Stress Detection Using Context-Aware Sensor Fusion From Wearable Devices

Wearable medical technology has become increasingly popular in recent years. One function of wearable health devices is stress detection, which relies on sensor inputs to determine a patient’s mental state. This continuous, real-time monitoring can provide healthcare professionals with vital physiological data and enhance the quality patient care. Current methods detection lack: (i) robustness—wearable sensors contain high levels measurement noise that degrades performance, (ii) adaptation—static architectures fail adapt changing contexts sensing conditions. We propose address these deficiencies SELF-CARE, generalized selective fusion method employs novel techniques context identification ensemble machine learning. SELF-CARE uses learning-based classifier process features model environmental variations conditions known as context. selectively fuse different combinations across an models perform robust classification. Our findings suggest for wrist-worn devices, measure motion are most suitable understand context, while chest-worn those detect muscle contraction. demonstrate SELF-CARE’s state-of-the-art performance WESAD dataset. Using wrist-based sensors, achieves 86.34% 94.12% accuracy 3-class 2-class classification problems, respectively. For chest-based 86.19% (3-class) 93.68% (2-class) accuracy. work demonstrates benefits utilizing selective, context-aware mobile be applied broadly Internet Things applications.