the initial data collection without further calibration, maintaining a low overall false alarm rate of 1.7% and swift response capabilities. Our research makes two significant contributions to the field. Firstly, it demonstrates the feasibility of integrating two temporal derivative features with an effect size-based feature selection strategy, particularly in online EEG-based BCIs. Secondly, our work introduces an innovative approach designed for continuous online error prediction, which includes a straightforward noise rejection technique to reduce false alarms. This study serves as a feasibility investigation into a methodology for seamless error detection that promises to transform practical applications in the domain of neuroadaptive technology and human-robot interaction.

Metastability indexes global changes in the dynamic working point of the brain following brain stimulation.

Several studies have shown that coordination among neural ensembles is a key to understand human cognition. A well charted path is to identify coordination states associated with cognitive functions from spectral changes in the oscillations of EEG or MEG. A growing number of studies suggest that the tendency to switch between coordination states, sculpts the dynamic repertoire of the brain and can be indexed by a measure known as metastability. In this article, we characterize perturbations in the metastability of global brain network dynamics following Transcranial Magnetic Stimulation that could quantify the duration for which information processing is altered. Thus allowing researchers to understand the network effects of brain stimulation, standardize stimulation protocols and design experimental tasks. We demonstrate the effect empirically using publicly available datasets and use a digital twin (a whole brain connectome model) to understand the dynamic principles that generate such observations. We observed a significant reduction in metastability, concurrent with an increase in coherence following single-pulse TMS reflecting the existence of a window where neural coordination is altered. The reduction in complexity was validated by an additional measure based on the Lempel-Ziv complexity of microstate labeled EEG data. Interestingly, higher frequencies in the EEG signal showed faster recovery in metastability than lower frequencies. The digital twin shed light on how the phase resetting introduced by the single-pulse TMS in local cortical networks can propagate globally, giving rise to changes in metastability and coherence.

A deep transfer learning network with two classifiers based on sample selection for motor imagery brain-computer interface

The non-stationary of Motor Imagery (MI) electroencephalogram (EEG) signals makes the traditional machine learning methods ineffective in EEG recognition across time, which limits the practicability of Motor Imagery Brain-computer Interface(MI BCI). Transfer learning makes the classification model of non-stationary MI EEG signals collected and trained at different times universal. In this paper, we proposed a new deep transfer neural network model that can effectively utilize labeled EEG data from previous times, thereby achieving good recognition performance for a small number of labeled EEG signals at the current time. Firstly, a filter bank was used to filter the original motor imagery EEG, and the filtered EEG retained the key features of each domain through the domain specific attention module. Then designed two domain adaptation modules simultaneously. The first domain adaptation module selected similar source domain data, and the second domain adaptation module minimized the difference between the source and target domains. At last, two adversarial classifiers were used to improve the accuracy and robustness of the classification. The results of average classification accuracy, feature visualization, and paired t-test on two public datasets all indicated that the classification performance of the proposed method was significantly higher than other comparison methods. The experimental results on two public datasets showed that the proposed method performed better than the comparison methods when classifying only a small number of labeled MI EEG data, and can achieve higher classification accuracy. This method has practical application value in the rehabilitation of stroke patients.

Unsupervised domain adaptation for cross-patient seizure classification

Objective. Epileptic seizure is a chronic neurological disease affecting millions of patients. Electroencephalogram (EEG) is the gold standard in epileptic seizure classification. However, its low signal-to-noise ratio, strong non-stationarity, and large individual difference nature make it difficult to directly extend the seizure classification model from one patient to another. This paper considers multi-source unsupervised domain adaptation for cross-patient EEG-based seizure classification, i.e. there are multiple source patients with labeled EEG data, which are used to label the EEG trials of a new patient. Approach. We propose an source domain selection (SDS)-global domain adaptation (GDA)-target agent subdomain adaptation (TASA) approach, which includes SDS to filter out dissimilar source domains, GDA to align the overall distributions of the selected source domains and the target domain, and TASA to identify the most similar source domain to the target domain so that its labels can be utilized. Main results. Experiments on two public seizure datasets demonstrated that SDS-GDA-TASA outperformed 13 existing approaches in unsupervised cross-patient seizure classification. Significance. Our approach could save clinicians plenty of time in labeling EEG data for epilepsy patients, greatly increasing the efficiency of seizure diagnostics.

DOMAIN-ADAPTIVE TSK FUZZY SYSTEM BASED ON MULTISOURCE DATA FUSION FOR EPILEPTIC EEG SIGNAL CLASSIFICATION

In recent years, machine learning methods based on epileptic signals have shown good results with brain-computer interfaces (BCIs). With the continuous expansion of their applications, the demand for labeled epileptic signals is increasing. For a large number of data-driven models, such signals are not suitable, as they extend the calibration cycle. Therefore, a new domain-adaptive TSK fuzzy system model based on multisource data fusion (DA-TSK) is proposed. The purpose of DA-TSK is to maintain high classification performance when the amount of labeled data is insufficient. The DA-TSK model not only has a strong learning ability to learn characteristic information from EEG data but is also interpretable, which aids in the understanding of the analytic process of the model for medical purposes. In particular, this model can make full use of a small amount of labeled EEG data in the source domain and target domain through domain adaptation. Therefore, the DA-TSK model can reduce data dependence to a certain extent and improve the generalization performance of the target classifier. Experiments are performed to evaluate the effectiveness of the DA-TSK model on public EEG datasets based on epileptic signals. The DA-TSK model can obtain satisfactory accuracy when the labeled data are insufficient in the target domain.

Deep stacked least square support matrix machine with adaptive multi-layer transfer for EEG classification

The recent extraordinary success of deep neural networks for electroencephalogram (EEG) decoding can be mainly attributed to the availability of large-scale labeled datasets. However, it requires prolonged EEG calibration, which is especially inconvenient for people with disabilities. Training deep neural networks with insufficient labeled EEG data will make the model suffer from limited generalization capability. In view of this, a new deep stacked network (DSN), called deep stacked transfer least square support matrix machine (DST-LSSMM), is proposed to alleviate the above issue. Following the philosophy of stacked generalization, the proposed DST-LSSMM utilizes least square support matrix machine (LSSMM) as its basic building unit and the random projection of previous layers as its stacking element. The adopted matrix classifier LSSMM directly models the matrix-form data, which is able to exploit the structural information between the columns or rows in EEG feature matrices to further improve the generalization capability of the model. Besides, we design an adaptive multi-layer model knowledge transfer learning scheme to guarantee consistency across the interrelated layers. The proposed adaptive multi-layer transfer scheme allows the use of model knowledge of previous related layers to facilitate the model construction at higher layers, consequently enhancing the generalization capability of the deep stacked model. DST-LSSMM is carried out in an efficient feed-forward way without parameter pre-training and fine-tuning, which can simplify the model optimization process. We extensively evaluate our method on three EEG datasets. Experimental results demonstrate that our method achieves promising performances on EEG classification.

Source-Free Domain Adaptation (SFDA) for Privacy-Preserving Seizure Subtype Classification.

Electroencephalogram (EEG) based seizure subtype classification is very important in clinical diagnostics. Source-free domain adaptation (SFDA) uses a pre-trained source model, instead of the source data, for privacy-preserving transfer learning. SFDA is useful in seizure subtype classification, which can protect the privacy of the source patients, while reducing the amount of labeled calibration data for a new patient. This paper introduces semi-supervised transfer boosting (SS-TrBoosting), a boosting-based SFDA approach for seizure subtype classification. We further extend it to unsupervised transfer boosting (U-TrBoosting) for unsupervised SFDA, i.e., the new patient does not need any labeled EEG data. Experiments on three public seizure datasets demonstrated that SS-TrBoosting and U-TrBoosting outperformed multiple classical and state-of-the-art machine learning approaches in cross-dataset/cross-patient seizure subtype classification.

Language Model-Guided Classifier Adaptation for Brain-Computer Interfaces for Communication.

Brain-computer interfaces (BCIs), such as the P300 speller, can provide a means of communication for individuals with severe neuromuscular limitations. BCIs interpret electroencephalography (EEG) signals in order to translate embedded information about a user's intent into executable commands to control external devices. However, EEG signals are inherently noisy and nonstationary, posing a challenge to extended BCI use. Conventionally, a BCI classifier is trained via supervised learning in an offline calibration session; once trained, the classifier is deployed for online use and is not updated. As the statistics of a user's EEG data change over time, the performance of a static classifier may decline with extended use. It is therefore desirable to automatically adapt the classifier to current data statistics without requiring offline recalibration. In an existing semi-supervised learning approach, the classifier is trained on labeled EEG data and is then updated using incoming unlabeled EEG data and classifier-predicted labels. To reduce the risk of learning from incorrect predictions, a threshold is imposed to exclude unlabeled data with low-confidence label predictions from the expanded training set when retraining the adaptive classifier. In this work, we propose the use of a language model for spelling error correction and disambiguation to provide information about label correctness during semi-supervised learning. Results from simulations with multi-session P300 speller user EEG data demonstrate that our language-guided semi-supervised approach significantly improves spelling accuracy relative to conventional BCI calibration and threshold-based semi-supervised learning.

Reducing the Calibration Effort of EEG Emotion Recognition using Domain Adaptation with Soft Labels.

Electroencephalogram (EEG)-based emotion recognition has made great progress in recent years. The current pipelines collect EEG training data in a long-time calibration session for each new subject, which is time consuming and user unfriendly. To reduce the time required for the calibration session, there have been many studies using domain adaptation (DA) approaches to transfer knowledge from existing subjects (source domain) to the new subject (target domain) for reducing the dependence on the calibration session. Existing DA methods usually require substantial unlabeled EEG data of the new subject. However, the real scenario is that there are a small number of labeled samples in the calibration session of the target. Motivated by this, we introduce a novel domain adaptation architecture based on adversarial training to learn domain-invariant feature representations across subjects. To improve the performance when there are few labeled EEG data in the calibration session, we add a soft label loss to the architecture, which can ensure that the inter-class relationships learned from the source domain are transferred to target domain. We evaluate the method on the SEED dataset, and the experimental results show that our method uses only 15 examples per trial in the calibration session to achieve an average accuracy of 87.28%, indicating the effectiveness of our framework.

A Novel Transfer Support Matrix Machine for Motor Imagery-Based Brain Computer Interface.

In recent years, emerging matrix learning methods have shown promising performance in motor imagery (MI)-based brain-computer interfaces (BCIs). Nonetheless, the electroencephalography (EEG) pattern variations among different subjects necessitates collecting a large amount of labeled individual data for model training, which prolongs the calibration session. From the perspective of transfer learning, the model knowledge inherent in reference subjects incorporating few target EEG data have the potential to solve the above issue. Thus, a novel knowledge-leverage-based support matrix machine (KL-SMM) was developed to improve the classification performance when only a few labeled EEG data in the target domain (target subject) were available. The proposed KL-SMM possesses the powerful capability of a matrix learning machine, which allows it to directly learn the structural information from matrix-form EEG data. In addition, the KL-SMM can not only fully leverage few labeled EEG data from the target domain during the learning procedure but can also leverage the existing model knowledge from the source domain (source subject). Therefore, the KL-SMM can enhance the generalization performance of the target classifier while guaranteeing privacy protection to a certain extent. Finally, the objective function of the KL-SMM can be easily optimized using the alternating direction method of multipliers method. Extensive experiments were conducted to evaluate the effectiveness of the KL-SMM on publicly available MI-based EEG datasets. Experimental results demonstrated that the KL-SMM outperformed the comparable methods when the EEG data were insufficient.

An artificial intelligence-based EEG algorithm for detection of epileptiform EEG discharges: Validation against the diagnostic gold standard

ObjectiveTo validate an artificial intelligence-based computer algorithm for detection of epileptiform EEG discharges (EDs) and subsequent identification of patients with epilepsy. MethodsWe developed an algorithm for automatic detection of EDs, based on a novel deep learning method that requires a low amount of labeled EEG data for training. Detected EDs are automatically grouped into clusters, consisting of the same type of EDs, for rapid visual inspection. We validated the algorithm on an independent dataset of 100 patients with sharp transients in their EEG recordings (54 with epilepsy and 46 with non-epileptic paroxysmal events). The diagnostic gold standard was derived from the video-EEG recordings of the patients’ habitual events. ResultsThe algorithm had a sensitivity of 89% for identifying EEGs with EDs recorded from patients with epilepsy, a specificity of 70%, and an overall accuracy of 80%. ConclusionsAutomated detection of EDs using an artificial intelligence-based computer algorithm had a high sensitivity. Human (expert) supervision is still necessary for confirming the clusters of detected EDs and for describing clinical correlations. Further studies on different patient populations will be needed to confirm our results. SignificanceThe automated algorithm we describe here is a useful tool, assisting neurophysiologist in rapid assessment of EEG recordings.

Double-Criteria Active Learning for Multiclass Brain-Computer Interfaces.

Recent technological advances have enabled researchers to collect large amounts of electroencephalography (EEG) signals in labeled and unlabeled datasets. It is expensive and time consuming to collect labeled EEG data for use in brain-computer interface (BCI) systems, however. In this paper, a novel active learning method is proposed to minimize the amount of labeled, subject-specific EEG data required for effective classifier training, by combining measures of uncertainty and representativeness within an extreme learning machine (ELM). Following this approach, an ELM classifier was first used to select a relatively large batch of unlabeled examples, whose uncertainty was measured through the best-versus-second-best (BvSB) strategy. The diversity of each sample was then measured between the limited labeled training data and previously selected unlabeled samples, and similarity is measured among the previously selected samples. Finally, a tradeoff parameter is introduced to control the balance between informative and representative samples, and these samples are then used to construct a powerful ELM classifier. Extensive experiments were conducted using benchmark and multiclass motor imagery EEG datasets to evaluate the efficacy of the proposed method. Experimental results show that the performance of the new algorithm exceeds or matches those of several state-of-the-art active learning algorithms. It is thereby shown that the proposed method improves classifier performance and reduces the need for training samples in BCI applications.

Cross-Subject EEG Signal Recognition Using Deep Domain Adaptation Network

Collecting sufficient labeled electroencephalography (EEG) data to build an individual classifier for each subject is extremely time-consuming and labor-intensive, especially for the disabled patients. A feasible way is to use labeled EEG data from other subjects (source domains) to train a model for classifying EEG data from the new subjects (target domains). However, the model trained using other subjects EEG data may degrade the classification performance of the target subject, when there exists the substantial inter-subject variability of EEG data. In this paper, to account for the domain shift between different subjects, we propose a novel deep domain adaptation network (DDAN) for cross-subject EEG signal recognition. Specifically, a special end-to-end convolutional neural network (CNN) is firstly adopted to automatically extract deep features from the raw EEG data. Then, maximum mean discrepancy (MMD) is used to minimize the distribution discrepancy of deep features between source and target subjects. Finally, a center-based discriminative feature learning (CDFL) method is used to force the deep features closer to their corresponding class centers and make the inter-class centers more separable, so that it is possible to further improve the recognition performance of target domain EEG data. Experiments on public EEG datasets prove the effectiveness of the proposed method. This study may promote the practical use of EEG signal processing technology and expand its application range.

A Comparison of Regularization Methods in Forward and Backward Models for Auditory Attention Decoding.

The decoding of selective auditory attention from noninvasive electroencephalogram (EEG) data is of interest in brain computer interface and auditory perception research. The current state-of-the-art approaches for decoding the attentional selection of listeners are based on linear mappings between features of sound streams and EEG responses (forward model), or vice versa (backward model). It has been shown that when the envelope of attended speech and EEG responses are used to derive such mapping functions, the model estimates can be used to discriminate between attended and unattended talkers. However, the predictive/reconstructive performance of the models is dependent on how the model parameters are estimated. There exist a number of model estimation methods that have been published, along with a variety of datasets. It is currently unclear if any of these methods perform better than others, as they have not yet been compared side by side on a single standardized dataset in a controlled fashion. Here, we present a comparative study of the ability of different estimation methods to classify attended speakers from multi-channel EEG data. The performance of the model estimation methods is evaluated using different performance metrics on a set of labeled EEG data from 18 subjects listening to mixtures of two speech streams. We find that when forward models predict the EEG from the attended audio, regularized models do not improve regression or classification accuracies. When backward models decode the attended speech from the EEG, regularization provides higher regression and classification accuracies.

Automatic Detection of Spike and Wave Discharges in the EEG of Genetic Absence Epilepsy Rats From Strasbourg

Genetic absence epilepsy rats from Strasbourg are a strain of Wistar rats in which all animals exhibit spontaneous occurrences of spike and wave discharges (SWDs) in the EEG. In this paper, we propose a novel method for the detection of SWDs, based on the key observation that SWDs are quasi-periodic signals. The method consists of the following steps: 1) calculation of the spectrogram; 2) estimation of the background spectrum and detection of stimulation artifacts; 3) harmonic analysis with continuity analysis to estimate the fundamental frequency; and 4) classification based on the percentage of power in the harmonics to the total power of the spectrum. We evaluated the performance of the novel detection method and six SWD/seizure detection methods from literature on a large database of labeled EEG data consisting of two datasets running to a total duration of more than 26 days of recording. The method outperforms all tested SWD/seizure detection methods, showing a sensitivity and selectivity of 96% and 97%, respectively, on the first test set, and a sensitivity and selectivity of 94% and 92%, respectively, on the second test set. The detection performance is less satisfactory (as for all other methods) for EEG fragments showing more irregular and less periodic SWDs.