Towards mental stress detection using wearable physiological sensors

Predicting stress levels using physiological data: Real-time stress prediction models utilizing wearable devices

Stress has emerged as a prominent and multifaceted health concern in contemporary society, manifesting detrimental effects on individuals' physical and mental health and well-being. The ability to accurately predict stress levels in real time holds significant promise for facilitating timely interventions and personalized stress management strategies. The increasing incidence of stress-related physical and mental health issues highlights the importance of thoroughly understanding stress prediction mechanisms. Given that stress is a contributing factor to a wide array of mental and physical health problems, objectively assessing stress is crucial for behavioral and physiological studies. While numerous studies have assessed stress levels in controlled environments, the objective evaluation of stress in everyday settings still needs to be explored, primarily due to contextual factors and limitations in self-report adherence. This short review explored the emerging field of real-time stress prediction, focusing on utilizing physiological data collected by wearable devices. Stress was examined from a comprehensive standpoint, acknowledging its effects on both physical and mental well-being. The review synthesized existing research on the development and application of stress prediction models, underscoring advancements, challenges, and future directions in this rapidly evolving domain. Emphasis was placed on examining and critically evaluating the existing research and literature on stress prediction, physiological data analysis, and wearable devices for stress monitoring. The synthesis of findings aimed to contribute to a better understanding of the potential of wearable technology in objectively assessing and predicting stress levels in real time, thereby informing the design of effective interventions and personalized stress management approaches.

Noninvasive monitoring technologies to identify discomfort and distressing symptoms in persons with limited communication at the end of life: a scoping review

BACKGROUND

Discomfort and distressing symptoms are common at the end of life, while people in this stage are often no longer able to express themselves. Technologies may aid clinicians in detecting and treating these symptoms to improve end-of-life care. This review provides an overview of noninvasive monitoring technologies that may be applied to persons with limited communication at the end of life to identify discomfort.

METHODS

A systematic search was performed in nine databases, and experts were consulted. Manuscripts were included if they were written in English, Dutch, German, French, Japanese or Chinese, if the monitoring technology measured discomfort or distressing symptoms, was noninvasive, could be continuously administered for 4 hours and was potentially applicable for bed-ridden people. The screening was performed by two researchers independently. Information about the technology, its clinimetrics (validity, reliability, sensitivity, specificity, responsiveness), acceptability, and feasibility were extracted.

RESULTS

Of the 3,414 identified manuscripts, 229 met the eligibility criteria. A variety of monitoring technologies were identified, including actigraphy, brain activity monitoring, electrocardiography, electrodermal activity monitoring, surface electromyography, incontinence sensors, multimodal systems, and noncontact monitoring systems. The main indicators of discomfort monitored by these technologies were sleep, level of consciousness, risk of pressure ulcers, urinary incontinence, agitation, and pain. For the end-of-life phase, brain activity monitors could be helpful and acceptable to monitor the level of consciousness during palliative sedation. However, no manuscripts have reported on the clinimetrics, feasibility, and acceptability of the other technologies for the end-of-life phase.

CONCLUSIONS

Noninvasive monitoring technologies are available to measure common symptoms at the end of life. Future research should evaluate the quality of evidence provided by existing studies and investigate the feasibility, acceptability, and usefulness of these technologies in the end-of-life setting. Guidelines for studies on healthcare technologies should be better implemented and further developed.

Sparse Multichannel Decomposition of Electrodermal Activity With Physiological Priors

Goal: Inferring autonomous nervous system (ANS) activity is a challenging issue and has critical applications in stress regulation. Sweat secretions caused by ANS activity influence the electrical conductance of the skin. Therefore, the variations in skin conductance (SC) measurements reflect the sudomotor nerve activity (SMNA) and can be used to infer the underlying ANS activity. These variations are strongly correlated with emotional arousal as well as thermoregulation. However, accurately recovering ANS activity and the corresponding state-space system from a single channel signal is difficult due to artifacts introduced by measurement noise. To minimize the impact of noise on inferring ANS activity, we utilize multiple channels of SC data. Methods: We model skin conductance using a second-order differential equation incorporating a time-shifted sparse impulse train input in combination with independent cubic basis spline functions. Finally, we develop a block coordinate descent method for SC signal decomposition by employing a generalized cross-validation sparse recovery approach while including physiological priors. Results: We analyze the experimental data to validate the performance of the proposed algorithm. We demonstrate its capacity to recover the ANS activations, the underlying physiological system parameters, and both tonic and phasic components. Finally, we present an overview of the algorithm's comparative performance under varying conditions and configurations to substantiate its ability to accurately model ANS activity. Our results show that our algorithm performs better in terms of multiple metrics like noise performance, AUC score, the goodness of fit of reconstructed signal, and lower missing impulses compared with the single channel decomposition approach. Conclusion: In this study, we highlight the challenges and benefits of concurrent decomposition and deconvolution of multichannel SC signals.

Assessing fatigue and sleep in chronic diseases using physiological signals from wearables: A pilot study

Problems with fatigue and sleep are highly prevalent in patients with chronic diseases and often rated among the most disabling symptoms, impairing their activities of daily living and the health-related quality of life (HRQoL). Currently, they are evaluated primarily via Patient Reported Outcomes (PROs), which can suffer from recall biases and have limited sensitivity to temporal variations. Objective measurements from wearable sensors allow to reliably quantify disease state, changes in the HRQoL, and evaluate therapeutic outcomes. This work investigates the feasibility of capturing continuous physiological signals from an electrocardiography-based wearable device for remote monitoring of fatigue and sleep and quantifies the relationship of objective digital measures to self-reported fatigue and sleep disturbances. 136 individuals were followed for a total of 1,297 recording days in a longitudinal multi-site study conducted in free-living settings and registered with the German Clinical Trial Registry (DRKS00021693). Participants comprised healthy individuals (N = 39) and patients with neurodegenerative disorders (NDD, N = 31) and immune mediated inflammatory diseases (IMID, N = 66). Objective physiological measures correlated with fatigue and sleep PROs, while demonstrating reasonable signal quality. Furthermore, analysis of heart rate recovery estimated during activities of daily living showed significant differences between healthy and patient groups. This work underscores the promise and sensitivity of novel digital measures from multimodal sensor time-series to differentiate chronic patients from healthy individuals and monitor their HRQoL. The presented work provides clinicians with realistic insights of continuous at home patient monitoring and its practical value in quantitative assessment of fatigue and sleep, an area of unmet need.

Psychophysiological Arousal in Young Children Who Stutter: An Interpretable AI Approach

The presented first-of-its-kind study effectively identifies and visualizes the second-by-second pattern differences in the physiological arousal of preschool-age children who do stutter (CWS) and who do not stutter (CWNS) while speaking perceptually fluently in two challenging conditions: speaking in stressful situations and narration. The first condition may affect children's speech due to high arousal; the latter introduces linguistic, cognitive, and communicative demands on speakers. We collected physiological parameters data from 70 children in the two target conditions. First, we adopt a novel modality-wise multiple-instance-learning (MI-MIL) approach to classify CWS vs. CWNS in different conditions effectively. The evaluation of this classifier addresses four critical research questions that align with state-of-the-art speech science studies' interests. Later, we leverage SHAP classifier interpretations to visualize the salient, fine-grain, and temporal physiological parameters unique to CWS at the population/group-level and personalized-level. While group-level identification of distinct patterns would enhance our understanding of stuttering etiology and development, the personalized-level identification would enable remote, continuous, and real-time assessment of stuttering children's physiological arousal, which may lead to personalized, just-in-time interventions, resulting in an improvement in speech fluency. The presented MI-MIL approach is novel, generalizable to different domains, and real-time executable. Finally, comprehensive evaluations are done on multiple datasets, presented framework, and several baselines that identified notable insights on CWSs' physiological arousal during speech production.

Singing and music making: physiological responses across early to later stages of dementia

Background: Music based interventions have been found to improve wellbeing for people with dementia. More recently there has been interest in physiological measures to provide additional information about how music and singing impact this population. Methods: This multiple-case study design explored physiological responses (heart rate-HR, electrodermal activity-EDA, movement, and skin temperature-ST) of nine people with mild-to-moderate using simulation modelling analysis.             Results: In study 1,  the singing group showed an increase in EDA (p < 0.01 for 8/9 participants) and HR (p < 0.01 for 5/9 participants) as the session began. HR (p < 0.0001 for 5/9 participants) and ST (p < 0.0001 for 6/9 participants) increased during faster tempos. EDA (p < 0.01 all), movement (p < 0.01 for 8/9 participants) and engagement were higher during singing compared to a baseline control. In study 2 EDA (p < 0.0001 for 14/18 data points [3 music conditions across 6 participants]) and ST (p < 0.001 for 10/18 data points) increased and in contrast to the responses during singing, HR decreased as the sessions began (p < 0.002 for 9/18 data points). EDA was higher during slower music (p < 0.0001 for 13/18 data points), however this was less consistent in more interactive sessions than the control. There were no consistent changes in HR and movement responses during different music genre.   Conclusions: Physiological measures provide valuable information about the experiences of people with dementia participating in musical activities, particularly for those with verbal communication difficulties. Future research should consider using physiological measures. video-analysis and observational measures to explore further how engagement in specific activities, wellbeing and physiology interact.

Medically-oriented design for explainable AI for stress prediction from physiological measurements

Background

In the last decade, a lot of attention has been given to develop artificial intelligence (AI) solutions for mental health using machine learning. To build trust in AI applications, it is crucial for AI systems to provide for practitioners and patients the reasons behind the AI decisions. This is referred to as Explainable AI. While there has been significant progress in developing stress prediction models, little work has been done to develop explainable AI for mental health.

Methods

In this work, we address this gap by designing an explanatory AI report for stress prediction from wearable sensors. Because medical practitioners and patients are likely to be familiar with blood test reports, we modeled the look and feel of the explanatory AI on those of a standard blood test report. The report includes stress prediction and the physiological signals related to stressful episodes. In addition to the new design for explaining AI in mental health, the work includes the following contributions: Methods to automatically generate different components of the report, an approach for evaluating and validating the accuracies of the explanations, and a collection of ground truth of relationships between physiological measurements and stress prediction.

Results

Test results showed that the explanations were consistent with ground truth. The reference intervals for stress versus non-stress were quite distinctive with little variation. In addition to the quantitative evaluations, a qualitative survey, conducted by three expert psychiatrists confirmed the usefulness of the explanation report in understanding the different aspects of the AI system.

Conclusion

In this work, we have provided a new design for explainable AI used in stress prediction based on physiological measurements. Based on the report, users and medical practitioners can determine what biological features have the most impact on the prediction of stress in addition to any health-related abnormalities. The effectiveness of the explainable AI report was evaluated using a quantitative and a qualitative assessment. The stress prediction accuracy was shown to be comparable to state-of-the-art. The contributions of each physiological signal to the stress prediction was shown to correlate with ground truth. In addition to these quantitative evaluations, a qualitative survey with psychiatrists confirmed the confidence and effectiveness of the explanation report in the stress made by the AI system. Future work includes the addition of more explanatory features related to other emotional states of the patient, such as sadness, relaxation, anxiousness, or happiness.

Singing and music making: physiological responses across early to later stages of dementia

Background: Music based interventions have been found to improve the wellbeing of people living with dementia. More recently there has been interest in physiological measures to provide additional information about how music and singing impact this population. Methods: This multiple-case study design explored physiological responses (heart rate-HR, electrodermal activity-EDA, movement, and skin temperature-ST) of nine people with mild-to-moderate dementia during a singing group, and six people in the later stages of dementia during an interactive music group. The interactive music group was also video recorded to provide information about engagement. Data were analysed using simulation modelling analysis.             Results: The singing group showed an increase in EDA (p < 0.01 for 8/9 participants) and HR (p < 0.01 for 5/9 participants) as the session began. HR (p < 0.0001 for 5/9 participants) and ST (p < 0.0001 for 6/9 participants) increased during faster paced songs. EDA (p < 0.01 all), movement (p < 0.01 for 8/9 participants) and engagement were higher during an interactive music group compared to a control session (music listening). EDA (p < 0.0001 for 14/18 participants) and ST (p < 0.001 for 10/18 participants) increased and in contrast to the responses during singing, HR decreased as the sessions began (p < 0.002 for 9/18 participants). EDA was higher during slower music (p < 0.0001 for 13/18 participants), however this was less consistent in more interactive sessions than the control. There were no consistent changes in HR and movement responses during different styles of music.   Conclusions: Physiological measures may provide valuable information about the experiences of people with dementia participating in arts and other activities, particularly for those with verbal communication difficulties. Future research should consider using physiological measures with video-analysis and observational measures to explore further how engagement in specific activities, wellbeing and physiology interact.

Quantitative Assessment of Stress Through EEG During a Virtual Reality Stress-Relax Session

Recent studies have addressed stress level classification via electroencephalography (EEG) and machine learning. These works typically use EEG-based features, like power spectral density (PSD), to develop stress classifiers. Nonetheless, these classifiers are usually limited to the discrimination of two (stress and no stress) or three (low, medium, and high) stress levels. In this study we propose an alternative for quantitative stress assessment based on EEG and regression algorithms. To this aim, we conducted a group of 23 participants (mean age 22.65 ± 5.48) over a stress-relax experience while monitoring their EEG. First, we stressed the participants via the Montreal imaging stress task (MIST), and then we led them through a 360-degree virtual reality (VR) relaxation experience. Throughout the session, the participants reported their self-perceived stress level (SPSL) via surveys. Subsequently, we extracted spectral features from the EEG of the participants and we developed individual models based on regression algorithms to predict their SPSL. We evaluated stress regression performance in terms of the mean squared percentage error (MSPE) and the correlation coefficient (R²). The results yielded from this evaluation (MSPE = 10.62 ± 2.12, R² = 0.92 ± 0.02) suggest that our approach predicted the stress level of the participants with remarkable performance. These results may have a positive impact in diverse areas that could benefit from stress level quantitative prediction. These areas include research fields like neuromarketing, and training of professionals such as surgeons, industrial workers, or firefighters, that often face stressful situations.

Singing and music making: physiological responses across early to later stages of dementia

Background: Music based interventions have been found to improve the wellbeing of people living with dementia. More recently there has been interest in physiological measures to provide additional information about how music and singing impact this population. Methods: This multiple-case study design explored physiological responses (heart rate-HR, electrodermal activity-EDA, movement, and skin temperature-ST) of nine people with mild-to-moderate dementia during a singing group, and six people in the later stages of dementia during an interactive music group. The interactive music group was also video recorded to provide information about engagement. Data were analysed using simulation modelling analysis.             Results: The singing group showed an increase in EDA (p < 0.01 for 8/9 participants) and HR (p < 0.01 for 5/9 participants) as the session began. HR (p < 0.0001 for 5/9 participants) and ST (p < 0.0001 for 6/9 participants) increased during faster paced songs. EDA (p < 0.01 all), movement (p < 0.01 for 8/9 participants) and engagement were higher during an interactive music group compared to a control session (music listening). EDA (p < 0.0001 for 14/18 participants) and ST (p < 0.001 for 10/18 participants) increased and in contrast to the responses during singing, HR decreased as the sessions began (p < 0.002 for 9/18 participants). EDA was higher during slower music (p < 0.0001 for 13/18 participants), however this was less consistent in more interactive sessions than the control. There were no consistent changes in HR and movement responses during different styles of music.   Conclusions: Physiological measures may provide valuable information about the experiences of people with dementia participating in arts and other activities, particularly for those with verbal communication difficulties. Future research should consider using physiological measures with video-analysis and observational measures to explore further how engagement in specific activities, wellbeing and physiology interact.

Examining Response to Negative Life Events Through Fitness Tracker Data

Negative life events, such as the death of a loved one, are an unavoidable part of life. These events can be overwhelmingly stressful and may lead to the development of mental health disorders. To mitigate these adverse developments, prior literature has utilized measures of psychological responses to negative life events to better understand their effects on mental health. However, psychological changes represent only one aspect of an individual's potential response. We posit measuring additional dimensions of health, such as physical health, may also be beneficial, as physical health itself may be affected by negative life events and measuring its response could provide context to changes in mental health. Therefore, the primary aim of this work was to quantify how an individual's physical health changes in response to negative life events by testing for deviations in their physiological and behavioral state (PB-state). After capturing post-event, PB-state responses, our second aim sought to contextualize changes within known factors of psychological response to negative life events, namely coping strategies. To do so, we utilized a cohort of professionals across the United States monitored for 1 year and who experienced a negative life event while under observation. Garmin Vivosmart-3 devices provided a multidimensional representation of one's PB-state by collecting measures of resting heart rate, physical activity, and sleep. To test for deviations in PB-state following negative life events, One-Class Support Vector Machines were trained on a window of time prior to the event, which established a PB-state baseline. The model then evaluated participant's PB-state on the day of the life event and each day that followed, assigning each day a level of deviance relative to the participant's baseline. Resulting response curves were then examined in association with the use of various coping strategies using Bayesian gamma-hurdle regression models. The results from our objectives suggest that physical determinants of health also deviate in response to negative life events and that these deviations can be mitigated through different coping strategies. Taken together, these observations stress the need to examine physical determinants of health alongside psychological determinants when investigating the effects of negative life events.

A Review of Biophysiological and Biochemical Indicators of Stress for Connected and Preventive Healthcare

Stress is a known contributor to several life-threatening medical conditions and a risk factor for triggering acute cardiovascular events, as well as a root cause of several social problems. The burden of stress is increasing globally and, with that, is the interest in developing effective stress-monitoring solutions for preventive and connected health, particularly with the help of wearable sensing technologies. The recent development of miniaturized and flexible biosensors has enabled the development of connected wearable solutions to monitor stress and intervene in time to prevent the progression of stress-induced medical conditions. This paper presents a review of the literature on different physiological and chemical indicators of stress, which are commonly used for quantitative assessment of stress, and the associated sensing technologies.

Pain and Stress Detection Using Wearable Sensors and Devices-A Review

Pain is a subjective feeling; it is a sensation that every human being must have experienced all their life. Yet, its mechanism and the way to immune to it is still a question to be answered. This review presents the mechanism and correlation of pain and stress, their assessment and detection approach with medical devices and wearable sensors. Various physiological signals (i.e., heart activity, brain activity, muscle activity, electrodermal activity, respiratory, blood volume pulse, skin temperature) and behavioral signals are organized for wearables sensors detection. By reviewing the wearable sensors used in the healthcare domain, we hope to find a way for wearable healthcare-monitoring system to be applied on pain and stress detection. Since pain leads to multiple consequences or symptoms such as muscle tension and depression that are stress related, there is a chance to find a new approach for chronic pain detection using daily life sensors or devices. Then by integrating modern computing techniques, there is a chance to handle pain and stress management issue.

Electronic Devices for Stress Detection in Academic Contexts during Confinement Because of the COVID-19 Pandemic

This article studies the development and implementation of different electronic devices for measuring signals during stress situations, specifically in academic contexts in a student group of the Engineering Department at the University of Pamplona (Colombia). For the research’s development, devices for measuring physiological signals were used through a Galvanic Skin Response (GSR), the electrical response of the heart by using an electrocardiogram (ECG), the electrical activity produced by the upper trapezius muscle (EMG), and the development of an electronic nose system (E-nose) as a pilot study for the detection and identification of the Volatile Organic Compounds profiles emitted by the skin. The data gathering was taken during an online test (during the COVID-19 Pandemic), in which the aim was to measure the student’s stress state and then during the relaxation state after the exam period. Two algorithms were used for the data process, such as Linear Discriminant Analysis and Support Vector Machine through the Python software for the classification and differentiation of the assessment, achieving 100% of classification through GSR, 90% with the E-nose system proposed, 90% with the EMG system, and 88% success by using ECG, respectively.

Protocol of the STRess at Work (STRAW) Project: How to Disentangle Day-to-Day Occupational Stress among Academics Based on EMA, Physiological Data, and Smartphone Sensor and Usage Data

Several studies have reported on increasing psychosocial stress in academia due to work environment risk factors like job insecurity, work-family conflict, research grant applications, and high workload. The STRAW project adds novel aspects to occupational stress research among academic staff by measuring day-to-day stress in their real-world work environments over 15 working days. Work environment risk factors, stress outcomes, health-related behaviors, and work activities were measured repeatedly via an ecological momentary assessment (EMA), specially developed for this project. These results were combined with continuously tracked physiological stress responses using wearable devices and smartphone sensor and usage data. These data provide information on workplace context using our self-developed Android smartphone app. The data were analyzed using two approaches: 1) multilevel statistical modelling for repeated data to analyze relations between work environment risk factors and stress outcomes on a within- and between-person level, based on EMA results and a baseline screening, and 2) machine-learning focusing on building prediction models to develop and evaluate acute stress detection models, based on physiological data and smartphone sensor and usage data. Linking these data collection and analysis approaches enabled us to disentangle and model sources, outcomes, and contexts of occupational stress in academia.

A Critical Review of Ultra-Short-Term Heart Rate Variability Norms Research

Heart rate variability (HRV) is the fluctuation in time between successive heartbeats and is defined by interbeat intervals. Researchers have shown that short-term (∼5-min) and long-term (≥24-h) HRV measurements are associated with adaptability, health, mobilization, and use of limited regulatory resources, and performance. Long-term HRV recordings predict health outcomes heart attack, stroke, and all-cause mortality. Despite the prognostic value of long-term HRV assessment, it has not been broadly integrated into mainstream medical care or personal health monitoring. Although short-term HRV measurement does not require ambulatory monitoring and the cost of long-term assessment, it is underutilized in medical care. Among the diverse reasons for the slow adoption of short-term HRV measurement is its prohibitive time cost (∼5 min). Researchers have addressed this issue by investigating the criterion validity of ultra-short-term (UST) HRV measurements of less than 5-min duration compared with short-term recordings. The criterion validity of a method indicates that a novel measurement procedure produces comparable results to a currently validated measurement tool. We evaluated 28 studies that reported UST HRV features with a minimum of 20 participants; of these 17 did not investigate criterion validity and 8 primarily used correlational and/or group difference criteria. The correlational and group difference criteria were insufficient because they did not control for measurement bias. Only three studies used a limits of agreement (LOA) criterion that specified a priori an acceptable difference between novel and validated values in absolute units. Whereas the selection of rigorous criterion validity methods is essential, researchers also need to address such issues as acceptable measurement bias and control of artifacts. UST measurements are proxies of proxies. They seek to replace short-term values which, in turn, attempt to estimate long-term metrics. Further adoption of UST HRV measurements requires compelling evidence that these metrics can forecast real-world health or performance outcomes. Furthermore, a single false heartbeat can dramatically alter HRV metrics. UST measurement solutions must automatically edit artifactual interbeat interval values otherwise HRV measurements will be invalid. These are the formidable challenges that must be addressed before HRV monitoring can be accepted for widespread use in medicine and personal health care.

Wearable sensor-based evaluation of psychosocial stress in patients with metabolic syndrome

The prevalence of metabolic disorders has increased rapidly as such they become a major health issue recently. Despite the definition of genetic associations with obesity and cardiovascular diseases, they constitute only a small part of the incidence of disease. Environmental and physiological effects such as stress, behavioral and metabolic disturbances, infections, and nutritional deficiencies have now revealed as contributing factors to develop metabolic diseases. This study presents a multivariate methodology for the modeling of stress on metabolic syndrome (MES) patients. We have developed a supporting system to cope with MES patients' anxiety and stress by means of several biosignals such as ECG, GSR, body temperature, SpO₂, glucose level, and blood pressure that are measured by a wearable device. We employed a neural network model to classify emotions with HRV analysis in the detection of stressor moments. We have accurately recognized the stressful situations using physiological responses to stimuli by utilizing our proposed affective state detection algorithm. We evaluated our system with a dataset of 312 biosignal records from 30 participants and the results showed that our proposed method achieved an average accuracy of 92% and 89% in distinguishing stress level in MES and other groups respectively. Both being the focus of an MES group and others proved to be highly arousing experiences which were significantly reflected in the physiological signal. Exposure to the stress in MES and cardiovascular heart disease patients increases the chronic symptoms. An early stage of comprehensive intervention may reduce the risk of general cardiovascular events in these particular groups. In this context, the use of e-health applications such as our proposed system facilitates these processes.

A Novel Filter for Tracking Real-World Cognitive Stress using Multi-Time-Scale Point Process Observations

Determining the relationship between neurocognitive stress and changes in physiological signals is an important aspect of wearable monitoring. We present a state-space approach for tracking stress from skin conductance and electrocardiography measurements. Individual skin conductance responses (SCRs) are a primary source of information in a skin conductance signal and their rate of occurrence is related to psychological arousal. Likewise, heart rate too varies with emotion. We model SCRs and heartbeats as two different stress-related point processes linked to the same sympathetic nervous system activation. We derive Kalman-like filter equations for tracking stress and use both expectation-maximization and maximum likelihood estimation for parameter recovery. Our preliminary results show that stress is high when a task is unfamiliar, but reduces gradually with familiarity, albeit in the presence of other external stressors. The method demonstrates the feasibility of tracking real-world stress using skin conductance and heart rate measurements. It also serves as a novel state estimation framework for multiple point process observations on different time scales.

Continuous Stress Detection Using Wearable Sensors in Real Life: Algorithmic Programming Contest Case Study

The negative effects of mental stress on human health has been known for decades. High-level stress must be detected at early stages to prevent these negative effects. After the emergence of wearable devices that could be part of our lives, researchers have started detecting extreme stress of individuals with them during daily routines. Initial experiments were performed in laboratory environments and recently a number of works took a step outside the laboratory environment to the real-life. We developed an automatic stress detection system using physiological signals obtained from unobtrusive smart wearable devices which can be carried during the daily life routines of individuals. This system has modality-specific artifact removal and feature extraction methods for real-life conditions. We further tested our system in a real-life setting with collected physiological data from 21 participants of an algorithmic programming contest for nine days. This event had lectures, contests as well as free time. By using heart activity, skin conductance and accelerometer signals, we successfully discriminated contest stress, relatively higher cognitive load (lecture) and relaxed time activities by using different machine learning methods.

Ultra-short term HRV features as surrogates of short term HRV: a case study on mental stress detection in real life

Background

This paper suggests a method to assess the extent to which ultra-short Heart Rate Variability (HRV) features (less than 5 min) can be considered as valid surrogates of short HRV features (nominally 5 min). Short term HRV analysis has been widely investigated for mental stress assessment, whereas the validity of ultra-short HRV features remains unclear. Therefore, this study proposes a method to explore the extent to which HRV excerpts can be shortened without losing their ability to automatically detect mental stress.

Methods

ECGs were acquired from 42 healthy subjects during a university examination and resting condition. 23 features were extracted from HRV excerpts of different lengths (i.e., 30 s, 1 min, 2 min, 3 min, and 5 min). Significant differences between rest and stress phases were investigated using non-parametric statistical tests at different time-scales. Features extracted from each ultra-short length were compared with the standard short HRV features, assumed as the benchmark, via Spearman’s rank correlation analysis and Bland-Altman plots during rest and stress phases. Using data-driven machine learning approaches, a model aiming to detect mental stress was trained, validated and tested using short HRV features, and assessed on the ultra-short HRV features.

Results

Six out of 23 ultra-short HRV features (MeanNN, StdNN, MeanHR, StdHR, HF, and SD2) displayed consistency across all of the excerpt lengths (i.e., from 5 to 1 min) and 3 out of those 6 ultra-short HRV features (MeanNN, StdHR, and HF) achieved good performance (accuracy above 88%) when employed in a well-dimensioned automatic classifier.

Conclusion

This study concluded that 6 ultra-short HRV features are valid surrogates of short HRV features for mental stress investigation.

Electronic supplementary material

The online version of this article (10.1186/s12911-019-0742-y) contains supplementary material, which is available to authorized users.

The Impact of Using Measurements of Electrodermal Activity in the Assessment of Problematic Behaviour in Dementia

Background

A major and complex challenge when trying to support individuals with dementia is meeting the needs of those who experience changes in behaviour and mood.

Aim

To explore how a sensor measuring electrodermal activity (EDA) impacts assistant nurses’ structured assessments of problematic behaviours amongst people with dementia and their choices of care interventions.

Methods

Fourteen individuals with dementia wore a sensor that measured EDA. The information from the sensor was presented to assistant nurses during structured assessments of problematic behaviours. The evaluation process included scorings with the instrument NPI-NH (Neuropsychiatric Inventory-Nursing Home version), the care interventions suggested by assistant nurses to decrease problematic behaviours, and the assistant nurses’ experiences obtained by focus group interviews.

Results

The information from the sensor measuring EDA was perceived to make behavioural patterns more visual and clear, which enhanced assistant nurses’ understanding of time-related patterns of behaviours. In turn, this enhancement facilitated timely care interventions to prevent the patterns and decrease the levels of problematic behaviour.

Conclusion

With the addition of information from the sensor, nursing staff could target causes and triggers in a better way, making care interventions more specific and directed towards certain times throughout the day to prevent patterns of problematic behaviours.

Stress Detection Using Low Cost Heart Rate Sensors

The automated detection of stress is a central problem for ambient assisted living solutions. The paper presents the concepts and results of two studies targeted at stress detection with a low cost heart rate sensor, a chest belt. In the device validation study (n = 5), we compared heart rate data and other features from the belt to those measured by a gold standard device to assess the reliability of the sensor. With simple synchronization and data cleaning algorithm, we were able to select highly (>97%) correlated, low average error (2.2%) data segments of considerable length from the chest data for further processing. The protocol for the clinical study (n = 46) included a relax phase followed by a phase with provoked mental stress, 10 minutes each. We developed a simple method for the detection of the stress using only three time-domain features of the heart rate signal. The method produced accuracy of 74.6%, sensitivity of 75.0%, and specificity of 74.2%, which is impressive compared to the performance of two state-of-the-art methods run on the same data. Since the proposed method uses only time-domain features, it can be efficiently implemented on mobile devices.

Objective detection of chronic stress using physiological parameters

The aim of this study was to design a system to diagnose chronic stress, based on blunted reactivity of the autonomic nervous system (ANS) to cognitive load (CL). The system concurrently measures CL-induced variations in pupil diameter (PD), heart rate (HR), pulse wave amplitude (PWA), galvanic skin response (GSR), and breathing rate (BR). Measurements were recorded from 58 volunteers whose stress level was identified using the State-Trait Anxiety Inventory. Number-multiplication questions were used as CLs. HR, PWA, GSR, and PD were significantly (p < 0.05) changed during CL. CL-induced changes in PWA (16.87 ± 21.39), GSR (- 13.71 ± 7.86), and PD (11.56 ± 9.85) for non-stressed subjects (n = 36) were significantly different (p < 0.05) from those in PWA (2.92 ± 12.89), GSR (- 6.87 ± 9.54), and PD (4.51 ± 10.94) for stressed subjects (n = 22). ROC analysis for PWA, GSR, and PD illustrated their usefulness to identify stressed subjects. By inputting all features to different classification algorithms, up to 91.7% of sensitivity and 89.7% of accuracy to identify stressed subjects were achieved using 10-fold cross-validation. This study was the first to document blunted CL-induced changes in PWA, GSR, and PD in stressed subjects, compared to those in non-stressed subjects. Preliminary results demonstrated the ability of our system to objectively detect chronic stress with good accuracy, suggesting the potential for monitoring stress to prevent dangerous stress-related diseases. Graphical abstract Chronic stress degrads the autonomic nervous system reaction to cognitive loads. Measurement of reduced changes in physiological signals during asking math questions was useful to identify people with high STAI score (stressed subjects).

A Trust Assessment mechanism for Ubiquitous Healthcare environment employing cloud theory

Mental healthcare domain highlights the significance of trustworthiness between patient and psychiatrist for treatment process. In this paper, the issue of assessing psychiatrist trustworthiness from patient perspective, within a Ubiquitous Healthcare (UH) environment, is addressed. To meet that challenge, a Trust Assessment mechanism mimicking human cognitive judgment, is proposed. The exploitation of innovative fuzzy-probabilistic transformation model, denoted as cloud, for mechanism deployment enables fuzziness as well as adhered randomness of cognitive perception and assessment to be captured. A set of simulations within MATLAB software environment verify the introduced mechanism efficiency.

Are ultra-short heart rate variability features good surrogates of short-term ones? State-of-the-art review and recommendations

Ultra-short heart rate variability (HRV) analysis refers to the study of HRV features in excerpts of length <5 min. Ultra-short HRV is widely growing in many healthcare applications for monitoring individual's health and well-being status, especially in combination with wearable sensors, mobile phones, and smart-watches. Long-term (nominally 24 h) and short-term (nominally 5 min) HRV features have been widely investigated, physiologically justified and clear guidelines for analysing HRV in 5 min or 24 h are available. Conversely, the reliability of ultra-short HRV features remains unclear and many investigations have adopted ultra-short HRV analysis without questioning its validity. This is partially due to the lack of accepted algorithms guiding investigators to systematically assess ultra-short HRV reliability. This Letter critically reviewed the existing literature, aiming to identify the most suitable algorithms, and harmonise them to suggest a standard protocol that scholars may use as a reference in future studies. The results of the literature review were surprising, because, among the 29 reviewed papers, only one paper used a rigorous method, whereas the others employed methods that were partially or completely unreliable due to the incorrect use of statistical tests. This Letter provides recommendations on how to assess ultra-short HRV features reliably and proposes an inclusive algorithm that summarises the state-of-the-art knowledge in this area.

On the Efficiency of Individualized Theta/Beta Ratio Neurofeedback Combined with Forehead EMG Training in ADHD Children

Background: Neurofeedback training (NFT) to decrease the theta/beta ratio (TBR) has been used for treating hyperactivity and impulsivity in attention deficit hyperactivity disorder (ADHD); however, often with low efficiency. Individual variance in EEG profile can confound NFT, because it may lead to influencing non-relevant activity, if ignored. More importantly, it may lead to influencing ADHD-related activities adversely, which may even result in worsening ADHD symptoms. Electromyogenic (EMG) signal resulted from forehead muscles can also explain the low efficiency of the NFT in ADHD from both practical and psychological point-of-view. The first aim of this study was to determine EEG and EMG biomarkers most related to the main ADHD characteristics, such as impulsivity and hyperactivity. The second aim was to confirm our hypothesis that the efficiency of the TBR NFT can be increased by individual adjustment of the frequency bands and simultaneous training on forehead muscle tension. Methods: We recruited 94 children diagnosed with ADHD (ADHD) and 23 healthy controls (HC). All participants were male and aged between six and nine. Impulsivity and attention were assessed with Go/no-Go task and delayed gratification task, respectively; and 19-channel EEG and forehead EMG were recorded. Then, the ADHD group was randomly subdivided into (1) standard, (2) individualized, (3) individualized+EMG, and (4) sham NFT (control) groups. The groups were compared based on TBR and EEG alpha activity, as well as hyperactivity and impulsivity three times: pre-NFT, post-NFT and 6 months after the NFT (follow-up). Results: ADHD children were characterized with decreased individual alpha peak frequency, alpha bandwidth and alpha amplitude suppression magnitude, as well as with increased alpha1/alpha2 (a1/a2) ratio and scalp muscle tension when c (η2 ≥ 0.212). All contingent TBR NFT groups exhibited significant NFT-related decrease in TBR not evident in the control group. Moreover, we detected a higher overall alpha activity in the individualized but not in the standard NFT group. Mixed MANOVA considering between-subject factor GROUP and within-subject factor TIME showed that the individualized+EMG group exhibited the highest level of clinical improvement, which was associated with increase in the individual alpha activity at the 6 months follow-up when comparing with the other approaches (post hoc t = 3.456, p = 0.011). Conclusions: This study identified various (adjusted) alpha activity metrics as biomarkers with close relationship with ADHD symptoms, and demonstrated that TBR NFT individually adjusted for variances in alpha activity is more successful and clinically more efficient than standard, non-individualized NFT. Moreover, these training effects of the individualized TBR NFT lasted longer when combined with EMG.

Measuring Electrodermal Activity to Improve the Identification of Agitation in Individuals with Dementia

Background

Understanding and interpreting the complexity of agitation in people with dementia is challenging.

Objective

To explore whether a sensor measuring electrodermal activity (EDA) can improve the identification of agitation in individuals with dementia.

Methods

Nine individuals with dementia wore a sensor that measured EDA. During the same time, assistant nurses annotated the observed behavior of the person with dementia. A binary logistic regression model was applied to assess the relationship between the sensor and the assistant nurses' structured observations of agitation.

Results

The sensor values correlated with the assistant nurses' observations both at the time of the observation and 1 and 2 h prior to the observation.

Conclusion

A sensor measuring EDA can support early detection of agitation in persons with dementia.

Wearable depression monitoring system with heart-rate variability

A wearable depression monitoring system is proposed with an application-specific system-on-chip (SoC) solution. The SoC is designed to accelerate the filtering and feature extraction of heart-rate variability (HRV) from the electrocardiogram (ECG). Thanks to the SoC solution and planar-fashionable circuit board (P-FCB), the monitoring system becomes a low-power wearable system. Its dimension is 14cm × 7cm with 5mm thickness covering the chest band for convenient usage. In addition, with 3.7V 500mAh battery, its lifetime is at least 10 hours. For user's convenience, the system is interfacing to smart phones through Bluetooth communication. With the features of the HRV and Beck depression inventory (BDI), the smart phone application trains and classifies the user's depression scale with 71% of accuracy.

A 259.6 μW HRV-EEG Processor With Nonlinear Chaotic Analysis During Mental Tasks

A system-on-chip (SoC) with nonlinear chaotic analysis (NCA) is presented for mental task monitoring. The proposed processor treats both heart rate variability (HRV) and electroencephalography (EEG). An independent component analysis (ICA) accelerator decreases the error of HRV extraction from 5.94% to 1.84% in the preprocessing step. Largest Lyapunov exponents (LLE), as well as linear features such as mean and standard variation and sub-band power, are calculated with NCA acceleration. Measurements with mental task protocols result in confidence level of 95%. Thanks to the hardware acceleration, the chaos-processor fabricated in 0.13 μm CMOS technology consumes only 259.6 μW.

Personalized Coaching Systems to support healthy behavior in people with chronic conditions

Chronic conditions cannot be cured but daily behavior has a major effect on the severity of secondary problems and quality of life. Changing behavior however requires intensive support in daily life, which is not feasible with a human coach. A new coaching approach - so-called Personal Coaching Systems (PCSs) - use on-body sensing, combined with smart reasoning and context-aware feedback to support users in developing and maintaining a healthier behavior. Three different PCSs will be used to illustrate the different aspects of this approach: (1) Treatment of neck/shoulder pain. EMG patterns of the Trapezius muscles are used to estimate their level of relaxation. Personal vibrotactile feedback is given, to create awareness and enable learning when muscles are insufficiently relaxed. (2) Promoting a healthy activity pattern. Using a 3D accelerometer to measure activity and a smartphone to provide feedback. Timing and content of the feedback are adapted real-time, using machine-learning techniques, to optimize adherence. (3) Management of stress during daily living. The level of stress is quantified using a personal model involving a combination of different sensor signals (EMG, ECG, skin conductance, respiration). Results show that Personal Coaching Systems are feasible and a promising and challenging way forward to coach people with chronic conditions.

Wearable mental-health monitoring platform with independent component analysis and nonlinear chaotic analysis

The wearable mental-health monitoring platform is proposed for mobile mental healthcare system. The platform is headband type of 50 g and consumes 1.1 mW. For the mental health monitoring two specific functions (independent component analysis (ICA) and nonlinear chaotic analysis (NCA)) are implemented into CMOS integrated circuits. ICA extracts heart rate variability (HRV) from EEG, and then NCA extracts the largest lyapunov exponent (LLE) as physiological marker to identify mental stress and states. The extracted HRV is only 1.84% different from the HRV obtained by simple ECG measurement system. With the help of EEG signals, the proposed headband mental monitoring system shows 90% confidence level in stress test, which is better than the test results of only HRV.