Machine Learning Based Solutions for Real-Time Stress Monitoring

Proceedings of the 2022 Workshop on Emerging Devices for Digital Biomarkers

Public health surveillance is typically done through self-report surveys. Personal smart devices that collect near real-time and zeroeffort health data can support traditional surveillance efforts by providing novel and diverse data, which can be used to predict the prevalence of conditions in a population using advanced analytics. Apple Health is one of the most popular sources of personal health data, supporting a variety of devices that collect a wide range of information from heart rate to blood pressure and sleep. This paper introduces a mobile health platform that extracts Apple Health data to support public health monitoring based on personal devices, as well as a protocol for a study that utilizes this platform to predict stress in a population. Preliminary results are also presented: Random Forests and Support Vector Machines are used to predict the participant's stress levels and achieved an accuracy of 85% and 70%, respectively. Implications for public health, challenges, limitations, and future work are also discussed. The system described in this paper is one of the first works to leverage health data from consumer-level personal devices for public health.

Sensors

Intelligent sociotechnical systems are gaining momentum in today’s information-rich society, where different technologies are used to collect data from such systems and mine this data to make useful insights about our daily activities [...]

The number of individuals in the modernworld experience elevated stress level, which is non-specific response on the body and plays a significant toll on health, productivity at work, relationships and also effect overall well-being. Many individuals are not aware of the stress triggers and potential health problems caused by prolonged stress. In order to effectively combat stress and its ill effects on health, stress triggers and responses to stress must be recognized and managed in real time. In this paper, applications of machine learning techniques are suggested to categorize and reduce stress is explored. The idea of monitoring stress and reducingstress usesmethods like personalized music, wallpaper themes, favorite games or favorite food ordering and so on. Activities which reduce stress and their degree of reduction are monitored in real time and based on customized stress reduction portfolio is designed using machine learning algorithms.

International Journal for Research in Applied Science and Engineering Technology, 2021

Stress is the part of life that is an unpleasant emotional state that individuals experience in situations like working for long hours ahead of a computer. Stress is often positive, but it can affect your health if it's chronic. Also Stress is a characteristic response to different pressure instigating factors which can prompt physiological and conduct changes. On the off chance that continues for a more extended period, stress can cause destructive consequences for our body. The body sensors alongside the idea of the Internet of Things can give rich data about one's psychological and actual wellbeing. The proposed work focuses the mind level and identifies enthusiastic changes that happened in an individual when he/she is under pressure, melancholy, or uneasiness. On recognizing, A hint message will be sent to their relatives so that they will assist that individual with emerging from his/her circumstances to fostering an IoT framework which can proficiently identify the anxiety of an individual and give an input which can help the individual to adapt to the stressors.

2021

Anxiety is your body's natural response to stress. People with anxiety disorders frequently have intense, excessive and persistent worry and fear about everyday situations. As we know stress is the main cause of long-term health disorder, Stress management is so necessary to maintain an individual's stress levels keep down and able to reduce health risks. In this article, we are finding out some measures and methods through which we employ low-cost wearable sensors collection of data and machine learning algorithms to predict an individual's stress level so as to manage his stress level. Researchers have found out that stress levels can also be detected using physiological tests such as pulse rate, heart rate amplitude and skin infections or skin diseases. This research paper' main aim is to offer a necessary analysis of various kinds of stress detection and to find out different methods to overcome it as well as a reliable checklist for more effective stress detecti...

International Journal of Computer Techniques, 2022

Due to a rapidly changing lifestyle and increasing workload, there is an urgent need to create new technologies to monitor the physical and mental health of people during their daily life. Continuous stress monitoring will help users better understand their stress patterns and provide physicians with more reliable data for interventions. Stress and fatigue can be monitored by measuring physiological parameters like Electrocardiogram (ECG), and Galvanic Skin Response (GSR) continuously over a period. Autonomic Nervous System (ANS) primarily depends on the emotional responses of the human body to the dynamic surrounding. As a result of this fact, bio-signal recordings reflecting the operating condition of the physiological systems can provide useful information representing the dynamic mental stress levels. In this paper, I gathered baseline physiological measurements of Electrocardiogram (ECG), and Galvanic Skin Response (GSR) signals while users were subjected to multiple mental stressors. Raw physiological signals available at the PHYSIONET website were used to train the classifiers for stress interference. I classified the affective states as "Low Stress", "Moderate Stress" and "High Stress" using features extracted from ECG and GSR. By using a combination of both ECG and GSR features I was able to obtain a prediction accuracy of more than 90 %.

ASIAN JOURNAL OF CONVERGENCE IN TECHNOLOGY

Stress is the mental condition of the human body that causes it's dis-functioning. It affects adversely on body parts resulting in health disorders. Traditional method of stress detection includes lab tests done by doctor. Besides traditional techniques, sensors are used to measure physiological signals, as these signals make it easy to detect stress. Based on techniques of data collection, this paper is divided into two types, one for In-lab experiment, in which participants wear various sensors on their body which is invasive for real time application while in second, data was collected from sensors which are already available in the handy devices of participant such as smartphone, wearable devices etc. Different types of sensors and their uses are explained in this paper. Automatic real time stress detection systems can be developed. This paper lists various algorithms used to gain more accuracy in detecting stress. This paper is helpful for the fellow researchers who will be working on automatic stress detection. Various studies in this domain have been reviewed and this is a primary effort in summarizing the highlights of the previous research done in stress detection domain.

International Journal of Electrical and Computer Engineering (IJECE), 2025

With the current technological advancements, particularly in sensing technologies, monitoring various health aspects, including heart rate, has become feasible. The problem addressed in this study is the need for effective stress detection methods to mitigate the significant consequences of high-intensity or long-term stress, which impacts safety and disrupts normal routines. We propose a stress detection system developed based on the convolutional neural network (CNN) method to address this. The study involves university students aged 20-22, focusing on mental stress. The dataset encompasses parameters such as heart rate, footsteps, and resting heart rate recorded through a smartwatch with 149,797-row data. Our results indicate that the CNN model achieves an 84.5% accuracy, 80.9% precision, 79.8% recall, and an 80.4% F1-score, confirming its efficacy in stress classification. The confusion matrix further validates the model's accuracy, particularly for classes 1 and 2. This research contributes significantly to the development of an effective and practical stress detection method, holding promise for enhancing well-being and preventing stress-related health issues.

IEEE Transactions on Information Technology in Biomedicine, 2000

Chronic stress is endemic to modern society. However, as it is unfeasible for physicians to continuously monitor stress levels, its diagnosis is nontrivial. Wireless body sensor networks offer opportunities to ubiquitously detect and monitor mental stress levels, enabling improved diagnosis, and early treatment. This article describes the development of a wearable sensor platform to monitor a number of physiological correlates of mental stress. We discuss tradeoffs in both system design and sensor selection to balance information content and wearability. Using experimental signals collected from the wearable sensor, we describe a selected number of physiological features that show good correlation with mental stress. In particular, we propose a new spectral feature that estimates the balance of the autonomic nervous system by combining information from the power spectral density of respiration and heart rate variability. We validate the effectiveness of our approach on a binary discrimination problem when subjects are placed under two psychophysiological conditions: mental stress and relaxation. When used in a logistic regression model, our feature set is able to discriminate between these two mental states with a success rate of 81% across subjects.

IEEE Access, 2017

Mental stress has become a social issue and could become a cause of functional disability during routine work. In addition, chronic stress could implicate several psychophysiological disorders. For example, stress increases the likelihood of depression, stroke, heart attack, and cardiac arrest. The latest neuroscience reveals that the human brain is the primary target of mental stress, because the perception of the human brain determines a situation that is threatening and stressful. In this context, an objective measure for identifying the levels of stress while considering the human brain could considerably improve the associated harmful effects. Therefore, in this paper, a machine learning (ML) framework involving electroencephalogram (EEG) signal analysis of stressed participants is proposed. In the experimental setting, stress was induced by adopting a well-known experimental paradigm based on the montreal imaging stress task. The induction of stress was validated by the task performance and subjective feedback. The proposed ML framework involved EEG feature extraction, feature selection (receiver operating characteristic curve, t-test and the Bhattacharya distance), classification (logistic regression, support vector machine and naïve Bayes classifiers) and tenfold cross validation. The results showed that the proposed framework produced 94.6% accuracy for two-level identification of stress and 83.4% accuracy for multiple level identification. In conclusion, the proposed EEG-based ML framework has the potential to quantify stress objectively into multiple levels. The proposed method could help in developing a computer-aided diagnostic tool for stress detection.

Grenze International Journal , 2024

Stress-related health issues are growing day by day. Therefore, it is important to find easier ways for accurate stress tracking on a daily basis. This research is an attempt to use photoplethysmography sensors to track stress levels in individuals. Photoplethysmography is a non-invasive optical technique used to detect blood volume changes in the microvascular bed of tissue. PPG sensors use various features like heart rate variability and pulse amplitude The main aim is to achieve real-time stress detection on a day-today basis. The system utilizes various machine learning algorithms to analyze signals and track stress patterns in people for stress management. The system also includes the use of different classifiers including Random Forest (RF), k-Nearest Neighbors (kNN), and Support Vector Machine (SVM). The paper focuses on providing an effective way to track and manage stress without interfering in the individual's dayto-day activities.

ACTA IMEKO, 2021

Developing automatic methods to measure psychological stress in everyday life has become an important research challenge. Here, we describe the design and implementation of a personalized mobile system for the detection of psychological stress episodes based on Heart-Rate Variability (HRV) indices. The system’s architecture consists of three main modules: a mobile acquisition module; an analysis-decision module; and a visualization-reporting module. Once the stress level is calculated by the mobile system, the visualization-reporting module of the mobile application displays the current stress level of the user. We carried out an experience-sampling study, involving 15 participants, monitored longitudinally, for a total of 561 ECG analyzed, to select the HRV features which best correlate with self-reported stress levels. Drawing on these results, a personalized classification system is able to automatically detect stress events from those HRV features, after a training phase in whic...

2015 International Conference on Computing, Control, Networking, Electronics and Embedded Systems Engineering (ICCNEEE), 2015

This is the era of modern life. The era of email, text messages, Facebook and Twitter, careers Crisis news coming from everywhere at any time. We (human) are assaulted with facts, pseudo facts, jibber-jabber, and rumour all posing as information. We text while we're walking across the street, catch up on email while standing in a queue. When people think they're multitasking, they're actually just switching from one task to another very rapidly. It has been found to increase the production of the stress that results overstimulate brains and cause mental fog or scrambled thinking. However, stress management should start far before the stress start causing illnesses. In this paper, a real-time personalized stress detection system from physiological signals is introduced. It is based on Pulse rate and temperature. That could record a person's stress levels.

Bulletin of Electrical Engineering and Informatics, 2023

Currently, medical experts use psychophysiological questionnaires to evaluate human stress levels during counseling or interviews. Typically, biochemical samples use urine, saliva, and blood samples to identify the effects of stress on the human body. This research explains that stress detection can be done by analyzing psychological signals and the importance of monitoring stress levels. The authors develop research on stress detection based on psychological signals. The system then processes the recorded data; the android application displays the calculation results. The database can also be accessed as a spreadsheet via a web application. The design of real-time stress detection and monitoring using internet of things (IoT) can work well.

Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2016

In today's society, working environments are becoming more stressful. The problem of occupational stress is generally recognized as one of the major factors leading to a wide spectrum of health problems. However work should, ideally, be a source of health, pride and happiness, in the sense of enhancing motivation and strengthening personal development. In this work, we present StayActive, a system which aims to detect stress and burnout risks by analyzing the behaviour of the users via their smartphone. The main purpose of StayActive is the use of the mobile sensor technology for detecting stress. Then a mobile service can recommend and present various relaxation activities "just in time" in order to allow users to carry out and solve everyday tasks and problems at work. In particular, we collect data from people's daily phone usage gathering information about the sleeping pattern, the social interaction and the physical activity of the user. We assign a weight factor to each of these three dimensions of wellbeing according to the user's personal perception and build a stress detection system. We evaluate our system in a real world environment with young adults and people working in the transportation company of Geneva. This paper highlights the architecture and model of this innovative stress detection system. The main innovation of this work is addressed in the fact that the way the stress level is computed is as less invasive as possible for the users.