System, method, and apparatus for predicting core temperature

Abstract

A system and method are described herein for collecting biometric data from a subject wearing personal protective equipment. Such data is collected while the subject is engaged in physical activity. The data is then used to generate a model for effectively predicting core body temperature.

Description

[0001] SYSTEM, METHOD, AND APPARATUS FOR PREDICTING CORE TEMPERATURE Inventor:

[0002] Cory Coehoorn

[0003] RELATED APPLICATION

[0004] This application claims the benefit of United States Application No. 63 / 600,933, filed November 20, 2023.

[0005] TECHNICAL FIELD

[0006] Embodiments are described herein relating to measurement of core temperature of a human body.

[0007] BACKGROUND

[0008] The direct measure of core body temperature (Tc) is typically performed through invasive techniques (rectal, esophageal, or intestinal). Existing predictive methods involve complex systems with multiple pieces of impractical equipment or are otherwise unsuitable for the work environment. It is herein hypothesized that a novel, non-invasive algorithm using variables collected from a single piece of commercially available equipment could effectively predict Tc.

[0009] INCORPORATION BY REFERENCE

[0010] Each patent, patent application, and / or publication mentioned in this specification is herein incorporated by reference in its entirety to the same extent as if each individual patent, patent application, and / or publication was specifically and individually indicated to be incorporated by reference.

[0011] BRIEF DESCRIPTION OF THE FIGURES

[0012] Figure 1A shows an apparatus for measuring biometric data, under an embodiment.

[0013] Figure IB shows measured and monitored sensor data, under an embodiment.

[0014] Figure 1C shows an apparatus for measuring biometric data, under an embodiment.

[0015] Figure ID shows and apparatus for measuring biometric data, under an embodiment. Figure 2 shows rate of thermal acquisition changes for Tc and Tc. Increases in each metric were observed between data collection points.

[0016] Figure 3 shows HR changes throughout the protocol. Increases were observed at each data collection point.

[0017] Figure 4 shows RR changes throughout the protocol. Increases were observed at each data collection point.

[0018] Figure 5 shows the relationship between measured and predicted Tc (°C). The predicted Tc values are the values produced for the left-out subject when performing the bootstrapping "leave-one-out" analysis. The vertical dashed line shows participants would need to be stopped at a Tc of 38.54 °C to prevent 97.5% of the population from exceeding 39 °C.

[0019] Figure 6 shows the Bland-Altman plot of agreement between predicted and measured Tc from the data following "leave-one-out" bootstrap method (robust equation) with bias (dashed line) and lower and upper 95% limits of agreement (dotted lines).

[0020] Figure 7 shows a firefighter on the treadmill in the environmental chamber during the experimental protocol. The subject is wearing a metabolic monitoring system to measure respiratory exchange ratio (RER) to ensure steady-state exercise.

[0021] Figure 8 shows correlation between measured and predicted Tc (°C), under an embodiment. The predicted Tc values are those produced for the left-out data when performing the bootstrapping "leave-one-out" analysis.

[0022] Figure 9 shows Tsk and measured Tc quadratic relationship (Spearman r = 0.86), under an embodiment.

[0023] Figure 10 shows measured and predicted Tc (°C) values, under an embodiment.

[0024] DETAILED DESCRIPTION

[0025] Example One

[0026] Methods: The participants performed a steady-state exercise protocol in an environmental chamber (35 °C, 45% humidity) while donning firefighter personal protective equipment. The variables collected were skin temperature (Tsk), heart rate (HR), time, respiratory rate (RR), and rate of skin temperature acquisition per minute (Tsk / min).

[0027] Results: Of the variables collected, all contributed to the multiple regression model, except

[0028] HR. Tsk / min was calculated using Tsk and time. The initial model created in this study predicted Tc with a standard error of the estimate (SEE) of 0.23 °C and an adjusted R2of 0.897. Following a "leave-one-out" bootstrap method, a robust equation was created using mean coefficients. This robust equation predicted Tc with a SEE of 0.23 and an R2of 0.902.

[0029] 1. Introduction

[0030] Each day at work, firefighters and other occupational workers are exposed to excessive heat that can result in heat illness. Heat illness is a significant threat to these individuals and exists on a spectrum that includes heat cramps, heat exhaustion, and heat stroke.

[0031] Heat exhaustion is common among firefighters and others who work in high ambient temperatures and wear personal protective equipment (PPE) and includes mild symptoms such as profuse sweating, headache, weakness, vertigo, heat cramps, chills, vomiting, and nausea (Coris et al., 2004). The PPE is cumbersome, heavy, and non-permeable, significantly contributing to overall heat stress. Heat stroke, resulting from core temperature (Tc) above 40.5 °C, can exhibit more severe symptoms, including hyperventilation, muscle incoordination, agitation, poor judgment, and confusion (Coris et al., 2004). These signs and symptoms are of serious concern for individuals with pre-existing pathologies and can be exacerbated with dehydration. Statistics show that 74.8% of firefighters experience heat-related illness, and 5% of total firefighters experience heat-related illness 20 times or more in one year (Kim et al., 2019). Tc's of 38 °C and up to 40.5 °C have been associated in the literature with heat exhaustion (Lee-Chi ong & Stitt, 1995).

[0032] Specific Tc points and Tc rate of acquisition have been identified in the literature as objective measures of adverse outcomes. It has been well documented that cognitive issues are associated with increased Tc while wearing PPE in a hyperthermic environment. A rapid Tc acquisition (0.04 °C / min) up to 39.5° is related to cognitive decline, subsequently impacting one's decision-making ability (Coehoorn et al., 2020). A Tc of 38 °C during exercise in a hyperthermic environment while wearing PPE is associated with a peak, plateau, and subsequent decrease in cerebral oxygenation and hemodynamics. At the same point, overall metabolic acidosis causes individuals to reach their respiratory compensation threshold (Coehoorn et al., 2022). In addition to the cognitive and metabolic effects, research has determined that rapid Tc acquisition of 0.04 °C per minute is associated with an increased stress response as evidenced by increased salivary cortisol secretion (Coehoorn et al., 2022). Despite the consequences of elevated core temperature, employers, coaches, and other supervisory personnel rely on subjective assessments such as visual cues to identify detrimental heat-related issues. These techniques can delay the identification of the problem and subsequent treatment (Casa et al., 2015; Glazer, 2005).

[0033] Direct measurement of Tc is possible but not typically feasible in occupational or athletic environments. Multiple sites and methods of Tc measurement have been proposed; however, many of them, such as the mouth, ear canal, armpit, and forehead, have demonstrated problems with validity (Casa et al., 2007). On the other hand, sites such as the pulmonary artery and the esophagus provide accurate measurements but are invasive and require medically trained personnel (Savoie et al., 2015). The "gold standard" measurement of Tc is via rectal probes (El-Radhi, 2014). This measure is invasive and requires cumbersome equipment that cannot be readily used in the occupational field. Ingested telemetry pills provide the most user-friendly solution for direct core temperature analysis in the work environment or field of play. Telemetry pills are restricted in that fluid and food ingestion can cause artifacts in the data (Wilkinson et al., 2008), and it is impossible to determine the exact location of the pill within the gastrointestinal tract, which could confound the Tc measurements (Byrne & Lim, 2007). These pills are also a logistical and financial burden in the field as they require activation and can cost approximately $40 per pill (Bongers et al., 2015).

[0034] Due to the lack of feasibility in using direct measurement of Tc in the field it is vital to provide non-invasive and accurate methods to predict Tc. This has been attempted by many authors (Fox & Solman, 1971; Gunga et al., 2009; Huang & Chen, 2010; Kimberger et al., 2009; Teunissen et al., 2011; Yamakage & Namiki, 2003), and in many cases, the results are generally accurate, valid, and reliable. However, many of these methods rely on multiple sites for skin temperature (Tsk) and / or use multiple pieces of equipment that would add substantial time to firefighters' timelimited preparation dressing time. For example, a study (Richmond et al., 2015) was recently performed that initially utilized 30 variables for its prediction model and ultimately included four variables (insulated Tsk, microclimate temperature, HR, and work) in its final equation. The model in this study predicted Tc with a standard error of 0.27 °C and an R2= 0.86. While four variables are considered satisfactory, three required their own equipment, some of which would be difficult to use during work. In another recent study (Niedermann et al., 2014), six variables (3 Tsk measurements, 2 skin heat fluxes, and HR) were used to predict the Tc. This study predicted Tc with an R2= 0.72 and a root mean square deviation (RMSD) range between 0.28 °C and 0.34 °C for all environmental conditions (10 °C, 30 °C). Like the previous study, however, this required multiple pieces of equipment, some of which were not readily available or could not be adequately used in the field. One of the most accurate studies (Richmond et al., 2013) for occupational workers using PPE in a hyperthermic environment used insulated Tsk and the microclimate temperature to predict Tc with an R2= 0.85 and a SEE= 0.2 °C. This study, however, only had valid results for insulated Tsk values over 36.5 °C, and relied on a system that would be difficult to deploy in live occupational scenarios.

[0035] Many other studies have created models to predict core temperature using various methods and analyses. One such study by Buller et al. (Buller et al., 2010) estimated core temperature using a Kalman filter to estimate a current Tc using knowledge of the previous Tc and HR, therefore requiring an initial direct measurement of Tc. Overall, this model estimated Tc with a root mean squared error (RMSE) of 0.30 ± 0.13 °C. Another study by Buller et al. (Buller et al., 2013) used both HR and a Kalman filter. That study used a Bland-Altman limits of agreement method and found that there was an overall bias of -0.03 ± 0.32 °C and that 95% of all Tc predictions fell within ±0.63 °C. A further Buller et al. (Buller et al., 2015) study also used HR measurements from first responders to estimate Tc while wearing different levels of PPE, resulting in a bias of 0.02 °C and an RMSE of 0.21 °C. These studies using the Kalman filter method rely on an initial directly measured Tc value. While obtaining an initial Tc value seems like a minimal requirement to predict Tc accurately, it would not be practical in firefighting. Finally, another method was developed by Nakada et al. (Nakada et al., 2017) that used temperature measurement from multiple locations inside the external auditory canal and regression models to estimate Tc. The R2values ranged from 0.370 to 0.904, and the RMSE values ranged from 0.113 to 0.341, depending on the combination of sites used for measurement. While this system is promising, the study allowed for little head movement, little wind, and a constant rate of radiating heat, all of which are not replicable in the occupational environment.

[0036] Studies predicting Tc have used many collection variables to create predictive algorithms. As mentioned previously, the Buller et al. studies (Buller et al., 2010, 2013, 2015, 2020) rely heavily on HR. Other studies include heat flux measurements (Buller et al., 2011; Eggenberger et al., 2018; Niedermann et al., 2014; Welles et al., 2018; Xu et al., 2013), which take into account energy flux onto or through a surface. Tsk measurements vary significantly between studies; some use multiple Tsk sites (Niedermann et al., 2014; Richmond et al., 2015), and others use insulated Tsk (Richmond et al., 2013, 2015) on single measurement sites. These sites include, among others, the face, head, upper arms, forearms, hands, fingers, back, chest, abdomen, medial thigh, lateral thigh, posterior thigh, anterior calves, posterior calves, feet, and toes (Lenhardt & Sessler, 2006). In addition, infrared Tsk is used in some studies (Kistemaker et al., 2006; Limpabandhu et al., 2022) to estimate Tc. Other variables include relative humidity in the clothing microclimate, temperature of the clothing microclimate, respiratory rate (RR), body mass, age, body fat, sex, type of clothing, thermal comfort, thermal sensation, and sweat rate (Richmond et al., 2015).

[0037] In the present study, we sought to create a novel, non-invasive predictive model to predict Tc using one piece of equipment that could be deployed during work or play. Participants performed a steady-state exercise protocol on a treadmill wearing full firefighter PPE while in an environmental chamber (35 °C, 45% humidity), creating a rapid heat stress (RHS) scenario. RHS is contained in the general definition of uncompensable heat stress (UHS), a state where the heat loss required from sweating exceeds the evaporative capacity of the ambient environment (Cheung et al., 2000). However, RHS is more specific because it considers the thermal acquisition rate. Research demonstrates that RHS results in double the rate (0.04 °C / min) of thermal acquisition compared to a non-PPE exercise environment (0.02 °C / min) (Coehoorn et al., 2020). The rate of thermal acquisition is vital to consider because research has demonstrated changes in cognition, cerebral oxygenation, and stress resulting from elevated rates (Coehoorn et al., 2020, 2022, 2023). In this study, we hypothesized that variables collected from a single piece of commercially available equipment would accurately predict Tc (R2> 0.86). This hypothesis was created because no other research predicting Tc has considered the rate of thermal acquisition in their models. The equations in this study consider Tsk / min as a critical and primary variable. Therefore, we expected a higher R2value than in previous research.

[0038] 2. Methods

[0039] 2.1. Participants

[0040] Twenty-four adult (age: 30.5 ± 7.4 years; resting HR: 66.8 ± 4.5 bpm; body fat %: 18.6 ± 5.9%) participants (20 male, 4 female) participated in the study (see Table 1). The inclusion criteria for the subjects were that they were 18 or older, fluent in English, physically active, and could swallow a Tc capsule (e-CELCIUS® Performance, BodyCap, France), as indicated by completing a physician-prepared esophageal restriction questionnaire. The exclusion criteria were subjects 17 years old or younger, pregnant women, non-fluent in English, non-physically active, and unable to swallow a Tc capsule. Seven participants were removed from the analysis, six due to the inability to reach the final termination Tc of 39 °C and one due to unusable Tsk data. This left 17 (age: 30.7 ± 7.5 years; resting HR: 65.6 ± 10.3 bpm; body fat%: 17.9 ± 6.0%) subjects (15 male, 2 female) for our analysis. The percentage of females in this study (11%) is similar to the statistics (9%) from the National Fire Protection Association (NEP A) in 2020 (Fahy et al., 2022). The sample size of this study is similar to previous research involving the prediction of core temperature among human subjects (Gribok et al., 2010; Niedermann et al., 2014; Richmond et al., 2015). Of the 17 subjects, 10 were professional firefighters, and seven were athletes from the general population. Resting HR indicates fitness status (Kang et al., 2017), and the subjects in this study had an overall HR similar to the firefighter population (Choi et al., 2017). In addition, the body fat percentage of the subject pool was similar to the firefighter population (Jitnarin et al., 2014). The subjects in this study completed the Physical Activity Readiness Questionnaire (PAR-Q), a standardized health screening tool to indicate one's ability to perform physical activity. The institutional review board at Louisiana State University - Shreveport approved this study under IRB # 2021-00037. All participants provided written informed consent.

[0041] 2.2. Experimental design

[0042] The subjects in this study completed an RHS session in an environmental chamber while wearing full firefighter PPE. All participants became familiarized with the PPE used in the study before participation in the RHS session. The participants were asked to complete a pre-testing standardization protocol, which included not eating 2 h before the RHS session; refraining from caffeine, alcohol, and physical activity 12 h before testing; and drinking 3.7 L of water in the 24 h before testing, 500 ml of which needed to be consumed in the 2 h preceding the test (Convertino et al., 1996). A urine-specific gravity test was done before the RHS session to validate euhydration. Height, body mass, and body composition (BodPod, Cosmed, USA) were measured before RHS testing. After these anthropometric measurements, the participants were equipped with an integrated physiological monitoring system (Eq Lifemonitor, Equivital, United Kingdom). During the RHS testing, the system measured live Tsk (medical -grade infrared thermometer via one lateral chest location), time, HR, and respiratory rate (RR). The validity of these measures, when compared to standard measurement devices, are as follows (mean ± SEE, Limits of Agreement (LoA), and correlation coefficient (r)): Tsk = 0.59 ± 0.04 °C, ±0.88 °C, 0.96, HR= 1.2 ± 0.54 beats / min, ±6.6 beats / min, 0.98, RR= 0.2 ± 0.19 breaths / min, ±2.4 breaths / min, 0.97 (Liu et al., 2013). These measurements were all collected from the physiological monitoring system every 15 s. During the test, the subjects also swallowed the Tc capsule for live Tc measurement. The Tc capsule has a systematic bias of - 0.038 °C ± 0.086 °C (p < .001) and a test-retest evaluation by way of minute overall difference of 0.0095 °C ± 0.048 °C (p < .001) (Service et al., 2023). The capsule was swallowed 50-60 min before the exercise test, and no fluids were ingested during this period. The 50-60 min time frame ensures that the capsule is not influenced by ingested water and has sufficient time to enter the small intestine (Mittal et al., 1991).

[0043] The treadmill (4Front, WOODWAY®, USA) exercise protocol in the environmental chamber (35 °C, 45% relative humidity) consisted of a steady-state protocol where subjects walked at three miles per hour (mph), and the grade was adjusted to maintain a steady state as indicated by a respiratory exchange ratio (RER) (See Fig. 1). A RER between 0.9 and 0.99 avoids crossing the anaerobic threshold and therefore maintaining steady-state (Laplaud et al., 2006). Immediately before and immediately following the treadmill protocol, the subjects rested for 10-min periods. The treadmill protocol was terminated when the subjects reached a Tc of 39 °C or volitional maximum. The mean duration of the treadmill exercise protocol was 40.36 ± 3.67 min. The purpose of the treadmill protocol was to induce heat stress on the subjects during exercise to simulate the physiological stress of the work environment.

[0044] The subjects wears a metabolic monitoring system to measure respiratory exchange ratio (RER) to ensure steady-state exercise.

[0045] Figures 1A-1D show devices for measuring biometric data of subjects. Such data are then used to generate one or more predictive models as described below.

[0046] Figure 1A shows an orange box 102 representing an insulated skin temperature monitor (see Figure 1C). Red dots 104 represent respiratory and electrocardiogram sensors. These are used to calculate respiratory rate through the impedance of the chest strap and the electrocardiogram signal. Time is marked as the point at which the physical work in the hyperthermic environment begins.

[0047] The respiratory sensor simply measures the mechanical impedance of the chest strap. Additionally, the ECG sensor measures / supports the respiratory rate found by the RR sensor.

[0048] The ECG sensor measures the R-R intervals to determine inhalation and exhalation. Inhalation increases heart rate (shorter R-R intervals), and exhalation decreases heart rate (longer R-R intervals). Additionally, the QRS complex is used to support the findings in that the amplitude of the QRS complex decreases during inhalation and increases during exhalation due to the position of the ECG sensor relative to the heart.

[0049] The chest device / strap (and / or arm band as described below) includes sensors described herein and measures and monitors data shown in Figure IB. Measured or monitored data includes insulated skin temperature 110, microclimate temperature 112, insulated skin temperature per minute 114, time 116, heart rate 118, and respiratory rate 120. Sensor data is used as described herein to predict core temperature 124. As indicated above time 116 is marked as the point at which the physical work in the hyperthermic environment begins. Under an embodiment, time is marked when an inbuilt accelerometer indicates movement. The accelerometer is incorporated into the chest strap. (Also note that the chest device / strap comprises the Equivital system, as further described elsewhere herein).

[0050] Figure 1C shows an insulated skin temperature monitor embedded in the chest gear worn by a human. The device includes a thermistor 130 for monitoring the microclimate temperature of the PPE, a block of closed-cell cross-linked polyethylene foam 132 (can be replaced after use), and an infrared skin temperature monitor 134 that measures temperature every 15 seconds.

[0051] Under an embodiment, a subject may wear an arm band 140 comprising a skin temperature monitor 142 (seen in Figure ID) analogous to the skin temperature monitor 102 shown in Figure 1C. Under an embodiment, chest strap and wrist band skin temperature monitors may simultaneously collect temperature data for the sake of redundancy. The interior of the arm bank 140 may include a heart rate sensor or detection of heart rate data 118, under an embodiment. (Note that the chest strap may also incorporate a heart rate sensor). 2.3. Statistical analysis

[0052] Multiple linear regression was used to develop the prediction equation for Tc, which was the dependent variable in the regression analysis. The independent variables included in the initial model were Tsk, HR, time, Tsk / min, and RR. A stepwise regression was performed to determine which variables would be used for the final model. In determining the significant contributing variables, the improvement of the SEE and the adjusted R2were examined. Root mean squared error (RMSE) was also reported for the final two regression analyses. Each variable was recorded at the start of the exercise and at each 0.5 °C increase in Tc. (Under alternative embodiments, variable data is captured at intervals ranging from 15 to 30 seconds). Tsk / min was recorded as the rate of Tsk acquisition per minute in the previous period that resulted in a 0.5 °C increase in measured Tc. For example, if the Tsk was 38.3 °C at Tc of 38 °C and the Tsk was 36.7 °C at Tc 37.5° and it took 10 min to make the 0.5 °C increase, then the Tsk / min would be (38.3 °C-36.7 °C) / 10 min = 0.16 °C / min. All of these variables were collected from the integrated physiological monitoring system (Equivital). The result of the analysis indicated that all variables significantly contributed to the final model except for HR.

[0053] Adherence to the assumptions of multiple regression was assessed. The independence of errors was determined using the Durbin Watson test (DURBIN & WATSON, 1951). The Durbin- Watson test examines a regression model for serial correlation. The assumption of multicollinearity was tested using the variance inflation factors (VIF) test. The assumption of the normality of residuals was tested using Q-Q plots. Lastly, the absence of outliers and influential cases was tested using Cook's distance (Cook, 1977).

[0054] Once the model was developed with the appropriate variables to be included, it was validated using the "leave-one-out" bootstrapping approach. Seventeen different equations were developed with 16 participant's data in each equation, leaving one out each time. The Tc values for the missing subject were then predicted using the equation generated from the other 16 participants. The mean of the 17 coefficients was used to create the final equation based on an independent sample (Richmond et al., 2015).

[0055] 3. Results

[0056] 3.1. Physiological changes and rate of thermal acquisition

[0057] The rate of thermal acquisition for Tc was 0.067 °C / min from Tc 37.5° to Tc 39 °C (Fig.

[0058] 2). The rate of thermal acquisition for Tsk was 0.078 °C / min from Tc 37.5° to Tc 39 °C (Fig. 2). The changes in HR and RR increased at each measurement point (37.5 °C, 38 °C, 38.5 °C, and 39 °C) (Figs. 3 and 4).

[0059] 3.2. Algorithm development

[0060] The initial exploratory stepwise regression model with Tsk, HR, time, Tsk / min, and RR resulted in an adjusted R2value of 0.845 and a SEE of 0.29 °C. Through this analysis, it was determined that HR did not significantly contribute to the model. Therefore, HR was removed from the subsequent analysis. The VIF test for Tsk and HR was 5.70 and 6.49, respectively, which indicates some correlation between the variables in the first regression model (Kutner, Nachtsheim, Neter, & Li, 2005). Five outliers were removed as a result of Cook's test. The resultant second regression analysis adjusted R2with HR, and the outliers removed were equal to 0.897 with a SEE of 0.23 °C and an RMSE of 0.23 °C, with all VIF scores under 5, indicating a lack of multicollinearity (Kutner, Nachtsheim, Neter, & Li, 2005). The range of Tc values used during the analysis was 36.48 °C-39.03 °C. This captured all of the results from our measured Tc data collection.

[0061] Initial Equation:

[0062] Tc= 32.665 + (0.111 x Tsk) + (0.001 x HR) (0.026 x Time) + (-1.215 x Tsk / min) + (0.014 x RR) R2= 0.845, SEE = 0.29 °C

[0063] Fig. 2 shows rate of thermal acquisition changes for Tc and Tc. Increases in each metric were observed between data collection points.

[0064] Fig. 3 shows HR changes throughout the protocol. Increases were observed at each data collection point.

[0065] Fig. 4 shows RR changes throughout the protocol. Increases were observed at each data collection point.

[0066] Fig. 5 shows the relationship between measured and predicted Tc (°C). The predicted Tc values are the values produced for the left-out subject when performing the bootstrapping "leave- one-out" analysis. The vertical dashed line shows participants would need to be stopped at a Tc of 38.54 °C to prevent 97.5% of the population from exceeding 39 °C.

[0067] Second Equation without HR and with outliers removed:

[0068] Tc = 34.013 + (0.075 x Tsk) + (0.032 x Time) + (-1.332 x Tsk / min) + (0.018 x RR)

[0069] R2= 0.897, SEE = 0.23 °C, RMSE = 0.23 °C Following the initial analysis, the results of the "leave-one-out" bootstrapping approach produced values for each subject left out of the analysis. The Pearson's correlation coefficient (r) between the values produced from the "leave-one-out" analysis and the actual measured Tc was 0.95 (Fig. 5). Following this process, the mean coefficients from the 79 generated equations were calculated to create a robust equation based on an independent sample. The adjusted R2produced using this equation was 0.90, with a SEE of 0.23 °C and an RMSE of 0.23 °C see Table 2). The sensitivity and specificity from this equation are 86% and 81% respectively. A Bland- Altman plot was created from the predicted (robust equation) and measured Tc (Fig. 6). The bias and 95% LoA calculated as: bias, 0.00 ± 0.23 °C; 95% LoA [-0.44 °C, 0.44 °C], Robust Equation from Mean Coefficients:

[0070] Tc = 34.013 + (0.075 x Tsk) + (0.032 x Time) + (-1.333 x Tsk / min) + (0.0180 x RR)

[0071] R2= 0.902, SEE= 0.23 °C, RMSE = 0.23 °C

[0072] 4. Discussion

[0073] This study used a single piece of physiological monitoring equipment (Equivital) to discover if Tc could be predicted during laboratory rapid heat stress in firefighters wearing PPE. The prediction model used Tsk, time, Tsk / min, and RRto predict Tc. The concept of Tc prediction using non-invasive methods is not novel. Previous research has developed systems that produce accurate results, but none can efficiently and effectively be deployed during live firefighting. Existing systems use multiple pieces of equipment that require precise placement on the body, conditions that are not realistic, or an initial Tc reading; these conditions are difficult to accommodate during live firefighting. This study used a single transcutaneous monitoring system (Equivital) attached to an elastic strap around the chest, which is already designed to be worn during intense exercise and is valid and reliable during all intensities (Liu et al., 2013). We propose this monitoring system could be easily added to the firefighter ensemble and would not add significant time to the preparation procedures for a fire emergency. The timing of preparation for a fire emergency is crucial. Response time can be the difference between life and death for those in essential situations. This timing is so vital that research has been done to determine ways to improve dressing time for emergency personnel (Yeh & Hsu, 2016, pp. 1-4).

[0074] Previous research has used the Equivital Lifemonitor to estimate core temperature and determine the instrument's validity. One study (Agostinelli et al., 2023) used the built-in Equivital HR-based Tc algorithm to predict core temperature and validate it with a rectal thermistor. Simulated firefighting tasks were completed with and without PPE. These authors also performed a Bland-Altman analysis and found a bias of -0.092 with a 95% LoA [-0.761 , 0.578], The algorithm in our study performed better than the in-built Equivital HR-based algorithm but was only tested using PPE. Another study (Pearson et al., 2022) performed a validation of the in-built Equivital HR-based algorithm with Tc capsules during training exercises on volunteer firefighters while wearing PPE. They compared the predicted and measured Tc using Lin's Concordance correlation coefficient and found pc = 0.74, which indicates poor agreement (Akoglu, 2018). Our study was performed in a laboratory setting with PPE, used Pearson's correlation coefficient, and found an r = 0.95, which indicates a strong to very strong correlation (Akoglu, 2018). The novelty of this study is that it used the Equivital system with a novel algorithm.

[0075] When comparing our results to previous research that performed regression analysis, our robust algorithm reached an R2= 0.902, a SEE = 0.23 °C, and RMSE = 0.23 °C. It thus performed better than any other existing algorithm in the literature, but our study only used one condition (35 °C, 45%). The Richmond et al. study (Richmond et al., 2015) reached an R2= 0.86 and a SEE= 0.27 °C. This study had subjects exercise in permeable and impermeable clothing in different conditions (25 °C, 50%; 35 °C, 35%; 40 °C, 25%). Another study (Niedermann et al., 2014) found an R2= 0.72, but also performed the research under two conditions (10 °C, 30 °C). Finally, a study (Richmond et al., 2013) found an R2= 0.85 and a SEE = 0.2 °C. This study researched police, firefighters, and ambulance workers in hot (30 °C) and neutral (18 °C) conditions. Future research using our algorithm must validate it in various work environments and situations.

[0076] Table 2

[0077] Performance of each of the equations generated in this study.

[0078] Equation R"' SEE CO 0.29

[0079] Second 0,897 0.23

[0080] Robust 0,902 0.23

[0081] Fig. 6 shows the Bland-Altman plot of agreement between predicted and measured Tc from the data following "leave-one-out" bootstrap method (robust equation) with bias (dashed line) and lower and upper 95% limits of agreement (dotted lines). Figure 7 shows a firefighter on the treadmill in the environmental chamber during the experimental protocol. The subject is wearing a metabolic monitoring system to measure respiratory exchange ratio (RER) to ensure steady-state exercise.

[0082] Figure 8 shows correlation between measured and predicted Tc (°C), under an embodiment. The predicted Tc values are those produced for the left-out data when performing the bootstrapping "leave-one-out" analysis.

[0083] Figure 9 shows Tsk and measured Tc quadratic relationship (Spearman r = 0.86), under an embodiment.

[0084] Figure 10 shows measured and predicted Tc (°C) values, under an embodiment.

[0085] This study included several parameters for Tc prediction. Tsk was very important as it was used as an objective measure of thermal load and a metric for thermal acquisition rate. Previous studies using non-invasive methods to predict Tc have found that insulated Tsk is the most important predictor for Tc (Richmond et al., 2013, 2015). Insulated Tsk is typically measured using a temperature thermistor insulated by a block of closed-cell cross-linked polyethylene foam or similar (Richmond et al., 2013). Our study did not use insulated skin temperature but found noninsulated Tsk, collected from a medical-grade infrared thermometer, contributed significantly to the regression model. The time point from the start of the RHS task was the most critical predictor in our model, followed by RR. Tsk / min was also a significant predictor. The fact that time and Tsk / min were essential predictors for our model accentuates the need to consider the thermal acquisition rate, which no other predictive models emphasized. This metric is essential because it results in complications not demonstrated with slower acquisition rates. Our previous research found that rapid heat stress (0.04 °C / min) or rapid thermal acquisition resulted in changes in cerebral oxygenation (Coehoom et al., 2023), neural function (Coehoorn et al., 2020), and the stress response (Coehoorn et al., 2022), while a slower rate of thermal acquisition (0.02 °C / min) did not. Monitoring the rate of thermal acquisition as a predictor of Tc could provide valuable information that can be used to keep firefighters safe.

[0086] A meaningful connection exists between minute ventilation and core temperature (Haldane, 1905). Therefore, ventilation was necessary for the prediction model in this study. RR is one component used to calculate minute ventilation alongside tidal volume. RR is a strong indicator of physical effort. The increase in RR during exercise resembles that of blood lactate, which indicates effort and exercise task difficulty (Nicolo et al., 2017). Therefore, considering RR as a predictor of Tc is important because a measure of intensity is valuable in situations where physiological stress increases over time.

[0087] Criticism of the use of RR to predict Tc has been mentioned in the literature. In one study (Richmond et al., 2015), a physiological monitor captured HR and RR as metrics to predict Tc. The authors criticized using the system for RR due to invalid measures that resulted from talking, drinking, and sporadic deep breaths. They also determined that the system was inaccurate because it predicted RR using the impedance of the chest strap and the electrocardiogram signal; these two methods produced different results, so RR could not be included in the model. We integrated these findings in our study, which did not allow subjects to talk or drink during the task, to enhance its accuracy. It should be mentioned that it is not possible to not speak in real-world firefighting situations. The speaking would create artifacts in the data but would not impede the overall data collection because each data point is recorded every 15 s on the Equivital system. Regarding the accuracy of the RR readings from the physiological monitor used in our study, previous research has determined that it is valid and reliable for collecting RR readings (Liu et al., 2013).

[0088] Applying our system in live fire suppression could provide a meaningful way to monitor the predicted Tc of firefighters. In these cases, there is the potential to mediate the consequences that result from high Tc and rapid rates of thermal acquisition. Collapse from heat stroke is associated with a Tc > 40 °C (ARMSTRONG, 2003). Based on this information, removing workers at -39 °C is crucial to mitigate the chance of developing heat stroke. The sensitivity and specificity plot (Fig. 5) demonstrated that to prevent 97.5% of the population from exceeding a true Tc of 39 °C, they would need to be stopped at 38.54 °C predicted Tc due to the SEE in our model (0.23 °C). At 38.54 °C, only 2.5% of the population would risk being above 39 °C, but another 2.5% of the population would stop at below 38.08 °C (Richmond et al., 2015). This would have some individuals being removed below the critical temperature (38.5 °C) for uncompensable heat stress in the work environment (Montain et al., 1994; Selkirk & Mclellan, 2001). While it could harm work productivity, it would improve on the World Health Organization recommendation of withdrawing at a Tc 38 °C (WHO, 1996).

[0089] 4.1. Conclusion

[0090] This study provided a valid, reliable, and deployable system that can be used to predict Tc during laboratory rapid heat stress. Future research using this algorithm must evaluate the system in other work situations and environmental conditions. Previous systems have used complex equipment and algorithms that cannot be successfully integrated into the live firefighting environment. Additionally, previous research has focused on addressing the consequences of heat- related complications. While this is important, preventing the complications from ever occurring is more important. This study demonstrates a valid system that could be added to firefighter PPE effectively and efficiently without adding significant time during the preparation period. This system could increase firefighters' health and safety by preventing excessive Tc acquisition. This would prevent heat illness, cognition decrements, and the elevated stress response. Thereby lengthening careers and increasing longevity post-career.

[0091] Example Two

[0092] Under another embodiment, the systems and methods described herein provide a method and system for predicting core temperature using a supervised machine-learning approach incorporating multiple regression. The model leverages physiological metrics (see Data Collection below) to produce accurate and timely predictions of core temperature.

[0093] 1. Methodology: o Data Collection:

[0094] ■ Collect physiological metrics from subjects, including:

[0095] ■ Skin temperature (°C)

[0096] ■ Insulated skin temperature (°C)

[0097] ■ Skin temperature rate of change (°C / min)

[0098] ■ Insulated skin temperature rate of change (°C / min)

[0099] ■ Microclimate / Ambient environment temperature (°C / min)

[0100] ■ Heart rate (beats per minute)

[0101] ■ Respiratory rate (breaths per minute)

[0102] ■ Time (in minutes since measurement began) o Data Preprocessing:

[0103] ■ Load and explore the dataset

[0104] ■ Handle missing values and outliers as necessary. o Model Development:

[0105] ■ Utilize a supervised machine learning approach with multiple regression techniques to establish a predictive model. ■ Train the model using a percentage of the historical data with known core temperature values, optimizing for accuracy. o Model Validation:

[0106] ■ Validate the model using a separate test dataset, calculating performance metrics such as Mean Absolute Error (MAE) and Root Mean Squared Error (RMSE) to assess prediction accuracy. o Implementation:

[0107] ■ Deploy the predictive model in real-time monitoring systems, enabling non- invasive core temperature estimation based on continuous input of the defined physiological metrics.

[0108] 2. System Components: o A data acquisition unit for collecting physiological metrics (as described above) o A processing unit capable of running the supervised machine learning model, o A user interface for displaying predicted core temperature and other relevant metrics, o The supervised machine learning model uses previously collected physiological and timebased data (predictors) and core temperature data (predicted) to create an initial multiple regression model. The data is then tested for outliers using the standardized residuals method (± 3). The outliers are removed from the original dataset and a new data set is created without the outliers. 80% of the data lines are used to create a training set for the predictive regression model. The model is tested on the remaining 20%. The performance of the model is measured using the coefficient of determination (R2), mean absolute error (MAE), and root mean squared error (RMSE). The model and predictive algorithms are continuously updated when new data is added. o The updated core temperature prediction algorithms are updated to the hardware with periodic firmware updates.

[0109] This systems described above provide a realistic, valid, reliable, and deployable system that can be used to predict core temperature in the live work environment during rapid heat stress. Previous systems have used complex equipment and algorithms that would not have success being integrated into the live firefighting environment. The system can easily be added to the firefighter ensemble without adding significant time during the preparation period. This system could increase firefighters' health and safety by preventing excessive core temperature acquisition. This would prevent heat illness, cognition decrements, and elevated stress response, lengthening careers and increasing longevity post-career.

[0110] A system is described herein comprising under an embodiment a wearable measurement device, wherein the wearable measurement device comprises a chest harness configured for wear by subjects in an environment, the measurement device comprising a plurality of sensors configured to measure first biometric data of first subjects and second biometric data of second subjects, wherein the plurality of sensors are communicatively coupled with one or more applications running on at least one processor of a remote server, the one or more applications configured to receive the first biometric data, wherein the first biometric data comprises data captured in real time from the first subjects during physical activity in the environment, receive real time core body temperature measurements of the first subjects, wherein the first biometric data comprises the real time core body temperature measurements, train a model to predict core body temperature of the first subjects using the first biometric data including the core body temperature measurements, wherein the model is trained using the first biometric data of a respiratory rate variable, skin temperature variables, and a time variable, receive the second biometric data, wherein the second biometric data comprises data captured in real time from the second subjects during physical activity in the environment, and apply the model to the second biometric data to predict core body temperature of the subject in the environment using the second biometric data of the respiratory rate variable, the skin temperature variables, and the time variable.

[0111] In embodiments, the training the model comprises capturing the first biometric data of the respiratory rate variable, the skin temperature variables, and the time variable at each interval increase of 0.5 degrees Celsius in core body temperature of the first subjects.

[0112] In embodiments, the time variable is marked relative to when physical activity begins.

[0113] In embodiments, the plurality of sensors comprises an accelerometer.

[0114] In embodiments, the accelerometer is configured to mark the time upon detecting movement of the first subjects.

[0115] In embodiments, the plurality of sensors comprises a respiratory sensor.

[0116] In embodiments, the plurality of sensors comprises an Electrocardiogram sensor.

[0117] In embodiments, respiratory rate variable is computed using information captured by the respiratory sensor and the electrocardiogram sensor. In embodiments, the plurality of sensors comprises an insulated skin temperature monitor.

[0118] In embodiments, the insulation comprises a block of closed-cell cross-linked polyethylene foam.

[0119] In embodiments, the insulated skin temperature monitor comprises an infrared skin temperature monitor for detecting skin temperature of the first subjects.

[0120] In embodiments, the skin temperature variables comprise the skin temperature of the first subjects.

[0121] In embodiments, the skin temperature variables comprise a time adjusted skin temperature metric of the first subjects.

[0122] In embodiments, the time adjusted skin temperature metric comprises a difference between immediately successive skin temperature measurements divided by a period of time passing between the measurements.

[0123] In embodiments, the insulated skin temperature monitor comprises a thermistor for monitoring the microclimate temperature of the first subjects in the environment.

[0124] In embodiments, the environment comprises a fully encapsulated personal protective equipment environment.

[0125] In embodiments, the one or more applications are communicatively coupled with a capsule ingested by the first subjects, wherein the capsule measures the core body temperature of the first subjects.

[0126] In embodiments, the plurality of sensors comprise a heart rate sensor.

[0127] In embodiments, the first biometric data and the second biometric data comprise a heart rate variable.

[0128] A device is described herein comprising a harness configured for wear by a subject in an environment, wherein the chest harness comprises a measurement device, wherein the measurement device comprises a plurality of sensors configured to measure first biometric data and additional biometric data of the subject, wherein the measurement device comprises one or more applications running on at least one processor, wherein the one or more applications are communicatively coupled with the plurality of sensors, wherein the one or more applications are configured to receive the first biometric data comprising data captured in real time from the subject during physical activity in the environment, receive real time core body temperature measurements of the subject, wherein the first biometric data comprises the real time core body temperature measurements, train a model to predict core body temperature of the subject using the first biometric data including the core body temperature measurements, wherein the model is trained using the first biometric data of a respiratory rate variable, skin temperature variables, and a time variable, receive the additional biometric data comprising data captured in real time from the subject during continued physical activity in the environment, and apply the model to the additional biometric data to predict core body temperature of the subject in the environment using the additional biometric data of the respiratory rate variable, the skin temperature variables, and the time variable.

[0129] In embodiments, the training the model comprises capturing the first biometric data of the respiratory rate variable, the skin temperature variables, and the time variable at each interval increase of 0.5 degrees Celsius in core body temperature of the subject.

[0130] In embodiments, the time variable is marked relative to when the physical activity begins. In embodiments, the plurality of sensors comprises an accelerometer.

[0131] In embodiments, the accelerometer is configured to mark the time upon detecting movement of the subject.

[0132] In embodiments, the plurality of sensors comprises a respiratory sensor.

[0133] In embodiments, the plurality of sensors comprises an electrocardiogram sensor.

[0134] In embodiments, the respiratory rate variable is computed using information captured by the respiratory sensor and the echocardiogram sensor.

[0135] In embodiments, the plurality of sensors comprises an insulated skin temperature monitor.

[0136] In embodiments, the insulation comprises a block of closed-cell cross-linked polyethylene foam.

[0137] In embodiments, the insulated skin temperature monitor comprises an infrared skin temperature monitor for detecting skin temperature of the subject.

[0138] In embodiments, the skin temperature variables comprise the skin temperature of the subject.

[0139] In embodiments, the skin temperature variables comprise a time adjusted skin temperature metric of the subject. In embodiments, the time adjusted skin temperature metric comprises a difference between immediately successive skin temperature measurements divided by a period of time passing between the measurements.

[0140] In embodiments, the insulated skin temperature monitor comprises a thermistor for monitoring the microclimate temperature of the subject in the environment.

[0141] In embodiments, the environment comprises a personal protective equipment environment.

[0142] In embodiments, the one or more applications are communicatively coupled with a capsule ingested by the subject, wherein the capsule measures the core body temperature of the subject.

[0143] In embodiments, the plurality of sensors comprise a heart rate sensor.

[0144] In embodiments, the first biometric data and the additional biometric data comprise a heart rate variable.

[0145] In the description herein, numerous specific details are introduced to provide a thorough understanding of, and enabling description for, embodiments of the systems and methods described herein. One skilled in the relevant art, however, will recognize that these embodiments can be practiced without one or more of the specific details, or with other components, systems, etc. In other instances, well-known structures or operations are not shown, or are not described in detail, to avoid obscuring aspects of the disclosed embodiments.

[0146] The systems and methods described herein include and / or run under and / or in association with a processing system. The processing system includes any collection of processor-based devices or computing devices operating together, or components of processing systems or devices, as is known in the art. For example, the processing system can include one or more of a portable computer, portable communication device operating in a communication network, and / or a network server. The portable computer can be any of a number and / or combination of devices selected from among personal computers, cellular telephones, personal digital assistants, portable computing devices, and portable communication devices, but is not so limited. The processing system can include components within a larger computer system.

[0147] The processing system of an embodiment includes at least one processor and at least one memory device or subsystem. The processing system can also include or be coupled to at least one database. The term “processor” as generally used herein refers to any logic processing unit, such as one or more central processing units (CPUs), digital signal processors (DSPs), application-specific integrated circuits (ASIC), etc. The processor and memory can be monolithically integrated onto a single chip, distributed among a number of chips or components of a host system, and / or provided by some combination of algorithms. The methods described herein can be implemented in one or more of software algorithm(s), programs, firmware, hardware, components, circuitry, in any combination.

[0148] System components embodying the systems and methods described herein can be located together or in separate locations. Consequently, system components embodying the systems and methods described herein can be components of a single system, multiple systems, and / or geographically separate systems. These components can also be subcomponents or subsystems of a single system, multiple systems, and / or geographically separate systems. These components can be coupled to one or more other components of a host system or a system coupled to the host system.

[0149] Communication paths couple the system components and include any medium for communicating or transferring files among the components. The communication paths include wireless connections, wired connections, and hybrid wireless / wired connections. The communication paths also include couplings or connections to networks including local area networks (LANs), metropolitan area networks (MANs), wide area networks (WANs), proprietary networks, interoffice or backend networks, and the Internet. Furthermore, the communication paths include removable fixed mediums like floppy disks, hard disk drives, and CD-ROM disks, as well as flash RAM, Universal Serial Bus (USB) connections, RS-232 connections, telephone lines, buses, and electronic mail messages.

[0150] Unless the context clearly requires otherwise, throughout the description, the words “comprise,” “comprising,” and the like are to be construed in an inclusive sense as opposed to an exclusive or exhaustive sense; that is to say, in a sense of “including, but not limited to.” Words using the singular or plural number also include the plural or singular number respectively. Additionally, the words “herein,” “hereunder,” “above,” “below,” and words of similar import refer to this application as a whole and not to any particular portions of this application. When the word “or” is used in reference to a list of two or more items, that word covers all of the following interpretations of the word: any of the items in the list, all of the items in the list and any combination of the items in the list.

[0151] The above description of embodiments of the systems and methods is not intended to be exhaustive or to limit the systems and methods described to the precise form disclosed. While specific embodiments of, and examples for, the systems and methods are described herein for illustrative purposes, various equivalent modifications are possible within the scope of other systems and methods, as those skilled in the relevant art will recognize. The teachings of the systems and methods provided herein can be applied to other systems and methods, not only for the systems and methods described above.

[0152] The elements and acts of the various embodiments described above can be combined to provide further embodiments. These and other changes can be made to the system in light of the above detailed description.

[0153] The elements and acts of the various embodiments described above can be combined to provide further embodiments. These and other changes can be made to the embodiments described above in light of the above detailed description.

[0154] References

[0155] Agostinelli, P. J., Linder, B. A., Frick, K. A., Bordonie, N. C., Neal, F. K., & Sefton, J. M. (2023). Validity of heart rate derived core temperature estimation during simulated firefighting tasks. Scientific Reports, 13(1), Article 22503.

[0156] Akoglu, H. (2018). User's guide to correlation coefficients. Turk J Emerg Med, 18(3), 91- 93.

[0157] Armstrong, L. E. (2003). Exertional heat illness. USA: Human Kinetics, Inc.

[0158] Bongers, C. C., Hopman, M. T., & Eijsvogels, T. M. (2015). Using an ingestible telemetric temperature pill to assess gastrointestinal temperature during exercise. Journal of Visualized Experiments: JoVE, (104).

[0159] Buller, M. J., Castellani, J., Roberts, W. S., Hoyt, R. W., & Jenkins, O. C. (2011). Human thermoregulatory system state estimation using non-invasive physiological sensors. Annu Int Conf IEEE Eng Med Biol Soc, 2011, 3290-3293.

[0160] Buller, M. J., Davey, T., Fallowfield, J. L., Montain, S. J., Hoyt, R. W., & Delves, S. K.

[0161] (2020). Estimated and measured core temperature responses to high-intensity warm weather military training: Implications for exertional heat illness risk assessment. Physiological Measurement, 41(6), Article 065011 .

[0162] Buller, M. J ., Tharion, W. J., Cheuvront, S. N., Montain, S. J., Kenefick, R. W., Castellani, J., ... Hoyt, R. W. (2013). Estimation of human core temperature from sequential heart rate observations. Physiological Measurement 34(7), 781-798.

[0163] Buller, M. J., Tharion, W. J., Duhamel, C. M., & Yokota, M. (2015). Real-time core body temperature estimation from heart rate for first responders wearing different levels of personal protective equipment. Ergonomics, 58(11), 1830-1841.

[0164] Buller, M., Tharion, W., Hoyt, R., & Jenkins, O. (Eds.). (2010). Estimation of human internal temperature from wearable physiological sensors. Proceedings of the AAAI Conference on Artificial Intelligence.

[0165] Byrne, C., & Lim, C. L. (2007). The ingestible telemetric body core temperature sensor: A review of validity and exercise applications. British Journal of Sports Medicine, 41(3), 126-133.

[0166] Casa, D. J ., Becker, S. M., Ganio, M. S., Brown, C. M., Yeargin, S. W., Roti, M. W., ... Armstrong, L. E. (2007). Validity of devices that assess body temperature during outdoor exercise in the heat. Journal of Athletic Training, 42(3), 333-342.

[0167] Casa, D. J., DeMartini, J. K., Bergeron, M. F., Csillan, D., Eichner, E. R., Lopez, R. M., ... Yeargin, S. W. (2015). National athletic trainers' association position statement: Exertional heat illnesses. Journal of Athletic Training, 50, 986-1000.

[0168] Cheung, S. S., & McLellan, T. M. (1998). Heat acclimation, aerobic fitness, and hydration effects on tolerance during uncompensable heat stress. Journal of Applied Physiology, 84(5), 1731-1739 (1985).

[0169] Cheung, S. S., McLellan, T. M., & Tenaglia, S. (2000). The thermophysiology of uncompensable heat stress. Physiological manipulations and individual characteristics. Sports Medicine, 29(5), 329-359.

[0170] Choi, B., Ko, S., & Kojaku, S. (2017). Resting heart rate, heart rate reserve, and metabolic syndrome in professional firefighters: A cross-sectional study. American Journal of Industrial Medicine, 60(10), 900-910.

[0171] Coehoorn, C. J., Neary, J. P., Krigolson, O. E., Service, T. W., & Stuart-Hill, L.A. (2022). Firefighter salivary cortisol responses following rapid heat stress. Journal of Thermal Biology, 108, Article 103305. Coehoom, C. J., Patrick Neary, J., Krigolson, O. E., & Stuart-Hill, L.A. (2023). Firefighter pre-frontal cortex oxygenation and hemodynamics during rapid heat stress. Brain Research, 1798, Article 148156.

[0172] Coehoom, C. J., Stuart-Hill, L. A., Abimbola, W., Neary, J. P., & Krigolson, O. E. (2020). Firefighter neural function and decision-making following rapid heat stress. Fire Safety Journal, 118, Article 103240.

[0173] Convertino, V. A., Armstrong, L. E., Coyle, E. F., Mack, G. W., Sawka, M. N., Senay, L. C., & Sherman, W. M. (1996). American College of Sports Medicine position stand. Exercise and fluid replacement. Medicine & Science in Sports & Exercise, 28(1), i-vii.

[0174] Cook, R. D. (1977). Detection of influential observation in linear regression. Technometrics, 19(1), 15-18.

[0175] Coris, E. E., Ramirez, A. M., & Van Durme, D. J. (2004). Heat illness in athletes: The dangerous combination of heat, humidity and exercise. Sports Medicine, 34(1), 9-16.

[0176] Durbin, J., & Watson, G. S. (1951). Testing for serial correlation in least squares regression. U. Biometrika, 38(1-2), 159-178.

[0177] Eggenberger, P., MacRae, B. A., Kemp, S., Biirgisser, M., Rossi, R. M., & Annaheim, S. (2018). Prediction of core body temperature based on skin temperature, heat flux, and heart rate under different exercise and clothing conditions in the heat in young adult males. Frontiers in Physiology, 9, 1780.

[0178] El-Radhi, A. S. (2014). Determining fever in children: The search for an ideal thermometer. British Journal ofNursing, 23(2), 91-94.

[0179] Fahy, R. F., Evarts, B., & Stein, G. P. (2022). US fire department profile 2020. Quincy, MA, USA: National Fire Protection Association.

[0180] Fox, R. H., & Solman, A. J. (1971). A new technique for monitoring the deep body temperature in man from the intact skin surface. J Physiol, 212(2), 8P-10P.

[0181] Glazer, J. L. (2005). Management of heatstroke and heat exhaustion. American Family Physician, 71(11), 2133-2140.

[0182] Gribok, A. V., Buller, M. J., Hoyt, R. W., & Reifman, J. (2010). A real-time algorithm for predicting core temperature in humans. IEEE Transactions on Information Technology in Biomedicine, 14(4), 1039-1045. Gunga, H. C., Werner, A., Stahn, A., Steinach, M., Schlabs, T., Koralewski, E., ... Koch, J. (2009). The Double Sensor-A non-invasive device to continuously monitor core temperature in humans on earth and in space. Respiratory Physiology & Neurobiology, 169(Suppl 1), S63-S68.

[0183] Haldane, J. S. (1905). The influence of high air temperatures: No. 1. Journal of Hygiene, 5(4), 494-513.

[0184] Huang, M., & Chen, W. (2010). Theoretical simulation of the dual -heat-flux method in deep body temperature measurements. Annu Int Conf IEEE Eng Med Biol Soc, 2010, 561-564.

[0185] Jitnarin, N., Poston, W. S., Haddock, C. K., Jahnke, S. A., & Day, R. S. (2014). Accuracy of body mass index-defined obesity status in US firefighters. Saf Health Work, 5 (3), 161-164.

[0186] Kang, S. J., Ha, G. C., & Ko, K. J. (2017). Association between resting heart rate, metabolic syndrome and cardiorespiratory fitness in Korean male adults. J Exerc Sci Fit 15(1), 27-31.

[0187] Kim, S., Kim, D.-H., Lee, H.-H., & Lee, J.-Y. (2019). Frequency of firefighters' heat- related illness and its association with removing personal protective equipment and working hours. Industrial Health, 57(3), 370-380.

[0188] Kimberger, O., Thell, R., Schuh, M., Koch, J ., Sessler, D. I., & Kurz, A. (2009). Accuracy and precision of a novel non-invasive core thermometer. British Journal of Anaesthesia, 103(2), 226-231.

[0189] Kistemaker, J. A., Den Hartog, E. A., & Daanen, H. A. (2006). Reliability of an infrared forehead skin thermometer for core temperature measurements. Journal of Medical Engineering & Technology, 30(4), 252-261.

[0190] Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (2005). Applied linear statistical models (Vol. 5). Boston: McGraw-Hill.

[0191] Laplaud, D., Guinot, M., Favre-Juvin, A., & Flore, P. (2006). Maximal lactate steady state determination with a single incremental test exercise. European Journal of Applied Physiology, 96(4), 446-452.

[0192] Lee-Chiong, T. L., & Stitt, J. T. (1995). Heatstroke and other heat-related illnesses. The maladies of summer. Postgraduate Medical Journal, 98(1), 26-28, 31-28.

[0193] Lenhardt, R., & Sessler, D. I. (2006). Estimation of mean body temperature from mean skin and core temperature. Anesthesiology, 105(6), 1117-1121. Limpabandhu, C., Hooper, F. S. W., Li, R., & Tse, Z. (2022). Regression model for predicting core body temperature in infrared thermal mass screening. IPEM Transl, 3, Article 100006.

[0194] Liu, Y., Zhu, S. H., Wang, G. H., Ye, F., & Li, P. Z. (2013). Validity and reliability of multiparameter physiological measurements recorded by the Equivital LifeMonitor during activities of various intensities. Journal of Occupational and Environmental Hygiene, 10(2), 78- 85.

[0195] Mittal, B. B., Sathiaseelan, V., Rademaker, A. W., Pierce, M. C., Johnson, P. M., & Brand, W. N. (1991). Evaluation of an ingestible telemetric temperature sensor for deep hyperthermia applications. International Journal of Radiation Oncology, Biology, Physics, 21(5), 1353-1361.

[0196] Montain, S. J., Sawka, M. N., Cadarette, B. S., Quigley, M. D., & McKay, J. M. (1994). Physiological tolerance to uncompensable heat stress: Effects of exercise intensity, protective clothing, and climate. Journal of Applied Physiology, 77(1), 216-222 (1985).

[0197] Nakada, H., Horie, S., Kawanami, S., Inoue, J., lijima, Y., Sato, K., & Abe, T. (2017). Development of a method for estimating oesophageal temperature by multi -locational temperature measurement inside the external auditory canal. International Journal of Biometeorology, 61(9), 1545-1554.

[0198] Nicolo, A., Massaroni, C., & Passfield, L. (2017). Respiratory frequency during exercise: The neglected physiological measure. Frontiers in Physiology, 8, 922.

[0199] Niedermann, R., Wyss, E., Annaheim, S., Psikuta, A., Davey, S., & Rossi, R. M. (2014). Prediction of human core body temperature using non-invasive measurement methods. International Journal of Biometeorology, 58(1), 7-15.

[0200] Pearson, S. J., Highlands, B., Jones, R., & Matthews, M. J. (2022). Comparisons of core temperature between a telemetric pill and heart rate estimated core temperature in firefighters. Saf Health Work, 13(1), 99-103.

[0201] Richmond, V. L., Davey, S., Griggs, K., & Havenith, G. (2015). Prediction of core body temperature from multiple variables. Annals of Occupational Hygiene, 59(9), 1168-1178.

[0202] Richmond, V. L., Wilkinson, D. M., Blacker, S. D., Horner, F. E., Carter, J., Havenith, G., & Rayson, M. P. (2013). Insulated skin temperature as a measure of core body temperature for individuals wearing CBRN protective clothing. Physiological Measurement 34(11), 1531-1543. Savoie, F. A., Dion, T., Asselin, A., Gariepy, C., Boucher, P. M., Berrigan, F., & Goulet, E. D. (2015). Intestinal temperature does not reflect rectal temperature during prolonged, intense running with cold fluid ingestion. Physiological Measurement 36(2), 259.

[0203] Selkirk, G. A., & McLellan, T. M. (2001). Influence of aerobic fitness and body fatness on tolerance to uncompensable heat stress. Journal of Applied Physiology, 91(5), 2055-2063 (1985).

[0204] Service, T. W., Junker, K., Service, B., Coehoorn, C. J., Harrington, M., Martin, S., & Stuart-Hill, L. A. (2023). An assessment of the validity and reliability of the P022-P version of e- Celsius core temperature capsules. Journal of Thermal Biology, 112, 103486.

[0205] Teunissen, L. P., Klewer, J., de Haan, A., de Koning, J. J., & Daanen, H. A. (2011). Non- invasive continuous core temperature measurement by zero heat flux. Physiological Measurement. 32(5), 559-570.

[0206] Welles, A. P., Xu, X., Santee, W. R., Looney, D. P., Buller, M. J., Potter, A. W., & Hoyt, R. W. (2018). Estimation of core body temperature from skin temperature, heat flux, and heart rate using a Kalman filter. Computers in Biology and Medicine, 99, 1-6.

[0207] WHO. (1996). Health factors involved in working under conditions of heat stress. Geneva, Switzerland: World Health Organization.

[0208] Wilkinson, D. M., Carter, J. M., Richmond, V. L., Blacker, S. D., & Rayson, M. P. (2008). The effect of cool water ingestion on gastrointestinal pill temperature. Medicine & Science in Sports & Exercise, 40(3), 523-528.

[0209] Xu, X., Karis, A. J., Buller, M. J., & Santee, W. R. (2013). Relationship between core temperature, skin temperature, and heat flux during exercise in heat. European Journal of Applied Physiology, 113(9), 2381-2389.

[0210] Yamakage, M., & Namiki, A. (2003). Deep temperature monitoring using a zero-heat-flow method. Journal of Anesthesia, 17(2), 108-115.

[0211] Yeh, C. W., & Hsu, U. K. (2016). The feasibility strategy to enhance rate of emergency rescue dispatch with optimized way of dressing. 2016 international conference on applied system innovation (ICASI). IEEE.

Claims

CLAIMS1. A system comprising, a wearable measurement device, wherein the wearable measurement device comprises a chest harness configured for wear by subjects in an environment; the measurement device comprising a plurality of sensors configured to measure first biometric data of first subjects and second biometric data of second subjects, wherein the plurality of sensors are communicatively coupled with one or more applications running on at least one processor of a remote server, the one or more applications configured to: receive the first biometric data, wherein the first biometric data comprises data captured in real time from the first subjects during physical activity in the environment; receive real time core body temperature measurements of the first subjects, wherein the first biometric data comprises the real time core body temperature measurements; train a model to predict core body temperature of the first subjects using the first biometric data including the core body temperature measurements, wherein the model is trained using the first biometric data of a respiratory rate variable, skin temperature variables, and a time variable; receive the second biometric data, wherein the second biometric data comprises data captured in real time from the second subjects during physical activity in the environment; apply the model to the second biometric data to predict core body temperature of the subject in the environment using the second biometric data of the respiratory rate variable, the skin temperature variables, and the time variable.

2. The system of claim 1, wherein the training the model comprises capturing the first biometric data of the respiratory rate variable, the skin temperature variables, and the time variable at intervals.

3. The system of claim 2, wherein the intervals comprise interval increases of 0.5 degrees Celsius in core body temperature of the first subjects.

4. The system of claim 2, wherein the intervals comprise time intervals.

5. The system of claim 2, wherein the time variable is marked relative to when physical activity begins.

6. The system of claim 5, wherein the plurality of sensors comprises an accelerometer.

7. The system of claim 6, wherein the accelerometer is configured to mark the time upon detecting movement of the first subjects.

8. The system of claim 2, wherein the plurality of sensors comprises a respiratory sensor.

9. The system of claim 8, wherein the plurality of sensors comprises an Electrocardiogram sensor.

10. The system of claim 9, wherein respiratory rate variable is computed using information captured by the respiratory sensor and the electrocardiogram sensor.

11. The system of claim 2, wherein the plurality of sensors comprises an insulated skin temperature monitor.

12. The system of claim 11, wherein the insulation comprises a block of closed-cell cross-linked polyethylene foam.

13. The system of claim 11, wherein the insulated skin temperature monitor comprises an infrared skin temperature monitor for detecting skin temperature of the first subjects.

14. The system of claim 13, wherein the skin temperature variables comprise the skin temperature of the first subjects.

15. The system of claim 14, wherein the skin temperature variables comprise a time adjusted skin temperature metric of the first subjects.

16. The system of claim 11, wherein the time adjusted skin temperature metric comprises a difference between immediately successive skin temperature measurements divided by a period of time passing between the measurements.

17. The system of claim 11, wherein the insulated skin temperature monitor comprises a thermistor for monitoring the microclimate temperature of the first subjects in the environment.

18. The system of claim 1, wherein the environment comprises a fully encapsulated personal protective equipment environment.

19. The system of claim 1, wherein the one or more applications are communicatively coupled with a capsule ingested by the first subjects, wherein the capsule measures the core body temperature of the first subjects.

20. The system of claim 1, wherein the plurality of sensors comprise a heart rate sensor.

21. The system of claim 20, wherein the first biometric data and the second biometric data comprise a heart rate variable.

22. A device comprising, a harness configured for wear by a subject in an environment, wherein the chest harness comprises a measurement device, wherein the measurement device comprises a plurality of sensors configured to measure first biometric data and additional biometric data of the subject,wherein the measurement device comprises one or more applications running on at least one processor, wherein the one or more applications are communicatively coupled with the plurality of sensors, wherein the one or more applications are configured to: receive the first biometric data comprising data captured in real time from the subject during physical activity in the environment; receive real time core body temperature measurements of the subject, wherein the first biometric data comprises the real time core body temperature measurements; train a model to predict core body temperature of the subject using the first biometric data including the core body temperature measurements, wherein the model is trained using the first biometric data of a respiratory rate variable, skin temperature variables, and a time variable; receive the additional biometric data comprising data captured in real time from the subject during continued physical activity in the environment; apply the model to the additional biometric data to predict core body temperature of the subject in the environment using the additional biometric data of the respiratory rate variable, the skin temperature variables, and the time variable.

23. The system of claim 22, wherein the training the model comprises capturing the first biometric data of the respiratory rate variable, the skin temperature variables, and the time variable at intervals.

24. The system of claim 23, wherein the intervals comprise interval increases of 0.5 degrees Celsius in core body temperature of the first subjects.

25. The system of claim 23, wherein the intervals comprise time intervals.

26. The device of claim 25, wherein the plurality of sensors comprises an accelerometer.

27. The device of claim 26, wherein the accelerometer is configured to mark the time upon detecting movement of the subject.

28. The device of claim 23, wherein the plurality of sensors comprises a respiratory sensor.

29. The device of claim 28, wherein the plurality of sensors comprises an electrocardiogram sensor.

30. The device of claim 29, wherein respiratory rate variable is computed using information captured by the respiratory sensor and the echocardiogram sensor.

31. The device of claim 23, wherein the plurality of sensors comprises an insulated skin temperature monitor.

32. The device of claim 31, wherein the insulation comprises a block of closed-cell cross-linked polyethylene foam.

33. The device of claim 31, wherein the insulated skin temperature monitor comprises an infrared skin temperature monitor for detecting skin temperature of the subject.

34. The device of claim 33, wherein the skin temperature variables comprise the skin temperature of the subject.

35. The device of claim 34, wherein the skin temperature variables comprise a time adjusted skin temperature metric of the subject.

36. The device of claim 35, wherein the time adjusted skin temperature metric comprises a difference between immediately successive skin temperature measurements divided by a period of time passing between the measurements.

37. The device of claim 31, wherein the insulated skin temperature monitor comprises a thermistor for monitoring the microclimate temperature of the subject in the environment.

38. The device of claim 22, wherein the environment comprises a personal protective equipment environment.

39. The device of claim 22, wherein the one or more applications are communicatively coupled with a capsule ingested by the subject, wherein the capsule measures the core body temperature of the subject.

40. The system of claim 22, wherein the plurality of sensors comprise a heart rate sensor.

41. The system of claim 20, wherein the first biometric data and the additional biometric data comprise a heart rate variable.

Images