Classification Of Event-related Potentials Research Articles

Improving zero-training brain-computer interfaces by mixing model estimators

Objective. Brain-computer interfaces (BCI) based on event-related potentials (ERP) incorporate a decoder to classify recorded brain signals and subsequently select a control signal that drives a computer application. Standard supervised BCI decoders require a tedious calibration procedure prior to every session. Several unsupervised classification methods have been proposed that tune the decoder during actual use and as such omit this calibration. Each of these methods has its own strengths and weaknesses. Our aim is to improve overall accuracy of ERP-based BCIs without calibration. Approach. We consider two approaches for unsupervised classification of ERP signals. Learning from label proportions (LLP) was recently shown to be guaranteed to converge to a supervised decoder when enough data is available. In contrast, the formerly proposed expectation maximization (EM) based decoding for ERP–BCI does not have this guarantee. However, while this decoder has high variance due to random initialization of its parameters, it obtains a higher accuracy faster than LLP when the initialization is good. We introduce a method to optimally combine these two unsupervised decoding methods, letting one method’s strengths compensate for the weaknesses of the other and vice versa. The new method is compared to the aforementioned methods in a resimulation of an experiment with a visual speller. Main results. Analysis of the experimental results shows that the new method exceeds the performance of the previous unsupervised classification approaches in terms of ERP classification accuracy and symbol selection accuracy during the spelling experiment. Furthermore, the method shows less dependency on random initialization of model parameters and is consequently more reliable. Significance. Improving the accuracy and subsequent reliability of calibrationless BCIs makes these systems more appealing for frequent use.

Generalized sparse discriminant analysis for event-related potential classification

A brain–computer interface (BCI) is a system which provides direct communication between the mind of a person and the outside world by using only brain activity (EEG). The event-related potential (ERP)-based BCI problem consists of a binary pattern recognition. Linear discriminant analysis (LDA) is widely used to solve this type of classification problems, but it fails when the number of features is large relative to the number of observations. In this work we propose a penalized version of the sparse discriminant analysis (SDA), called generalized sparse discriminant analysis (GSDA), for binary classification. This method inherits both the discriminative feature selection and classification properties of SDA and it also improves SDA performance through the addition of Kullback–Leibler class discrepancy information. The GSDA method is designed to automatically select the optimal regularization parameters. Numerical experiments with two real ERP-EEG datasets show that, on one hand, GSDA outperforms standard SDA in the sense of classification performance, sparsity and required computing time, and, on the other hand, it also yields better overall performances, compared to well-known ERP classification algorithms, for single-trial ERP classification when insufficient training samples are available. Hence, GSDA constitute a potential useful method for reducing the calibration times in ERP-based BCI systems.

Brain Network Activation Analysis Utilizing Spatiotemporal Features for Event Related Potentials Classification.

The purpose of this study was to introduce an improved tool for automated classification of event-related potentials (ERPs) using spatiotemporally parcellated events incorporated into a functional brain network activation (BNA) analysis. The auditory oddball ERP paradigm was selected to demonstrate and evaluate the improved tool.Methods: The ERPs of each subject were decomposed into major dynamic spatiotemporal events. Then, a set of spatiotemporal events representing the group was generated by aligning and clustering the spatiotemporal events of all individual subjects. The temporal relationship between the common group events generated a network, which is the spatiotemporal reference BNA model. Scores were derived by comparing each subject's spatiotemporal events to the reference BNA model and were then entered into a support vector machine classifier to classify subjects into relevant subgroups. The reliability of the BNA scores (test-retest repeatability using intraclass correlation) and their utility as a classification tool were examined in the context of Target-Novel classification.Results: BNA intraclass correlation values of repeatability ranged between 0.51 and 0.82 for the known ERP components N100, P200, and P300. Classification accuracy was high when the trained data were validated on the same subjects for different visits (AUCs 0.93 and 0.95). The classification accuracy remained high for a test group recorded at a different clinical center with a different recording system (AUCs 0.81, 0.85 for 2 visits).Conclusion: The improved spatiotemporal BNA analysis demonstrates high classification accuracy. The BNA analysis method holds promise as a tool for diagnosis, follow-up and drug development associated with different neurological conditions.

A periodic spatio-spectral filter for event-related potentials

With respect to single trial detection of event-related potentials (ERPs), spatial and spectral filters are two of the most commonly used pre-processing techniques for signal enhancement. Spatial filters reduce the dimensionality of the data while suppressing the noise contribution and spectral filters attenuate frequency components that most likely belong to noise subspace. However, the frequency spectrum of ERPs overlap with that of the ongoing electroencephalogram (EEG) and different types of artifacts. Therefore, proper selection of the spectral filter cutoffs is not a trivial task. In this research work, we developed a supervised method to estimate the spatial and finite impulse response (FIR) spectral filters, simultaneously. We evaluated the performance of the method on offline single trial classification of ERPs in datasets recorded during an oddball paradigm. The proposed spatio-spectral filter improved the overall single-trial classification performance by almost 9% on average compared with the case that no spatial filters were used. We also analyzed the effects of different spectral filter lengths and the number of retained channels after spatial filtering.

Multilinear Discriminant Analysis With Subspace Constraints for Single-Trial Classification of Event-Related Potentials

The classification accuracy of a brain–computer interface (BCI) frequently suffers from ill-posed and overfitting problems. To avoid and alleviate these problems, we propose a method of a multilinear discriminant analysis with constraints to augment parameter reduction, regularization, and additional prior information for event-related potential (ERP)-based BCIs. The method reduces the number of parameters by multilinearization, regularizes the ill-posedness via subspaces that constrain the parameter spaces, and incorporates a brain functional connectivity through the constraints. The experimental results show that the proposed method improved the classification accuracy rates in a single-trial ERP processing.

Comparing EEG, its time-derivative and their joint use as features in a BCI for 2-D pointer control.

Efficient and accurate classification of event related potentials is a core task in brain-computer interfaces (BCI). This is normally obtained by first extracting features from the voltage amplitudes recorded via EEG at different channels and then feeding them into a classifier. In this paper we evaluate the relative benefits of using the first order temporal derivatives of the EEG signals, not the EEG signals themselves, as inputs to the BCI: an area that has not been thoroughly examined. Specifically, we compare the classification performance of features extracted from the first derivative, with those derived from the amplitude, as well as their combination using data from a P300-based BCI mouse. Features were selected based on the absolute difference of medians of the target and non-target classes. Classification was carried out by an ensemble of linear support vector machines which were optimised using the mutual information criterion. Comparisons were based on the area under the receiver operating characteristics. The Mann-Whitney one-tailed test was used to study significance. Results show that EEG amplitudes are outperformed by both the first derivative and the combined feature vector and that derivatives are better than the combined vector.

Improving Accuracy of Event-Related Potentials Classification by Channel Selection Using Independent Component Analysis and Least Square Methods

This paper proposes a method for achieving a high performance of N200 and P300 classification by applying independent component analysis to select the channels, which deliver brain signals with large N200 and P300 potentials and small artifacts. In this study, the authors find out the relationship between the highest accuracy and the weights of the independent components and use this relationship to predict the optimal channels of each individual subject. They compare five channel selection methods: the ICA-based method and the curve-fitting-based method proposed in this paper, the amplitude-based method, the experiential optimal 8 channel combination and all 30 channel combination methods. The comparative studies show that the ICA-based method achieves an average accuracy of 99.3% across four subjects, which is superior to the other four methods.

A new approach for the classification of event related potentials for valid and paradox reasoning

In the present paper, a novel approach is introduced for comparing and classifying recorded ERP signals from subjects applying valid (Aristotle's) and paradox (Zeno's) syllogisms. In fact, the authors conceived and realized a corresponding experiment, as well as a new method for processing, fitting and classifying the corresponding captured ERP signals into groups according to their similarity. Subsequently, for each such group, an ideal curve that represents all signals of the group has been evaluated for valid and paradox reasoning separately. These “ideal representatives” manifest essential statistical differences per subject for a considerable number of electrodes (5 electrodes with 99% level of confidence, 14 electrodes with 95% level of confidence, 17 electrodes with 90% level of confidence). These results support the assumption that the obtained ideal representatives may indeed reflect essential differences in the underlying brain functions which generated the obtained ERPs. Equivalently, one may claim that the performed experiment and the associated results manifest statistically essential differences between the mental functions during valid and paradox reasoning.

A framework for a real time intelligent and interactive Brain Computer Interface

This research proposes a framework for a real time implementation of a Brain Computer Interface (BCI). This interface is designed with a future objective of providing a testing platform as an interactive and intelligent Image Search and Retrieval tool that allows users, disabled or otherwise, to browse and search for images using non-invasive electroencephalography (EEG) signals. As a proof of concept, a prototype system was designed and implemented to test real time data collection and navigation through the interface by detection and classification of event-related potentials (ERPs). A comparison of different feature extraction methods and classifiers for the detection of ERPs is performed on a standard data set to determine the best fit for the real time implementation of the BCI. The preliminary results of the real time implementation of the prototype BCI system are quantified by the success rate of the user/subject in navigating through the interface and spelling a search keyword using the mental-typewriter Hex-O-Speller.

Images from the mind: BCI image evolution based on rapid serial visual presentation of polygon primitives

This paper provides a proof of concept for an EEG-based reconstruction of a visual image which is on a user’s mind. Our approach is based on the rapid serial visual presentation (RSVP) of polygon primitives and brain-computer interface (BCI) technology. The presentation of polygons that contribute to building a target image (because they match the shape and/or color of the target) trigger attention-related EEG patterns. Accordingly, these target primitives can be determined using BCI classification of event-related potentials (ERPs). They are then accumulated in the display until a satisfactory reconstruction is reached. Selection steps have an average classification accuracy of 75%. Twenty-five percent of the images could be reconstructed completely, while more than 65% of the available visual details could be captured on average. Most of the misclassifications were not misinterpretations of the BCI concerning users’ intent; rather, users tried to select polygons that were different than what was intended by the experimenters. Open problems and alternatives to develop a practical BCI-based image reconstruction application are discussed.

Single-trial classification of event-related potentials in rapid serial visual presentation tasks using supervised spatial filtering.

Accurate detection of single-trial event-related potentials (ERPs) in the electroencephalogram (EEG) is a difficult problem that requires efficient signal processing and machine learning techniques. Supervised spatial filtering methods that enhance the discriminative information in EEG data are commonly used to improve single-trial ERP detection. We propose a convolutional neural network (CNN) with a layer dedicated to spatial filtering for the detection of ERPs and with training based on the maximization of the area under the receiver operating characteristic curve (AUC). The CNN is compared with three common classifiers: 1) Bayesian linear discriminant analysis; 2) multilayer perceptron (MLP); and 3) support vector machines. Prior to classification, the data were spatially filtered with xDAWN (for the maximization of the signal-to-signal-plus-noise ratio), common spatial pattern, or not spatially filtered. The 12 analytical techniques were tested on EEG data recorded in three rapid serial visual presentation experiments that required the observer to discriminate rare target stimuli from frequent nontarget stimuli. Classification performance discriminating targets from nontargets depended on both the spatial filtering method and the classifier. In addition, the nonlinear classifier MLP outperformed the linear methods. Finally, training based AUC maximization provided better performance than training based on the minimization of the mean square error. The results support the conclusion that the choice of the systems architecture is critical and both spatial filtering and classification must be considered together.

P300 variability during target detection in natural images: implications for single-trial classification

References The P300 is one of the most prominent and well-studied event-related potentials (ERPs) in the literature. The P300 is also a primary discriminant signal for many brain-computer interface (BCI) systems. It has been well established that the P300 amplitude and latency, along with reaction time, are directly linked to target discriminability. However, since the majority of P300 studies using an odd-ball or target-detection paradigm have a fixed or unquantified level of target discriminability, it remains unclear how the P300 is systematically modulated by stimulus properties affecting target detection. In this study we quantified visual properties of target objects within a large ensemble of natural images. Using a rapid serial visual presentation (RSVP) paradigm, we were able to systematically measure the effect of these properties on the behavioral reaction time (RT), P300, and single-trial classification accuracy. ERP & Single-trial Classification Target Size & P300 Group Effects

Early differential processing of material images: Evidence from ERP classification.

Investigating the temporal dynamics of natural image processing using event-related potentials (ERPs) has a long tradition in object recognition research. In a classical Go-NoGo task two characteristic effects have been emphasized: an early task independent category effect and a later task-dependent target effect. Here, we set out to use this well-established Go-NoGo paradigm to study the time course of material categorization. Material perception has gained more and more interest over the years as its importance in natural viewing conditions has been ignored for a long time. In addition to analyzing standard ERPs, we conducted a single trial ERP pattern analysis. To validate this procedure, we also measured ERPs in two object categories (people and animals). Our linear classification procedure was able to largely capture the overall pattern of results from the canonical analysis of the ERPs and even extend it. We replicate the known target effect (differential Go-NoGo potential at frontal sites) for the material images. Furthermore, we observe task-independent differential activity between the two material categories as early as 140 ms after stimulus onset. Using our linear classification approach, we show that material categories can be differentiated consistently based on the ERP pattern in single trials around 100 ms after stimulus onset, independent of the target-related status. This strengthens the idea of early differential visual processing of material categories independent of the task, probably due to differences in low-level image properties and suggests pattern classification of ERP topographies as a strong instrument for investigating electrophysiological brain activity.

EEG channel selection using particle swarm optimization for the classification of auditory event-related potentials.

Brain-machine interfaces (BMI) rely on the accurate classification of event-related potentials (ERPs) and their performance greatly depends on the appropriate selection of classifier parameters and features from dense-array electroencephalography (EEG) signals. Moreover, in order to achieve a portable and more compact BMI for practical applications, it is also desirable to use a system capable of accurate classification using information from as few EEG channels as possible. In the present work, we propose a method for classifying P300 ERPs using a combination of Fisher Discriminant Analysis (FDA) and a multiobjective hybrid real-binary Particle Swarm Optimization (MHPSO) algorithm. Specifically, the algorithm searches for the set of EEG channels and classifier parameters that simultaneously maximize the classification accuracy and minimize the number of used channels. The performance of the method is assessed through offline analyses on datasets of auditory ERPs from sound discrimination experiments. The proposed method achieved a higher classification accuracy than that achieved by traditional methods while also using fewer channels. It was also found that the number of channels used for classification can be significantly reduced without greatly compromising the classification accuracy.

Toward brain-computer interface based wheelchair control utilizing tactually-evoked event-related potentials.

BackgroundPeople with severe disabilities, e.g. due to neurodegenerative disease, depend on technology that allows for accurate wheelchair control. For those who cannot operate a wheelchair with a joystick, brain-computer interfaces (BCI) may offer a valuable option. Technology depending on visual or auditory input may not be feasible as these modalities are dedicated to processing of environmental stimuli (e.g. recognition of obstacles, ambient noise). Herein we thus validated the feasibility of a BCI based on tactually-evoked event-related potentials (ERP) for wheelchair control. Furthermore, we investigated use of a dynamic stopping method to improve speed of the tactile BCI system.MethodsPositions of four tactile stimulators represented navigation directions (left thigh: move left; right thigh: move right; abdomen: move forward; lower neck: move backward) and N = 15 participants delivered navigation commands by focusing their attention on the desired tactile stimulus in an oddball-paradigm.ResultsParticipants navigated a virtual wheelchair through a building and eleven participants successfully completed the task of reaching 4 checkpoints in the building. The virtual wheelchair was equipped with simulated shared-control sensors (collision avoidance), yet these sensors were rarely needed.ConclusionWe conclude that most participants achieved tactile ERP-BCI control sufficient to reliably operate a wheelchair and dynamic stopping was of high value for tactile ERP classification. Finally, this paper discusses feasibility of tactile ERPs for BCI based wheelchair control.

AGGREGATION OF SPARSE LINEAR DISCRIMINANT ANALYSES FOR EVENT-RELATED POTENTIAL CLASSIFICATION IN BRAIN-COMPUTER INTERFACE

Two main issues for event-related potential (ERP) classification in brain-computer interface (BCI) application are curse-of-dimensionality and bias-variance tradeoff, which may deteriorate classification performance, especially with insufficient training samples resulted from limited calibration time. This study introduces an aggregation of sparse linear discriminant analyses (ASLDA) to overcome these problems. In the ASLDA, multiple sparse discriminant vectors are learned from differently l1-regularized least-squares regressions by exploiting the equivalence between LDA and least-squares regression, and are subsequently aggregated to form an ensemble classifier, which could not only implement automatic feature selection for dimensionality reduction to alleviate curse-of-dimensionality, but also decrease the variance to improve generalization capacity for new test samples. Extensive investigation and comparison are carried out among the ASLDA, the ordinary LDA and other competing ERP classification algorithms, based on different three ERP datasets. Experimental results indicate that the ASLDA yields better overall performance for single-trial ERP classification when insufficient training samples are available. This suggests the proposed ASLDA is promising for ERP classification in small sample size scenario to improve the practicability of BCI.

Offline analysis of context contribution to ERP-based typing BCI performance

Objective. We aim to increase the symbol rate of electroencephalography (EEG) based brain–computer interface (BCI) typing systems by utilizing context information. Approach. Event related potentials (ERP) corresponding to a stimulus in EEG can be used to detect the intended target of a person for BCI. This paradigm is widely utilized to build letter-by-letter BCI typing systems. Nevertheless currently available BCI typing systems still require improvement due to low typing speeds. This is mainly due to the reliance on multiple repetitions before making a decision to achieve higher typing accuracy. Another possible approach to increase the speed of typing while not significantly reducing the accuracy of typing is to use additional context information. In this paper, we study the effect of using a language model (LM) as additional evidence for intent detection. Bayesian fusion of an n-gram symbol model with EEG features is proposed, and a specifically regularized discriminant analysis ERP discriminant is used to obtain EEG-based features. The target detection accuracies are rigorously evaluated for varying LM orders, as well as the number of ERP-inducing repetitions. Main results. The results demonstrate that the LMs contribute significantly to letter classification accuracy. For instance, we find that a single-trial ERP detection supported by a 4-gram LM may achieve the same performance as using 3-trial ERP classification for the non-initial letters of words. Significance. Overall, the fusion of evidence from EEG and LMs yields a significant opportunity to increase the symbol rate of a BCI typing system.

A Two-Level Predictive Event-Related Potential-Based Brain–Computer Interface

Increasing the freedom of communication using conventional row/column (RC) P300 paradigm by naive way (increasing matrix size) may deteriorate inherent distraction effect and interaction speed. In this paper, we propose a two-level predictive (TLP) paradigm by integrating a 3×3 two-level matrix paradigm with a statistical language model. The TLP paradigm is evaluated using offline and online data from ten healthy subjects. Significantly larger event-related potentials (ERPs) are evoked by the TLP paradigm compared with the classical 6×6 RC. During an online task (correctly spell an English sentence with 57 characters), accuracy and information transfer rate for the TLP are increased by 14.45% and 29.29%, respectively, when compared with the 6×6 RC. Time to complete the task is also decreased by 24.61% using TLP. In sharp contrast, an 8×8 RC (naive extension of the 6×6 RC) consumed 19.18% more time than the classical 6×6 RC. Furthermore, the statistical language model is also exploited to improve classification accuracy in a Bayesian approach. The proposed Bayesian fusion method is tested offline on data from the online spelling tasks. The results show its potential improvement on single-trial ERP classification.

Spatial-Temporal Discriminant Analysis for ERP-Based Brain-Computer Interface

Linear discriminant analysis (LDA) has been widely adopted to classify event-related potential (ERP) in brain-computer interface (BCI). Good classification performance of the ERP-based BCI usually requires sufficient data recordings for effective training of the LDA classifier, and hence a long system calibration time which however may depress the system practicability and cause the users resistance to the BCI system. In this study, we introduce a spatial-temporal discriminant analysis (STDA) to ERP classification. As a multiway extension of the LDA, the STDA method tries to maximize the discriminant information between target and nontarget classes through finding two projection matrices from spatial and temporal dimensions collaboratively, which reduces effectively the feature dimensionality in the discriminant analysis, and hence decreases significantly the number of required training samples. The proposed STDA method was validated with dataset II of the BCI Competition III and dataset recorded from our own experiments, and compared to the state-of-the-art algorithms for ERP classification. Online experiments were additionally implemented for the validation. The superior classification performance in using few training samples shows that the STDA is effective to reduce the system calibration time and improve the classification accuracy, thereby enhancing the practicability of ERP-based BCI.

Interval feature extraction for classification of event-related potentials (ERP) in EEG data analysis

Event-related potential data can be used to index perceptual and cognitive operations. However, they are typically high-dimensional and noisy. This study examines the original raw data and six feature-extraction methods as a pre-processing step before classification. Four traditionally used feature-extraction methods were considered: principal component analysis, independent component analysis, auto-regression, and wavelets. We add to these a less well-known method called interval feature extraction. It overproduces features from the ERP signal and then eliminates irrelevant and redundant features by the fast correlation-based filter. To make the comparisons fair, the other feature-extraction methods were also run with the filter. An experiment on two EEG datasets (four classification scenarios) was carried out to examine the classification accuracy of four classifiers on the extracted features: support vector machines with linear and perceptron kernel, the nearest neighbour classifier and the random forest ensemble method. The interval features led to the best classification accuracy in most of the configurations, specifically when used with the Random Forest classifier ensemble.