Body Core Temperature Estimation Using New Compartment Model With Vital Data From Wearable Devices

Abstract

With increasing heat-wave frequency, the prevention and public awareness of heat-related illnesses has become an essential topic. In the standard for heat strain and stress, empirical guidelines to prevent excess core temperature rise above 1C have been prescribed for workers. However, measuring core temperature change in our daily life or working place is not straightforward. The estimation of core temperature from measured vital signals in a non-invasive manner is thus essential for the management of heat stress or strain. Here, we propose an estimation method for core temperature change by a simplified thermodynamics model with the measured heart rate and ambient conditions (temperature and relative humidity). Our proposed model is based on a two-layer two-compartment model with tuned parameters, which were derived from comparison between the computations using high-resolution anatomical human body model. Our model exhibited good agreement with the measured core temperature rise; the computed and measured core temperature rise for the naked trial were 0.54 ºC and 0.53 ºC, whereas those for the clothed trial were 0.70 ºC, and 0.71 ºC, respectively. Furthermore, our compartment model with vital data measured from a wearable device achieved good estimation in real time for field measurement in addition to computational replication with a previous study.

Computed and Measured Core Temperature of Patients With Heatstroke Transported From Their Homes via Ambulance

“…Furthermore, these findings suggested the necessity to consider monitoring or tracking heat stress for several days rather than during a single day for the elderly in their homes, which is not considered in the current monitoring system. Future study includes the improvement of and application to the monitoring system [48].…”

“…The computational approach that we used was identical to that used in our previous studies [44]; thus, only the outline has been mentioned below. Previous studies have offered validations of our computational code for healthy participants [44]- [48].…”

Prospective Validation of 2B-Cool: Integrating Wearables and Individualized Predictive Analytics to Reduce Heat Injuries

“…Recently, a handful of approaches have been proposed to integrate vital signs collected from COTS wearable devices with customized algorithms to estimate T C as an alternative to invasive measurements (10)(11)(12)(13)(14)(15). These approaches use a variety of noninvasive physiological variables collected from wearables, ranging from a single heart rate (HR) variable (11) to a few vital signs (HR as well as skin temperature and skin heat flux at multiple body locations) (14) to systems requiring these plus environmental conditions (10,12,15) and anthropometric data (13).…”

“…Recently, a handful of approaches have been proposed to integrate vital signs collected from COTS wearable devices with customized algorithms to estimate T C as an alternative to invasive measurements (10)(11)(12)(13)(14)(15). These approaches use a variety of noninvasive physiological variables collected from wearables, ranging from a single heart rate (HR) variable (11) to a few vital signs (HR as well as skin temperature and skin heat flux at multiple body locations) (14) to systems requiring these plus environmental conditions (10,12,15) and anthropometric data (13). In terms of algorithms, some approaches use data-driven, machine-learning (ML) algorithms, such as regression analysis (14,15), others use ML algorithms in the form of an extended Kalman filter, where the relationships between vital signs and T C are either represented by empirical correlations (11,13) or by physics-based, energy-balance mathematical models with varying complexities (10,12).…”

“…Although these approaches show promise, yielding rootmean-square errors (RMSE) between algorithm-estimated T C and rectal-or gastrointestinal-measured temperatures of <0.50°C, the studies used for their validation have certain limitations: 1) only involved men and used very small sample sizes (10,11); 2) consisted of physical work intensities with short durations (~120 min or less), resulting in nearly monotonic increases in T C (10)(11)(12)(13)(14)(15), which are less challenging to estimate than sharp rises and drops in T C ; 3) the bulk or all measured T C never exceed 38.50°C (10,11,14), limiting or negating the relevance of the study for heat stress management (15); 4) they assessed the results using a leave-one-out procedure (13,15), which overestimates performance; 5) with one exception (12), they cannot account for between-subject variability; 6) all studies, except for one (11), involved retrospective data analysis, as opposed to a prospective, real-time assessment that more closely resembles actual operational use; and 7) all studies provide estimates of instantaneous T C (T ̂C), lacking the ability to forecast temperature values into the near future to allow for proactive interventions.…”