Estimation of human core temperature from heart rate: a preliminary study for application in occupational field

Continuous monitoring of the physical state of workers is a very valuable tool to prevent the onset of possible critical situations. Technologically advanced wearable smart systems have been recently developed, which are able to collect physiological parameters and give an alert to workers, if necessary. The body core temperature (CT) represents a key parameter to be monitored, in order to follow the time evolution of the body thermal state for workers exposed to heat or cold stress. State-of-the-art estimation models of CT are based on processing heart rate (HR) data, combining non-invasiveness and prediction accuracy. Because such models make use of a database derived for a sample of army volunteers, the extension of the model to a population of workers must be investigated. In this preliminary study, we present and discuss results found for three healthy women performing three consecutive tasks (descending and ascending a work ladder, lifting a 5-kg-box and walking on a treadmill), in a climate chamber at an air temperature of 32 °C, for a total test duration of 24 minutes. Oxygen consumption (VO2), HR and CT data were collected during the tests. Results reveal a different impact of the three tasks on the CT and HR time evolutions. A good overlap was found between the time evolution of VO2 and HR, while CT, due to the body large thermal inertia, showed a much slower increase, resulting in a significant time delay and much smoother overall trend. Model predictions of the CT appear to provide a reasonable approximation to data measured during the work phases, but not during the recovery phase. Future actions are discussed that could improve the model performance in view of a possible application in workplaces.