Electroencephalography Data Research Articles

Relation between cortical thickness and frontal lobe P50 response in neurotypical aging: MEG study

AbstractBackgroundThe auditory Event Related Potential (ERP) with latency at peak at 50 ms is a potential biomarker of prodromal Alzheimer’s disease (AD). However, the relation between P50 ERP and the thickness of the gray matter over frontal brain regions in healthy individuals over 55 y.o is not well studied. We examined the relation between frontal lobe generator of auditory P50 and degree of the gray matter volume lost in healthy individuals. We hypothesized that reduced thickness in the cingulate cortex is associated with aberrant function of the frontal lobe generator of the P50 response.MethodSeven healthy subjects (ages 58‐72 years, 4 women) participated in the ongoing study. P50 response was calculated from magnetoencephalographic and electroencephalographic data during presentation of 7 tones (30 ms duration) ranging from 500 to 986 Hz. Each tone was preceded by 850 ms period of silence. Free‐surfer computed the cortical thickness for each subject as the distance between the gray/white matter boundary and the pial surface for each pair of vertices. 110 T1‐MPRAGE scans of elderly healthy controls (over 65 yo) from a Neurocognitive aging dataset were used as an average‐control group. To calculate the difference in thickness between averaged control and individual subject, the z‐value was applied. The template of 110‐averaged data was transformed into the subject’s space, and for each vertices we subtracted the mean and divide by the standard deviationResultIn this study we show, that in subjects with less thickness in the cingulate cortex the amplitude of P50 response was enhanced with respect to those subjects who had more cortical thickness (Z value >5 vs. Z value <2). The aberrant function of the frontal lobe generator of P50 response was associated with enchanted amplitude of the P50 in MEG and EEG data.ConclusionOur results suggest that age‐related deficiency in frontal lobe generators of auditory P50 in healthy aging participants is associated with a reduction of gray matter volume in cingulate cortex. Present results support the notion about age‐related changes in modulatory role of frontal lobe generator of auditory P50 that are located at the cingulate cortex.

Enhancing pain modulation: the efficacy of synchronous combination of virtual reality and transcutaneous electrical nerve stimulation

IntroductionVirtual reality (VR) and transcutaneous electrical nerve stimulation (TENS) have emerged as effective interventions for pain reduction. However, their standalone applications often yield limited analgesic effects, particularly in certain painful...

Identifying Individualized Neurophysiological Causal Features for Working Memory Performance: Implications for Non‐Invasive Brain Stimulation

AbstractBackgroundNon‐invasive brain stimulation (NIBS) has been extensively used to target specific neural oscillations with the aim of improving working memory performance. However, current NIBS paradigms have two main limitations: first, lack of personalization and second, a need for a randomized controlled trial (RCT) to infer causality between targets of NIBS and subsequent performance. RCT is a powerful measure to infer causality, but its implementation is expensive, time consuming, and sometimes simply not possible. Therefore, our aim was to use computational models and observational data to discover causal relations between neural oscillations and working memory performance to identify potential targets and stimulation parameters for personalized NIBS paradigms.MethodWe used electroencephalography (EEG) data of 66 young healthy participants collected while performing a 3‐back working memory task. Using graphical causal modeling, we extracted individualized causal brain oscillations of 3‐back performance and compared the causal features between two groups: high and low performers.ResultTotal number of causal features in high performers was higher than low performers. Among the causal features, right temporal gamma oscillation was ∼5 times (z‐score = 3.87, p = 0.003) more frequently a causal feature among high performers than low performers. However, the power of causal temporal gamma oscillation was not different between the two groups.ConclusionOur findings suggest that a potential approach to improve working memory performance using NIBS is to induce more causal gamma oscillations by, for example, generating more local gamma entrainment over the right temporal cortex and not necessarily by increasing gamma power.

Semi-synthetic EEG Data for the Evaluation of Linear EEG Cleaning Methods

Electroencephalography (EEG) data recordings can be contaminated by artefacts that reduce the quality and make analysis difficult, and therefore cleaning methods are essential for accurate analysis of EEG data. It is not yet well established how to measure performance based on measured contaminated data since there is no established benchmark for comparison. Here we use “clean” EEG data synthetically contaminated by electrocardiography (ECG), electrooculography (EOG) and electromyography (EMG). This introduces fewer assumptions to the comparison between various cleaning methods, providing a clear datum for comparison. Further contamination is controlled, adding artefacts individually and also as a combination of artefacts. The results show that signal to noise ratio (SNR) of the simulated artefacts was within the same ranges as found with measured artefacts from literature. Popular linear cleaning methods were evaluated on the dataset, showing similar results to those in the literature, further validating the usefulness and accuracy of the semi-synthetic dataset. The semi-synthetic dataset showed comparable characteristics to real measured EEG data and proved useful in the assessment of EEG cleaning methods. The cleaning methods showed varied results when performance was evaluated on individual artefacts.

Network-motif delay differential analysis of brain activity during seizures.

Delay Differential Analysis (DDA) is a nonlinear method for analyzing time series based on principles from nonlinear dynamical systems. DDA is extended here to incorporate network aspects to improve the dynamical characterization of complex systems. To demonstrate its effectiveness, DDA with network capabilities was first applied to the well-known Rössler system under different parameter regimes and noise conditions. Network-motif DDA, based on cortical regions, was then applied to invasive intracranial electroencephalographic data from drug-resistant epilepsy patients undergoing presurgical monitoring. The directional network motifs between brain areas that emerge from this analysis change dramatically before, during, and after seizures. Neural systems provide a rich source of complex data, arising from varying internal states generated by network interactions.

Neurophysiological mental fatigue assessment for developing user-centered Artificial Intelligence as a solution for autonomous driving.

The human factor plays a key role in the automotive field since most accidents are due to drivers' unsafe and risky behaviors. The industry is now pursuing two main solutions to deal with this concern: in the short term, there is the development of systems monitoring drivers' psychophysical states, such as inattention and fatigue, and in the medium-long term, there is the development of fully autonomous driving. This second solution is promoted by recent technological progress in terms of Artificial Intelligence and sensing systems aimed at making vehicles more and more accurately aware of their "surroundings." However, even with an autonomous vehicle, the driver should be able to take control of the vehicle when needed, especially during the current transition from the lower (SAE < 3) to the highest level (SAE = 5) of autonomous driving. In this scenario, the vehicle has to be aware not only of its "surroundings" but also of the driver's psychophysical state, i.e., a user-centered Artificial Intelligence. The neurophysiological approach is one the most effective in detecting improper mental states. This is particularly true if considering that the more automatic the driving will be, the less available the vehicular data related to the driver's driving style. The present study aimed at employing a holistic approach, considering simultaneously several neurophysiological parameters, in particular, electroencephalographic, electrooculographic, photopletismographic, and electrodermal activity data to assess the driver's mental fatigue in real time and to detect the onset of fatigue increasing. This would ideally work as an information/trigger channel for the vehicle AI. In all, 26 professional drivers were engaged in a 45-min-lasting realistic driving task in simulated conditions, during which the previously listed biosignals were recorded. Behavioral (reaction times) and subjective measures were also collected to validate the experimental design and to support the neurophysiological results discussion. Results showed that the most sensitive and timely parameters were those related to brain activity. To a lesser extent, those related to ocular parameters were also sensitive to the onset of mental fatigue, but with a delayed effect. The other investigated parameters did not significantly change during the experimental session.

Disturbed laterality of non-rapid eye movement sleep oscillations in post-stroke human sleep: a pilot study.

Sleep is known to promote recovery post-stroke. However, there is a paucity of data profiling sleep oscillations in the post-stroke human brain. Recent rodent work showed that resurgence of physiologic spindles coupled to sleep slow oscillations (SOs) and concomitant decrease in pathological delta (δ) waves is associated with sustained motor performance gains during stroke recovery. The goal of this study was to evaluate bilaterality of non-rapid eye movement (NREM) sleep-oscillations (namely SOs, δ-waves, spindles, and their nesting) in post-stroke patients vs. healthy control subjects. We analyzed NREM-marked electroencephalography (EEG) data in hospitalized stroke-patients (n = 5) and healthy subjects (n = 3). We used a laterality index to evaluate symmetry of NREM oscillations across hemispheres. We found that stroke subjects had pronounced asymmetry in the oscillations, with a predominance of SOs, δ-waves, spindles, and nested spindles in affected hemisphere, when compared to the healthy subjects. Recent preclinical work classified SO-nested spindles as restorative post-stroke and δ-wave-nested spindles as pathological. We found that the ratio of SO-nested spindles laterality index to δ-wave-nested spindles laterality index was lower in stroke subjects. Using linear mixed models (which included random effects of concurrent pharmacologic drugs), we found large and medium effect size for δ-wave nested spindle and SO-nested spindle, respectively. Our results in this pilot study indicate that considering laterality index of NREM oscillations might be a useful metric for assessing recovery post-stroke and that factoring in pharmacologic drugs may be important when targeting sleep modulation for neurorehabilitation post-stroke.

A role for retro-splenial cortex in the task-related P3 network

ObjectiveThe P3 is an event-related response observed in relation to task-relevant sensory events. Despite its ubiquitous presence, the neural generators of the P3 are controversial and not well identified. MethodsWe compared source analysis of combined magneto- and electroencephalography (M/EEG) data with functional magnetic resonance imaging (fMRI) and simulation studies to better understand the sources of the P3 in an auditory oddball paradigm. ResultsOur results suggest that the dominant source of the classical, postero-central P3 lies in the retro-splenial cortex of the ventral cingulate gyrus. A second P3 source in the anterior insular cortex contributes little to the postero-central maximum. Multiple other sources in the auditory, somatosensory, and anterior midcingulate cortex are active in an overlapping time window but can be functionally dissociated based on their activation time courses. ConclusionsThe retro-splenial cortex is a dominant source of the parietal P3 maximum in EEG. SignificanceThese results provide a new perspective for the interpretation of the extensive research based on the P3 response.

Prolonged neural encoding of visual information in autism.

Autism spectrum disorder (ASD) is associated with a hyper-focused visual attentional style, impacting higher-order social and affective domains. The understanding of such peculiarity can benefit from the use of multivariate pattern analysis (MVPA) of high-resolution electroencephalography (EEG) data, which has proved to be a powerful technique to investigate the hidden neural dynamics orchestrating sensory and cognitive processes. Here, we recorded EEG in typically developing (TD) children and in children with ASD during a visuo-spatial attentional task where attention was exogenously captured by a small (zoom-in) or large (zoom-out) cue in the visual field before the appearance of a target at different eccentricities. MVPA was performed both in the cue-locked period, to reveal potential differences in the modulation of the attentional focus, and in the target-locked period, to reveal potential cascade effects on stimulus processing. Cue-locked MVPA revealed that while in the TD group the pattern of neural activity contained information about the cue mainly before the target appearance, the ASD group showed a temporally sustained and topographically diffuse significant decoding of the cue neural response even after the target onset, suggesting a delayed extinction of cue-related neural activity. Crucially, this delayed extinction positively correlated with behavioral measures of attentional hyperfocusing. Results of target-locked MVPA were coherent with a hyper-focused attentional profile, highlighting an earlier and stronger decoding of target neural responses in small cue trials in the ASD group. The present findings document a spatially and temporally overrepresented encoding of visual information in ASD, which can constitute one of the main reasons behind their peculiar cognitive style.

Leveraging Unlabeled Electroencephalographic Data to Predict Neurological Recovery for Comatose Patients Following Cardiac Arrest.

In response to the 2023 George B. Moody PhysioNet Challenge, we propose an automated, unsupervised pre-training approach to boost the performance of models that predict neurologic outcomes after cardiac arrest. Our team, (BrownBAI), developed a model architecture consisting of three parts: a pre-processor to convert raw electroencephalograms (EEGs) into two-dimensional spectrograms, a three-layer convolutional neural network (CNN) encoder for unsupervised pre-training, and a time series transformer (TST) model. We trained the CNN encoder on unlabeled five-minute EEG samples from the Temple University EEG Corpus (TUEG), which included more than 20x the patients available in the PhysioNet competition training dataset. We then incorporated the pre-trained encoder into the TST as a base layer and trained the composite model as a classifier on EEGs from the 2023 PhysioNet Challenge dataset. Our team was not able to submit an official competition entry and was therefore not scored on the test set. However, in a side-by-side comparison on the competition training dataset, our model performed better with a pretrained (competition score 0.351), rather than randomly initialized (competition score 0.211) CNN encoder layer. These results show the potential benefits of leveraging unlabeled data to boost task-specific performance of predictive EEG models.

Deep Learning Approaches for Electroencephalography (EEG)-Based User Response Prediction

In deep learning, finding the best algorithms for time series data can be challenging due to its stochastic and nonlinear nature. This study endeavours to address the challenges posed by a 10-class classification and binary classification problem employing deep learning algorithms. We collected Electroencephalography (EEG) data from participants engaged in the Corsi Block Tapping Task, utilizing various combinations of Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and Bidirectional LSTM (BiLSTM) models across multiple layers to achieve the highest accuracy across different frequency bands - namely, beta band (14-30 Hz), alpha band (8-13 Hz), theta band (4-8 Hz), and delta band (0.5-4.35 Hz). Our findings in the context of the 10-class classification problem highlight the superior performance of the 1 CNN + 4 linear layers model, boasting an accuracy of 64.47%. In the realm of binary classification, the 1 LSTM + 4 linear layers model emerged as the top performer, achieving an impressive accuracy of 93.30%. Notably, the beta wave exhibited enhanced predictive capabilities. These results hold promising implications for the design of brain-computer interface experiments, where specific brain regions can predict responses with heightened accuracy. Furthermore, future applications may encompass the development of cognitive systems where both time and accuracy play pivotal roles.

Solving large-scale MEG/EEG source localisation and functional connectivity problems simultaneously using state-space models

State-space models are widely employed across various research disciplines to study unobserved dynamics. Conventional estimation techniques, such as Kalman filtering and expectation maximisation, offer valuable insights but incur high computational costs in large-scale analyses. Sparse inverse covariance estimators can mitigate these costs, but at the expense of a trade-off between enforced sparsity and increased estimation bias, necessitating careful assessment in low signal-to-noise ratio (SNR) situations. To address these challenges, we propose a three-fold solution: (1) Introducing multiple penalised state-space (MPSS) models that leverage data-driven regularisation; (2) Developing novel algorithms derived from backpropagation, gradient descent, and alternating least squares to solve MPSS models; (3) Presenting a K-fold cross-validation extension for evaluating regularisation parameters. We validate this MPSS regularisation framework through lower and more complex simulations under varying SNR conditions, including a large-scale synthetic magneto- and electro-encephalography (MEG/EEG) data analysis. In addition, we apply MPSS models to concurrently solve brain source localisation and functional connectivity problems for real event-related MEG/EEG data, encompassing thousands of sources on the cortical surface. The proposed methodology overcomes the limitations of existing approaches, such as constraints to small-scale and region-of-interest analyses. Thus, it may enable a more accurate and detailed exploration of cognitive brain functions.

BrainGridNet: A two-branch depthwise CNN for decoding EEG-based multi-class motor imagery

Brain–computer interfaces (BCIs) based on motor imagery (MI) enable the disabled to interact with the world through brain signals. To meet demands of real-time, stable, and diverse interactions, it is crucial to develop lightweight networks that can accurately and reliably decode multi-class MI tasks. In this paper, we introduce BrainGridNet, a convolutional neural network (CNN) framework that integrates two intersecting depthwise CNN branches with 3D electroencephalography (EEG) data to decode a five-class MI task. The BrainGridNet attains competitive results in both the time and frequency domains, with superior performance in the frequency domain. As a result, an accuracy of 80.26 percent and a kappa value of 0.753 are achieved by BrainGridNet, surpassing the state-of-the-art (SOTA) model. Additionally, BrainGridNet shows optimal computational efficiency, excels in decoding the most challenging subject, and maintains robust accuracy despite the random loss of 16 electrode signals. Finally, the visualizations demonstrate that BrainGridNet learns discriminative features and identifies critical brain regions and frequency bands corresponding to each MI class. The convergence of BrainGridNet’s strong feature extraction capability, high decoding accuracy, steady decoding efficacy, and low computational costs renders it an appealing choice for facilitating the development of BCIs.

The effect of acupuncture at the Taiyang acupoint on visual function and EEG microstates in myopia.

Acupuncture has certain effects to improve myopia visual function, but its neural mechanism is unclear. In this study, we acupunctured at the right Taiyang acupoint of myopic patients to analyze the effects of acupuncture on visual function and electroencephalographic activity and to investigate the correlation between improvements in visual function and changes in the brain. In this study, a total of 21 myopic patients were recruited. The contrast sensitivity (CS) of the subjects was examined before and after acupuncture, and electroencephalography (EEG) data of the entire acupuncture process were recorded. The study found that compared with before acupuncture, the CS of both eyes in myopic patients at each spatial frequency was increased after acupuncture; compared with the resting state, the contribution of microstate C was decreased during the post-acupuncture state, and the transition probability between microstate A and microstate C was reduced; in addition, the contribution of microstate C was negatively correlated with CS at both 12 and 18 cpd. The contrast sensitivity of myopic patients was improved after acupuncture at the Taiyang acupoint (20 min), which may be related to microstate C.

The emergence of task-relevant representations in a nonlinear decision-making task

This paper describes the relationship between performance in a decision-making task and the emergence of task-relevant representations. Participants learnt two tasks in which the appropriate response depended on multiple relevant stimuli and the underlying stimulus-outcome associations were governed by a latent feature that participants could discover. We divided participants into good and bad performers based on their overall classification rate and computed behavioural accuracy for each feature value. We found that participants with better performance had a better representation of the latent feature space. We then used representation similarity analysis on Electroencephalographic (EEG) data to identify when these representations emerge. We were able to decode task-relevant representations in a time window emerging 700 ms after stimulus presentation, but only for participants with good task performance. Our findings suggest that, in order to make good decisions, it is necessary to create and extract a low-dimensional representation of the task at hand.

Decoding the content of working memory in school-aged children

Developmental improvements in working memory (WM) maintenance predict many real-world outcomes, including educational attainment. It is thus critical to understand which WM mechanisms support these behavioral improvements, and how WM maintenance strategies might change through development. One challenge is that specific WM neural mechanisms cannot easily be measured behaviorally, especially in a child population. However, new multivariate decoding techniques have been designed, primarily in adult populations, that can sensitively decode the contents of WM. The goal of this study was to deploy multivariate decoding techniques known to decode memory representations in adults to decode the contents of WM in children. We created a simple computerized WM game for children, in which children maintained different categories of information (visual, spatial or verbal). We collected electroencephalography (EEG) data from 20 children (7–12-year-olds) while they played the game. Using Multivariate Pattern Analysis (MVPA) on children's EEG signals, we reliably decoded the category of the maintained information during the sensory and maintenance period. Across exploratory reliability and validity analyses, we examined the robustness of these results when trained on less data, and how these patterns generalized within individuals throughout the testing session. Furthermore, these results matched theory-based predictions of WM across individuals and across ages. Our proof-of-concept study proposes a direct and age-appropriate potential alternative to exclusively behavioral WM maintenance measures in children. Our study demonstrates the utility of MVPA to measure and track the uninstructed representational content of children's WM. Future research could use our technique to investigate children's WM maintenance and strategies.

40-Hz auditory steady-state response deficits are correlated with the severity of persistent auditory verbal hallucination in patients with schizophrenia

BackgroundAbnormal 40 Hz auditory steady-state response (ASSR) has been observed in some psychiatric disorders. Nevertheless, the role of 40 Hz ASSR in persistent auditory verbal hallucinations (pAVHs) schizophrenia (SCZ) is still unknown. This study aims to investigate whether the 40 Hz ASSR impairment is related to pAVHs and can detect pAVHs severity. MethodsWe analyzed high-density electroencephalography data that from 43 pAVHs patients (pAVH group), 20 moderate auditory verbal hallucinations patients (mid-AVH group), and 24 without auditory verbal hallucinations patients (non-AVH group). Event-related spectral perturbation and inter-trial phase coherence (ITPC) were calculated to quantify dynamic changes of the 40 Hz ASSR power and ITPC, respectively. ResultsFrontal-central, the 40 Hz ASSR power, and ITPC were significantly lower in the pAVH group than in the non-AVH group; There was no significant difference between the pAVH and mid-AVH group. The 40 Hz ASSR was significantly negatively correlated with the severity of pAVHs. The 40 Hz ASSR power, and ITPC could be used as a combinational marker to detect SCZ patients with and without pAVHs. ConclusionOur findings have shed light on the pathological mechanism of pAVHs in SCZ patients. These results can provide potential avenues for therapeutic intervention of pAVHs.

Analysis of Electroencephalography (EEG) Electrode-Pair Correlations

We measured electroencephalography (EEG) data streams from participants wearing a wireless EEG headset in two modes: eyes open and eyes closed. Then we analyzed the data by computing the correlation coefficients for a pair of electrodes in each measurement mode. We also plotted and visually inspected the associated scatter plots. We observed that for the electrodes selected, the signals were more strongly correlated in the eyes closed mode and relatively weakly correlated in the eyes open mode. In most measurements, the signals were dissimilar. These observations could be harnessed to inform expedient placement of EEG electrodes and efficient selection of data stream channels for further analysis.

Impaired proactive cognitive control in Parkinson's disease.

Adaptive control has been studied in Parkinson's disease mainly in the context of proactive control and with mixed results. We compared reactive- and proactive control in 30 participants with Parkinson's disease to 30 age matched healthy control participants. The electroencephalographic activity of the participants was recorded over 128 channels while they performed a numerical Stroop task, in which we controlled for confounding stimulus-response learning. We assessed effects of reactive- and proactive control on reaction time-, accuracy- and electroencephalographic time-frequency data. Behavioural results show distinct impairments of proactive- and reactive control in participants with Parkinson's disease, when tested on their usual medication. Compared to healthy control participants, participants with Parkinson's disease were impaired in their ability to adapt cognitive control proactively and were less effective to resolve conflict using reactive control. Successful reactive and proactive control in the healthy control group was accompanied by a reduced conflict effect between congruent and incongruent items in midline-frontal theta power. Our findings provide evidence for a general impairment of proactive control in Parkinson's disease and highlight the importance of controlling for the effects of S-R learning when studying adaptive control. Evidence concerning reactive control was inconclusive, but we found that participants with Parkinson's disease were less effective than healthy control participants in resolving conflict during the reactive control task.

61 Robustness of Attention Networks Across Multiple Sessions: Behavioral and ERP Findings

Objective:Attention is the backbone of cognitive systems and is requisite for many cognitive processes vital to everyday functioning, including memory, problem solving, and the cognitive control of behavior. Attention is commonly impaired following traumatic brain injury and is a critical focus of rehabilitation efforts. The development of reliable methods to assess rehabilitation-related changes are paramount. The Attention Network Test (ANT) has been used previously to identify 3 independent, yet interactive attention networks—alerting, orienting, and executive control (EC). We examined the behavioral and neurophysiological robustness and temporal stability of these networks across multiple sessions to assess the ANT’s potential utility as an effective measure of change during attention rehabilitative interventions.Participants and Methods:15 healthy young adults completed 4 sessions of the ANT (1 session/7-day period). ANT networks were assessed within the task by contrasting opposing stimulus conditions: cued vs. non-cued trials probed alerting, valid vs. invalid spatial cues probed orienting, and congruent vs. incongruent targets probed EC. Differences in median correct-trial reaction times (RTs) and error rates (ERs) between the condition pairs were assessed to determine attention network scores; robustness of networks effects, as determined by one-sample t-tests at each session, against a mean of 0, determining the presence of significant network effects at each session. Sixty-four-channel electroencephalography (EEG) data were acquired concurrently and processed using Matlab to create condition-related event-related potentials (ERPs)—particularly the cue- and probe-related P1, N1, and P3 deflection amplitudes, measured by using signed-area calculation in regions of interest (ROIs) determined by observation of spherical-spline voltages. This enabled us to examine the robustness of cue- and probe-attention-network ERPs.Results:All three attention networks showed robust effects. However, only the EC RT and ER network scores remained significantly robust [t(14)s&gt;13.9,ps&lt;.001] across all sessions, indicating that EC is robust in the face of repeated exposure. Session 1 showed the greatest EC-RT robustness effect which became smaller during the subsequent sessions per ANOVAs on Session x Congruency [F(3,42)=10.21,p&lt;.0001], reflecting persistence despite practice effects. RT robustness of the other networks varied across sessions. Alerting and EC ERs were similarly robust across all 4 sessions, but were more variable for the orienting network. ERP results: The cue-locked P1-orienting (valid vs. invalid) was generally larger to valid- than invalid-cues, but the robustness across sessions was variable (significant in only sessions 1 and 4 [t(14)s&gt;2.13,ps&lt;.04], as reflected in a significant main effect of session [p=.0042]. Next, target-locked EC P3s were generally smaller to congruent than incongruent targets [F(1,14)=9.40,p=.0084], showing robust effects only in sessions 3 and 4 [ps&lt;.005].Conclusions:The EC network RT and ER scores were consistently robust across all sessions, suggesting that this network may be less vulnerable to practice effects across session than the other networks and may be the most reliable probe of attentional rehabilitation. ERP measures were more variable across attention networks with respect to robustness. Behavioral measures of EC-network may be most reliable for assessing progress related to attentional-rehabilitation efforts.