Development of an IoT-Based Wearable Device for Monitoring Psychological Stress and Cardiovascular Responses

Psychological stress has a profound impact on mental and physical well-being, yet reliable, real-time monitoring tools remain scarce. In this study, we present the design and development of an IoT-based wearable device that integrates Galvanic Skin Response (GSR) and Photoplethysmogram (PPG) sensors to monitor psychological stress, mood, and risk-taking behavior. Powered by an ESP32-based microcontroller, the device processes physiological signals, including cardiovascular and skin responses, and transmits data wirelessly via the MQTT protocol. The system incorporates advanced signal processing techniques implemented in MATLAB, including filtering, feature extraction, and decomposition of the GSR signal into tonic (SCL) and phasic (SCR) components. To evaluate the system's performance, data were collected from 26 participants under three distinct conditions - relaxed, cognitive stress, and emotional stress - demonstrating that the device can effectively differentiate between these states based on physiological markers. Multiple supervised machine learning algorithms, including Logistic Regression, SVM, Neural Networks, and XGBoost, were trained and achieved average test accuracies of up to 92.5% in distinguishing stressed versus relaxed states. A custom graphical user interface (GUI) was developed to facilitate feature selection, signal visualization, and automated report generation, making the system suitable for clinical or research applications. The findings of this work provide the stage for the development of wearable devices aimed at improving mental health and stress management in daily life.