Accuracy of Algorithm to Non-Invasively Predict Core Body Temperature Using the Kenzen Wearable Device

Area of Science:

• Environmental Health
• Biomedical Engineering
• Occupational Safety

Background:

• Rising global temperatures due to climate change increase occupational heat exposure risks.
• Continuous core body temperature (Tc) monitoring is crucial for preventing heat injury but current methods are invasive or costly.
• Developing accessible and accurate wearable technology for real-time Tc monitoring is essential for worker safety.

Purpose of the Study:

• To validate the accuracy of a novel wearable device in predicting core body temperature (Tc) against gold-standard measurements.
• To assess the performance of a machine learning algorithm integrated into a wearable device for real-time Tc estimation.
• To determine the feasibility of using wearable technology for continuous, non-invasive worker heat stress monitoring.

Main Methods:

• Developed and validated a machine learning algorithm using real-time physiological and anthropometric data from a wearable device.
• Collected data from 52 trials across 27 unique subjects, totaling over 4000 minutes of data.
• Compared wearable device Tc predictions against gold-standard measurements (rectal probe or gastrointestinal pill).

Main Results:

• The wearable device's algorithm demonstrated high accuracy in predicting Tc, with a mean absolute error of 0.25 °C and Pearson correlation of 0.94.
• The algorithm met pre-established accuracy criteria for Tc prediction across various environmental conditions and heart rate zones.
• Accurate Tc predictions were achieved across a wide range of temperatures (13-43 °C) and for both biological sexes.

Conclusions:

• The Kenzen wearable device accurately predicts core body temperature (Tc) in real-time, offering a significant advancement in heat stress monitoring.
• This technology provides a non-invasive, cost-effective solution for continuous worker monitoring, crucial for mitigating heat injury risks.
• This study represents the first demonstration of a wearable device accurately predicting Tc in real-time, paving the way for enhanced occupational safety in hot environments.