The proposed IoT-enabled multimodal mental stress monitoring method demonstrates improved accuracy in stress-level detection compared to existing solutions.
Stress is a body's natural way of responding to any kind of demand or challenge that everyone experiences from time to time. Although short-term stress typically does not impose a health burden, exposure to prolonged stress can lead to significant adverse physiological and behavioral changes. Coping with the impact of stress is a challenging task and in this context, stress assessment is essential in preventing detrimental long-term effects. The public embracement of connected wearable Internet of Things (IoT) devices, as well as the proliferation of Artificial Intelligence (AI) and Machine Learning (ML) technologies, have generated new opportunities for personalized stress tracking and management. Despite the advantages of this paradigm shift - including availability and accessibility, cost-effective delivery, and proactive intervention - still, many challenges need to be addressed to be able to develop ubiquitous solutions. In this paper, we present a comprehensive and generalizable IoT-based stress-level detection method with the following key attributes: (i) Connected: deploying vigilant IoT-based wearables and sensing technologies for continuous stress-related data collection; (ii) Data-driven: combining multimodal and heterogeneous data sources from sensor readouts; (iii) Hierarchical: consisting of device/sensor, data, intelligence, and service layers. Experimental results based on real-life stress datasets highlight the accuracy of the proposed approach for assessing the stress-level compared to state-of-the-art solutions.
Mozafari et al. (Sat,) studied this question.