Biodynamics under Dual Load: Physiological variability of Stress and Exercise under Wearable Sensor analysis - A Review

The congruence of wearable sensor technology and physiological monitoring has revolutionized our perceptions of human bio-dynamics to diverse physiological loads. This is a summary of intricate trends in variability of physiological responses in conditions of co-existing stress and exercise, with particular attention to wearable sensor analysis technologies. We perform a systematic survey of literature on the topics of multi-modal physiological signal acquisition i.e. electrocardiogram (ECG), photoplethysmography (PPG), electrodermal activity (EDA) and accelerometry measurements. The article describes state of the art signal processing tools on feature extraction and machine learning tools on classifying physiological conditions in dual-load conditions. The latest developments in the deep learning feature extraction models utilizing transformers network and multimodal fusion attention ways are thoroughly examined. Such critical issues as signal quality problems, personal differences, real-time processing limitations, and barriers to clinical validation are examined. Future research priorities are on an individualized bio-dynamic model, multi-sensor fusion designs, edge-computing, and clinical applicability of wearable-based stress management and exercise optimization monitoring system. The review is a synthesis of 85 recent studies and reveals some of the research gaps they are important and forms an outline of where future research needs to go.