Event-related Potential Research Research Articles

Detection of Unfocused EEG Epochs by the Application of Machine Learning Algorithm.

Electroencephalography (EEG) is a non-invasive method used to track human brain activity over time. The time-locked EEG to an external event is known as event-related potential (ERP). ERP can be a biomarker of human perception and other cognitive processes. The success of ERP research depends on the laboratory conditions and attentiveness of the test subjects. Specifically, the inability to control experimental variables has reduced ERP research in the real world. This study collected EEG data under various experimental circumstances within an auditory oddball paradigm experiment to enable the use of ERP as an active biomarker in normal laboratory conditions. Then, ERP epochs were analyzed to identify unfocused epochs, affected by typical artifacts and external distortion. For the initial comparison, the ability of four unsupervised machine learning algorithms (MLAs) was evaluated to identify unfocused epochs. Then, their accuracy was compared with the human inspection and a current EEG analysis tool (EEGLab). All four MLAs were typically 95-100% accurate. In summary, our analysis finds that humans might miss subtle differences in the regular ERP patterns, but MLAs could efficiently identify those. Thus, our analysis suggests that unsupervised MLAs perform better for detecting unfocused ERP epochs compared with the other two standard methods.

Blunted stimulus-preceding negativity during reward anticipation in major depressive disorder

BackgroundReward processing dysfunction is a core characteristic of major depressive disorder (MDD), yet event-related potential (ERP) research in MDD has predominantly focused on reward receipt as opposed to anticipation. The stimulus-preceding negativity (SPN) ERP reflects anticipatory brain processing. This study examines whether individuals with MDD exhibit deficits during reward anticipation as evidenced by altered SPN amplitude. MethodsWe assessed prefeedback-SPN amplitudes during a monetary incentive delay (MID) task in individuals with MDD (n = 142, 99 with comorbid anxiety disorders [MDD + ANX]) compared to Controls (n = 37). A mixed analysis of variance was performed on prefeedback-SPN amplitude and behavioral measures, with group (MDD, MDD + ANX, Control) as the between-subjects factor, and feedback (gain, loss) and electrode (F3, F4, Fz, C3, C4, Cz, P3, P4, Pz) as within-subjects factors. ResultsA group main effect revealed faster reaction times for the Control group than MDD and MDD + ANX groups. A group x feedback interaction indicated that the MDD subgroup had smaller prefeedback-SPN amplitudes than MDD + ANX and Control groups when anticipating gain feedback. Additionally, individuals with current MDD, irrespective of past MDD and comorbid anxiety, exhibited smaller SPN amplitudes than Controls prior to gain feedback. LimitationsThe MID paradigm, designed for functional magnetic resonance imaging (fMRI) data acquisition, lacks optimization for ERP analysis. Moreover, the clinical groups included more females than the Control group. ConclusionsReduced resource allocation to reward anticipation may differentiate MDD from MDD + ANX and Control groups. Further investigation of the neural mechanisms of distinct MDD phenotypes is warranted.

A Wearable Device-Based System: The Potential Role in Real-time and Remote EEG Monitoring

Purpose: In recent years, the Electroencephalography (EEG)-based Brain-Computer Interface (BCI) application has been growing rapidly and has emerged as a technology with high translational potential since it allows disabled people to translate human intentions into control signals and to interact with the external environment without any kinesthetic movement. Individuals with significant health problems can benefit from this technology to improve their independence and facilitate participation in activities, thus improving general well-being and preventing impairments. The rapid advances in technology have led to optimal innovation in the context of wearable health monitoring and remote control solutions. Many wearable devices for capturing EEG signals in daily life have recently been released on the market. Our paper aims to present a wearable de-vice-based system for real-time and remote EEG monitoring, to describe the proposed system modules and the signal processing algorithms, to explore the functionalities of this wearable EEG solution, and to suggest its potential application for daily brain data recordings in the home environment. Materials and Methods: The validity of the Emotiv Epoc+ device in the continuous and real-time EEG signals acquisition and monitoring was demonstrated by means of preliminary test measurements during a resting state paradigm performed by a healthy control subject. Results: Our findings confirmed the Epoc+ reliability for event-related brain potential research and in measuring acceptable spectral EEG data. Conclusion: The described approach, focusing on the emerging area of remote monitoring sensors and wearable device applications, might be leveraged to measure complex health outcomes in non-specialist and remote settings.

A mapping-knowledge-domain analysis of ERP research on language processing.

The event-related potentials (ERPs) technique represents a newly developed methodology in cognitive neuroscience and has significantly extended the scope of linguistic studies, offering valuable insights into cognitive processes related to language. While extant literature reviews have addressed specific facets of ERP research on language processing, a comprehensive overview of this domain remains notably absent. This study aims to fill this gap by pioneering a mapping-knowledge-domain analysis of ERP research on language processing using Citespace, a visualized bibliometric software. The current study conducted a meticulous survey and evaluation of relevant literature extracted from the Web of Science core collection. Initially, this study outlines the spatial-temporal distribution within this domain. Subsequently, employing document co-citation analysis, keyword co-occurrence analysis, cluster analysis, and burst detection analysis, this study delved deeper into the research landscape. Findings reveal that key areas in ERP research on language processing predominantly focus on sentence comprehension, reading comprehension, and mismatch negativity, with notable emphasis on topics such as speech perception, temporal dynamics, and working memory. The current study advocates for future investigations to concentrate on larger linguistic units, explore the integration of ERP components and their functional significance, and scrutinize individual differences among participants. These directions are imperative for advancing the understanding of language processing mechanisms.

How variable are the classic ERP effects during sentence processing? A systematic resampling analysis of the N400 and P600 effects

Although event-related potential (ERP) research on language processing has capitalized on key, theoretically influential components such as the N400 and P600, their measurement properties—especially the variability in their temporal and spatial parameters—have rarely been examined. The current study examined the measurement properties of the N400 and P600 effects elicited by semantic and syntactic anomalies, respectively, during sentence processing. We used a bootstrap resampling procedure to randomly draw many thousands of resamples varying in sample size and stimulus count from a larger sample of 187 participants and 40 stimulus sentences of each type per condition. Our resampling investigation focused on three issues: (a) statistical power; (b) variability in the magnitudes of the effects; and (c) variability in the temporal and spatial profiles of the effects. At the level of grand averages, the N400 and P600 effects were both robust and substantial. However, across resamples, there was a high degree of variability in effect magnitudes, onset times, and scalp distributions, which may be greater than is currently appreciated in the literature, especially for the P600 effects. These results provide a useful basis for designing future studies using these two well-established ERP components. At the same time, the results also highlight challenges that need to be addressed in future research (e.g., how best to analyze the ERP data without engaging in such questionable research practices as p-hacking).

Exploration of the influence of the quantification method and reference scheme on feedback-related negativity and standardized measurement error of feedback-related negativity amplitudes in a trust game

Various approaches have been taken over the years to quantify event-related potential (ERP) responses and these approaches may vary in their utility connecting empirical research and scientific claims. In this work we compared different quantification methods as well as the influence of three reference methods (linked mastoids, average reference, and current source density) on the resulting ERP amplitude. We use the experimental effects and effect sizes (Cohen's d) to evaluate the different methodological variants and we calculate intraclass correlation coefficients (ICC). In addition, the bootstrapped standard error of the means (SME, Luck et al., 2021), which was recently suggested as a quality criterion for ERP research, is used for this purpose. Our example for an ERP is the feedback-related negativity (FRN) to feedback about trustee behavior in a trust game with participants in the trustor position. We found that the quantification methods concerning the FRN influenced the absolute value of condition effects in the experimental paradigm. Yet, the patterns of effects were detected by all chosen methods, except for the ‘individual difference wave’-based peak window approach. In addition, our findings stress the importance of checking the reference electrodes concerning effects of the experimental conditions. Furthermore, interactions of topographical distribution and reference choice should be considered. Finally, we were able to show that the SME is lower for more datapoints that are given in the quantification period of the FRN, and higher for more negative FRN amplitudes. These biases may lead to divergence of SME and effect size detection. Therefore, if the SME was used to compare different processing choices one should consider controlling for these important aspects of the data and possibly include other quality criteria like effect sizes.

Optimal filters for ERP research I: A general approach for selecting filter settings.

Filtering plays an essential role in event-related potential (ERP) research, but filter settings are usually chosen on the basis of historical precedent, lab lore, or informal analyses. This reflects, in part, the lack of a well-reasoned, easily implemented method for identifying the optimal filter settings for a given type of ERP data. To fill this gap, we developed an approach that involves finding the filter settings that maximize the signal-to-noise ratio for a specific amplitude score (or minimizes the noise for a latency score) while minimizing waveform distortion. The signal is estimated by obtaining the amplitude score from the grand average ERP waveform (usually a difference waveform). The noise is estimated using the standardized measurement error of the single-subject scores. Waveform distortion is estimated by passing noise-free simulated data through the filters. This approach allows researchers to determine the most appropriate filter settings for their specific scoring methods, experimental designs, subject populations, recording setups, and scientific questions. We have provided a set of tools in ERPLAB Toolbox to make it easy for researchers to implement this approach with their own data.

Optimal filters for ERP research II: Recommended settings for seven common ERP components.

In research with event-related potentials (ERPs), aggressive filters can substantially improve the signal-to-noise ratio and maximize statistical power, but they can also produce significant waveform distortion. Although this tradeoff has been well documented, the field lacks recommendations for filter cutoffs that quantitatively address both of these competing considerations. To fill this gap, we quantified the effects of a broad range of low-pass filter and high-pass filter cutoffs for seven common ERP components (P3b, N400, N170, N2pc, mismatch negativity, error-related negativity, and lateralized readiness potential) recorded from a set of neurotypical young adults. We also examined four common scoring methods (mean amplitude, peak amplitude, peak latency, and 50% area latency). For each combination of component and scoring methods, we quantified the effects of filtering on data quality (noise level and signal-to-noise ratio) and waveform distortion. This led to recommendations for optimal low-pass and high-pass filter cutoffs. We repeated the analyses after adding artificial noise to provide recommendations for data sets with moderately greater noise levels. For researchers who are analyzing data with similar ERP components, noise levels, and participant populations, using the recommended filter settings should lead to improved data quality and statistical power without creating problematic waveform distortion.

Evaluating the effectiveness of artifact correction and rejection in event-related potential research.

Eyeblinks and other large artifacts can create two major problems in event-related potential (ERP) research, namely confounds and increased noise. Here, we developed a method for assessing the effectiveness of artifact correction and rejection methods in minimizing these two problems. We then used this method to assess a common artifact minimization approach, in which independent component analysis (ICA) is used to correct ocular artifacts, and artifact rejection is used to reject trials with extreme values resulting from other sources (e.g., movement artifacts). This approach was applied to data from five common ERP components (P3b, N400, N170, mismatch negativity, and error-related negativity). Four common scoring methods (mean amplitude, peak amplitude, peak latency, and 50% area latency) were examined for each component. We found that eyeblinks differed systematically across experimental conditions for several of the components. We also found that artifact correction was reasonably effective at minimizing these confounds, although it did not usually eliminate them completely. In addition, we found that the rejection of trials with extreme voltage values was effective at reducing noise, with the benefits of eliminating these trials outweighing the reduced number of trials available for averaging. For researchers who are analyzing similar ERP components and participant populations, this combination of artifact correction and rejection approaches should minimize artifact-related confounds and lead to improved data quality. Researchers who are analyzing other components or participant populations can use the method developed in this study to determine which artifact minimization approaches are effective in their data.

The effect of hair type and texture on electroencephalography and event-related potential data quality.

Research utilizing event-related potential (ERP) methods is generally biased with regard to sample representativeness. Among the myriad of factors that contribute to sample bias are researchers' assumptions about the extent to which racial differences in hair texture, volume, and style impact electrode placement, and subsequently, study eligibility. The current study examines these impacts using data collected from n = 213 individuals ages 17-19 years, and offers guidance on collection of ERP data across the full spectrum of hair types. Individual differences were quantified for hair texture using a visual scale, and for hair volume by measuring the amount of gel used in cap preparation. Electroencephalography data quality was assessed with multiple metrics at the preprocessing, post-processing, and variable generation stages. Results indicate that hair volume is associated with small, but systematic differences in signal quality and signal amplitude. Such differences are highly problematic as they could be misattributed to cognitive differences among groups. However, inclusion of gel volume as a covariate to account for individual differences in hair volume significantly reduced, and in most cases eliminated, group differences. We discuss strategies for overcoming real and perceived technical barriers for researchers seeking to achieve greater inclusivity and representativeness in ERP research.

Prefrontal event-related potential markers in association with mild cognitive impairment.

Alzheimer's disease (AD) is among the leading contributors of dementia globally with approximately 60-70% of its cases. Current research is focused on the mild cognitive impairment (MCI), which is associated with cognitive decline but does not disrupt routine activities. Event-related potential (ERP) research is essential in screening patients with MCI. Low-density channel electroencephalography (EEG) is frequently used due to its convenience, portability, and affordability, making it suitable for resource-constrained environments. Despite extensive research on neural biomarkers for cognitive impairment, there is a considerable gap in understanding the effects on early stages of cognitive processes, particularly when combining physiological and cognitive markers using portable devices. The present study aimed to examine cognitive shortfalls and behavioral changes in patients with MCI using prefrontal selective attention ERP recorded from a prefrontal two-channel EEG device. We assessed cognitive decline using the Mini-Mental State Examination (MMSE) and the Seoul Neuropsychological Screening Battery (SNSB). We administered auditory selective attention tasks to 598 elderly participants, including those with MCI (160) and cognitively normal (CN) individuals (407). We conducted statistical analyses such as independent t-tests, Pearson's correlations, and univariate and multiple logistic regression analyses to assess group differences and associations between neuropsychological tests, ERP measures, behavioral measures, and MCI prevalence. Our findings revealed that patients with MCI demonstrated slower information-processing abilities, and exhibited poorer task execution, characterized by reduced accuracy, increased errors, and higher variability in response time, compared to CN adults. Multiple logistic regression analyses confirmed the association between some ERP and behavioral measures with MCI prevalence, independent of demographic and neuropsychological factors. A relationship was observed between neuropsychological scores, ERP, and behavioral measures. The slower information processing abilities, and poor task execution in the MCI group compared to the CN individuals suggests flawed neurological changes and reduced attentional maintenance during cognitive processing, respectively. Hence, the utilization of portable EEG devices to capture prefrontal selective attention ERPs, in combination with behavioral assessments, holds promise for the identification of mild cognitive deficits and neural alterations in individuals with MCI. This approach could potentially augment the traditional neuropsychological tests during clinical screening for MCI.

Language Experience Modulates the Visual N200 Response for Disyllabic Chinese Words: An Event-Related Potential Study.

Prior event-related potential (ERP) research on how the brain processes non-alphabetic scripts like Chinese has identified an N200 component related to early visual processing of Chinese disyllabic words. An enhanced N200 response was observed when similar prime-target pairs were presented, but it was not elicited when native Chinese speakers read Korean Hangul, a script resembling Chinese characters. This led to the proposal that N200 was not a universal marker for orthographic processing but rather specific and unique to Chinese. However, there was uncertainty due to the absence of Korean participants in the previous research. The impact of language experience on N200 remains unclear. To address this, the present pilot ERP study included three adult groups (totaling 30 participants) with varying language proficiency levels. The participants judged if randomly presented words were Chinese or Korean, while the ERP responses were recorded. The behavioral data showed high accuracy across the groups. The reaction times differed between the groups with the native speakers responding faster. The N200 patterns varied across the groups. Both Chinese native speakers and Chinese-as-second-language learners showed stronger N200 responses for Chinese words compared to Korean words regardless of whether an adaptive or a fixed-time window was used for the N200 quantification, but this was not the case for Korean native speakers. Our cross-linguistic study suggests that N200 is not exclusive to Chinese orthography. Instead, it reflects general visual processing sensitive to both orthographic features and learning experience.

Of ugly gains and happy losses: An event-related potential study of interactions of the intrinsic and acquired valence of emotional pictures

During the last decades, event-related potential research on the processing of intrinsic and acquired valence has made great progress, but the two dimensions rarely varied simultaneously. Only that way, however, can we investigate whether the acquisition of extrinsic valence varies with intrinsic valence and whether intrinsic and acquired valence share the same brain mechanisms. Forty-five participants performed associative learning of gains and losses, using pictures varying on intrinsic valence (positive, negative) and outcome (90 % gain, 50 %/50 %, 90 % loss). 64-channel EEG was recorded. During acquisition, one picture from each valence/outcome combination was repeatedly presented, followed by abstract outcome information (+10 ct, −10 ct) at the predefined probability. In the test phase, participants pressed buttons to earn the real gains and avoid the real losses associated with the pictures. Here, effects of outcome and/or its congruence with intrinsic valence were observed for RT, error rate, frontal theta power, posterior P2, P300, and LPP. Moreover, outcome systematically affected post-test valence and arousal ratings. During acquisition, a contingency effect (90 % > 50 %) on amplitude of a frontal negative slow wave accompanied the progress of learning, independently of outcome, valence, and congruence. The relative absence of outcome effects during acquisition suggests “cold” semantic rather than genuinely affective processing of gains and losses. However, with real gains and losses in the test phase, “hot” affective processing took place, and outcome and its congruence with intrinsic valence influenced behavior and neural processing. Finally, the data suggest both shared and distinct brain mechanisms of intrinsic and acquired valence.

Estimating statistical power for ERP studies using the auditory N1, Tb, and P2 components.

The N1, Tb, and P2 components of the event-related potential (ERP) are thought to reflect the sequential processing of auditory stimuli in the human brain. Despite their extensive use in biological, cognitive, and clinical neuroscience, there are no guidelines for how to appropriately power ERP studies using these components. In the present study, we investigated how the number of trials, number of participants, effect magnitude, and study design influenced statistical power. Using Monte Carlo simulations of ERP data from a passive listening task, we determined the probability of finding a statistically significant effect in 58,900 experiments repeated 1,000 times each. We found that as the number of trials, number of participants, and effect magnitude increased, so did statistical power. We also found that increasing the number of trials had a bigger effect on statistical power for within-subjectdesigns than for between-subjectdesigns, and that within-subjectdesigns required a smaller number of trials and participants to provide the same level of statistical power for a given effect magnitude than between-subjectdesigns. These results show that it is important to carefully consider these factors when designing ERP studies, rather than relying on tradition or anecdotal evidence. To improve the robustness and reproducibility of ERP research, we have built an online statistical power calculator (https://bradleynjack.shinyapps.io/ErpPowerCalculator), which we hope will allow researchers to estimate the statistical power of previous studies, as well as help them design appropriately-powered studies in the future.

Learning new words via feedback-Association between feedback-locked ERPs and recall performance-An exploratory study.

Feedback learning is thought to involve the dopamine system and its projection sites in the basal ganglia and anterior cingulate cortex (ACC), regions associated with procedural learning. Under certain conditions, such as when feedback is delayed, feedback-locked activation is pronounced in the medial temporal lobe (MTL), which is associated with declarative learning. In event-related potential research, the feedback-related negativity (FRN) has been linked to immediate feedback processing, while the N170, possibly reflecting MTL activity, has been related to delayed feedback processing. In the current study, we performed an exploratory investigation on the relation between N170 and FRN amplitude and memory performance in a test for declarative memory (free recall), also exploring the role of feedback delay. To this end, we adapted a paradigm in which participants learned associations between non-objects and non-words with either immediate or delayed feedback, and added a subsequent free recall test. We indeed found that N170, but not FRN amplitudes, depended on later free recall performance, with smaller amplitudes for later remembered non-words. In an additional analysis with memory performance as dependent variable, the N170, but not the FRN amplitude predicted free recall, modulated by feedback timing and valence. This finding shows that the N170 reflects an important process during feedback processing, possibly related to expectations and their violation, but is distinct from the process reflected by the FRN.

Quantity Processing in Object Counting Task Based on Event-Related Potentials

In the preliminary event-dependent potentials study, our team found that the neural mechanism of Chinese singular and plural picture naming is different, and explained that the differences might lie in the different numbers of quantitative processing. However, previous studies have lacked separate neural processing of objects in singular-plural pictures. This study intends to explore the neural differences in quantity processing in singular and plural pictures by using an object counting task and combining Event-related potential (ERP). It is expected to prove that the neural differences of naming Chinese single and plural pictures found in our previous ERP research are neural processing differences of different quantities, further demonstrating that the quantity processing in Chinese singular and plural picture naming is automatic. The experimental results showed no significant differences in object counting accuracy (P=0.139) and reaction time (P=0.231) between the singular and plural groups. However, there was a noticeable difference in ERPs between the two groups (P<0.05) and statistical analysis showed that the P1 effect of the parieto-occipital lobe was greater in plural pictures than in singular pictures (140-180ms) and the P2 effect of the parieto-occipital lobe was more significant in singular pictures than in plural pictures (200-320ms). There are differences in the quantity processing of healthy adults in object counting tasks. These differences are similar to the neural differences in Chinese singular-plural picture naming, indicating that the quantity is automatically processed in this process.

RELAX part 2: A fully automated EEG data cleaning algorithm that is applicable to Event-Related-Potentials

ObjectiveElectroencephalography (EEG) is often used to examine neural activity time-locked to stimuli presentation, referred to as Event-Related Potentials (ERP). However, EEG is influenced by non-neural artifacts, which can confound ERP comparisons. Artifact cleaning reduces artifacts, but often requires time-consuming manual decisions. Most automated methods filter frequencies <1 Hz out of the data, so are not recommended for ERPs (which contain frequencies <1 Hz). Our aim was to test the RELAX (Reduction of Electroencephalographic Artifacts) pre-processing pipeline for use on ERP data. MethodsThe cleaning performance of multiple versions of RELAX were compared to four commonly used EEG cleaning pipelines across both artifact cleaning metrics and the amount of variance in ERPs explained by different conditions in a Go-Nogo task.ResultsRELAX with Multi-channel Wiener Filtering (MWF) and wavelet-enhanced independent component analysis applied to artifacts identified with ICLabel (wICA_ICLabel) cleaned data most effectively and produced amongst the most dependable ERP estimates. RELAX with wICA_ICLabel only or MWF_only may detect effects better for some ERPs. ConclusionsRELAX shows high artifact cleaning performance even when data is high-pass filtered at 0.25 Hz (applicable to ERP analyses). SignificanceRELAX is easy to implement via EEGLAB in MATLAB and freely available on GitHub. Given its performance and objectivity we recommend RELAX to improve artifact cleaning and consistency across ERP research.

A two for one special: EEG hyperscanning using a single-person EEG recording setup

EEG hyperscanning refers to recording electroencephalographic (EEG) data from multiple participants simultaneously. Many hyperscanning experimental designs seek to mimic naturalistic behavior, relying on unpredictable participant-generated stimuli. The majority of this research has focused on neural oscillatory activity that is quantified over hundreds of milliseconds or more. This contrasts with traditional event-related potential (ERP) research in which analysis focuses on transient responses, often only tens of milliseconds in duration. Deriving ERPs requires precise time-locking between stimuli and EEG recordings, and thus typically relies on pre-set stimuli that are presented to participants by a system that controls stimulus timing and synchronization with an EEG system. EEG hyperscanning methods typically use separate EEG amplifiers for each participant, increasing cost and complexity — including challenges in synchronizing data between systems. Here, we describe a method that allows for simultaneous acquisition of EEG data from a pair of participants engaged in conversation, using a single EEG system with simultaneous audio data collection that is synchronized with the EEG recording. This allows for the post-hoc insertion of trigger codes so that it is possible to analyze ERPs time-locked to specific events. We further demonstrate methods for deriving ERPs elicited by another person's spontaneous speech, using this setup.•EEG hyperscanning method using a single EEG amplifier•EEG hyperscanning method allowing simultaneous recording of audio data directly into the EEG data file for perfect synchronization•EEG method for naturalistic language and human interaction studies that allows the study of event-related potentials time-locked to spontaneous speech

A Guided Tutorial on Modelling Human Event-Related Potentials with Recurrent Neural Networks

In cognitive neuroscience research, computational models of event-related potentials (ERP) can provide a means of developing explanatory hypotheses for the observed waveforms. However, researchers trained in cognitive neurosciences may face technical challenges in implementing these models. This paper provides a tutorial on developing recurrent neural network (RNN) models of ERP waveforms in order to facilitate broader use of computational models in ERP research. To exemplify the RNN model usage, the P3 component evoked by target and non-target visual events, measured at channel Pz, is examined. Input representations of experimental events and corresponding ERP labels are used to optimize the RNN in a supervised learning paradigm. Linking one input representation with multiple ERP waveform labels, then optimizing the RNN to minimize mean-squared-error loss, causes the RNN output to approximate the grand-average ERP waveform. Behavior of the RNN can then be evaluated as a model of the computational principles underlying ERP generation. Aside from fitting such a model, the current tutorial will also demonstrate how to classify hidden units of the RNN by their temporal responses and characterize them using principal component analysis. Statistical hypothesis testing can also be applied to these data. This paper focuses on presenting the modelling approach and subsequent analysis of model outputs in a how-to format, using publicly available data and shared code. While relatively less emphasis is placed on specific interpretations of P3 response generation, the results initiate some interesting discussion points.

Phase Analysis of Event-Related Potentials Based on Dynamic Mode Decomposition

Real-time detection of event-related potentials (ERPs) and exploration of ERPs generation mechanisms are vital to practical application of brain–computer interfaces (BCI). Traditional methods for ERPs analysis often fall into time domain, time–frequency domain, or spatial domain. Methods which can reveal spatiotemporal interactions by simultaneously analyzing multi-channel EEG signals may provide new insights into ERP research and is highly desired. Additionally, although phase information has been investigated to describe the phase consistency of a certain frequency component across different ERP trials, it is of research significance to analyze the phase reorganization across different frequency components that constitute a single-trial ERP signal. To address these problems, dynamic mode decomposition (DMD) was applied to decompose multi-channel EEG into a series of spatial–temporal coherent DMD modes, and a new metric, called phase variance distribution (PVD) is proposed as an index of the phase reorganization of DMD modes during the ERP in a single trial. Based on the PVD, a new error-related potential (ErrP) detection method based on symmetric positive defined in Riemann manifold is proposed to demonstrate the significant PVD differences between correct and error trials. By including the phase reorganization index, the 10-fold cross-validation results of an ErrP detection task showed that the proposed method is 4.98%, 27.99% and 7.98% higher than the counterpart waveform-based ErrP detection method in the terms of weighted accuracy rate, precision and recall of the ErrP class, respectively. The resulting PVD curve shows that with the occurrence of ERP peaks, the phases of different frequency rhythms are getting to aligned and yields a significant smaller PVD. Since the DMD modes of different frequencies characterize spatiotemporal coherence of multi-channel EEG at different functional regions, the new phase reorganization index, PVD, may indicate the instantaneous phase alignment of different functional networks and sheds light on a new interpretation of ERP generation mechanism.