P300 response modulation reflects breaches of non-probabilistic expectations

Optimization of transfer learning based on source sample selection in Euclidean space for P300-based brain-computer interfaces

Introduction

Event-related potentials (ERPs), such as P300, are widely utilized for non-invasive monitoring of brain activity in brain-computer interfaces (BCIs) via electroencephalogram (EEG). However, the non-stationary nature of EEG signals and different data distributions among subjects create significant challenges for implementing real-time P300-based BCIs. This requires time-consuming calibration and a large number of training samples.

Methods

To address these challenges, this study proposes a transfer learning-based approach that uses a convolutional neural network for high-level feature extraction, followed by Euclidean space data alignment to ensure similar distributions of extracted features. Furthermore, a source selection technique based on the Euclidean distance metric was applied to measure the distance between each source feature sample and a reference point from the target domain. The samples with the lowest distance were then chosen to increase the similarity between source and target datasets. Finally, the transferred features are applied to a discriminative restricted Boltzmann machine classifier for P300 detection.

Results

The proposed method was evaluated on the state-of-the-art BCI Competition III dataset II and rapid serial visual presentation dataset. The results demonstrate that the proposed technique achieves an average accuracy of 97% for both online and offline after 15 repetitions, which is comparable to the state-of-the-art methods. Notably, the proposed approach requires <½ of the training samples needed by previous studies.

Discussion

Therefore, this technique offers an efficient solution for developing ERP-based BCIs with robust performance against reduced a number of training data.

Multisensory integration deficits in Schizophrenia and Autism evidenced in behaviour but not event related potentials

The process of integrating information from different sensory channels, known as multisensory integration (MSI) was assessed in two disorders, Autism Spectrum Disorder (ASD) and Schizophrenia (SCZ). 32 healthy controls (HC), 35 SCZ patients, and 23 ASD patients performed an audiovisual (AV) synchronous target detection task while reaction time (RT) and scalp recorded electrophysiological (EEG) activity were measured. MSI in the AV condition resulted in faster and less variable RTs compared to the unimodal conditions. Using our novel bootstrap method, MSI gain was observed in 78 % of HC, 26 % of ASD, and 48 % of SCZ patients. At the neural level, MSI in the AV condition resulted in larger amplitude of sensory evoked responses and cognitive P3 response compared to the corresponding unimodal conditions. These neural effects of MSI were not related to the behavioural MSI gain identified at the individual level and could not explain the deficits in behavioural MSI of patient groups. In conclusion, a robust MSI gain deficit in RT was observed in both patient groups that was not reflected in early perceptual and cognitive electro-cortical responses, suggesting that behavioural MSI deficits in ASD and SCZ may arise at late processing stages such as response selection.

Event Probabilities Have a Different Impact on Early and Late Electroencephalographic Measures Regarded as Metrics of Prediction

The oddball protocol has been used to study the neural and perceptual consequences of implicit predictions in the human brain. The protocol involves presenting a sequence of identical repeated events that are eventually broken by a novel "oddball" presentation. Oddball presentations have been linked to increased neural responding and to an exaggeration of perceived duration relative to repeated events. Because the number of repeated events in such protocols is circumscribed, as more repeats are encountered, the conditional probability of a further repeat decreases-whereas the conditional probability of an oddball increases. These facts have not been appreciated in many analyses of oddballs; repeats and oddballs have rather been treated as binary event categories. Here, we show that the human brain is sensitive to conditional event probabilities in an active, visual oddball paradigm. P300 responses (a relatively late component of visually evoked potentials measured with EEG) tended to be greater for less likely oddballs and repeats. By contrast, P1 responses (an earlier component) increased for repeats as a goal-relevant target presentation neared, but this effect occurred even when repeat probabilities were held constant, and oddball P1 responses were invariant. We also found that later, more likely oddballs seemed to last longer, and this effect was largely independent of the number of preceding repeats. These findings speak against a repetition suppression account of the temporal oddball effect. Overall, our data highlight an impact of event probability on later, rather than earlier, electroencephalographic measures previously related to predictive processes-and the importance of considering conditional probabilities in sequential presentation paradigms.

Epigenetic Effects in HPA Axis Genes Associated with Cortical Thickness, ERP Components and SUD Outcome

Association between familial loading for alcohol use disorders (AUD) and event-related potentials (ERPs) suggests a genetic basis for these oscillations though much less is known about epigenetic pathways influenced by environmental variation. Early life adversity (ELA) influences negative outcomes much later in life. The stress-activated neuropeptide corticotropin-releasing hormone (CRH) contributes to the deleterious effects of ELA on brain structure and function in animals. Accordingly, we hypothesized that ELA would be related to cortical thickness and electrophysiological characteristics through an epigenetic effect on CRH receptor type-1 (CRHR1) methylation. A total of 217 adolescent and young adult participants from either multiplex alcohol dependence or control families were scanned using magnetic resonance imaging (MRI) at 3T and cortical thickness was determined. Longitudinal follow-up across childhood, adolescence, and young adulthood provided developmental ERP data and measures of adversity. Blood samples for genetic and epigenetic analyses were obtained in childhood. Cortical thickness and visual ERP components were analyzed for their association and tested for familial risk group differences. Visual P300 amplitude at Pz and cortical thickness of the left lateral orbitofrontal region (LOFC), were significantly related to risk group status. LOFC cortical thickness showed a negative correlation with CRHR1 methylation status and with childhood total stress scores from the Life Stressors and Social Resources Inventory (LISRES). Stress scores were also significantly related to P300 amplitude recorded in childhood. The present results suggest that early life adversity reflected in greater total LISRES stress scores in childhood can impact the methylation of the CRHR1 gene with implications for brain development as seen in cortical thickness and electrophysiological signals emanating from particular brain regions.

Impaired selective attention in patients with severe primary monosymptomatic nocturnal enuresis: An event-related potential study

Objectives

Primary monosymptomatic nocturnal enuresis (PMNE) is a very common problem in school age children. It is thought that PMNE represents a maturational lag in the central nervous system of those children. We did this case control study to assess the selective attention and resource allocation in those children using the P300 wave of the Event-Related Potentials (ERPs) and its relation to disease severity.

Methods

Forty four patients with PMNE and twenty three healthy controls were included in this study. Patients were diagnosed according to the criteria of international children's continence society and were classified into two groups; patients with frequent wetting (≥4 episodes/week), and patients with infrequent wetting (<4 episodes/week). ERPs were recorded at Fz, Cz, and Pz locations using odd-ball paradigm. N200 and P300 peak latencies (ms), and N200/P300 peak to peak amplitudes (µV) were measured.

Results

We found significant increase of P300 and N200/P300 interpeak latencies, and significant decrease of P300 amplitudes in frequent wetting group "severe" PMNE compared to healthy controls and infrequent wetting group.

Conclusion

Abnormal selective attention and resource allocation were found in patients with severe PMNE. Measures to improve selective attention might be helpful in treatment of patients with severe PMNE.

Significance

Impaired selective attention might play a role in pathogenesis of severe PMNE and the need for the various measures to improve selective attention may be further studied as a therapeutic tool for patients with severe PMNE.

Neural Encoding of the Reliability of Directional Information During the Preparation of Targeted Movements

Visual information about the location of an upcoming target can be used to prepare an appropriate motor response and reduce its reaction time. Here, we investigated the brain mechanisms associated with the reliability of directional information used for motor preparation. We recorded brain activity using magnetoencephalography (MEG) during a delayed reaching task in which a visual cue provided valid information about the location of the upcoming target with 50, 75, or 100% reliability. We found that reaction time increased as cue reliability decreased and that trials with invalid cues had longer reaction times than trials with valid cues. MEG channel analysis showed that during the late cue period the power of the beta-band from left mid-anterior channels, contralateral to the responding hand, correlated with the reliability of the cue. This effect was source localized over a large motor-related cortical and subcortical network. In addition, during invalid-cue trials there was a phasic increase of theta-band power following target onset from left posterior channels, localized to the left occipito-parietal cortex. Furthermore, the theta-beta cross-frequency coupling between left mid-occipital and motor cortex transiently increased before responses to invalid-cue trials. In conclusion, beta-band power in motor-related areas reflected the reliability of directional information used during motor preparation, whereas phasic theta-band activity may have signaled whether the target was at the expected location or not. These results elucidate mechanisms of interaction between attentional and motor processes.

Neural Processing of Cognitive Control in an Emotionally Neutral Context in Anxiety Patients

Impaired cognitive control plays a crucial role in anxiety disorders and is associated with deficient neural mechanisms in the fronto-parietal network. Usually, these deficits were found in tasks with an emotional context. The present study aimed at investigating electrophysiological and vascular signatures from event-related brain potentials (ERPs) and functional near-infrared spectroscopy (fNIRS) in anxiety patients versus healthy controls during an inhibition task integrated in an emotionally neutral context. Neural markers were acquired during the completion of a classical Eriksen flanker task. The focus of data analysis has been the ERPs N200 and P300 and fNIRS activations in addition to task performance. No behavioral or neural group differences were identified. ERP findings showed a larger N2pc and a delayed and reduced P300 for incongruent stimuli. The N2pc modulation suggests the reorienting of attention to salient stimuli, while the P300 indicates longer lasting stimulus evaluation processes due to increased task difficulty. FNIRS did not result in any significant activation potentially suggesting a contribution from deeper brain areas not measurable with fNIRS. The missing group difference in our non-emotional task indicates that no generalized cognitive control deficit but rather a more emotionally driven deficit is present in anxiety patients.