Human thermoregulatory system state estimation using non-invasive physiological sensors

Area of Science:

• Physiology
• Environmental Health
• Biotechnology

Background:

• High exertion work in challenging environments poses significant thermal work strain risks for emergency workers and military personnel.
• Unmitigated thermal strain can lead to heat exhaustion, heat stroke, and fatalities.
• Current methods for assessing thermal work strain often ignore thermoregulatory context and require problematic internal temperature (IT) measurements in ambulatory settings.

Purpose of the Study:

• To develop a physiology-based Dynamic Bayesian Network (DBN) model for estimating internal temperature, heat production, and heat transfer.
• To overcome limitations of existing thermal work strain indices by incorporating thermoregulatory context.
• To enable practical, real-time monitoring of thermal work strain using physiological data.

Main Methods:

• A Dynamic Bayesian Network (DBN) model was developed to estimate internal temperature, heat production, and heat transfer.
• The model utilizes physiological observations including heart rate, accelerometry, and skin heat flux.
• Model parameters were learned from data collected during a 48-hour military field training exercise involving seven volunteers.

Main Results:

• The DBN model's minute-to-minute heat production estimates showed a strong correlation with total daily energy expenditure (TDEE) measured via the doubly labeled water technique (r² = 0.73).
• The model accurately inferred internal temperature (IT) within ±0.5 °C for over 85% of new datasets.
• The DBN approach demonstrated improved estimation of critical high internal temperatures by incorporating additional thermoregulatory context compared to previous methods.

Conclusions:

• The proposed Dynamic Bayesian Network (DBN) model shows significant promise for real-time thermal work strain monitoring.
• This approach can be integrated with existing physiological monitoring systems for practical field applications.
• The DBN model offers a viable solution for preventing heat-related injuries in high-risk occupational groups.