Background/Objectives: The Human Performance Envelope (HPE) is a multidimen-sional model that represents the range in which an individual operator's performance is acceptable or begins to become dangerous. Although several alternative models have been proposed, HPE currently remains primarily a theoretical concept. The goal of the study was therefore to translate this theoretical concept into practical applications, seeking to characterise and measure how HPE manifests itself in real-world contexts. Methods: Multivariate Autoregressive Models (MVAR) have been used in the analysis of complex systems in which variables are interdependent and mutually influence their dynamics over time. Professional Air Traffic Controllers (ATCOs) were involved in the study and asked to deal with realistic traffic scenarios while their behavioural, subjec-tive and neurophysiological data were collected. Partial Information Decomposition - Least Absolute Shrinkage and Selection Operator (PID – LASSO) model was then em-ployed to estimate the interactions among ATCO’s Human Factors (HFs) and identify the most appropriate characterisation of the HPE. Results: The results showed high and significant correlations among each ATCO’s performance and the corresponding neu-rophysiological – based HPE values. Furthermore, high-performance conditions (Best) were characterized by a significantly higher HPE values and a higher inter-HFs con-nections compared to low-performance (Worst) states. This suggested that a densely interconnected network of HFs is a prerequisite for operational resilience. Conclusions: The study provides the first application of a neurophysiological framework to model the causal interactions between HFs, translating the theoretical HPE into a quantifiable model validated against operator performance.

BACKGROUND/OBJECTIVES:The human performance envelope (HPE) is a multidimensional model that represents the range in which an individual operator's performance is acceptable or begins to become dangerous. Although several alternative models have been proposed, HPE currently remains primarily a theoretical concept. The goal of the study was therefore to translate this theoretical concept into practical applications, seeking to characterize and measure how HPE manifests itself in real-world contexts. METHODS:Multivariate Autoregressive (MVAR) models and conditional transfer entropy (cTE) have been used in the analysis of complex systems in which processes are interdependent and mutually influence their dynamics over time. Professional Air Traffic Controllers were involved in the study and asked to deal with realistic traffic scenarios while their behavioural, subjective and neurophysiological data were collected. MVAR-cTE models were then employed to estimate the interactions among controller human factors and to identify the most appropriate characterization of the HPE. RESULTS:The results showed high and significant correlations among each controller's performance and the corresponding neurophysiological-based HPE values. Furthermore, high-performance conditions (best) were characterized by significantly higher HPE values and higher inter-human factor connections compared to the low-performance (worst) status. This evidence suggested that a densely interconnected network of Human Factors is a prerequisite for operational resilience. CONCLUSIONS:The study provided the first application of a neurophysiological framework to model the directed interactions between human factors, translating the theoretical HPE into a quantifiable model validated against operator performance.

Machine learning, particularly deep learning, typically achieves high facial emotion image recognition accuracy benefiting from multiple labeled data. However, the datasets usually contain insufficient labeled samples and numerous unlabeled data since human labeling is a costly endeavor. For semi-supervised learning of these datasets, self-training procedure solely based on the visual features of images fails to comprehensively understand the intricate high-level semantic features. Since EEG signals contain not only visual information related to the visual stimulus but also emotional information related to brain activity, they are highly suitable as supervisory signals for labeling unlabeled facial emotion images. In this study, we specifically employ EEG signals evoked by visual image stimuli in conjunction with EEGNet3D to learn a discriminative EEG class representation manifold of brain activity. The one-hot class label is replaced with the EEG class representation as the supervisory to train the base model. Then, better pseudo-labeling is achieved using the base model in the EEG class representation manifold. Based on pseudo-labeling results, the utilization of unlabeled data is further improved. Interestingly, our findings reveal that when utilizing EEG class representations as supervisory information for the base model, the base model demonstrates a learning pattern that involves focusing more on the eye area when making judgments about emotions. This behavior closely resembles how the human brain decodes emotions. Experiments show that the performance of the proposed method can be effectively enhanced by combining labeled and pseudo-labeled images. Further experiments demonstrate that our method exhibits strong generalization abilities when applied to new image datasets and other visual networks.

Featured Application The proposed multimodal framework can be applied to the objective evaluation and optimization of e-learning authoring platforms, supporting the identification of usability bottlenecks that are not detectable through traditional questionnaires alone. By combining neurophysiological, eye tracking, and interaction-based metrics, the approach enables designers and developers to perform data-driven, user-centred improvements of complex web interfaces. This methodology can be adopted in iterative UX design processes to enhance platform usability, reduce cognitive load during content creation, and ultimately facilitate the adoption of digital educational technologies by instructors and non-expert users.Abstract Background: Digital learning platforms increasingly leverage semantic web technologies to support interoperable and adaptive e-learning. However, the usability and cognitive impact of web-based authoring tools are still mainly assessed through subjective questionnaires and interaction logs, which provide limited time resolution and weak diagnostic power for identifying specific interface bottlenecks. Methods: We propose a multimodal evaluation of SOULSS, a semantic web-oriented platform for creating and optimizing digital learning contents. Eighteen participants completed an authoring workflow organized into three macro-segments (tutorial, initialization, module creation) while wearable electroencephalography, electrodermal activity, photoplethysmography, and eye tracking were recorded; objective metrics were analyzed both across macro-segments and within predefined micro-activities, whereas subjective engagement was collected after each macro-segment using the UES-SF. Results: Objective measures indicated increased EEG-derived mental workload and stress, higher tonic sympathetic arousal, and greater visual search and interaction effort during initialization and module creation, while UES-SF scores were lower during initialization. Fine-grained analyses localized critical elements to tutorial navigation options, the new course entry point, and spoiler-related controls. Repeated-measures correlations linked subjective scores with objective markers and supported an association between stress-related activation and delayed visual discovery. Conclusions: Integrating neurophysiological and eye tracking measures enables a more diagnostic assessment of semantic web-based authoring platforms than questionnaires alone, providing actionable evidence for iterative UX optimization and supporting a more user-centred design of digital educational tools.

Recent advancements in wearable, non-invasive neurophysiological sensors have increased interest in applying Human factors (HF) research beyond controlled laboratory settings. HF research aims to objectively quantify individuals' (alone or working in team) mental and emotional states to ensure safety, maintain good performances and prevent risks. These devices enable real time, unobtrusive monitoring of individuals in real-world environments, overcoming the limitations of traditional subjective assessments. Unlike self-reports, neurophysiological signals provide objective, real-time data, allowing for a more accurate and continuous understanding of mental states. This review examines the latest advancements over the past ten years in real-time monitoring of the most studied and operationally relevant mental states, including mental workload (MW), stress, attention, fatigue, drowsiness, and teamwork, by analyzing studies that contribute to research advancing toward real-world application direction. While wearable biosensors were used as one criterion for selecting studies, this review extends beyond device usage to explore crucial methodological aspects relevant to real-world applications, such as data quality, customized processing steps, and the robustness of results compared to traditional laboratory-grade devices. Thus, this review analyzes 123 articles exploring current advancements in real-time monitoring of MW, stress, attention, mental fatigue, drowsiness, and teamwork using wearable devices. The findings provide a comprehensive overview of methodology advancement toward real-time, objective individuals monitoring in real world settings.

Electroencephalography (EEG) and functional near-infrared spectroscopy (fNIRS) exhibit complementary advantages in temporal and spatial resolution for brain activity monitoring, and their integration has the potential to improve motor imagery (MI) decoding performance. However, in EEG–fNIRS multimodal MI decoding, modality heterogeneity and differences in signal-to-noise ratio and convergence speed often cause joint training to bias toward the modality that is easier to learn, leading to modality dominance, degraded fused representations, and reduced generalization. To address this issue, we propose a dynamic re-initialization framework for EEG–fNIRS multimodal decoding, which achieves cross-modal balanced learning through a diagnosis–adjustment–re-initialization mechanism. The proposed method uses the separability difference of unimodal features between the training and validation sets as a diagnostic signal, and integrates network hierarchical priors with modality-specific gradient statistics to derive adjustment factors. Guided by these signals, the framework periodically performs soft re-initialization of the EEG and fNIRS encoder parameters during training, promoting re-learning of weaker modalities, suppressing single-modality dominance, and preserving the stability of converged representations. As a result, multimodal imbalance is alleviated and the discriminative capability of fused representations is enhanced. Experiments on a publicly available EEG–fNIRS motor imagery dataset demonstrate that the proposed method achieves an average classification accuracy of 92.42 ± 4.54

PurposeThe surgical operating room is a high-stakes environment where stress can impact performance and patient safety. While hormonal and neurophysiological markers are established stress indicators, integrative studies in real-world surgical settings are scarce. This study aimed to provide a comprehensive, multimodal characterization of stress in surgical teams during live operations, comparing neurophysiological, biological, and behavioral responses across different levels of expertise and surgical phases. The goal was to validate a multi-method approach and identify objective markers for monitoring stress in real-time.MethodSurgical teams, each composed of four members, were categorized as “Expert” or “Novice” based on the lead surgeon's experience. All teams performed a standardized inguinal hernia repair. Continuous electroencephalography (EEG) and electrodermal activity (EDA) were recorded throughout the procedure to derive stress indices. Blood samples were collected pre- and post-surgery to measure Adrenocorticotropic Hormone (ACTH) and cortisol levels. Subjective stress was assessed via questionnaires, and team performance was quantified using a Combined Behavioral Teamwork Index (CBTI) based on surgical time, materials used, and patient outcomes.FindingNeurophysiological data showed that the EEG-based stress index was significantly higher in Novice surgeons compared to Experts, particularly during the final and most demanding phase of the surgery (p = 0.008). This effect was most pronounced for the lead Novice surgeon (p = 0.01). Similarly, the EDA-based stress index was higher overall in Novices (p = 0.02). Post-surgery, ACTH levels increased significantly in Novices while decreasing in Experts (p = 0.008), indicating a sustained endocrine stress response in the less experienced group. Strong positive correlations were found between the EEG-stress index and both ACTH levels (R = 0.67) and subjective stress (R = 0.63), validating the multimodal assessment.ConclusionThis study demonstrates that a multimodal approach can effectively characterize stress dynamics in a real-world surgical environment. The EEG-derived metric emerged as the most sensitive indicator, capable of discriminating stress levels with high temporal and role-specific precision. Novice surgeons exhibit significantly greater neurophysiological and endocrine stress responses, underscoring the need for targeted support and advanced training protocols. These findings lay the groundwork for developing real-time, objective stress monitoring systems to enhance surgical performance, training, and patient safety.

The emotional characterization is a crucial application in the field of industrial neurosciences, where the objective is to objectively assess human experience through the employment of neurophysiological signals. In this scenario, the present study developed and validated a physiology-driven indicator of the user's emotional experience, namely, the emotional index, computed as the synthetic combination of the user's heart rate and the skin conductance level. The index was validated on a sample of 25 participants looking at a set of emotional videos taken from a previously validated affective database, thereby highlighting a substantial effect of valence and arousal labels in terms of heart rate and skin conductance level modulations. The results pointed out a significant effect of valence on heart rate values, chi(2)(4) = 31.14, p < .001, and an effect of arousal on the skin conductance level, F(24, 2.35) = 4.95, p = .007. Additionally, an effect of the stimuli emotional content was also found on values of the emotional index, chi(2)(4) = 29.57, p < .001. At the same time, the index demonstrated its capacity to overcome the limitations affecting a previous version of its definition while maintaining its correlation with it, r(s) (418) = .81, p < .001. In conclusion, the new computation of the emotional index was validated for the emotion characterization of video stimuli, and it was demonstrated to be more representative of the physiological response.

In Air Traffic Management (ATM), Air Traffic Control (ATC) is becoming increasingly challenging as controllers handle increasing volumes of air traffic while simultaneously adapting to novel technological systems. This surge in complexity places greater cognitive demands on Air Traffic Controllers (ATCOs), making their tasks more mentally demanding. This paper presents some research works coming from Single European Sky ATM Research (SESAR) projects and delivered by a large European Consortium over the past decade. Based on this work, this article seeks to present a neurophysiologically based approach to strengthening Human–Machine Teaming (HMT) in the frame of ATC/ATM by integrating neurophysiological data. By leveraging multi-modal signals, including Electroencephalography (EEG), Electrocardiography (ECG), Electrodermal Activity (EDA), Galvanic Skin Response (GSR), and eye-tracking, the framework can be based on the operator’s mental states. By targeting specific controller’s cognitive and affective states, real-time system adaptations can be used in order to decrease controller’s workload, stress and fatigue while enhancing situational awareness, vigilance, and trust. The five projects being discussed here (NINA, MOTO, ARTIMATION, TRUSTY and CODA) collectively demonstrate the growing relevance of the neurophysiology-based approach to developing the new ATC/ATM tools of the years to come.

Explainability is crucial for establishing user trust in Artificial Intelligence (AI), particularly within safety-critical domains such as Air Traffic Management (ATM) and Air Traffic Control (ATC). This study empirically investigates the effects of Explainable AI (XAI), specifically HeatMap-based visual explanations, on cognitive workload, user acceptance, and intention to use AI-driven decision-support systems among Air Traffic Control Officers (ATCOs). Despite significant theoretical advancements in the broader XAI domain, empirical evidence addressing the specific impact of visual explanations on human-AI interactions in safety-critical environments like ATC remains limited. To address these critical gaps, an experimental comparison was conducted between explainable (HeatMap) and non-explainable (BlackBox) AI conditions, involving two user groups: expert and student ATCOs. Both objective neurophysiological measures (Electroencephalography) and subjective questionnaires were employed to capture comprehensive user responses. Key findings revealed that the presence of visual explanations significantly reduced cognitive workload and enhanced users’ willingness to adopt the AI system, regardless of participants’ level of expertise. However, explicit perceptions of AI’s impact on work performance were predominantly influenced by expertise, with less experienced controllers reporting a greater perceived impact than their expert counterparts. By combining objective neurometrics with subjective user assessments, this research advances methodological rigor in evaluating human-AI interactions and highlights the importance of tailored, user-centric explanations. These findings directly contribute to practical guidelines for designing cognitively compatible and trustworthy AI tools in ATC, providing nuanced insights for targeted training and deployment strategies based on user expertise.

IntroductionThis study investigated the impact of chronic tinnitus on auditory perception, text comprehension, and physiological stress responses, with a focus on sex-related differences. The main objectives were to assess the influence of sex and stress on tinnitus severity, examine neurophysiological indicators of listening effort, and evaluate the effects of background noise on perceived difficulty and listening pleasantness.Materials and methodsForty-seven participants (24 with tinnitus, 23 controls) performed a listening task involving audiobook excerpts presented at different signal-to-noise ratios. Subjective ratings, comprehension scores, and physiological data were collected, including salivary alpha-amylase, electrodermal activity, heart rate, and EEG-based measures of listening pleasantness.ResultsControl participants outperformed tinnitus participants during the initial quiet condition (p = 0.020), with male controls scoring significantly higher than males with tinnitus (p = 0.008). Tinnitus participants rated listening as less pleasant in both quiet (p = 0.036) and high-noise conditions (p = 0.012). Female participants reported greater difficulty under moderate noise (p = 0.030), while EEG data showed higher enjoyment in males (p = 0.005). Salivary amylase increased post-task (p = 0.016), electrodermal activity differed between the initial and final quiet phases (p < 0.001), and heart rate varied according to noise levels (p = 0.008). Negative correlation emerged between subjective and EEG-based pleasantness in the quiet condition.DiscussionThese findings suggest that tinnitus imposes a measurable cognitive and emotional burden, influenced by both sex and stress responses. They emphasize the need for multimodal, personalized, and gender-sensitive approaches in the assessment and management of tinnitus.

This study addresses limitations in EEG-based stress detection research by developing a novel approach to differentiate multiple mental states in different stress baseline population samples. Utilizing EEG signals, graph convolutional neural networks (GCNs), and binaural beats stimulation (BBs), the research investigates stress detection and reduction in two population sample groups with distinct baselines (group 1: low daily baseline, and group 2: stressed daily baseline). The experiment comprises four phases: rest state, control alertness, stress induction, and stress mitigation. Mental states were assessed using behavioral data: reaction time to stimuli (RT) and target detection accuracy, subjective reports: Perceived Stress Scale scores (PSS-10), biochemical indicators: salivary cortisol levels, and neurophysiological measure: EEG effective connectivity via Partial Directed Coherence (PDC). BBs significantly improved target detection accuracy by 31.6% and 22.8% for low and high-stress groups, respectively. PDC connectivity showed a shift to the temporal region during mitigation, indicating a return to a more balanced state. GCN classification achieved accuracies of 76.43 +/- 9.01 % and 76.32 +/- 7.79 % for each group, and 76.37 +/- 8.40 % for a common baseline. While 16-Hz BBs enhanced focusing abilities they did not significantly reduce subjective stress scores. This study highlights the complex relationship between cognitive performance, perceived stress, and neurophysiological measures, emphasizing the need for multifaceted stress research and management approaches.

Person identification method based on electroencephalograms (EEG) signals, or so called brainprint recognition is a novel way to distinguish identities with advantages of high security. However, existing methods neglect the distribution difference between training and test data, and the large distance between projected features in the latent space makes the performance of the model degrade in the unseen domain data. In this paper, we propose channel aggregated based generalized contrastive learning framework, which combines multiple modules to overcome this challenge. To capture features from different granularities, we involve multi-scale convolution with channel attention block. In face of distribution of unseen domain, we introduce feature enhancement-based generalized contrast learning to improve the model generalization ability. In the generalized contrast learning module, taking the difficulty of reconstructing EEG signals into consideration, we augment the source domain data at the feature level to improve the generalization ability of the model on the unseen domain data. Extensive experiments on two multi-session datasets shows that our model outperformed other baseline methods, demonstrating its capability of better generalization performance to unseen domain.

Background/Objectives: Since high frequencies are susceptible to disruption in various types of hearing loss, a symptom which is common in people with tinnitus, the aim of the study was to investigate EEG cortical auditory evoked and P300 responses to both a high- and low frequency-centered oddball paradigm to begin to establish the most suitable cognitive physiologic testing conditions for those with both unimpaired hearing and those with hearing impairments. Methods: Cortical auditory evoked potential (CAEP) P1, N1, P2 and P300 (subtraction wave) peaks were identified in response to high- (standard: 6000 Hz, deviant: 8000 Hz) and low frequency (Standard: 375 Hz, Deviant: 500 Hz) oddball paradigms. Each paradigm was presented at various intensity levels. Latencies and amplitudes were then computed for each condition to assess the effects of frequency and intensity. Results: Stimulus intensity had no effect on either the high- or low frequency paradigms of P300 characteristics. In contrast, for the low frequency paradigm, intensity influenced the N1 latency and P2 amplitude, while for the high frequency paradigm intensity influenced P1 and P2 latency and P2 amplitude. Conclusions: Obligatory CAEP components responded more readily to stimulus frequency and intensity changes, and one possible consideration is that higher frequencies could play a role in the response characteristics exhibited by N1 (except for N1 amplitude) and P2, given their involvement in attentional processes linked to the detection of warning cues. P300 latency and amplitude were not influenced by such factors. These findings support the hypothesis that disentangling the cognitive from the more sensory-based response is possible, even in those with hearing loss, provided that the patient’s hearing loss is considered when determining the presentation level. While the present study was performed in participants with unimpaired hearing, these data set up future studies investigating the effectiveness of using similar methods in hearing-impaired persons.

Ocular artifacts, particularly blinks, significantly affect the integrity of electroencephalographic (EEG) signals, posing a challenge for real-time applications. Traditional correction methods often require a calibration phase or additional electrooculogram (EOG) channels, limiting their applicability in mobile and real-world settings. This study presents a novel detection and correction method, designed for online ocular artifact correction without the need for prior calibration: the CFo-CLEAN. The proposed method integrates an Enhanced Adaptive Data-driven Algorithm (eADA) for real-time identification and correction of ocular artifacts directly from EEG signals. Unlike conventional approaches, this implementation adapts dynamically to ongoing EEG variations, enhancing flexibility and performance. The study evaluates the CFo-CLEAN method using EEG data recorded from 38 participants during real-world driving scenarios. Performance comparisons were conducted against established correction techniques, including Independent Component Analysis (ICA), regression-based methods, and subspace reconstruction approaches. The evaluation considered both artifact removal efficiency and EEG signal preservation across different experimental conditions. Results demonstrated that the method effectively reduced ocular artifact contamination while preserving neurophysiological content. Specifically, two implementations of the method, utilizing 60-second and 90-second time windows, were analyzed, revealing that longer windows provided superior EEG signal preservation, particularly in higher frequency bands. These findings validate the effectiveness of the CFo-CLEAN method for real-time applications, making it a valuable tool for brain-computer interfaces (BCIs), neuroergonomics, and cognitive state monitoring. By avoiding the need for a calibration phase and incorporating adaptive processing, this method represents a significant advancement in real-time EEG artifact correction, facilitating its deployment in dynamic, real-world environments.

IntroductionFatigue is a major factor contributing to road accidents, and extensive research has focused on its physiological and behavioral characterization. Due to safety and economic constraints, studies on driving fatigue are commonly conducted in simulated environments, where fatigue is typically induced through prolonged tasks and assessed using a Time-on-Task (ToT) approach. However, ToT-based labeling may not accurately reflect individual variations in fatigue onset.MethodsThis study compared fatigue onset in matched simulated and real driving conditions by evaluating two labeling approaches: the traditional ToT-driven method and a novel physiology-driven method based on electroencephalographic (EEG) parameters. Experimental periods of Low and High Fatigue were defined using both approaches, and physiological and behavioral responses were analyzed through ocular and cardiac activity.ResultsWhen using the ToT-driven approach, no significant differences emerged between low and high fatigue periods across the two environments. In contrast, the EEG-driven labeling revealed clear physiological responses to fatigue onset, as evidenced by changes in ocular and heart activity.DiscussionThe findings demonstrate that the method used to define fatigue substantially influences the detection of fatigue onset. The results highlight the importance of physiology-based labeling for capturing individual fatigue dynamics and provide novel insights into how fatigue manifests differently in simulated and real driving contexts.

This pilot study investigates the impact of transcranial alternating current stimulation (tACS) on psychological stress using functional near-infrared spectroscopy (fNIRS). Forty volunteers were randomly assigned to two groups: the tACS and the control. The experiment was divided into three distinct stages: pre-stimulation, stimulation, and post-stimulation. The Stroop Color-Word Task (SCWT) was employed as a validated stress-inducing paradigm to assess pre- and post-stimulation changes. During the initial phase, the participants completed the SCWT. This was followed by either tACS or sham. In the third session, the individuals solved the task again. The anode and cathode for the transcranial tACS were placed on the dorsolateral prefrontal cortex (DLPFC). tACS, was applied with current intensity of 1.5 mA at 16 Hz over the dorsolateral prefrontal cortex (DLPFC), aimed to modulate cortical activation and mitigate stress. Sham included 5-second ramp periods. Physiological data using alpha amylase and the NASA Task Load Index (NASA-TLX) were utilized. The results revealed significant hemodynamic changes and reduced stress levels in the tACS group compared to the sham group (p < 0.001). The connectivity network changed significantly (p < 0.001) following tACS. In addition, the NASA-TLX results showed a statistically significant difference between the pre-and post-tACS sessions. In contrary, no statistical significance was noticed for the sham control group. An increase in the blood flow in the prefrontal cortex region of the brain was observed, demonstrating the potential of tACS as a non-invasive neuromodulation technique for stress mitigation.

In the context of Industry 5.0 and human-robot interaction, ensuring the safety of operators by avoiding human errors is crucial. Monitoring vigilance decrement is an essential aspect of this effort, aimed at mitigating safety risks and enhancing productivity. A potentially promising solution to this challenge is using a passive braincomputer interface (BCI) based on electroencephalography (EEG) recordings. However, its application in industrial settings has yet to be explored in-depth. This study uses EEG data to introduce a novel experimental protocol and analysis pipeline to predict vigilance degradation in an industrial research laboratory. The dataset was gathered from ten healthy volunteers who observed a robotic arm for 23 min. The EEG power spectrum over time was computed using the continuous wavelet transform (CWT). After confirming growth in power for the alpha band using a linear regression model, we forecast its trend using four models. As a conventional approach, we used the vector autoregressive (VAR) model, serving as a reference for comparison with three deep learning architectures: a temporal convolutional network (TCN), a gated recurrent unit (GRU) and an encoder-decoder (ED)-GRU. The proposed ED-GRU model outperformed the others showing accurate forecasts (mean absolute error = 0.048, R2 = 0.726) up to 5.5 s. The findings suggest that monitoring vigilance degradation in Industry 5.0 is a feasible strategy to prevent human accidents and reduced performance during repetitive tasks.

PurposeTeamwork involves intricate interactions among individuals or groups with shared goals. It necessitates effective communication, defined roles, decision-making processes, and the allocation of cognitive and emotional resources. Objective teamwork assessment demands a comprehensive set of metrics. Although subjective and behavioral metrics, such as self-evaluation and task completion time, are generally applied, they are prone to bias and a lack of objectivity, highlighting the inherent limitations of capturing the unconscious processes of human behavior.MethodsTo mitigate these limitations, the present study proposed a novel approach to teamwork evaluation based on neurophysiological signals (electroencephalograms, EEGs) compatible with real-world applications, i.e., surgical teams engaged in real-world surgeries. To the best of our knowledge, there is no scientific evidence of an objective teamwork measure performed among more than two members and relying on neurophysiological signals in real-world environments. Therefore, the present work aimed at i) developing and investigating the reliability of an objective EEG-based teamwork index using mutual information (MI) methods and ii) providing additional and objective insights for surgeons’ supervisors in healthcare training.FindingsThe results demonstrated the capability of the EEG-based training index to provide additional and objective information, along with its added value and reliability compared to conventional measures (all R > 0.62, all p < 0.002). Furthermore, the EEG-based teamwork index allowed the determination (all p < 0.001) of surgeons’ experience levels (expert vs novice) in terms of cooperative behavior.ConclusionThe results pave the way for targeted interventions, adaptive training sessions, and optimizations in team dynamics and open up opportunities for applying neurophysiological measurements for teamwork evaluation in all operational fields, where proper and granular teamwork optimization could play a crucial role in terms of safety.

Brainprint recognition technology, regarded as a promising biometric technology, encounters challenges stemming from the time-varied, low signal-to-noise ratio of brain signals, such as electroencephalogram (EEG). Steady-state visual evoked potentials (SSVEP) exhibit high signal-to-noise ratio and frequency locking, making them a promising paradigm for brainprint recognition. Consequently, the extraction of time-invariant identity information from SSVEP EEG signals is essential. In this paper, we propose an Attentive Multi-sub-band Depth Identity Embedding Learning Network for stable cross-session SSVEP brainprint recognition. To address the issue of low recognition accuracy across sessions, we introduce the Sub-band Attentive Frequency mechanism, which integrates the frequency-domain relevant characteristics of the SSVEP paradigm and focuses on exploring depth-frequency identity embedding information. Also, we employ Attentive Statistic Pooling to enhance the stability of frequency domain feature distributions across sessions. Extensive experimentation and validation were conducted on two multi-session SSVEP benchmark datasets. The experimental results show that our approach outperforms other state-of-art models on 2-second samples across sessions and has the potential to serve as a benchmark in multi-subject biometric recognition systems.