Nattapong Thammasan
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Nattapong Thammasan has a total of 37 co-authors. Here are the top co-authors by number of shared publications:
Mannes Poel
Faculty of Electrical Engineering, Mathematics and Computer Science, University of Twente, Enschede, Netherlands
4 shared publications- Anne-Marie Brouwer
TNO (Netherlands Organization for Applied Scientific Research), Utrecht, The Netherlands
3 shared publications - Ivo Vincent Stuldreher
Affiliation Human Performance, Netherlands Organization for Applied Scientific Research (TNO), Soesterberg, The Netherlands
3 shared publications - Boonserm Kijsirikul
Chulalongkorn University
2 shared publications - Jan B. F. van Erp
HMI/ EEMCS lab, University of Twente, Enschede, Netherlands and TNO, Soesterberg, Netherlands
2 shared publications - Masayuki Numao
Department of Information and Computer Science, Faculty of Engineering, Kyoto Tachibana University, Kyoto 607-8175, Japan
2 shared publications
Latest Publications
- 2 November 2026
- Published by Wiley in International Journal of Eating Disorders
Beyond Self‐Report: Adding Wearable Physiological Data to Predict Dietary Restriction in Anorexia Nervosa
Objective: Restricting food intake during meals is the core behavioral mechanism maintaining anorexia nervosa (AN) yet predicting when restriction occurs in daily life remains largely unexplored. Ecological momentary assessment (EMA) captures negative affect preceding restriction but relies on intermittent self‐report, whereas wearables can passively measure physiological signals (EPA). We investigated whether combining EMA and EPA improves prediction of restriction beyond EMA alone, hypothesizing that EMA + EPA models would outperform EMA‐only models, which would outperform EPA‐only models. Methods: Individuals with AN ( n = 78) completed a 2‐week EMA protocol while wearing an Empatica watch, completing 7151 surveys and reporting 4005 meals. Features derived from heart rate, electrodermal activity, and skin temperature were extracted across the hour before each meal. As preregistered, machine learning models were developed to predict both the occurrence and severity of self‐reported restriction (logistic/linear regression, random forest, support vector machines) using EMA predictors, EPA predictors, or their combination. Results: Contrary to our hypothesis, adding EPA predictors did not consistently improve performance. Combined models outperformed EMA‐only models for some classifiers but not for logistic regression and provided no benefit for predicting severity. Models containing EMA predictors consistently outperformed EPA‐only models. The best classification model was EMA‐only logistic regression (AUC = 0.734), while the best regression model was EMA‐only linear regression (RMSE = 1.894). Conclusion: Self‐reported negative affect provided more information for predicting restriction than the physiological signals examined in this study. These findings suggest that self‐report currently offers greater utility than wearable‐derived stress physiology for identifying affect‐related risk for restriction.
- 2 November 2026
- Published by Wiley in International Journal of Eating Disorders
Beyond Self‐Report: Adding Wearable Physiological Data to Predict Dietary Restriction in Anorexia Nervosa
Objective: Restricting food intake during meals is the core behavioral mechanism maintaining anorexia nervosa (AN) yet predicting when restriction occurs in daily life remains largely unexplored. Ecological momentary assessment (EMA) captures negative affect preceding restriction but relies on intermittent self‐report, whereas wearables can passively measure physiological signals (EPA). We investigated whether combining EMA and EPA improves prediction of restriction beyond EMA alone, hypothesizing that EMA + EPA models would outperform EMA‐only models, which would outperform EPA‐only models. Methods: Individuals with AN ( n = 78) completed a 2‐week EMA protocol while wearing an Empatica watch, completing 7151 surveys and reporting 4005 meals. Features derived from heart rate, electrodermal activity, and skin temperature were extracted across the hour before each meal. As preregistered, machine learning models were developed to predict both the occurrence and severity of self‐reported restriction (logistic/linear regression, random forest, support vector machines) using EMA predictors, EPA predictors, or their combination. Results: Contrary to our hypothesis, adding EPA predictors did not consistently improve performance. Combined models outperformed EMA‐only models for some classifiers but not for logistic regression and provided no benefit for predicting severity. Models containing EMA predictors consistently outperformed EPA‐only models. The best classification model was EMA‐only logistic regression (AUC = 0.734), while the best regression model was EMA‐only linear regression (RMSE = 1.894). Conclusion: Self‐reported negative affect provided more information for predicting restriction than the physiological signals examined in this study. These findings suggest that self‐report currently offers greater utility than wearable‐derived stress physiology for identifying affect‐related risk for restriction.
- Open Access
- 8 July 2026
- Published by Elsevier in Current Research in Behavioral Sciences
- Vol. 11
Multi-Modal Human Digital Twin for Hydration
- 8 July 2026
- Published by Elsevier in Current Research in Behavioral Sciences
- Vol. 11
Open AccessMulti-Modal Human Digital Twin for Hydration
- Open Access
- 2 July 2026
- Published by IOS Press
Digital Twins for Health Management: From Sensing to Actionable Insights
Digital Twins offer a promising pathway toward preventive and personalized health by enabling virtual experimentation on continuously updated representations of individuals. We present a human-centered, hybrid intelligence approach in which AI augments self-understanding and decision-making rather than replacing human judgment. By integrating multi-modal sensing, including ingestible sensors, urine sensing, wearables, and self-reports, Digital Twins can infer interpretable physiological and behavioural health states. We outline how hybrid modelling and continuous AI pipelines operationalize these twins to deliver actionable, context-aware feedback while preserving agency, trust, and autonomy.
- 2 July 2026
- Published by IOS Press
Open AccessDigital Twins for Health Management: From Sensing to Actionable Insights
Digital Twins offer a promising pathway toward preventive and personalized health by enabling virtual experimentation on continuously updated representations of individuals. We present a human-centered, hybrid intelligence approach in which AI augments self-understanding and decision-making rather than replacing human judgment. By integrating multi-modal sensing, including ingestible sensors, urine sensing, wearables, and self-reports, Digital Twins can infer interpretable physiological and behavioural health states. We outline how hybrid modelling and continuous AI pipelines operationalize these twins to deliver actionable, context-aware feedback while preserving agency, trust, and autonomy.
- 11 October 2023
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 242, 107859
Simulation of ambulatory electrodermal activity and the handling of low-quality segments
- 11 October 2023
- Published by Elsevier in Computer Methods and Programs in Biomedicine
- Vol. 242, 107859
Simulation of ambulatory electrodermal activity and the handling of low-quality segments
- Open Access
- 17 January 2022
- Published by Frontiers Media SA in Frontiers in Computer Science
- Vol. 3
Detecting Fear of Heights Response to a Virtual Reality Environment Using Functional Near-Infrared Spectroscopy
To enable virtual reality exposure therapy (VRET) that treats anxiety disorders by gradually exposing the patient to fear using virtual reality (VR), it is important to monitor the patient's fear levels during the exposure. Despite the evidence of a fear circuit in the brain as reflected by functional near-infrared spectroscopy (fNIRS), the measurement of fear response in highly immersive VR using fNIRS is limited, especially in combination with a head-mounted display (HMD). In particular, it is unclear to what extent fNIRS can differentiate users with and without anxiety disorders and detect fear response in a highly ecological setting using an HMD. In this study, we investigated fNIRS signals captured from participants with and without a fear of height response. To examine the extent to which fNIRS signals of both groups differ, we conducted an experiment during which participants with moderate fear of heights and participants without it were exposed to VR scenarios involving heights and no heights. The between-group statistical analysis shows that the fNIRS data of the control group and the experimental group are significantly different only in the channel located close to right frontotemporal lobe, where the grand average oxygenated hemoglobin Δ[HbO] contrast signal of the experimental group exceeds that of the control group. The within-group statistical analysis shows significant differences between the grand average Δ[HbO] contrast values during fear responses and those during no-fear responses, where the Δ[HbO] contrast values of the fear responses were significantly higher than those of the no-fear responses in the channels located towards the frontal part of the prefrontal cortex. Also, the channel located close to frontocentral lobe was found to show significant difference for the grand average deoxygenated hemoglobin contrast signals. Support vector machine-based classifier could detect fear responses at an accuracy up to 70% and 74% in subject-dependent and subject-independent classifications, respectively. The results demonstrate that cortical hemodynamic responses of a control group and an experimental group are different to a considerable extent, exhibiting the feasibility and ecological validity of the combination of VR-HMD and fNIRS to elicit and detect fear responses. This research thus paves a way toward the a brain-computer interface to effectively manipulate and control VRET.
- 17 January 2022
- Published by Frontiers Media SA in Frontiers in Computer Science
- Vol. 3
Open AccessDetecting Fear of Heights Response to a Virtual Reality Environment Using Functional Near-Infrared Spectroscopy
To enable virtual reality exposure therapy (VRET) that treats anxiety disorders by gradually exposing the patient to fear using virtual reality (VR), it is important to monitor the patient's fear levels during the exposure. Despite the evidence of a fear circuit in the brain as reflected by functional near-infrared spectroscopy (fNIRS), the measurement of fear response in highly immersive VR using fNIRS is limited, especially in combination with a head-mounted display (HMD). In particular, it is unclear to what extent fNIRS can differentiate users with and without anxiety disorders and detect fear response in a highly ecological setting using an HMD. In this study, we investigated fNIRS signals captured from participants with and without a fear of height response. To examine the extent to which fNIRS signals of both groups differ, we conducted an experiment during which participants with moderate fear of heights and participants without it were exposed to VR scenarios involving heights and no heights. The between-group statistical analysis shows that the fNIRS data of the control group and the experimental group are significantly different only in the channel located close to right frontotemporal lobe, where the grand average oxygenated hemoglobin Δ[HbO] contrast signal of the experimental group exceeds that of the control group. The within-group statistical analysis shows significant differences between the grand average Δ[HbO] contrast values during fear responses and those during no-fear responses, where the Δ[HbO] contrast values of the fear responses were significantly higher than those of the no-fear responses in the channels located towards the frontal part of the prefrontal cortex. Also, the channel located close to frontocentral lobe was found to show significant difference for the grand average deoxygenated hemoglobin contrast signals. Support vector machine-based classifier could detect fear responses at an accuracy up to 70% and 74% in subject-dependent and subject-independent classifications, respectively. The results demonstrate that cortical hemodynamic responses of a control group and an experimental group are different to a considerable extent, exhibiting the feasibility and ecological validity of the combination of VR-HMD and fNIRS to elicit and detect fear responses. This research thus paves a way toward the a brain-computer interface to effectively manipulate and control VRET.
- Open Access
- 7 May 2021
- Published by Frontiers Media SA in Frontiers in Computer Science
- Vol. 3
Exploring the Brain Activity Related to Missing Penalty Kicks: An fNIRS Study
At vital moments in professional soccer matches, penalties were often missed. Psychological factors, such as anxiety and pressure, are among the critical causes of the mistakes, commonly known aschoking under pressure. Nevertheless, the factors have not been fully explored. In this study, we used functional near-infrared spectroscopy (fNIRS) to investigate the influence of the brain on this process. Anin-situstudy was set-up (N= 22), in which each participant took 15 penalties under three different pressure conditions: without a goalkeeper, with an amiable goalkeeper, and with a competitive goalkeeper. Both experienced and inexperienced soccer players were recruited, and the brain activation was compared across groups. Besides, fNIRS activation was compared between sessions that participants felt anxious against sessions without anxiety report, and between penalty-scoring and -missing sessions. The results show that the task-relevant brain region, the motor cortex, was more activated when players were not experiencing performance anxiety. The activation of task-irrelevant areas was shown to be related to players experiencing anxiety and missing penalties, especially the prefrontal cortex (PFC). More particularly, an overall higher activation of the PFC and an increase of PFC lateral asymmetry were related to anxious players and missed penalties, which can be caused by players' worries about the consequences of scoring or missing the penalty kicks. When experienced players were feeling anxious, their left temporal cortex activation increased, which could be an indication that experienced overthink the situation and neglect their automated skills. Besides, the left temporal cortex activation is higher when inexperienced players succeeded to score a penalty. Overall, the results of this study are in line with the neural efficiency theory and demonstrate the feasibility and ecological validity to detect neurological clues relevant to anxiety and performance from fNIRS recordingsin the field.
- 7 May 2021
- Published by Frontiers Media SA in Frontiers in Computer Science
- Vol. 3
Open AccessExploring the Brain Activity Related to Missing Penalty Kicks: An fNIRS Study
At vital moments in professional soccer matches, penalties were often missed. Psychological factors, such as anxiety and pressure, are among the critical causes of the mistakes, commonly known aschoking under pressure. Nevertheless, the factors have not been fully explored. In this study, we used functional near-infrared spectroscopy (fNIRS) to investigate the influence of the brain on this process. Anin-situstudy was set-up (N= 22), in which each participant took 15 penalties under three different pressure conditions: without a goalkeeper, with an amiable goalkeeper, and with a competitive goalkeeper. Both experienced and inexperienced soccer players were recruited, and the brain activation was compared across groups. Besides, fNIRS activation was compared between sessions that participants felt anxious against sessions without anxiety report, and between penalty-scoring and -missing sessions. The results show that the task-relevant brain region, the motor cortex, was more activated when players were not experiencing performance anxiety. The activation of task-irrelevant areas was shown to be related to players experiencing anxiety and missing penalties, especially the prefrontal cortex (PFC). More particularly, an overall higher activation of the PFC and an increase of PFC lateral asymmetry were related to anxious players and missed penalties, which can be caused by players' worries about the consequences of scoring or missing the penalty kicks. When experienced players were feeling anxious, their left temporal cortex activation increased, which could be an indication that experienced overthink the situation and neglect their automated skills. Besides, the left temporal cortex activation is higher when inexperienced players succeeded to score a penalty. Overall, the results of this study are in line with the neural efficiency theory and demonstrate the feasibility and ecological validity to detect neurological clues relevant to anxiety and performance from fNIRS recordingsin the field.
- Open Access
A Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN
Emotion recognition based on electroencephalograms has become an active research area. Yet, identifying emotions using only brainwaves is still very challenging, especially the subject-independent task. Numerous studies have tried to propose methods to recognize emotions, including machine learning techniques like convolutional neural network (CNN). Since CNN has shown its potential in generalization to unseen subjects, manipulating CNN hyperparameters like the window size and electrode order might be beneficial. To our knowledge, this is the first work that extensively observed the parameter selection effect on the CNN. The temporal information in distinct window sizes was found to significantly affect the recognition performance, and CNN was found to be more responsive to changing window sizes than the support vector machine. Classifying the arousal achieved the best performance with a window size of ten seconds, obtaining 56.85% accuracy and a Matthews correlation coefficient (MCC) of 0.1369. Valence recognition had the best performance with a window length of eight seconds at 73.34% accuracy and an MCC value of 0.4669. Spatial information from varying the electrode orders had a small effect on the classification. Overall, valence results had a much more superior performance than arousal results, which were, perhaps, influenced by features related to brain activity asymmetry between the left and right hemispheres.
Open AccessA Comparative Study of Window Size and Channel Arrangement on EEG-Emotion Recognition Using Deep CNN
Emotion recognition based on electroencephalograms has become an active research area. Yet, identifying emotions using only brainwaves is still very challenging, especially the subject-independent task. Numerous studies have tried to propose methods to recognize emotions, including machine learning techniques like convolutional neural network (CNN). Since CNN has shown its potential in generalization to unseen subjects, manipulating CNN hyperparameters like the window size and electrode order might be beneficial. To our knowledge, this is the first work that extensively observed the parameter selection effect on the CNN. The temporal information in distinct window sizes was found to significantly affect the recognition performance, and CNN was found to be more responsive to changing window sizes than the support vector machine. Classifying the arousal achieved the best performance with a window size of ten seconds, obtaining 56.85% accuracy and a Matthews correlation coefficient (MCC) of 0.1369. Valence recognition had the best performance with a window length of eight seconds at 73.34% accuracy and an MCC value of 0.4669. Spatial information from varying the electrode orders had a small effect on the classification. Overall, valence results had a much more superior performance than arousal results, which were, perhaps, influenced by features related to brain activity asymmetry between the left and right hemispheres.
- Open Access
Cross-Frequency Power-Power Coupling Analysis: A Useful Cross-Frequency Measure to Classify ICA-Decomposed EEG
Magneto-/Electro-encephalography (M/EEG) commonly uses (fast) Fourier transformation to compute power spectral density (PSD). However, the resulting PSD plot lacks temporal information, making interpretation sometimes equivocal. For example, consider two different PSDs: a central parietal EEG PSD with twin peaks at 10 Hz and 20 Hz and a central parietal PSD with twin peaks at 10 Hz and 50 Hz. We can assume the first PSD shows a mu rhythm and the second harmonic; however, the latter PSD likely shows an alpha peak and an independent line noise. Without prior knowledge, however, the PSD alone cannot distinguish between the two cases. To address this limitation of PSD, we propose using cross-frequency power–power coupling (PPC) as a post-processing of independent component (IC) analysis (ICA) to distinguish brain components from muscle and environmental artifact sources. We conclude that post-ICA PPC analysis could serve as a new data-driven EEG classifier in M/EEG studies. For the reader’s convenience, we offer a brief literature overview on the disparate use of PPC. The proposed cross-frequency power–power coupling analysis toolbox (PowPowCAT) is a free, open-source toolbox, which works as an EEGLAB extension.
Open AccessCross-Frequency Power-Power Coupling Analysis: A Useful Cross-Frequency Measure to Classify ICA-Decomposed EEG
Magneto-/Electro-encephalography (M/EEG) commonly uses (fast) Fourier transformation to compute power spectral density (PSD). However, the resulting PSD plot lacks temporal information, making interpretation sometimes equivocal. For example, consider two different PSDs: a central parietal EEG PSD with twin peaks at 10 Hz and 20 Hz and a central parietal PSD with twin peaks at 10 Hz and 50 Hz. We can assume the first PSD shows a mu rhythm and the second harmonic; however, the latter PSD likely shows an alpha peak and an independent line noise. Without prior knowledge, however, the PSD alone cannot distinguish between the two cases. To address this limitation of PSD, we propose using cross-frequency power–power coupling (PPC) as a post-processing of independent component (IC) analysis (ICA) to distinguish brain components from muscle and environmental artifact sources. We conclude that post-ICA PPC analysis could serve as a new data-driven EEG classifier in M/EEG studies. For the reader’s convenience, we offer a brief literature overview on the disparate use of PPC. The proposed cross-frequency power–power coupling analysis toolbox (PowPowCAT) is a free, open-source toolbox, which works as an EEGLAB extension.
- Open Access
- 3 December 2020
- Published by Frontiers Media SA in Frontiers in Neuroscience
- Vol. 14
Physiological Synchrony in EEG, Electrodermal Activity and Heart Rate Detects Attentionally Relevant Events in Time
Interpersonal physiological synchrony (PS), or the similarity of physiological signals between individuals over time, may be used to detect attentionally engaging moments in time. We here investigated whether PS in the electroencephalogram (EEG), electrodermal activity (EDA), heart rate and a multimodal metric signals the occurrence of attentionally relevant events in time in two groups of participants. Both groups were presented with the same auditory stimulus, but were instructed to attend either to the narrative of an audiobook (audiobook-attending: AA group) or to interspersed emotional sounds and beeps (stimulus-attending: SA group). We hypothesized that emotional sounds could be detected in both groups as they are expected to draw attention involuntarily, in a bottom-up fashion. Indeed, we found this to be the case for PS in EDA or the multimodal metric. Beeps, that are expected to be only relevant due to specific “top-down” attentional instructions, could indeed only be detected using PS among SA participants, for EDA, EEG and the multimodal metric. We further hypothesized that moments in the audiobook accompanied by high PS in either EEG, EDA, heart rate or the multimodal metric for AA participants would be rated as more engaging by an independent group of participants compared to moments corresponding to low PS. This hypothesis was not supported. Our results show that PS can support the detection of attentionally engaging events over time. Currently, the relation between PS and engagement is only established for well-defined, interspersed stimuli, whereas the relation between PS and a more abstract self-reported metric of engagement over time has not been established. As the relation between PS and engagement is dependent on event type and physiological measure, we suggest to choose a measure matching with the stimulus of interest. When the stimulus type is unknown, a multimodal metric is most robust.
- 3 December 2020
- Published by Frontiers Media SA in Frontiers in Neuroscience
- Vol. 14
Open AccessPhysiological Synchrony in EEG, Electrodermal Activity and Heart Rate Detects Attentionally Relevant Events in Time
Interpersonal physiological synchrony (PS), or the similarity of physiological signals between individuals over time, may be used to detect attentionally engaging moments in time. We here investigated whether PS in the electroencephalogram (EEG), electrodermal activity (EDA), heart rate and a multimodal metric signals the occurrence of attentionally relevant events in time in two groups of participants. Both groups were presented with the same auditory stimulus, but were instructed to attend either to the narrative of an audiobook (audiobook-attending: AA group) or to interspersed emotional sounds and beeps (stimulus-attending: SA group). We hypothesized that emotional sounds could be detected in both groups as they are expected to draw attention involuntarily, in a bottom-up fashion. Indeed, we found this to be the case for PS in EDA or the multimodal metric. Beeps, that are expected to be only relevant due to specific “top-down” attentional instructions, could indeed only be detected using PS among SA participants, for EDA, EEG and the multimodal metric. We further hypothesized that moments in the audiobook accompanied by high PS in either EEG, EDA, heart rate or the multimodal metric for AA participants would be rated as more engaging by an independent group of participants compared to moments corresponding to low PS. This hypothesis was not supported. Our results show that PS can support the detection of attentionally engaging events over time. Currently, the relation between PS and engagement is only established for well-defined, interspersed stimuli, whereas the relation between PS and a more abstract self-reported metric of engagement over time has not been established. As the relation between PS and engagement is dependent on event type and physiological measure, we suggest to choose a measure matching with the stimulus of interest. When the stimulus type is unknown, a multimodal metric is most robust.
- Open Access
A Usability Study of Physiological Measurement in School Using Wearable Sensors
Measuring psychophysiological signals of adolescents using unobtrusive wearable sensors may contribute to understanding the development of emotional disorders. This study investigated the feasibility of measuring high quality physiological data and examined the validity of signal processing in a school setting. Among 86 adolescents, a total of more than 410 h of electrodermal activity (EDA) data were recorded using a wrist-worn sensor with gelled electrodes and over 370 h of heart rate data were recorded using a chest-strap sensor. The results support the feasibility of monitoring physiological signals at school. We describe specific challenges and provide recommendations for signal analysis, including dealing with invalid signals due to loose sensors, and quantization noise that can be caused by limitations in analog-to-digital conversion in wearable devices and be mistaken as physiological responses. Importantly, our results show that using toolboxes for automatic signal preprocessing, decomposition, and artifact detection with default parameters while neglecting differences between devices and measurement contexts yield misleading results. Time courses of students’ physiological signals throughout the course of a class were found to be clearer after applying our proposed preprocessing steps.
Open AccessA Usability Study of Physiological Measurement in School Using Wearable Sensors
Measuring psychophysiological signals of adolescents using unobtrusive wearable sensors may contribute to understanding the development of emotional disorders. This study investigated the feasibility of measuring high quality physiological data and examined the validity of signal processing in a school setting. Among 86 adolescents, a total of more than 410 h of electrodermal activity (EDA) data were recorded using a wrist-worn sensor with gelled electrodes and over 370 h of heart rate data were recorded using a chest-strap sensor. The results support the feasibility of monitoring physiological signals at school. We describe specific challenges and provide recommendations for signal analysis, including dealing with invalid signals due to loose sensors, and quantization noise that can be caused by limitations in analog-to-digital conversion in wearable devices and be mistaken as physiological responses. Importantly, our results show that using toolboxes for automatic signal preprocessing, decomposition, and artifact detection with default parameters while neglecting differences between devices and measurement contexts yield misleading results. Time courses of students’ physiological signals throughout the course of a class were found to be clearer after applying our proposed preprocessing steps.