Electroencephalography (EEG) offers a noninvasive, high-temporal-resolution modality for estimating mental workload. However, session-to-session variability limits the generalizability of workload classifiers, and few systematic cross-session evaluations are reported in the literature. This study systematically evaluates domain adaptation methods for cross-session mental workload classification using the publicly available COG-BCI dataset within an evaluation framework that may guide future studies on EEG-based classification models. We make four contributions: (i) integration of Optimal Transport (OT) with Graph Neural Networks (GNNs) to model spatial relationships and align feature distributions under strict session-wise separation; (ii) a data-centric evaluation pipeline incorporating Self-Organizing Map (SOM) visualizations for data exploration and a heuristic loss function for model selection; (iii) a strict cross-session protocol examining the effects of graph construction, feature selection, and data splits; and (iv) comparison of OT with CORrelation ALignment (CORAL) and GNN with EEGNet. Incorporating OT improved test accuracies across all experimental configurations. SOM visualizations confirmed enhanced feature alignment after OT. Our results highlight the potential of OT for mitigating session-to-session variability and underscore the importance of a data-centric approach and rigorous cross-session evaluation when developing classifiers for complex cognitive state estimation. Future work should explore semi-supervised OT strategies and scalable implementations for real-time applications.

BackgroundThe human body radiates multiple biological fields which can be measured with sensors directly on or off the body. These electromagnetic fields are hypothesized to be altered by mind and body practices. The primary goal of the current study was to evaluate the feasibility of continuous multi-sensor monitoring during meditation and breathwork practices. It also explored methods to investigate whether simultaneous biological field measures correlate as parts of a dynamic system of biofields responsive to changes in states of consciousness.MethodsTwenty-three adults were recruited to participate. The intervention consisted of a guided loving kindness meditation followed by a guided breathwork exercise while equipped with multiple biofield sensors to measure heart rate (HR), heart rate variability (HRV), skin conductance (SCR), alpha waves with electroencephalography (EEG), infrared radiation (IR), and ultraweak photon emission (UPE). Participants completed self-report measures of emotional affect and states of consciousness. Aggregate and individual differences in participants’ responses to the meditation and breathwork were assessed. Exploratory analyses of within-subject correlations between biofields were also conducted.ResultsThe study procedure was feasible with 100% recruitment and retention and the intervention was acceptable to participants. Meditation significantly increased IR of the nose (p = 0.003). Breathwork significantly increased HR (p = 0.002) and decreased IR of the nose (p < 0.001), and left hand UPE showed a near-significant decreasing trend (p = 0.057). Biofield measures changed as expected for some participants. Post-meditation, participants reported lower arousal and increased control, boundarylessness, and non-duality. Post-breathwork, participants reported increased arousal and decreased boundarylessness, connectedness, and non-duality. There were strong correlations (r > 0.5) between UPE from both hands and moderate correlations (r > 0.4) between IR nose temperature and left hand UPE.ConclusionThe current study demonstrated the feasibility of simultaneous measurement using multiple on- and off-body biological field sensors. Meditation and breathwork produced nearly opposite effects on self-report and biofield measures. Preliminary analysis indicated intra-subject correlations between different biofield measures. Future work with a larger sample size and appropriate control groups is required to draw conclusions about the systemic nature of biofield measures and their relationship with states of consciousness.

IntroductionDigital technologies now mediate a substantial proportion of human collaboration, reshaping how individuals coordinate attention, share information, and jointly act on goals. These digitally mediated interactions engage neural, physiological, and behavioral processes differently compared to face-to-face settings. Mobile hyperscanning, i.e., simultaneous (neuro-)physiological measures of two or more individuals, offers a unique window into these multidimensional dynamics. Yet, the existing literature is highly fragmented in design, modality, and analytic rigor, making it difficult to accumulate knowledge. This review systematically synthesizes hyperscanning research investigating collaboration involving digital components and identifies key methodological and conceptual gaps that must be addressed to advance the field.MethodsWe searched Scopus, PubMed, and Web of Science (April 2025) for mobile hyperscanning studies on digital collaboration. Forty-five eligible studies involving simultaneous measurements of at least two healthy adults engaged in collaborative tasks with a digital interaction component were included. Studies were categorized across 13 dimensions, including modality, task design, interaction type, analysis method, and cognitive domain. To ensure transparency and support cumulative synthesis, we created a continuously updated online resource (“InterBrainDB”).ResultsMost studies relied on unimodal neuroimaging, predominantly electroencephalography (EEG) or functional near-infrared spectroscopy (fNIRS), with only seven studies implementing multimodal combinations. Study designs favored cooperative tasks or naturalistic scenarios with symmetrical roles, typically using same-sex dyads of unfamiliar individuals. Non-verbal interaction was studied slightly more often than verbal. Analytically, functional connectivity dominated, whereas effective connectivity, multimodal fusion, and machine learning were scarcely used. Executive and social cognition were more frequently investigated than creativity, memory, and language.DiscussionResearch on digital collaboration through hyperscanning is growing, yet progress is limited by methodological heterogeneity, narrow use of modalities, and analytical conservatism. Future advances will require: (1) multimodal integration to fully capture neural, physiological, and behavioral dynamics; (2) systematic comparisons across varying degrees of digitalization to understand how technology shapes interaction; (3) physiology-informed analysis frameworks capable of modeling high-dimensional interpersonal dynamics; and (4) clearer reporting standards to enable reproducibility and large-scale synthesis. Resources like our InterBrainDB can structure a community-driven progress toward ecologically grounded models of digitally mediated collaboration, a domain of increasing scientific and societal relevance.

The context in which food is selected and consumed is an important factor in its choice, consumption, and acceptability. This study assessed the effect of information and multisensory contexts on meat-related food choices and taste perception. In total, 224 participants first watched one of two pitches, either discussing the implications of consuming animal meat (sustainable pitch) or promoting body movement (control pitch). Participants were then exposed to one of three multisensory contexts: a ‘sustainable’ context with natural green colours, nature sounds and a flower fragrance, a ‘meat’ context with red colours, the sounds of country music and a smokey BBQ smell, and a monotone off-white ‘neutral’ context with neutral background music and no additional smell. Participants were instructed to choose one of two presented hotdogs (animal meat hotdog or plant-based meat hotdog) and to taste and rate the chosen one on liking and taste attributes. Results showed that multisensory sustainable contextual cues combined with information on sustainability beforehand increased the likelihood of choosing plant-based meat hotdogs over animal meat hotdogs. In addition, while tasting the plant-based meat hotdog, multisensory contextual cues that are inspired by a meat context appeared to enhance taste perception, even for vegans and vegetarians. These findings provide further evidence for the importance of context in food choice and acceptance: the context where people choose plant-based meat should preferably be separated and different from the context of consumption. The findings also imply that information can change behaviour, not just attitudes as previous research indicated, but only if combined with multisensory cues in the context.

In this article, we transition from the theoretical and experimental groundwork of manipulating and measuring Sense of Embodiment (SoE) to addressing a fundamental question: What is the purpose of optimizing the SoE in a teleoperation system? This exploration centers on investigating the potential positive effects of SoE on motor adaptation, the acceleration of motor learning, and the potential enhancement of task performance. The article delves into this investigation by focusing on two critical research questions: (1) what is the effect of SoE on task performance in a perceptual-motor task? (2) What is the effect of SoE on the asymptote of the learning curve in a perceptual-motor task? Drawing insights from the existing literature, the hypothesis emerges that enhancing SoE yields positive effects not only on task performance (H1) but also on the overall embodiment experience (H2). An additional layer of exploration is introduced through an exploratory research question: Are these results consistent across diverse scenarios and tasks? The study design encompasses two distinct user studies, each set in different applications and featuring various avatars, yet all anchored in similar tasks, specifically a modified peg-in-hole task: (1) in the first experiment, participants operated a robotic arm with a human-like hand as end-effector, and they were required to perform a classic peg-in-hole task; (2) in user study 2, the task is transformed into a variation that we called "peg-on-button," wherein participants use a robotic arm with a gripper as the end-effector to press a lit button. In both studies, a consistent pattern emerges: a setup that fosters embodiment has a positive impact on motor learning and adaptation, resulting in improved task performance. A supportive setup also reduces the perception of the surrogate as a mere mediator between the operator and the remote environment, especially when contrasted with a setup that suppresses embodiment. The positive effects on motor learning and task performance advocate for the incorporation of embodiment-supportive designs in teleoperational setups. However, the nuanced relationship between SoE and long-term task performance prompts a call for further exploration and consideration of various factors influencing teleoperation outcomes across diverse scenarios and tasks.

IntroductionEducational practice increasingly makes use of technology to improve teaching and learning. New wearable technology is being developed that measures mental states like attention and stress, through neurophysiological signals like electroencephalography (EEG), electrodermal activity (EDA) and heart rate. However, little is known about the ethical aspects of this technology.MethodologyWe provide an overview of current ethical considerations on such wearable technologies in classroom settings and analyze these critically. We distinguished three ethical angles to analyze new technologies: epistemic, principle-based, and Foucauldian. We focus on a Foucauldian analysis, outlining how such technologies affect power relationships and self-understanding, but also which responses people develop to evade power. In addition, a focus group of high school students was set up to identify young people's views on such wearable technology and to initiate a reflection on the theory-based ethical considerations.ResultsOur study shows that although wearables may provide information on learning and attention, and even though possible users are enthusiastic about the potential, there are several risks of applying such technologies in educational settings. These risks concern governance and surveillance, normalization and exclusion, placing technology before pedagogy, stimulating neoliberal values and quantified self-understanding, and possible negative impact on identity for those who think they are outside of the norm. High school students highlighted that people are not only subjected to new technologies, but also subject these technologies to their own goals.DiscussionWe end with a discussion on the perils of implementing new technologies, and provide an alternative to prohibition in the form of co-creating and educating. Any potential future implementation of mental state tracking technology is to be accompanied by normative discussions on legitimate aims, on rights, interests and needs of both pupils, teachers, and educational institutions, taking broader debates on what should count as a good pedagogical climate into account.

Current sensors offering passive and continuous monitoring of behavioral patterns potentially enable real-time affective state monitoring. Previous research on affective state prediction with multimodal sensing in daily life has shown only small-to-moderate effects. One reason for this limited success might be the large variability across individuals. Current research is often of short duration, preventing proper within-individual modeling. With an extensive longitudinal data collection of nine months, this research focuses on individual-level predictions of valence and arousal in daily life. Sixteen PhD candidates from The Netherlands provided data about their affective states (self-reported valence and arousal), physiology (Oura rings) and behavioral patterns (AWARE framework for mobile phone data). Supporting our hypothesis, subject-dependent random forest (RF) models significantly outperformed subject-independent leave-one-subject-out (LOSO) models in predicting self-reported valence and arousal. The subject-dependent models achieved an average Spearman's rho correlation of 0.30 [0.14-0.60] for valence and 0.36 [0.16-0.69] for arousal. In many cases, participants' a priori indicated informative sources matched with the feature importance. Making use of participants' self-knowledge might thus help to reduce the amount of data to be collected. For future work, longer-term changes in affective state and combinations of features for estimating real behavioral patterns should be explored.

Is an audience captured by a speech or lecture? At what times especially? Do different groups in an audience experience the same speech in different ways? Insight into attentional engagement of individuals can be valuable but difficult to quantify using self-report. Physiological synchrony, the degree to which physiological measurements such as electrodermal activity (EDA) of multiple people uniformly change, has been shown to covary with attentional engagement in lab settings. In this study, we moved out of the lab and monitored EDA of 30 individuals attending a real-life inaugural lecture. These individuals were labeled as belonging to either the personal or professional group, based on their relation with the speaker. We expected these groups to differ in their attentional engagement. We computed physiological synchrony between the participants and investigated how well this metric distinguished between the professional and personal groups, how well it marked predefined engaging events in the lecture, and its relation with levels of engagement as self-reported afterwards. Where possible, we compared physiological synchrony results to results based on individuals' EDA. We found that physiological synchrony in EDA can distinguish between the two groups. Individuals' EDA can also distinguish between the groups, if the occurrence and timing is known of an event that is expected to elicit different levels of engagement for the two groups. We further found that both synchrony and individuals' EDA measures mark predefined engaging events with above-chance accuracies. Neither was reliably related to self-reported levels of attentional engagement, highlighting the complementary value of EDA. Our work shows the sensitivity of EDA measures in real-life conditions, where low-level sensory effects, movement and speech cannot be the explanatory factor. Ultimate applications may be in educational and entertainment domains, exploring potential differences in attentional engagement patterns between experts and novices, or different target groups in entertainment.

With the purpose of identifying a sensitive, robust, and easy-to-measure set of biomarkers to assess stress reactivity, we here study a large set of relatively easy to obtain markers reflecting subjective, autonomic nervous system (ANS), endocrine, and inflammatory responses to acute social stress (n=101). A subset of the participants was exposed to another social stressor the next day (n=48) while being measured in the same way. Acute social stress was induced following standardized procedures. The markers investigated were self-reported positive and negative affect, heart rate, electrodermal activity, salivary cortisol, and ten inflammatory markers both in capillary plasma and salivary samples, including IL-22 which has not been studied in response to acute stress in humans before. Robust effects (significant effect in the same direction for both days) were found for self-reported negative affect, heart rate, electrodermal activity, plasma IL-5, plasma IL-22, salivary IL-8 and salivary IL-10. Of these seven markers, the participants’ IL-22 responses on the first day were positively correlated to those on the second day. We found no correlations between salivary and capillary plasma stress responses for any of the ten cytokines and somewhat unexpectedly, cytokine responses in saliva seemed more pronounced and more in line with previous literature than cytokines in capillary plasma. In sum, seven robust and easy to obtain biomarkers to measure acute stress response were identified and should be used in future stress research to detect and examine stress reactivity. This includes IL-22 in plasma as a promising novel marker.

This study addresses concerns about reproducibility in scientific research, focusing on the use of electroencephalography (EEG) and machine learning to estimate mental workload. We established guidelines for reproducible machine learning research using EEG and used these to assess the current state of reproducibility in mental workload modeling. We first started by summarizing the current state of reproducibility efforts in machine learning and in EEG. Next, we performed a systematic literature review on Scopus, Web of Science, ACM Digital Library, and Pubmed databases to find studies about reproducibility in mental workload prediction using EEG. All of this previous work was used to formulate guidelines, which we structured along the widely recognized Cross-Industry Standard Process for Data Mining (CRISP-DM) framework. By using these guidelines, researchers can ensure transparency and comprehensiveness of their methodologies, therewith enhancing collaboration and knowledge-sharing within the scientific community, and enhancing the reliability, usability and significance of EEG and machine learning techniques in general. A second systematic literature review extracted machine learning studies that used EEG to estimate mental workload. We evaluated the reproducibility status of these studies using our guidelines. We highlight areas studied and overlooked and identify current challenges for reproducibility. Our main findings include limitations on reporting performance on unseen test data, open sharing of data and code, and reporting of resources essential for training and inference processes.

Automatically detecting mental state such as stress from video images of the face could support evaluating stress responses in applicants for high risk jobs or contribute to timely stress detection in challenging operational settings (e.g., aircrew, command center operators). Challenges in automatically estimating mental state include the generalization of models across contexts and across participants. We here aim to create robust models by training them using data from different contexts and including physiological features. Fifty-one participants were exposed to different types of stressors (cognitive, social evaluative and startle) and baseline variants of the stressors. Video, electrocardiogram (ECG), electrodermal activity (EDA) and self-reports (arousal and valence) were recorded. Logistic regression models aimed to classify between high and low arousal and valence across participants, where "high" and "low" were defined relative to the center of the rating scale. Accuracy scores of different models were evaluated: models trained and tested within a specific context (either a baseline or stressor variant of a task), intermediate context (baseline and stressor variant of a task), or general context (all conditions together). Furthermore, for these different model variants, only the video data was included, only the physiological data, or both video and physiological data. We found that all (video, physiological and video-physio) models could successfully distinguish between high- and low-rated arousal and valence, though performance tended to be better for (1) arousal than valence, (2) specific context than intermediate and general contexts, (3) video-physio data than video or physiological data alone. Automatic feature selection resulted in inclusion of 3-20 features, where the models based on video-physio data usually included features from video, ECG and EDA. Still, performance of video-only models approached the performance of video-physio models. Arousal and valence ratings by three experienced human observers scores based on part of the video data did not match with self-reports. In sum, we showed that it is possible to automatically monitor arousal and valence even in relatively general contexts and better than humans can (in the given circumstances), and that non-contact video images of faces capture an important part of the information, which has practical advantages.

IntroductionUnderstanding how food neophobia affects food experience may help to shift toward sustainable diets. Previous research suggests that individuals with higher food neophobia are more aroused and attentive when observing food-related stimuli. The present study examined whether electrodermal activity (EDA), as index of arousal, relates to food neophobia outside the lab when exposed to a single piece of food.MethodsThe EDA of 153 participants was analyzed as part of a larger experiment conducted at a festival. Participants completed the 10-item Food Neophobia Scale. Subsequently, they saw three lids covering three foods: a hotdog labeled as “meat”, a hotdog labeled as “100% plant-based”, and tofu labeled as “100% plant-based”. Participants lifted the lids consecutively and the area-under-the-curve (AUC) of the skin conductance response (SCR) was captured between 20 s before and 20 s after each food reveal.ResultsWe found a significant positive correlation between food neophobia and AUC of SCR during presentation of the first and second hotdog and a trend for tofu. These correlations remained significant even when only including the SCR data prior to the food reveal (i.e., an anticipatory response).DiscussionThe association between food neophobia and EDA indicates that food neophobic individuals are more aroused upon the presentation of food. We show for the first time that the anticipation of being presented with food already increased arousal for food neophobic individuals. These findings also indicate that EDA can be meaningfully determined using wearables outside the lab, in a relatively uncontrolled setting for single-trial analysis.

Stress is one of the most pressing problems in society as it severely reduces the physical and mental wellbeing of people. It is therefore of great importance to accurately monitor stress levels, especially in work environments. However, contemporary stress assessments, such as questionnaires and physiological measurements, have practical limitations, mostly related to their subjective or contact-based nature. To assess stress objectively and conveniently, we developed an automated model that detects biomarkers in webcam-recorded facial behavior indicative of heightened stress levels, using computer vision, artificial intelligence, and machine learning techniques. Heart-rate induced skin pulsations and facial muscle activity were extracted from videos of 264 participants that performed an online mental capacity test under considerable time pressure. The model could successfully use these facial biomarkers to explain a significant proportion of individual differences in scores on a self-perceived stress scale. Next, we used the model to objectively score stress levels of 63 military candidates (pre-hiring) and 69 military personnel (post-hiring) that also performed the mental capacity test. Results showed that military personnel expressed facial behavior indicative of significantly higher stress levels than military candidates. This suggests that joining the military heightens overall stress levels. With this study we take the first steps towards a non-contact, automated, and objective measure of stress that is easily applicable in a variety of health and work contexts.

Individuals that pay attention to narrative stimuli show synchronized heart rate (HR) and electrodermal activity (EDA) responses. The degree to which this physiological synchrony occurs is related to attentional engagement. Factors that can influence attention, such as instructions, salience of the narrative stimulus and characteristics of the individual, affect physiological synchrony. The demonstrability of synchrony depends on the amount of data used in the analysis. We investigated how demonstrability of physiological synchrony varies with varying group size and stimulus duration. Thirty participants watched six 10 min movie clips while their HR and EDA were monitored using wearable sensors (Movisens EdaMove 4 and Wahoo Tickr, respectively). We calculated inter-subject correlations as a measure of synchrony. Group size and stimulus duration were varied by using data from subsets of the participants and movie clips in the analysis. We found that for HR, higher synchrony correlated significantly with the number of answers correct for questions about the movie, confirming that physiological synchrony is associated with attention. For both HR and EDA, with increasing amounts of data used, the percentage of participants with significant synchrony increased. Importantly, we found that it did not matter how the amount of data was increased. Increasing the group size or increasing the stimulus duration led to the same results. Initial comparisons with results from other studies suggest that our results do not only apply to our specific set of stimuli and participants. All in all, the current work can act as a guideline for future research, indicating the amount of data minimally needed for robust analysis of synchrony based on inter-subject correlations.

This study examined the extent to which adolescent peer victimization predicted acute inflammatory responses to stress, and whether both resting parasympathetic nervous system (PNS) activity and PNS stress reactivity moderated this association. 83 adolescents (Mage = 14.89, SDage = 0.52, 48

Humans differ strongly in their willingness to try novel foods. Hesitance to try new foods is referred to as food neophobia. Understanding food neophobia is important, as it can be a significant barrier to adopt a healthy, balanced or plant-based diet. We here use electroencephalogram (EEG) recordings to obtain insight in the early attentional processes towards food stimuli as a function of food neophobia. 43 Dutch participants completed the food neophobia scale after which they were presented with pictures of familiar and unfamiliar foods and a 15minute movie about the origin and production of an unfamiliar food. We extracted two EEG-based metrics of attention: the late positive potential (LPP) amplitude in response to the food pictures, and inter-subject correlations (ISC-EEG) during the movie. The latter is a novel metric, based on similarities in EEG over time between individuals who are presented with the same stimulus, and suitable for examining attention towards continuous stimuli such as movies. Additionally, participants were asked to taste familiar and unfamiliar soups, and they were asked to rate the pictures and soups for valence and arousal. ISC-EEG and the LPP amplitude increased and sip size decreased with food neophobia, not only for unfamiliar food pictures, but also for familiar food pictures. Self-reported emotional experience was affected by food neophobia for unfamiliar food pictures or soups, but not for familiar ones. We conclude that food neophobia is associated with increased attentional processing and immediate implicit behavior, for all food stimuli and not only for unfamiliar food stimuli. This indicates that all food-related stimuli are of high importance to food neophobic individuals and that self-reported emotion does not capture the entire experience of food. The results also indicates that, unlike the name suggest, food neophobia does not only affect processing of novel foods, but of any food regardless of familiarity.

In this paper, we present a review of how the various aspects of any study using an eye tracker (such as the instrument, methodology, environment, participant, etc.) affect the quality of the recorded eye-tracking data and the obtained eye-movement and gaze measures. We take this review to represent the empirical foundation for reporting guidelines of any study involving an eye tracker. We compare this empirical foundation to five existing reporting guidelines and to a database of 207 published eye-tracking studies. We find that reporting guidelines vary substantially and do not match with actual reporting practices. We end by deriving a minimal, flexible reporting guideline based on empirical research (Section “ An empirically based minimal reporting guideline ”).

We examined the effects of an informative pitch and multisensory contexts as potential factors influencing individuals’ experience of tofu with soy sauce and the amount consumed outside the lab. Two hundred and sixteen participants watched one of two pitches (promoting either vegetarian diets or exercise) and were guided into one of three multisensory contexts (‘sustainable’, ‘meat’, or ‘neutral’ theme). Participants rated the aroma and appearance of soy sauce and the taste of tofu dipped in it using the intuitive ‘one touch’ EmojiGrid valence and arousal measuring tool. Our results showed that the ‘meat’ context increased arousal ratings for soy sauce and the tendency to consume more tofu relative to the other contexts. Pitch did not influence affective ratings or amounts consumed. We conclude that the multisensory context has the potential to positively affect peoples’ choices and perceptions of plant-based and sustainable food and promote its consumption.