Heat stroke early warning system based on wearable biosensor

Abstract

The invention provides a heat stroke early warning system based on a wearable biosensor, and relates to the technical field of biosensors, and the system comprises a data collection module which obtains a temperature parameter and a heart rate signal based on the wearable biosensor, carries out the preprocessing of the temperature parameter and the heart rate signal, and obtains a temperature feature component and a heart rate feature component; the environment risk analysis module obtains environment parameters, and obtains environment risk factors based on the environment parameters; the physiological risk analysis module obtains user physiological parameters and obtains dynamic risk factors based on the user physiological parameters; a prediction correction module obtains a heat stroke risk prediction result based on the temperature feature component and the heart rate feature component, corrects the heat stroke risk prediction result, and obtains a corrected risk prediction result; and the fusion early warning module performs heat stroke early warning based on the environmental risk factor, the dynamic risk factor and the corrected risk prediction result. The technical problem that heat stroke early warning is inaccurate and unreliable in the prior art is solved.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of biosensors, in particular to a heat stroke early warning system based on wearable biosensors. BACKGROUND

[0002] As a serious acute illness in high-temperature environment, heat stroke has high morbidity and mortality, and the risk is particularly prominent in outdoor work, sports competition, elderly care and other scenarios, posing a major threat to human health and life safety.

[0003] However, traditional heat stroke early warning methods rely on manual inspection, fixed monitoring equipment or single physiological indicators, which have poor real-time performance, limited coverage and difficulty in continuously tracking individual status. At the same time, it is difficult to accurately judge the actual risk of different groups, which may lead to delayed or misjudged early warning, resulting in missed best intervention opportunities.

[0004] With the development of wearable biosensor technology, it can collect physiological data in real time and conveniently, providing a technical basis for heat stroke risk early warning. Therefore, there is an urgent need for a heat stroke early warning system based on wearable biosensors to address the shortcomings of traditional early warning methods. SUMMARY

[0005] The present application provides a heat stroke early warning system based on wearable biosensors to solve the technical problems of inaccurate and unreliable heat stroke early warning in the prior art.

[0006] The technical solution of the present application to solve the above technical problems is as follows: The present application provides a heat stroke early warning system based on wearable biosensors, comprising: A data acquisition module for acquiring temperature parameters and heart rate signals based on wearable biosensors, preprocessing the temperature parameters and heart rate signals, and acquiring temperature feature components and heart rate feature components; An environmental risk analysis module for acquiring environmental parameters and obtaining environmental risk factors based on the environmental parameters; A physiological risk analysis module for acquiring user physiological parameters and obtaining dynamic risk factors based on the user physiological parameters; A prediction correction module for obtaining heat stroke risk prediction results based on the temperature feature components and the heart rate feature components, and correcting the heat stroke risk prediction results to obtain corrected risk prediction results; A fusion early warning module for heat stroke early warning based on the environmental risk factors, the dynamic risk factors and the corrected risk prediction results.

[0007] The present application has the following advantages: Compared with the prior art, firstly, the temperature parameter and the heart rate signal are acquired based on the wearable biological sensor through the data acquisition module, the temperature parameter and the heart rate signal are preprocessed, the temperature characteristic component and the heart rate characteristic component are acquired, and high-quality and high-reliability core input data are provided for subsequent heat stroke risk prediction. Secondly, the environmental risk analysis module acquires the environmental parameter, and based on the environmental parameter, the environmental risk factor is acquired, and the environmental parameter affecting the occurrence of heat stroke is accurately captured, which provides scientific and intuitive environmental dimension quantitative basis for subsequent comprehensive warning of heat stroke based on physiological data of human body. Thirdly, the physiological risk analysis module acquires the user physiological parameter, and based on the user physiological parameter, the dynamic risk factor is acquired, which effectively avoids the false alarm or missed alarm caused by the unified standard, and provides accurate individual risk basis for personalized warning of heat stroke. Further, the prediction correction module acquires the heat stroke risk prediction result based on the temperature characteristic component and the heart rate characteristic component, and corrects the heat stroke risk prediction result to acquire the corrected risk prediction result, generates the heat stroke risk prediction result conforming to the physiological law, and effectively improves the accuracy and reliability of the heat stroke warning by correcting the temperature data distortion caused by environmental interference. Finally, the fusion warning module performs heat stroke warning based on the environmental risk factor, the dynamic risk factor and the corrected risk prediction result, effectively avoids the false alarm or missed alarm caused by a single index, realizes accurate warning and guides corresponding intervention measures, and improves the effectiveness of heat stroke prevention and response.

[0008] Through the above technical solutions, the heat stroke warning is performed based on the wearable biological sensor, the false alarm or missed alarm caused by the traditional warning relying on a single index and ignoring the cooperative influence of environment and individual is effectively avoided, the real-time, personalized and accurate warning of heat stroke risk is realized, the timeliness and reliability of heat stroke risk prevention and control are improved, and reliable technical support is provided for heat stroke prevention and intervention in different scenarios. BRIEF DESCRIPTION OF DRAWINGS

[0009] Figure 1 A structure schematic diagram of the heat stroke warning system based on the wearable biological sensor provided by the application is shown in the figure. Figure 2 A structure schematic diagram of the prediction correction module in the heat stroke warning system based on the wearable biological sensor provided by the application is shown in the figure.

[0010] In the figure, the components represented by each number are as follows: Data acquisition module 11, environmental risk analysis module 12, physiological risk analysis module 13, prediction correction module 14, fusion warning module 15. DETAILED DESCRIPTION

[0011] With reference to the drawings of the embodiments of the present application, the technical solutions in the embodiments of the present application will be clearly and completely described. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without creative work fall within the scope of the present application.

[0012] In the description of the present application, the terms "first", "second" are used only for descriptive purposes, and cannot be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Therefore, the features defined as "first", "second" can explicitly or implicitly include one or more of the features. In the description of the present application, the meaning of "a plurality of" is two or more, unless otherwise specifically limited.

[0013] In the description of the present application, the term "for example" is used to indicate "as an example, illustration or explanation". Any embodiment described as "for example" in the present application is not necessarily interpreted as more preferred or more advantageous than other embodiments. The following description is given in order to enable any person skilled in the art to implement and use the present application. In the following description, details are listed for the purpose of explanation. It should be understood that those skilled in the art can recognize that the present application can be implemented without using these specific details. In other examples, well-known structures and processes will not be described in detail to avoid unnecessary details making the description of the present application obscure. Therefore, the present application is not intended to be limited to the shown embodiments, but is consistent with the broadest scope in accordance with the principles and characteristics disclosed.

[0014] Embodiments, such as Figure 1 As shown, the embodiments of the present application provide a heat stroke early warning system based on wearable biosensors, which comprises a data acquisition module 11, an environmental risk analysis module 12, a physiological risk analysis module 13, a prediction correction module 14, and a fusion early warning module 15.

[0015] The data acquisition module 11 is used to acquire temperature parameters and heart rate signals based on wearable biosensors, and to preprocess the temperature parameters and heart rate signals to obtain temperature feature components and heart rate feature components.

[0016] In the heat stroke early warning scenario, wearable biosensors can continuously acquire human body temperature parameters and heart rate signals, but such raw physiological signals are prone to outliers due to sensor loosening, sweating and other interferences, and there is data redundancy among multi-dimensions of temperature parameters and multi-indices of heart rate signals, and there may also be mixed environmental, motion and other non-heat stroke related interferences, which cannot be directly used for risk judgment.

[0017] Therefore, it is necessary to eliminate abnormal data through preprocessing, and then extract feature components to accurately reflect the real heat accumulation state and the autonomic nervous regulation state of the heart, so as to provide high-quality input data for subsequent heatstroke risk prediction.

[0018] To solve the above problems, the temperature parameter and the heart rate signal are acquired based on the wearable biological sensor, the temperature parameter and the heart rate signal are preprocessed, and the temperature feature component and the heart rate feature component are acquired.

[0019] Specifically, the data acquisition module 11 comprises: a user parameter acquisition unit configured to acquire a temperature parameter and a heart rate signal based on a wearable biological sensor; a deviation elimination unit configured to preprocess the temperature parameter and the heart rate signal, eliminate significant deviation values, and acquire a temperature sequence and a heart rate signal sequence; a component generation unit configured to perform principal component analysis dimension reduction on the temperature sequence and the heart rate signal sequence, and acquire a temperature feature component and a heart rate feature component through independent component analysis.

[0020] In the embodiment of the present application, first, the temperature parameter and the heart rate signal are acquired based on the wearable biological sensor. The wearable biological sensor refers to a portable device integrating a temperature sensing module and a heart rate sensing module, such as a smart bracelet, a chest patch monitor, etc. The wearable biological sensor can be close to the human skin to collect the temperature parameter and the heart rate signal in real time and continuously. The temperature parameter refers to the data related to the heat state of the human body collected by the wearable biological sensor, such as the body surface temperature, the estimated core body temperature, etc. The heart rate signal refers to the signal reflecting the heart beat rhythm collected by the wearable biological sensor, such as the instantaneous heart rate, the heart rate waveform, etc.

[0021] For example, when the target user performs physical labor outdoors, the wrist smart bracelet worn by the user collects the wrist body surface temperature at a collection frequency of 1 time per minute, such as the real-time temperature parameter of 37.3°C, and collects the heart rate signal at a collection frequency of 1 time per second, such as the real-time heart rate signal of 100 times per minute.

[0022] Secondly, since the wearable biological sensor may be affected by external interference such as sweat soaking, temporary loosening, and accidental touch due to violent action, there are significant deviation values in the collected temperature parameter and heart rate signal. Therefore, the temperature parameter and the heart rate signal need to be preprocessed to eliminate the significant deviation values, and the temperature sequence and the heart rate signal sequence are acquired. The significant deviation value refers to the jump data in the temperature parameter or the heart rate signal caused by external interference or device error. Such data cannot reflect the real physiological state of the human body, for example, the temperature parameter suddenly jumps from 37°C to 45°C, and the heart rate signal suddenly drops from 100 times per minute to 30 times per minute.

[0023] Exemplarily, the physiological range threshold can be set, such as the physiological range threshold of the temperature parameter is 35-42°C, the physiological range threshold of the heart rate signal is 40-200 times per minute, the significant deviation value exceeding the physiological range threshold is eliminated, then the linear interpolation method can be used to fill in the missing data points after the significant deviation value, and finally arranged in the order of collection time to obtain continuous and reliable temperature sequence and heart rate signal sequence.

[0024] For example, if the physiological range threshold of the temperature parameter is set as 35-42°C, when the wearable biosensor collects the body surface temperature of 43°C at a certain time, which exceeds the physiological range threshold of the temperature parameter, it will be determined as a significant deviation value, and after deleting the value of 43°C, the current body surface temperature is calculated as 37.6°C by linear interpolation combined with the body surface temperature of the two time points before and after the current time, to ensure the continuity of the temperature sequence, avoid errors in subsequent analysis due to data rupture, and obtain the temperature sequence and heart rate signal sequence according to the same logic and method.

[0025] Finally, the temperature sequence and heart rate signal sequence are subjected to principal component analysis for dimension reduction, and the temperature feature component and heart rate feature component are obtained by independent component analysis. The principal component analysis is a data dimension reduction algorithm, which can convert multiple original dimensions with correlation in the temperature sequence or heart rate signal sequence into a few principal components through linear transformation. The principal components can retain more than 90% of the effective information of the original data, while eliminating the redundant correlation between dimensions. The independent component analysis is a feature separation algorithm, which separates the independent source signals from the principal components subjected to the principal component analysis for dimension reduction, eliminates the influence of environmental interference and human state interference on the core physiological signal, and extracts the pure features directly related to heat stroke.

[0026] Exemplarily, if the temperature sequence contains two dimensions of wrist surface temperature (such as 37.2℃) and estimated core body temperature (such as 37.8℃), and the correlation of the two is 0.92 due to the influence of the heat conduction characteristics of the human body, the dimensionality reduction processing is performed by principal component analysis, the characteristic information of the wrist surface temperature and the estimated core body temperature is combined by linear combination, and a temperature principal component is generated, which can retain more than 95% of the effective information of the temperature sequence. Then, the temperature principal component is input into the independent component analysis model, the interference signal is identified and filtered, such as the low surface temperature caused by wrist sweating, and the temperature characteristic component closer to the real heat state in the body is extracted by the signal separation algorithm, such as 37.9℃. The temperature characteristic component excludes the environmental interference of the surface and can accurately reflect the real heat accumulation level in the body, providing reliable temperature dimension input for subsequent risk prediction. Similarly, after the principal component analysis and independent component analysis of the heart rate signal sequence, the heart rate characteristic component is obtained, such as 82 times / minute. The heart rate characteristic component can reflect the abnormal degree of autonomic nervous regulation of the heart, and provide reliable heart rate dimension input for heatstroke risk judgment.

[0027] In summary, compared with the prior art, the present application obtains temperature parameters and heart rate signals based on wearable biosensors, pre-processes the temperature parameters and heart rate signals, and obtains temperature characteristic components and heart rate characteristic components. In this way, abnormal data caused by sensor loosening, sweating and other interference can be effectively eliminated, data redundancy in the temperature parameters and heart rate signals can be removed, and temperature characteristic components and heart rate characteristic components that can accurately reflect the real heat accumulation state and autonomic nervous regulation state of the human body can be obtained, providing high-quality and reliable core input data for subsequent heatstroke risk prediction.

[0028] The environmental risk analysis module 12 is configured to obtain environmental parameters and obtain environmental risk factors based on the environmental parameters.

[0029] The occurrence of heatstroke is highly related to external environmental conditions. The environmental temperature determines the heat exchange direction between the human body and the external environment. When the environmental temperature is higher than the body temperature, the human body cannot dissipate heat through radiation and conduction. The humidity determines the evaporation efficiency of sweat. The higher the humidity, the slower the evaporation of sweat, and the heat is easy to accumulate in the body. Both of them are key environmental stress factors that induce heatstroke.

[0030] To solve the above problems, the present application obtains environmental parameters and obtains environmental risk factors based on the environmental parameters.

[0031] Specifically, the environmental risk analysis module 12 comprises: The environmental parameter acquisition unit is configured to obtain environmental parameters, wherein the environmental parameters at least include environmental temperature and humidity; The environmental risk acquisition unit is configured to obtain environmental risk factors based on the environmental parameters.

[0032] In this embodiment, environmental parameters are first acquired, including at least ambient temperature and humidity. Ambient temperature directly determines the direction of heat exchange between the human body and the external environment, while ambient humidity directly affects the efficiency of sweat evaporation. The combined effect of these two parameters determines the intensity of thermal stress in the environment. Optionally, in addition to the two core environmental parameters of ambient temperature and humidity, environmental parameters such as wind speed and solar radiation intensity can also be collected according to the requirements of early warning accuracy.

[0033] For example, when a target user is engaged in physical labor outdoors, if the wearable biosensor integrates a temperature and humidity sensor, it can collect ambient temperature and humidity data at a fixed sampling frequency as environmental parameters. If the wearable biosensor does not integrate a temperature and humidity sensor, it can connect to a weather monitoring instrument deployed on-site via Bluetooth to obtain ambient temperature and humidity data.

[0034] Secondly, environmental risk factors are obtained based on environmental parameters. These environmental risk factors are quantitative values ​​that reflect the degree of environmental thermal stress, obtained through specific calculations of environmental parameters. A higher environmental risk factor indicates a higher environmental thermal risk. The core function of environmental risk factors is to transform multi-dimensional environmental parameters into a unified risk measurement standard.

[0035] For example, environmental risk factors can be calculated by pre-setting mapping relationships, such as assigning values ​​to ambient temperature and humidity.

[0036] For example, the synergistic effect of human body heat dissipation efficiency and environmental temperature and humidity can be used as the fundamental basis, while referring to the wet-bulb sphere temperature (WBGT) risk grading standard in the "Guidelines for the Prevention and Treatment of Heatstroke" and human physiological tolerance experimental data to construct an environmental risk mapping relationship. For instance, the environmental temperature can be divided into the ranges of 30-32℃ (low stress), 33-35℃ (medium stress), and ≥36℃ (high stress), with corresponding basic environmental risk factors of 0.2, 0.4, and 0.6; the humidity can be divided into the ranges of 60%-70% (weak impact), 71%-80% (medium impact), and ≥81% (strong impact), with corresponding basic environmental risk factors of 0.1, 0.3, and 0.5. The environmental risk factors are obtained by combining the basic environmental risk factors for different environmental temperatures and humidity levels.

[0037] For example, if a wearable biosensor integrating temperature and humidity sensors collects data showing an ambient temperature of 34°C and humidity of 75%, and queries a pre-built environmental risk mapping relationship, the basic environmental risk factors are found to be 0.4 and 0.3, respectively. The calculated environmental risk factor is 0.4 + 0.3 = 0.7. The environmental risk factor intuitively quantifies the intensity of thermal stress in the current environment.

[0038] In summary, compared to existing technologies, this application obtains environmental parameters and, based on these parameters, acquires environmental risk factors. This accurately captures the environmental parameters affecting the occurrence of heatstroke and transforms them into quantifiable environmental risk factors, providing a scientific and intuitive quantitative basis for subsequent comprehensive early warning of heatstroke by integrating human physiological data.

[0039] The physiological risk analysis module 13 is used to obtain user physiological parameters and, based on the user physiological parameters, obtain dynamic risk factors.

[0040] The risk of heatstroke varies significantly among individuals. Users of different ages and with different underlying medical conditions have varying tolerance to heat stress. For example, the elderly, obese individuals, and those with underlying medical conditions have lower heat tolerance and are at higher risk of heatstroke. Furthermore, existing early warning methods often ignore users' physiological parameters, resulting in a lack of personalized warnings and a high risk of misjudgment or missed diagnoses.

[0041] To address the aforementioned issues, this application obtains user physiological parameters and, based on these parameters, acquires dynamic risk factors.

[0042] Specifically, the physiological risk analysis module 13 includes: A physiological parameter acquisition unit is used to acquire user physiological parameters, wherein the user physiological parameters include basic physiological parameters and historical sensor parameters; The basic component generation unit is used to obtain basic risk factors based on the basic physiological parameters, and to obtain basic temperature components and basic heart rate components based on the historical sensing parameters. The risk factor analysis unit is used to obtain a temperature risk factor based on the deviation between the basic temperature component and the temperature characteristic component, and to obtain a heart rate risk factor based on the deviation between the basic heart rate component and the heart rate characteristic component. The dynamic risk factor determination unit is used to perform weighted calculations based on the basic risk factor, the temperature risk factor, and the heart rate risk factor to obtain the dynamic risk factor.

[0043] In this embodiment, user physiological parameters are first acquired, including basic physiological parameters and historical sensor parameters. Basic physiological parameters refer to relatively fixed and slowly changing physiological attributes of the user, such as age, body mass index (BMI), underlying diseases, and a history of heatstroke. These parameters directly determine the user's innate tolerance to heatstroke and can be manually entered by the user or imported from medical data. Historical sensor parameters refer to historical temperature parameters and historical heart rate signals collected by the wearable biosensor when the user is in a healthy state, such as in a non-high-temperature environment or during non-exercise. These parameters can serve as a benchmark for subsequently determining whether the current physiological state is abnormal and can be retrieved from the local storage of the wearable biosensor or from a cloud database.

[0044] For example, when a user uses the wearable biosensor for the first time, basic physiological parameters are collected, such as the user manually entering data like age, body mass index (BMI), underlying diseases, and history of heatstroke. These historical sensor parameters are then automatically updated daily during subsequent use to ensure data timeliness and accuracy. For instance, a target user might manually enter their age as 55, have hypertension, a BMI of 28, and no history of heatstroke as basic physiological parameters upon first use. Over the next 7 days, temperature and heart rate signals (e.g., an average of 36.8°C and 72 beats per minute) under resting conditions at home are collected as historical sensor parameters.

[0045] Secondly, based on basic physiological parameters, basic risk factors are obtained, and based on historical sensor parameters, basic temperature and basic heart rate components are obtained. The basic risk factor quantifies the user's congenital risk of heatstroke based on basic physiological parameters, typically ranging from 0 to 1. A higher basic risk factor indicates a higher risk. Optionally, preset rules can be set based on the influence of various physiological dimensions on heatstroke risk in medical research. The basic risk factor is determined according to these preset rules: for example, adding 0.2 points for age ≥ 65 years, 0.2 points for underlying diseases, 0.15 points for BMI ≥ 30, 0.3 points for a history of heatstroke, and 0 points for no corresponding condition. The scores of each dimension are summed to obtain the basic risk factor, while ensuring that the sum of the scores for each dimension is limited to a value between 0 and 1.

[0046] For example, if the target user's basic physiological parameters are age 55, hypertension, BMI 28, and no history of heatstroke, the basic risk factor is calculated as 0 + 0.2 + 0 + 0 = 0.3 according to the preset rules. The basic risk factor accurately reflects the target user's congenital heatstroke risk level due to underlying diseases.

[0047] The basic temperature component and basic heart rate component are feature-processed historical sensor parameters. They can be extracted from these parameters using principal component analysis and independent component analysis methods described in the environmental risk analysis module. These two components serve as the benchmark for determining whether the current feature components are abnormal. For example, after feature extraction, the target user's historical sensor parameters yield a basic temperature component of 36.9℃ and a basic heart rate component of 73 beats / minute.

[0048] Next, based on the deviations of the baseline temperature component and the temperature characteristic component, a temperature risk factor is obtained, and based on the deviations of the baseline heart rate component and the heart rate characteristic component, a heart rate risk factor is obtained. Specifically, the temperature risk factor = |baseline temperature component - temperature characteristic component| / baseline temperature component, and the heart rate risk factor = |baseline heart rate component - heart rate characteristic component| / baseline heart rate component, quantifying the degree of abnormality in the target user's current physiological state. For example, if the target user's baseline temperature component is 36.9℃, the baseline heart rate component is 73 beats / minute, the temperature characteristic component is 37.9℃, and the heart rate characteristic component is 82 beats / minute, then the temperature risk factor = |36.9℃ - 37.9℃| / 36.9℃ ≈ 0.027, and the heart rate risk factor = |73 - 82| / 73 ≈ 0.12.

[0049] Finally, a weighted calculation is performed based on the basic risk factor, temperature risk factor, and heart rate risk factor to obtain the dynamic risk factor. The dynamic risk factor is calculated as follows: Dynamic Risk Factor = (Basic Risk Factor × Weight 1) + (Temperature Risk Factor × Weight 2) + (Heart Rate Risk Factor × Weight 3), where Weight 1 + Weight 2 + Weight 3 = 1. For example, if weights 1, 2, and 3 are determined to be 0.3, 0.35, and 0.35 respectively based on basic physiological parameters, and the target user's basic risk factor is 0.3, temperature risk factor is 0.027, and heart rate risk factor is 0.12, then the dynamic risk factor = (0.3 × 0.3) + (0.027 × 0.35) + (0.12 × 0.35) ≈ 0.14. This indicates that the target user's current risk of heatstroke is at a low level, providing an individualized risk basis for subsequent early warning.

[0050] Thus, by weighted fusion of innate basic risks and current physiological abnormality risks, a dynamic risk factor that can reflect the user's heatstroke risk in real time is obtained. The dynamic risk factor takes a value of 0-1, and the higher the dynamic risk factor, the higher the risk of heatstroke.

[0051] Furthermore, the weights for the weighted calculation are obtained based on basic physiological parameters.

[0052] In this embodiment of the application, basic physiological parameters can reflect the risk factors. Therefore, by adjusting the weights of basic risk factors, temperature risk factors, and heart rate risk factors through basic physiological parameters, the problem of general weights ignoring the risk focus of different populations can be avoided.

[0053] For example, the weights of the weighted calculation can be set according to the priority of the impact of basic physiological parameters on the risk of heatstroke: if the basic physiological parameters show that the user is a high-risk group with an age ≥65 years, serious underlying diseases, and a history of heatstroke, the weight of the basic risk factors should be increased, such as 0.8~1, while the weights of temperature risk factors and heart rate risk factors should be decreased. This is because even if the current physiological deviation is not large, the congenitally high-risk group is prone to heatstroke due to their weak tolerance, and the basic risk factors need to be given special attention; if the basic physiological parameters show that the user is 20-40 years old, has no underlying diseases, and has a history of heatstroke, the weight of the basic risk factors should be increased, such as 0.8~1, while the weights of temperature risk factors and heart rate risk factors should be decreased. This is because even if the current physiological deviation is not large, the congenitally high-risk group is prone to heatstroke due to their weak tolerance, and the basic risk factors need to be given special attention. For individuals with normal health and no history of heatstroke who are congenitally low-risk, the weight of basic risk factors can be appropriately reduced, while the weights of temperature and heart rate risk factors can be increased. This is because the risk for congenitally low-risk individuals mainly stems from abnormalities in their current physiological state, such as a sudden increase in heart rate due to high temperature, requiring close monitoring of real-time deviations. If basic physiological parameters indicate that the user is a congenitally intermediate-risk individual aged 41-64 years with mild underlying diseases and a slightly higher BMI, a balanced weighting can be adopted, such as setting the weights of basic risk factors, temperature risk factors, and heart rate risk factors to 0.3, 0.35, and 0.35, respectively.

[0054] For example, the target user's basic physiological parameters are: age 55, hypertension, BMI 28, no history of heatstroke. Because of the underlying disease, but other physiological indicators are basically normal, the target user is classified as a congenitally medium-risk group, and the weights of the basic risk factor, temperature risk factor and heart rate risk factor are configured as 0.3, 0.35 and 0.35, respectively.

[0055] In summary, compared to existing technologies, this application obtains user physiological parameters and, based on these parameters, acquires dynamic risk factors. This effectively avoids misjudgments or omissions in early warnings caused by standardized approaches, providing accurate individual risk data for personalized early warning of heatstroke.

[0056] The prediction correction module 14 is used to obtain a heatstroke risk prediction result based on the temperature feature component and the heart rate feature component, and to correct the heatstroke risk prediction result to obtain a corrected risk prediction result.

[0057] Since temperature and heart rate components are easily affected by environmental factors, such as low body temperature masking true heat accumulation in low-temperature environments and sensor distortion due to environmental temperature and humidity, it is necessary to correct the heatstroke risk prediction results to solve the problems of difficulty in capturing nonlinear correlations and data distortion caused by environmental interference.

[0058] To address the aforementioned issues, this application obtains heatstroke risk prediction results based on the temperature feature components and the heart rate feature components, and then corrects the heatstroke risk prediction results to obtain corrected risk prediction results.

[0059] Specifically, such as Figure 2 As shown, the prediction correction module 14 includes: The sample data acquisition unit is used to acquire historical temperature component sets and historical heart rate component sets based on historical sensing parameters. The sample data annotation unit is used to annotate the historical temperature component set and the historical heart rate component set to obtain a heatstroke risk result set. Model building unit, used to build risk predictors; The model training unit is used to train the risk predictor until convergence using the historical temperature component set, the historical heart rate component set, and the heatstroke risk result set. The prediction output unit is used to input the temperature feature component and the heart rate feature component into the risk predictor to obtain the heatstroke risk prediction result.

[0060] In this embodiment, historical temperature and heart rate component sets are first obtained based on historical sensor parameters. For example, based on historical sensor parameters, a large amount of historical sensor parameters from the past 30 days of the target user are collected. After principal component analysis and independent component analysis, the historical temperature component set is [36.7℃, 36.8℃, 36.9℃, ..., 37.1℃], and the historical heart rate component set is [71 beats / minute, 72 beats / minute, 73 beats / minute, ..., 74 beats / minute]. These historical temperature and heart rate component sets will be used as input samples for training the risk predictor.

[0061] Secondly, the historical temperature component set and historical heart rate component set are labeled to obtain a heatstroke risk result set. For example, the target user's historical actual status is collected, such as whether heatstroke symptoms have occurred or whether they are in a state of heat stress danger. The probability of heatstroke occurrence corresponding to different historical temperature component sets and historical heart rate component sets is calculated to obtain multiple heatstroke risk results, forming a heatstroke risk result set.

[0062] Next, a risk predictor is constructed. For example, to address the nonlinear correlation between temperature and heart rate components in heatstroke risk prediction, a shallow neural network can be used to construct a risk predictor, primarily consisting of an input layer, one hidden layer, and an output layer. Specifically: The input layer has two neurons, corresponding to the temperature and heart rate feature components respectively. The input layer converts these features into numerical inputs that the neural network can recognize. For example, normalization is used to map the values ​​to the [0,1] interval to avoid the impact of dimensional differences on training results. The hidden layer has 16 neurons using the ReLU activation function. The number of neurons can be dynamically adjusted based on the sample size. The ReLU activation function effectively simulates the non-linear relationship between physiological characteristics and risk. The hidden layer linearly combines the input features using a weight matrix (16×2 dimensional) and a bias term (16×1 dimensional), and then outputs 16 non-linear feature mapping values ​​after ReLU activation, achieving deep extraction of the synergistic effect of temperature and heart rate. The output layer has one neuron, corresponding to the heatstroke risk result.

[0063] Furthermore, the risk predictor is trained until convergence using historical temperature component sets, historical heart rate component sets, and heatstroke risk result sets. For example, the risk predictor can be trained using the following technical path: 1. Data Preparation: Divide the historical temperature component sets, historical heart rate component sets, and corresponding heatstroke risk result sets into training, validation, and test sets in a 7:1.5:1.5 ratio. 2. Model Training: Using the historical temperature and historical heart rate components from the training set as input features, and the corresponding heatstroke risk results as supervision labels, quantify the error between the model's predicted values ​​and the true labels using the cross-entropy loss function. Iteratively update the model weights and biases using an Adam optimizer with a learning rate of 0.01. Calculate the validation set loss after each iteration until the validation set loss does not decrease for five consecutive iterations, and the difference between the training set loss and the validation set loss is less than 0.02. The model is then considered converged, and the trained risk predictor is obtained.

[0064] Finally, the temperature and heart rate features are input into the risk predictor to obtain the heatstroke risk prediction result. For example, the temperature feature (e.g., 37.9℃) and heart rate feature (e.g., 82 beats / minute) are input into the pre-trained risk predictor to obtain the heatstroke risk prediction result. If the result is 40%, it indicates that the target user's current physiological characteristics correspond to a moderate risk of heatstroke.

[0065] Furthermore, the prediction correction module 14 also includes: The temperature confidence assessment unit is used to obtain temperature confidence levels based on environmental parameters and temperature sequences. The historical benchmark acquisition unit is used to acquire historical environmental parameters and, based on historical sensing parameters, to acquire the historical average temperature confidence level. A similarity calculation unit is used to calculate the similarity between the temperature confidence level and the historical average temperature confidence level, and to configure correction parameters; The parameter correction unit is used to correct the heatstroke risk prediction result using the correction parameters, and obtain the corrected risk prediction result.

[0066] In this embodiment, the temperature confidence level is first obtained based on environmental parameters and temperature sequence. For example, when the trend of human body temperature change is consistent with the trend of environmental temperature change, such as when the human body temperature rises synchronously with the environmental temperature rise, it indicates that the temperature sequence is more likely to reflect the true heat accumulation; otherwise, it may be affected by other factors and the reliability is not high. Therefore, the similarity between the changing trends of environmental parameters and temperature sequence within the same time window can be used as the temperature confidence level.

[0067] For example, within the same time window, select six consecutive data collection times and extract the environmental parameter sequence as [32℃, 33℃, 34℃, 35℃, 36℃, 37℃] and the human body temperature sequence as [36.5℃, 36.7℃, 36.9℃, 37.1℃, 37.3℃, 37.5℃]. Calculate the cycle change rate of both, such as the cycle change rate of the environmental parameter sequence being [+1℃, +1℃, +1℃, +1℃] and the cycle change rate of the temperature sequence being [+0.2℃, +0.2℃, +0.2℃, +0.2℃, +0.2℃]. Then, use the Pearson correlation coefficient to calculate the similarity between the two sets of cycle change rates. If the calculated similarity is 1.0, indicating a perfect positive correlation, this similarity can be directly used as the temperature confidence score, indicating that the current environmental parameters and temperature sequence are consistent and the data is reliable.

[0068] Secondly, historical environmental parameters are obtained, and based on these historical sensor parameters, historical average temperature confidence levels are calculated. For example, historical environmental parameters and corresponding historical temperature sequences are retrieved from a historical database, and multiple historical temperature confidence levels are calculated using the same method as described above. Then, the arithmetic mean is calculated to obtain the historical average temperature confidence level, such as 0.85, which serves as a benchmark for determining whether the current temperature confidence level deviates from the normal level.

[0069] Next, calculate the similarity between the temperature confidence level and the historical average temperature confidence level, and configure the correction parameters. The similarity can be calculated using a Gaussian function method that fits the data distribution characteristics, accurately characterizing the difference between the current temperature confidence level and the historical average temperature confidence level through nonlinear mapping. For example, the formula for calculating the similarity is: Similarity = exp[-k × (Temperature Confidence Level - Historical Average Temperature Confidence Level)²], where k is an adjustment factor, typically 5-10, used to control the rate of change of similarity with deviation; exp is the natural exponential function, ensuring that the similarity value always stays within the (0,1) interval.

[0070] For example, if the temperature confidence level is 1.0 and the historical average temperature confidence level is 0.85, and k=6, then the similarity = exp[-6×(1.0-0.85)²]≈0.87, indicating that the similarity remains at a high level. Similarity reflects the reliability of the current temperature data compared to the historical baseline; a higher similarity indicates a smaller difference from the historical baseline, and therefore a smaller correction force.

[0071] The correction parameter is a coefficient used to adjust the heatstroke risk prediction results, ranging from 0 to 1. The lower the similarity, the less reliable the current data, and the larger the correction parameter. For example, 1 minus the similarity can be used as the correction parameter. For instance, if the similarity is 0.87, it indicates that the difference between the temperature confidence level and the historical average temperature confidence level is small, and the reliability is high. In this case, the correction parameter = 1 - 0.87 = 0.13, indicating that the correction to the heatstroke risk prediction results is weak, and only a small adjustment is needed to avoid over-correction masking the original prediction trend based on reliable data.

[0072] Finally, correction parameters are used to correct the heatstroke risk prediction results, obtaining the corrected risk prediction results. The corrected risk prediction result is calculated as: Heatstroke risk prediction result × (1 + correction parameters). By appropriately amplifying the heatstroke risk prediction results, the risk underestimation that might be caused by insufficient reliability of temperature data is compensated for. For example, if a temperature signal is misjudged as normal due to environmental interference, the actual risk may be higher.

[0073] For example, if the predicted risk of heatstroke is 40% and the correction parameter is 0.13, then the corrected risk prediction result = 40% × (1 + 0.13) = 45.2%. This indicates that the current temperature data has high reliability, and only a slight amplification of the heatstroke risk prediction result is needed to compensate for possible minor data deviations and avoid over-adjustment leading to misjudgment of risk. At the same time, when the correction parameter is increased, the corrected result for the heatstroke risk prediction will be significantly improved, more effectively compensating for the underestimation of risk caused by unreliable data, ensuring that the corrected risk prediction result more closely reflects the actual risk level.

[0074] In summary, compared to existing technologies, this application obtains heatstroke risk prediction results based on the temperature feature components and the heart rate feature components, and then corrects these results to obtain corrected risk prediction results. This approach not only captures the nonlinear correlation between the temperature and heart rate feature components through the risk predictor, generating heatstroke risk prediction results that conform to physiological patterns, but also effectively improves the accuracy and reliability of heatstroke early warning by correcting and compensating for temperature data distortion caused by environmental interference.

[0075] The integrated early warning module 15 is used to provide early warning of heatstroke based on the environmental risk factors, the dynamic risk factors, and the corrected risk prediction results.

[0076] Traditional heatstroke early warning often relies on a single risk indicator, which is prone to misjudgment. However, the occurrence of heatstroke is essentially the result of an imbalance between environmental heat stress and individual physiological tolerance. It is necessary to integrate the environmental risk factors, dynamic risk factors and corrected risk prediction results obtained from the aforementioned steps to fully characterize the real risk and then implement comprehensive and accurate early warning.

[0077] To address the aforementioned issues, this application provides a heatstroke early warning system based on the aforementioned environmental risk factors, the aforementioned dynamic risk factors, and the aforementioned modified risk prediction results.

[0078] Specifically, the fusion early warning module 15 includes: The parameter fusion unit is used to obtain the heatstroke risk prediction level based on the environmental risk factor, the dynamic risk factor, and the corrected risk prediction result. The scheme acquisition unit is used to acquire heatstroke risk warning schemes. The output execution unit is used to execute the heatstroke risk warning according to the heatstroke risk warning scheme based on the heatstroke risk warning level.

[0079] In this embodiment, heatstroke occurs as a result of an imbalance between external environmental stress and individual physiological tolerance. Therefore, it is necessary to obtain a heatstroke risk prediction level based on environmental risk factors, dynamic risk factors, and modified risk prediction results to comprehensively reflect the true risk level. For example, weights can be assigned to environmental risk factors, dynamic risk factors, and modified risk prediction results. The weight allocation can be dynamically set according to the actual scenario. Preferably, since the modified risk prediction result is a physiological risk signal corrected for environmental interference, it directly reflects the current body's heat tolerance state and is the most crucial judgment criterion, thus receiving the highest weight. For example, a weight of 0.6 can be set, with environmental risk factors and dynamic risk factors assigned weights of 0.2 and 0.2 respectively. Those skilled in the art can dynamically adjust these weights according to actual needs.

[0080] For example, if the weights of the environmental risk factor, dynamic risk factor, and modified risk prediction result are 0.2, 0.2, and 0.6 respectively, and the environmental risk factor is 0.7, the dynamic risk factor is 0.14, and the modified risk prediction result is 45.2%, then the comprehensive risk factor = 0.2 × 0.7 + 0.2 × 0.14 + 0.6 × 45.2% = 0.4392.

[0081] For example, the comprehensive risk factor can be mapped to a four-level heatstroke risk prediction level according to medical standards: 0-0.2 is no risk, 0.2-0.5 is low risk, 0.5-0.8 is medium risk, and 0.8-1.0 is high risk. For instance, a calculated comprehensive risk factor of 0.4392 belongs to low risk and is used as the heatstroke risk prediction level.

[0082] Secondly, obtain a heatstroke risk warning plan. This plan is determined based on the predicted heatstroke risk level. It can be based on historical data, authoritative research papers, and the characteristics of different application scenarios to ensure that each predicted heatstroke risk level has a clear warning plan.

[0083] For example, the heatstroke risk warning scheme can be set up as follows: for no risk, risk data is only recorded in the system background without actively pushing notifications; for low risk, only mild reminders are needed, such as reaching the user through slight vibration of wearable biosensors and pop-up text prompts; for medium risk, stronger intervention is needed, and SMS prompts are sent to preset emergency contacts, such as prompts to immediately move to a cool place to rest for 15 minutes, drink electrolyte drinks, and monitor body temperature after 10 minutes; for high risk, the wearable biosensor's audible and visual alarm is triggered, and the emergency contact number is automatically dialed and the nearest emergency station is contacted.

[0084] Finally, based on the heatstroke risk warning level, a heatstroke risk warning is issued according to the heatstroke risk warning plan. For example, if the predicted heatstroke risk level is low, the corresponding heatstroke risk warning plan is to use a wearable biosensor to gently vibrate and alert the target user.

[0085] In summary, compared to existing technologies, this application provides heatstroke early warning based on the aforementioned environmental risk factors, dynamic risk factors, and modified risk prediction results. By integrating parameters from three dimensions, it comprehensively characterizes heatstroke risk, effectively avoiding misjudgments or omissions caused by single indicators, achieving accurate early warning and guiding corresponding intervention measures, thus improving the effectiveness of heatstroke prevention and response.

[0086] In summary, the embodiments of this application have at least the following technical effects: Compared to existing technologies, this application first acquires temperature parameters and heart rate signals based on wearable biosensors, and then preprocesses these parameters and signals to obtain temperature and heart rate feature components. This effectively eliminates abnormal data caused by sensor loosening, sweating, and other interferences, removes data redundancy from the temperature parameters and heart rate signals, and yields temperature and heart rate feature components that accurately reflect the true state of heat accumulation and autonomic nervous system regulation of the heart. This provides high-quality, highly reliable core input data for subsequent heatstroke risk prediction.

[0087] Secondly, this application acquires environmental parameters and, based on these parameters, obtains environmental risk factors. In this way, environmental parameters affecting the occurrence of heatstroke are accurately captured and transformed into quantifiable environmental risk factors, providing a scientific and intuitive quantitative basis for subsequent comprehensive early warning of heatstroke by integrating human physiological data.

[0088] Furthermore, this application obtains user physiological parameters and, based on these parameters, acquires dynamic risk factors. This effectively avoids misjudgments or omissions in early warnings caused by standardized approaches, providing accurate individual risk data for personalized early warnings of heatstroke.

[0089] Furthermore, this application obtains heatstroke risk prediction results based on the temperature feature components and the heart rate feature components, and corrects the heatstroke risk prediction results to obtain corrected risk prediction results. In this way, it can both capture the nonlinear correlation between the temperature feature components and the heart rate feature components through the risk predictor to generate heatstroke risk prediction results that conform to physiological laws, and effectively improve the accuracy and reliability of heatstroke early warning by correcting and compensating for temperature data distortion caused by environmental interference.

[0090] Finally, this application uses the aforementioned environmental risk factors, dynamic risk factors, and modified risk prediction results to provide early warning of heatstroke. By integrating parameters from three dimensions, it comprehensively characterizes the risk of heatstroke, effectively avoiding misjudgments or omissions caused by a single indicator, achieving accurate early warning and guiding corresponding intervention measures, and improving the effectiveness of heatstroke prevention and response.

[0091] Through the above technical solution, this application uses wearable biosensors for heatstroke early warning, which effectively avoids the problems of misjudgment or missed judgment caused by traditional early warning relying on single indicators and ignoring the synergistic influence of environment and individual. It realizes real-time, personalized and accurate early warning of heatstroke risk, improves the timeliness and effectiveness of heatstroke risk prevention and control, and provides reliable technical support for heatstroke prevention and intervention in different scenarios.

[0092] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0093] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0094] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0095] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0096] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0097] Although preferred embodiments of the invention have been described, those skilled in the art, once they have learned the basic inventive concept, can make other changes and modifications to these embodiments.

[0098] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of this invention and its equivalents, this invention also intends to include these modifications and variations.

Claims

1. A heatstroke early warning system based on wearable biosensors, characterized in that, The system includes: The data acquisition module is used to acquire temperature parameters and heart rate signals based on wearable biosensors, and to preprocess the temperature parameters and heart rate signals to obtain temperature feature components and heart rate feature components. The environmental risk analysis module is used to acquire environmental parameters and, based on these parameters, to obtain environmental risk factors. The physiological risk analysis module is used to acquire user physiological parameters and, based on these parameters, obtain dynamic risk factors. The prediction correction module is used to obtain heatstroke risk prediction results based on the temperature feature components and the heart rate feature components, and to correct the heatstroke risk prediction results to obtain corrected risk prediction results. The integrated early warning module is used to provide early warning of heatstroke based on the environmental risk factors, the dynamic risk factors, and the corrected risk prediction results.

2. The heatstroke early warning system based on wearable biosensors as described in claim 1, characterized in that, The data acquisition module includes: The user parameter acquisition unit is used to acquire temperature parameters and heart rate signals based on wearable biosensors. The deviation removal unit is used to preprocess the temperature parameters and the heart rate signal, remove significant deviation values, and obtain the temperature sequence and heart rate signal sequence. The component generation unit is used to perform principal component analysis to reduce the dimensionality of the temperature sequence and the heart rate signal sequence, and to obtain the temperature feature component and the heart rate feature component through independent component analysis.

3. The heatstroke early warning system based on wearable biosensors as described in claim 1, characterized in that, The environmental risk analysis module includes: An environmental parameter acquisition unit is used to acquire environmental parameters, wherein the environmental parameters include at least ambient temperature and humidity; An environmental risk acquisition unit is used to acquire environmental risk factors based on the environmental parameters.

4. The heatstroke early warning system based on wearable biosensors as described in claim 1, characterized in that, The physiological risk analysis module includes: A physiological parameter acquisition unit is used to acquire user physiological parameters, wherein the user physiological parameters include basic physiological parameters and historical sensor parameters; The basic component generation unit is used to obtain basic risk factors based on the basic physiological parameters, and to obtain basic temperature components and basic heart rate components based on the historical sensing parameters. The risk factor analysis unit is used to obtain a temperature risk factor based on the deviation between the basic temperature component and the temperature characteristic component, and to obtain a heart rate risk factor based on the deviation between the basic heart rate component and the heart rate characteristic component. The dynamic risk factor determination unit is used to perform weighted calculations based on the basic risk factor, the temperature risk factor, and the heart rate risk factor to obtain the dynamic risk factor.

5. The heatstroke early warning system based on wearable biosensors according to claim 4, characterized in that, The dynamic risk factor determination unit also includes: The weighting determination unit is used to perform weighted calculations based on the basic risk factor, the temperature risk factor, and the heart rate risk factor, wherein the weights for the weighted calculations are obtained based on the basic physiological parameters.

6. The heatstroke early warning system based on wearable biosensors according to claim 1, characterized in that, The prediction correction module includes: The sample data acquisition unit is used to acquire historical temperature component sets and historical heart rate component sets based on historical sensing parameters. The sample data annotation unit is used to annotate the historical temperature component set and the historical heart rate component set to obtain a heatstroke risk result set. Model building unit, used to build risk predictors; The model training unit is used to train the risk predictor until convergence using the historical temperature component set, the historical heart rate component set, and the heatstroke risk result set. The prediction output unit is used to input the temperature feature component and the heart rate feature component into the risk predictor to obtain the heatstroke risk prediction result.

7. The heatstroke early warning system based on wearable biosensors according to claim 1, characterized in that, The prediction correction module further includes: The temperature confidence assessment unit is used to obtain temperature confidence levels based on environmental parameters and temperature sequences. The historical benchmark acquisition unit is used to acquire historical environmental parameters and, based on historical sensing parameters, to acquire the historical average temperature confidence level. A similarity calculation unit is used to calculate the similarity between the temperature confidence level and the historical average temperature confidence level, and to configure correction parameters; The parameter correction unit is used to correct the heatstroke risk prediction result using the correction parameters, and obtain the corrected risk prediction result.

8. The heatstroke early warning system based on wearable biosensors according to claim 1, characterized in that, The fusion early warning module includes: The parameter fusion unit is used to obtain the heatstroke risk prediction level based on the environmental risk factor, the dynamic risk factor, and the corrected risk prediction result. The scheme acquisition unit is used to acquire heatstroke risk warning schemes. The output execution unit is used to execute the heatstroke risk warning according to the heatstroke risk warning scheme based on the heatstroke risk warning level.

Images