Multi-Feature Classification of Physiological Stress in Cardiovascular and Cardiorespiratory Interactions

Abstract

Physiological stress influences cardiovascular (CV) and cardiorespiratory (CR) control mechanisms. Accurate classification methods incorporating multiple features from the CV and CR physiological systems are necessary to better distinguish among different stress states. In this work, a novel multi-feature classification approach based on local information theory was used to differentiate postural and mental stress evoked in 127 healthy young subjects. The results evidenced a better classification performance of postural stress based on CV and CR dynamics, suggesting that these dynamics are affected more strongly by the postural than the mental challenge.