Goal: This study presents the development and implementation of a mobile application designed to monitor stress in medical students using a biometric wearable device, IoT technologies, and Machine Learning algorithms. The application captures real-time physiological data, including heart rate, heart rate variability, and electrodermal activity, utilizing these parameters to predict stress levels. Data is collected using Empatica E4 wristband to then be processed by a Python-based server using signal processing techniques, which is then fed into a predictive model to classify and provide real-time stress levels into three categories: no stress, moderate stress, and high stress. Two machine learning models were employed: Standard Random Forest and Balanced Random Forest. These models achieved accuracies of 91.18% and 88.57%, respectively, demonstrating reliable classification of stress levels. Additionally, the application offers stress management tools such as articles, podcasts, and meditation guides to enhance the well-being and academic performance of medical students. The system's effectiveness was validated through a trial involving two medical science students during their clinical shifts, showcasing its potential to bolster mental health support in clinical settings.
