A Coupling System for Prediction of Physiological Parameters in an Immersed Condition

Abstract

This study’s major aim is to use a coupling system to predict the physiological parameters of a person wearing a life preserver in an immersion condition. The coupling system is made up of a thermal manikin that can simulate the human body’s thermoregulatory response in an immersion environment, coupled to a multi-segment human thermal model. By comparing the results with those of the subjects, the coupled system’s predictions were shown to be accurate. The core temperature, skin temperature, and heat flow density of ten healthy men were all continuously measured while they wore life preservers and were exposed to the same temperature of air and water. The coupling system demonstrated accurate predictions of heat production and core temperature compared to real measures, with RMSD of 18.91 W and 0.12 °C, respectively. The majority of the predicted values for mean skin temperature were within SD of the measured values, and the paired sample t-test with a significance level of 0.05 showed good agreement with a maximum temperature difference of 1 °C. Also, the coupling system predicted results and the measured results showed some good accuracy for predicting local skin temperatures.

1. Introduction

The issue of drowning has received a lot of attention recently, both domestically and overseas, along with the growth of society and the demands of human endeavors. Since humans first entered civilization, saving lives at sea has been a regular public health issue, and studies on how humans react physiologically to submerged environments have received a great deal of interest from academics [1,2]. According to studies, exposure to cold water below 15 °C can cause “cold shock”, which can cause hypothermia in humans within 3~5 min and result in shock, comas, and even life-threatening illnesses [3,4]. In fact, allowing an unprepared individual to quickly immerse themselves in cold water is quite harmful. The body will inevitably cool down to the same temperature as the water when submerged in water that is below one’s own body temperature. The rate at which this happens depends on a variety of factors, including the temperature difference, clothing insulation, rate of water agitation, body heat production, body surface area, thickness of subcutaneous fat, level of physical fitness, and body posture in the water. Therefore, understanding how the human body regulates its temperature and physiological reactions in an immersed condition is crucial for risk assessment and life preserver design [5,6].

The physiological responses and thermal regulation of the human body in various environments, such as heat production, core temperature, skin temperature, and heat exchange between the human body and the environment, have been studied in recent decades using human experiments, thermal manikin experiments, and human thermal model simulation experiments [7,8,9,10]. A thermal manikin is a crucial instrument for measuring thermal comfort and evaluating local heat exchange in clothing today. They allow for repeated testing and eliminate inaccuracies caused by subject variability while avoiding the high expenses and ethical obligations associated with real-life testing [11]. Thermal manikin investigations are frequently pessimistic compared to actual test results because a thermal manikin cannot replicate the thermoregulatory response of the human body [12], and it is challenging to generate vasoconstriction and cold fibrillation responses. Based on this, researchers have also created numerous human thermal models, most of which have been focused on applications in air exposure conditions to replicate human physiological reactions [13,14,15]. When using water as the heat transfer medium in watery situations, the methods with which human, clothing, and the environment transmit heat are more complicated [11]. Due to its 3500 times greater volumetric specific heat capacity and 25 times stronger thermal conductivity, water can absorb heat more effectively than air [4]. Montgomery et al. [16], Xu et al. [17], and Wu et al. [18] proposed thermal models to simulate the physiological responses under low-temperature immersion environments. Kim et al. [19] and Yang et al. [11] conducted tests with a thermal manikin and used empirical equations to predict the thermal comfort of people wearing life preservers in water. The truth is most thermal manikins are more inclined for passively studying the heat exchange from the human body to the environment through clothing and are unable to make active adjustments to the complex environment, while the human thermal model can better simulate the thermal regulation response of the human body; however, considering that the heat transfer mechanism of the life jacket in the submerged environment is very complex, the model may be inaccurate in predicting the amount of heat transfer from the human body through the life jacket in the submerged environment [18]. Although research on the thermophysiological response to immersion environments has advanced, it is still difficult to estimate the thermal strain of the majority of the existing human thermal models. By controlling the thermal manikin with the established human thermal model and developing a system coupling the thermal manikin with the thermal model, one can combine the benefits of active regulation of the thermal model with the benefits of accurate heat transfer measurement of the thermal manikin while taking into account the limitations of both human thermal models and the thermal manikin [8,20]. The coupling system’s thermal manikin is outfitted with a thermal regulation system so that it can perceive environmental changes and adapt to them like a human would, while on the other hand, the thermal manikin’s real-time testing increases the model’s predictive accuracy [21].

The major aim of this study is to develop a coupling system model that can be used to predict human physiological parameters and assess how well life preservers defend against heat in an immersed condition. Two life preservers with intrinsic thermal resistances between 0.09 and 0.41 °C·m²/W were chosen for the study, and practical tests were carried out in a climate chamber under two different environmental conditions, with the water immersion tests taking place in a water tank that was built into the climate chamber. To determine the coupling system’s accuracy in predicting human physiological parameters while wearing a life preserver, the measured heat production, core temperature, and skin temperature data were statistically compared with the predicted results. Afterward, some recommendations for the coupling system’s improvement were made.

2. Methods

2.1. Coupling System

2.1.1. Thermal Manikin

The study used the “Walter” thermal manikin as the research object (PolyU, Hong Kong, China), because the “Walter” thermal manikin itself is composed of special fabric wrapped with water, and can work normally under the water-immersed environment. The “Walter” thermal manikin is designed according to the average Chinese adult male figure, whose height is 174 cm tall, weight is 68 kg, and body surface area is 1.81 m² [22]. The chest is equipped with four heating pumps, through the LabVIEW software for independent control. When the core temperature is set, the temperature of each part of the body can be adjusted by adjusting the speed of the water pump built into the body to deliver hot water from the core to the limbs faster or slower.

2.1.2. Thermal Model

The simulation of human physiological responses under transient conditions was performed using a multi-segmented human thermal model [18]. The thermal model divides the human torso and extremities into 15 segments, each of which has a passive system and an active system. In addition, the passive system of the model includes a wet clothing system that goes into great depth about how heat moves through wet clothing. Only subject characteristics, environment parameters, and process variables, such as air temperature, water temperature, relative humidity, wind speed, clothing thermal resistance, and others, are used in the model’s operation. The physiological response of the human body in a water immersion setting, as well as its thermal regulation, may be reasonably predicted and simulated with the thermal model.

2.1.3. Coupling Procedures

A multi-segmented human thermal model of the thermal manikin Walter controls the coupling system. The manikin is used to measure the heat exchanged from the human body to the environment through the garment, and the thermal model is used for active thermophysiological regulation in response to environmental changes. The coupling system between the manikin and the model can combine the advantages of the manikin to accurately measure the heat exchange volume and the model to better simulate the physiological regulation function of the human body, so that the manikin can feel the changes of the environmental conditions in real time and make active adjustments. According to the human body characteristics, clothing thermal resistance and environmental parameters are input into the model to obtain the core temperature and skin temperature; then, the core temperature and skin temperature are input parameters, through LabVIEW software to control the manikin, so that the manikin applies to the thermostatic mode of operation; when the manikin temperature reaches the target set value and remains stable, at this point, the heat transfer measured with the manikin is equal to the sum of the heat transfer between the body and the environment with convection and radiation. Record the heat flow and temperature of each part of the manikin at this point in time; the amount of heat transfer will be recorded for the human body thermal response model as the input value of the model, and the model calculates the core temperature and skin temperature; repeat the iterative process of the above to realize the coupled simulation (see Figure 1).

2.1.4. Experiment Procedures

The experiments were carried out in accordance with ISO 15831 (2004)’s standards [23] (thermal manikin wearing LP1 or LP2 (see Figure 2)). The experimenter arranged 6 groups of heat flow densitometers and 6 groups of thermocouples on the thermal manikin; heat flow densitometers and thermocouples were located in the front chest, back, left arm, right lower arm, left thigh, and right leg of the thermal manikin, and then the heating pump inside the thermal manikin was turned on through the LabVIEW software on the computer, and we waited for the manikin to enter the steady state (indicating that the heat production and heat dissipation of the manikin were equal at this time), and started the 20 min pre-experiment. The water tank was then filled with water until the water was above the neck of the manikin and the filling was stopped, and then the water immersion experiment was started, which lasted for 40 min. When the thermal manikin entered a steady state and began to collect data, heat flow and skin temperature were recorded every minute and transmitted to the LabVIEW system, which reset the core temperature of the thermal manikin according to the results of the thermal manikin’s calculations, and repeat the above process to achieve iterative calculations, so that the physiological parameters of the human and the environment in the transient or steady state environment can be obtained.

2.2. Subjects’ Test

2.2.1. Subjects

Ten untrained male volunteers agreed to take part in the experiment. The characteristics of the subjects were as follows (mean ± SD): age was 25.8 ± 1.3 years (range, 24 to 27), height was 1.75 ± 0.04 m (range, 1.72 to 1.82), weight was 70.6 ± 10.4 kg (range, 58.3 to 82.4), and body surface area was 1.86 ± 0.15 m² (range, 1.56 to 2.05). The relevant university officials approved the experimental protocol and ethics; all subjects voluntarily agreed to participate in the study, signed an informed consent form, and promised not to smoke, drink alcohol, or drink caffeinated beverages such as coffee before the experiment.

2.2.2. Life Preservers

As shown in Figure 3, two common types of life preservers (LP1 and LP2) available in the market were selected for the study. LP1 is made of TPU nylon composite fabric, 0.78 mm thick, weighing 0.5 kg; LP2 is made of Oxford polyester fabric, 1.06 mm thick, weighing 0.87 kg. Both life preservers comply with the ISO 12402 (2006) [24] international standard for life preservers. The study evaluated the thermal resistance of the two life preservers in dry and wet conditions according to ISO 15831 (2004) [23] and ASTM F2370 (2010) [25] standards. In addition, uniform swim trunks were provided for the test subjects as required in CAN/CGSB 65.16 [26].

2.2.3. Physiological Measurements

The study used a swallowable telemetry capsule with a capsule thermometer from e-Celsius Performance Systems (BodyCap, Caen, France; accuracy: ±0.2 °C) with a temperature range of 30 to 40 °C for continuous monitoring of core body temperature. The system was capable of uploading a set of core temperature data every 60 s after the subject swallowed the capsule. An 18-channel temperature and heat flow density tester (JTNT-1, China; temperature accuracy: ±0.1 °C; heat flow accuracy: ±0.1 W/m²) was used to monitor skin temperatures and heat flow densities on the chest, back, upper arm, arm, thigh, and leg. The four-point modeling method was used to determine the mean skin temperature [27].

$$MST=0.3T_{chest}+0.3T_{upperarm}+0.2T_{thigh}+0.2T_{leg}$$

(1)

All experiments started at 9:00 a.m. each day to eliminate the effects of different experimental times on individual subjective perception and circadian rhythms. At the beginning of the experiments, subjects first sat in the climate chamber wearing life preservers for 20 min to stabilize their physiological parameters, and then they entered the water tank and kept their bodies immersed in the water from the neck down for 40 min (see Figure 4). Subjects wore LP1 and LP2 for two separate experiments.

2.3. Statistical Analysis

The study assessed the concordance between the coupling system’s predictions and the subjects’ measurements using standard deviation (SD) and root mean square deviation (RMSD) in order to confirm the accuracy and dependability of the manikin coupling system [17,28]. By contrasting the root mean square deviation values with the experimental data, the model coupling system’s predictions’ accuracy was assessed. If the root mean square deviation is lower than the standard deviation, the predicted value is compatible with the tested value, and if it is inconsistent with the tested value and contains unacceptable values, the anticipated value is inconsistent with the tested value. The following is a definition of the RMSD:

$$RMSD=\sqrt{\frac{1}{n}{\displaystyle \sum _{i=1}^n{\left(T_{measure}-T_{predict}\right)}^2}}$$

(2)

where T_measure is the measured temperature of the experimental observation, °C; T_predict is the simulated temperature of the thermal manikin, °C; and n is the number of tests.

By using a paired-samples t-test with a significance level of 0.05, the study used Statistical Package for the Social Sciences (SPSS) software (version 18.0, IBM Corp., New York, NY, USA) to determine whether there was a significant correlation between the predicted and measured values.

3. Results

The experiment simulates the brief thermophysiological response of a human body entering water. The amount of heat released from the body to the outside world is directly determined with metabolic heat production (Q_s), which is the main source of body heat. The body’s heat balance can be determined with tiny changes in core temperature (T_c), which is the most significant signal of the body’s physiological response [10]. Changes in skin temperature (T_s), which is sensitive to environmental changes and is in close contact with the outside world, are frequently used to illustrate how the human body regulates its internal body temperature [29,30]. Therefore, it is necessary to accurately predict the changes in human heat production and core and skin temperatures in an immersed environment.

3.1. Heat Production

The coupling system can predict human heat production more accurately in both air exposure and water immersion environments, and all the predicted mean values are within SD of the measured values, according to a comparison of the predicted heat production results of the thermal manikin and the human thermal model coupling system with the experimental test results while wearing the two life preservers of LP1 and LP2 (see Figure 5). No significant difference between predicted and measured heat production was found using paired-sample t-tests (p > 0.05). Additionally, while wearing LP1, the maximum deviation between measured and predicted values under conditions of air exposure and water immersion did not exceed 9.59 W and 27.51 W, respectively, and while wearing LP2, the maximum deviation between measured and predicted values under conditions of air exposure and water immersion did not exceed 6.79 W and 23.51 W, respectively. The results show that the RMSD during the test period was 18.91 W, which is smaller than the maximum value of the test result SD of 49.59 W. According to the statistical relationship between RMSD and SD, the accuracy of the core temperature prediction of the coupling system is high. For the majority of the water immersion processes, the predicted values were within the SD of the measured values, but the predicted values indicated with the residual analysis were still skewed. The model tended to predict LP2, and for LP1, the model tended to exceed the measured mean, suggesting that the predicted values underestimated the tested values. This phenomenon will be discussed in detail in the Discussion. In addition, the correlation coefficients R² between the predicted and experimental values of heat production for wearing LP1 and LP2 were 0.952 and 0.946 (i.e., the closer the R² is to 1, the stronger the correlation). According to the R² test, it can be shown that the predicted and measured values are highly correlated.

3.2. Core Temperature

The coupling system can predict the human core temperature more accurately in both air exposure and water immersion environments, and all the predicted mean values are within SD of the measured values, according to a comparison of the core temperature prediction results of the thermal manikin and human thermal model coupling system with the experimental test results when wearing two life preservers, LP1 and LP2 (see Figure 6). No significant differences were measured between the predicted and measured core temperatures according to paired-sample t-tests (p > 0.05). Additionally, when wearing LP1 under conditions of air exposure and water immersion, and when wearing LP2 under conditions of air exposure and water immersion, respectively, the maximum deviation between measured and predicted values did not exceed 0.13 °C and 0.29 °C. The results demonstrate that all differences between the coupling system’s projected and actual core temperatures were less than 0.5 °C. In fact, a discrepancy of 0.5 °C between the predicted and measured values for the core temperature was still regarded as having good agreement [31]. Moreover, the R² for wearing LP1 and LP2 was 0.984 and 0.993; it can be shown that the predicted and measured values are highly correlated. The RMSD and SD for wearing LP1 and LP2 were 0.12 °C and 0.22 °C, and 0.13 °C and 0.24 °C, respectively. It is clear that RMSD is smaller than SD, indicating that the coupled system predicts the core temperature with high accuracy.

3.3. Mean Skin Temperature

The coupling system can predict the mean skin temperature more accurately in both air exposure and water immersion environments, and the majority of the predicted mean values are within SD of the measured values, according to a comparison of the predicted mean skin temperature of the thermal manikin and human thermal model coupling system with the experimental test results while wearing two life preservers of LP1 and LP2 (see Figure 7). There was no discernible difference between the predicted and measured mean skin temperatures, according to a paired-samples t-test (p > 0.05). When wearing LP1, the maximum difference between measured and predicted values under air exposure and water immersion conditions did not exceed 0.28 °C and 0.57 °C, respectively. When wearing LP2, the maximum difference did not exceed 0.31 °C and 0.47 °C, respectively. The skin temperature was more sensitive than the core temperature, and the overall fluctuation range actually outperformed the core temperature, with a safe minimum temperature of 15 °C for the mean skin temperature (54% lower than the normal mean skin temperature of 33 °C) [32], and the safe minimum core temperature is 34 °C (only 8% below the normal core temperature of 37 °C) [33]. Although the deviation of the predicted and actual mean skin temperature values of the coupling system exceeded the good agreement criterion of 0.5 °C, the predicted deviation of the mean skin temperature is within 1 °C, demonstrating that the expected value is still within the range where the measured value is acceptable [31]. Moreover, the RMSD and SD for wearing LP1 and LP2 were, respectively, 0.42 °C and 0.65 °C, and 0.35 °C and 0.48 °C, and the R² for wearing LP1 and LP2 was 0.974 and 0.985; it can be shown that the predicted and measured values are highly correlated. In conclusion, the coupling system shows good agreement for mean skin temperature between the predicted and measured values.

3.4. Local Skin Temperature

We compared the experimental test results when wearing the two life preservers of LP1 and LP2 with the local skin temperature prediction results of the thermal manikin and the human thermal model coupling system (see Figure 8a–f). There is variation in the coupling system’s prediction for several body parts in both air exposure and water immersion situations. While some of the predictions for the other locations, particularly for the thigh and leg (p < 0.05), were beyond the range of SD of the measured values, the mean predicted values for the chest and back were within SD of the measured values. When wearing LP1, the expected deviation values for the chest, thigh, and leg did not surpass 1 °C; when wearing LP2, they were 1.04 °C, 1.16 °C, and 1.27 °C, respectively, and did not exceed 1 °C, with the exception of the maximum predicted deviation value of 1.13 °C for the thigh. Furthermore, when wearing LP1, the RMSD was less than SD for the chest, upper arm, and arm, and when wearing LP2, it was less than SD for all locations. In conclusion, the coupling model could not accurately anticipate changes in local skin temperature, possibly as a result of the thermal manikin’s constraints. However, compared to Yang’s coupling system [21], the coupled model’s predicted values were still within an acceptable limit.

4. Discussion

4.1. The Effect of External Environment

The coupled system is predicted to work better in air than in water because in thermally neutral air environments, the body’s heat production and heat dissipation tend to form a thermal equilibrium, and the body’s active regulatory function plays only a minimal role. In this case, the human body is in a thermally comfortable steady state, and the local temperature is maintained as constant [34]. The coupled system can effectively simulate the state of thermal balance of the human body. However, in water, water as a heat transfer medium is different from air, and a large amount of the body’s heat will be dissipated through conduction and convection in water at the same temperature. In the water immersion experiments, the mean skin temperature predicted with the coupled system is often lower than the actual measurements of the subjects, which is mainly due to the reliability and accuracy of the model in the coupled system. The human body will increase the heat production through the shivering response in cold water, and the thermal model determines that the cold shivering response will be triggered when the skin temperature is lower than 20 °C [35]. However, it was found in the subjects’ experiments that the trigger of the cold shivering response of some of the subjects involved differences in the initial conditions, and some of the subjects experienced the cold shivering response earlier due to the cold shock of entering the water (T_w = 25 °C), and the skin temperature of this group was significantly higher than the average, which resulted in an overall higher than predicted average skin temperature test value. Consequently, this bias resulted in the coupled system having lower predicted values than tested values at subsequent times. This phenomenon would not be present in a much colder water environment, as any healthy person would trigger a cold tremor response in such an environment [29]. In addition, the initial value of the coupling system is supposed to be the initial value of the anthropometric measurements; however, the initial value of the coupling system in the water immersion environment is predicted with the coupling system in the air environment, so there is a deviation between the initial value of the coupling system and the anthropometric measurements in the high-temperature environment at the moment of t = 20 min, which leads to a bigger deviation in the water immersion environment.

4.2. The Effect of Life Preserver Type

The difference in local skin temperature obtained with the coupling system under the same environmental working condition is more obvious. Therefore, the coupled system can be used to evaluate the thermal protection performance of life preservers. The coupling system predicts the local skin temperature of wearing LP2 better than that of wearing LP1 because LP1 is a head-worn life preserver and LP2 is a vest-type life preserver, and the weight of LP1 is only 57.5% of LP2, and the volume is also about 25% of LP2. Compared to LP1, LP2 has a more significant improvement in thermal protection with the addition of wrapping around the chest and back of the body. Based on the difference between the test and prediction results, the coupled system predicts the core temperature of LP2 more accurately than that of LP1, and the predicted values of the coupled system in both environments are significantly lower than the measured values when LP1 is worn (p = 0.35), which is mainly due to the fact that there is no segmentation of the neck and shoulders in the node division of the human body thermal model, considering that LP1 wraps around the neck and shoulders only, which leads to a significant improvement in the thermal protection performance of LP1. The reason for this phenomenon is mainly due to the fact that the neck and shoulders are not subdivided in the node division of the human thermal model, considering that LP1 only wraps around the neck and shoulders, which results in the only thermal resistance of LP1 also belonging to the ineffective thermal resistance, and the body wearing LP1 is equivalent to the naked state in the model. However, in reality, wearing LP1 provides a little thermal protection to the subject, and thus the coupling system underestimates the actual measurements. The coupling system of LP2 more accurately predicts the actual measurements due to the fact that it effectively protects the important parts of the body, which are the nodes of the model. In addition, the difference in heat production between wearing the two life preservers suggests that the life preservers differ in their thermal protection, with those wearing LP2 (LP2 = 201.16 W) producing less heat on average in the water than those wearing LP1 (LP1 = 213.74 W), and a similar trend was found in the coupled system.

4.3. The Limitation of the Coupling System

Subject experiments showed that the coupled system, which inputs a human thermal model for calculations based on the actual measured heat exchange with the environment of a thermal manikin, outperforms simulations performed directly using a human thermal model and has the advantages of both a thermal manikin and a human thermal model. However, there are still some difficulties in using the coupled model. On the one hand, due to the limitations of the thermal human body model, it cannot fully simulate the thermal regulatory responses of real human beings, and there are discrepancies in the details, which leads to the gradual amplification of such discrepancies during the simulation process. On the other hand, due to the technical limitations of the thermal human model, the Walter was designed to be used only in an air environment, and the Walter’s internal heating pump is underpowered in water and takes longer to reach the desired value in cold water.

For the manikin coupled system, to improve its prediction accuracy, this can be realized from two aspects: first, improve the prediction accuracy of the human thermal model, and second, improve the accuracy of the sensor measurement of the thermal manikin and increase the number of manikin heating pumps, so as to better simulate the human body heat production in the cold water environment when using the manikin; the accuracy of the heating system determines the overall simulation performance, to ensure that the manikin is in a cold water environment when the skin temperature can still be maintained at the set target.

5. Conclusions

This paper establishes a coupling system between a thermal manikin and a thermal response model of the human body, and verifies the accuracy and reliability of the coupling system with experimental data from subjects. The validation experiments involved air and cold water environments, as well as the wearing of two different styles of life preservers, and the heat production, core temperature, and skin temperature values obtained from the coupling system and subject experiments were compared. The following conclusions were drawn:

(1)

The simulated values of the coupled system were compared with the experimental measurements of the subjects, and the statistical analysis based on the paired samples t-test and RMSD indicated that the coupled system could predict typical physiological parameters such as core temperature, mean skin temperature, and local skin temperature with a relatively high accuracy. Therefore, the developed coupled system can be used for physiological parameter prediction.

(2)

The established coupling system makes use of the advantages of the thermal manikin and the human thermal model, so that the manikin can be actively adjusted according to the changes in the environment, which is closer to the actual situation than the traditional thermal manikin that only characterizes the warmth of the garment with the thermal resistance of the garment, and at the same time compensates for the defects of the human thermal model that cannot be simulated in the complex environment, which enhances the applicability of the human thermal model to satisfy the needs of life preserver testing and research and development.

(3)

By comparing the core and skin temperature values obtained from different life preservers worn by dummies in the same environment, the thermal protectiveness of different life preservers can be evaluated, and on this basis, it can also be used for human thermal comfort and cold shock assessment.