IoT-Based Mental Health Monitoring System Using Machine Learning Stress Prediction Algorithm in Real-Time Application

With the primary focus of healthcare technologies being on the physical health of a person, mental health issues sometimes go unattended. Stress, anxiety, and depression are becoming increasingly common problems in our community leading to serious heart-related problems such as high blood pressure, episodes of heart attack and can even lead to chronic illness. Prediction of stress or depression at an earlier stage can prevent serious consequences as sometimes patients suffering from mental illness are not aware of the severity of their condition or do not keep up with counseling for a longer period of time. In this context this paper proposes a stress prediction method using machine learning to detect the development of stress or anxiety problems at an early stage. Our proposed method observes any changes in the human body under stress or depression by monitoring the ECG values and other physiological factors to predict any kind of possible stress or depression. The proposed model provided high accuracy of 98% in predicting stress. On detecting stress, appropriate actions such as informing the patient's guardian and doctor are taken. As compared with other models, our model outperforms the other state of the art models, making it a real-world predication model.