Depression Detection by Using Wearable Sensors

This research introduces an innovative approach to depression detection utilizing wearable health sensor systems integrated with machine learning models. By integrating self-reported PHQ-9 results and physiological measurements from wearable sensors, a comprehensive dataset was assembled. Evaluation of supervised and semi-supervised ML techniques displayed notable success, with Hybrid Supervised and Semi-Supervised Approach achieving a remarkable 98% accuracy in detecting Major Depressive Disorder (MDD), marking significant advancement in mental health technology. The study highlights the potential of wearable devices in proactive mental health monitoring and intervention. Insights from exploratory data analysis (EDA) informed algorithmic development, enhancing accuracy and robustness in depression detection. This research aims to make mental health support more accessible and helpful. Our results show that this wearable system could be useful in real-life situations for detecting depression.