Method and device for evaluating results of combined detection of biomarkers for early warning of heat stroke

Abstract

This invention discloses a method and apparatus for evaluating the results of biomarker combination detection for early warning of heatstroke, belonging to the field of trace detection technology. The method includes: an optimization method for sampling time points based on multi-time-point monitoring data of in vitro biological samples; obtaining concentration data of a first biomarker combination at an optimized sampling time point set, and calculating an early inflammation-driven index based on this data; obtaining concentration data of a second biomarker and a third biomarker for the in vitro samples; constructing an organ damage surrogate model to perform hierarchical regression decomposition on the third biomarker concentration data to obtain residual components in the third biomarker that cannot be explained by the organ damage surrogate; merging the early inflammation-driven index with the residual components to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score, and classifying the risk level based on the score.

Description

Technical Field

[0001] This invention relates to the field of trace detection technology, and in particular to a method and apparatus for evaluating the results of combined biomarker detection for early warning of heatstroke. Background Technology

[0002] Heatstroke is a severe heat stress response caused by exposure to high temperatures, characterized by a rapid rise in body temperature (usually exceeding 40°C) accompanied by multiple organ failure. It is an acute and critical illness that often occurs in extreme heat environments, particularly when exposed to high temperatures and humidity in hot weather, or during strenuous physical activity. Heatstroke is a fatal disease and can lead to death or long-term disability if not treated promptly. Therefore, to combat this disease, researchers have been developing technologies that can detect and provide early warnings for heatstroke.

[0003] Currently, heatstroke detection primarily relies on body temperature monitoring, heart rate changes, and assessment of other clinical symptoms, most of which depend on in vitro testing. Some studies have also proposed using combinations of common biomarkers in patients' blood, such as ALT (alanine aminotransferase), AST (aspartate aminotransferase), and IL-6 (interleukin-6), to assess the risk of inflammation caused by heatstroke based on changes in their concentrations. Furthermore, existing techniques generally establish mouse models and use statistical analysis to evaluate the correlation between changes in biomarker concentrations at different time points and indicators of multi-organ damage, thereby validating the effectiveness of biomarker combinations in early warning of heatstroke.

[0004] However, current technologies often lack obvious external symptoms in the early stages of heatstroke, and most early warning methods rely on clinical symptoms, which can easily lead to the loss of early intervention opportunities. When detecting common blood biomarker combinations, their concentrations only show significant changes after 24 hours, and the positive feedback regulation between ALT, AST, and IL-6 makes it difficult to distinguish between heatstroke and other inflammations, resulting in inaccurate warnings in clinical practice and affecting the final warning effect. Furthermore, inappropriate blood testing timing can lead to incomplete or redundant data collection, failing to effectively capture key changes in biomarker concentrations.

[0005] It is evident that existing technologies cannot effectively and accurately provide early warnings of heatstroke. Summary of the Invention

[0006] To address the technical problem of existing technologies failing to effectively and accurately provide early warnings of heatstroke, this invention provides a method and apparatus for evaluating the results of combined biomarker detection for early warning of heatstroke. The technical solution is as follows: On the one hand, a method for evaluating the results of combined biomarker detection for early warning of heatstroke is provided. This method is implemented by electronic devices and includes: A sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples is used to obtain the concentration data of the first biomarker combination at the optimized sampling time point set, wherein the optimized sampling time point set is a time point set generated by the sampling time point optimization process. The early inflammation-driving index was calculated based on the concentration data of the first biomarker combination at the optimized sampling time point set. Obtain the concentration data of the second and third biomarkers for in vitro samples, and construct an organ damage surrogate model; Based on the concentration data of the second biomarker and the organ damage proxy model, the concentration data of the third biomarker is subjected to hierarchical regression decomposition to obtain the residual components of the third biomarker that cannot be explained by the organ damage proxy. The early inflammation-driven index is combined with the residual component to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score; Risk levels are determined based on the inflammation-driven index score and the target component.

[0007] Optionally, the first biomarker combination is a combination of multiple biomarkers used to characterize inflammation and cell damage-related states; the second biomarker is at least one biomarker used to characterize organ damage-related states; and the third biomarker is at least one biomarker used to characterize inflammatory responses and immune activity.

[0008] Optionally, the first combination of biomarkers includes CitH3, NE, and HSP90;

[0009] The second biomarker includes at least one of ALT and AST; The third biomarker includes IL-6.

[0010] Optionally, the sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples, the step of obtaining the concentration data of the first biomarker combination at the optimized sampling time point set includes: The concentration data of the first biomarker combination at multiple candidate sampling time points of the same in vitro biological sample are obtained, and batch alignment and quality control are performed on the first biomarker combination concentration data to obtain comparable time series data. Based on the time series data, a smoothed time curve of the first biomarker combination is constructed, the rate of change of each candidate sampling time point is calculated, and the information content index of each candidate sampling time point is calculated. Under the premise of meeting the preset sampling constraints, the optimized sampling time point set is selected from the candidate sampling time point set and output according to the information content index.

[0011] Optionally, the sampling constraints include at least an upper limit constraint on the number of samplings, a minimum time interval constraint between adjacent sampling time points, and a phased coverage constraint; The phased coverage constraint is used to limit the set of sampling time points to cover at least two of the early, middle and late phases.

[0012] Optionally, the step of calculating the early inflammation-driving index based on the concentration data of the first biomarker combination at the optimized sampling time point set includes: The concentration data of the first biomarker combined on the optimized sampling time point set are standardized and then input into a mapping function to obtain the early inflammation driving index, wherein the mapping function is any one of a weighted summation function, a nonlinear function, or a piecewise function.

[0013] Optionally, the step of performing hierarchical regression decomposition on the third biomarker concentration data based on the second biomarker concentration data and the organ damage surrogate model to obtain the residual components of the third biomarker that cannot be explained by the organ damage surrogate model includes: The organ damage surrogate model constructs a damage surrogate quantity using the concentration data of the second biomarker as input, and uses the damage surrogate quantity as an explanatory variable to perform hierarchical regression decomposition on the concentration change represented by the concentration data of the third biomarker to obtain the regression fit value. The difference between the observed value of the third biomarker and the regression fitted value is determined as the residual component of the third biomarker that cannot be explained by organ damage proxy. Optionally, the hierarchical regression decomposition includes a first-level regression and a second-level regression; The first layer regression was used to explain the concentration changes of the third biomarker using the damage proxy to obtain intermediate residuals; The second-level regression interprets the intermediate residuals using the early inflammation-driven index to obtain the contribution index of the early inflammation-driven index to the changes in the third biomarker residuals.

[0014] Optionally, the step of combining the early inflammation-driven index with the residual component to determine the target portion of the third biomarker affected by the first biomarker includes: The early inflammation-driving index and residual components are gated and fused to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driving index score. The gating conditions used in the gating fusion are determined by the coverage integrity of the sampling time point set and / or the validity determination result of the second biomarker.

[0015] On the other hand, a biomarker combination detection result evaluation device for early warning of heatstroke is provided, the device comprising: The first acquisition module is used to acquire the concentration data of the first biomarker combination on the optimized sampling time point set based on the sampling time point optimization method of multi-time point monitoring data of in vitro biological samples, wherein the optimized sampling time point set is a time point set generated by the sampling time point optimization process. The calculation module is used to calculate the early inflammation-driven index based on the concentration data of the first biomarker combination at the optimized sampling time point set; The second acquisition module is used to acquire the concentration data of the second biomarker and the concentration data of the third biomarker for in vitro samples, and to construct an organ damage proxy model. The residual component determination module is used to perform hierarchical regression decomposition on the third biomarker concentration data based on the second biomarker concentration data and the organ damage proxy model to obtain the residual components of the third biomarker that cannot be explained by the organ damage proxy. The merging module is used to merge the early inflammation-driven index with the residual components to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score. The risk classification module is used to classify the risk level based on the inflammation-driven index score and the target component.

[0016] On the other hand, an electronic device is provided, comprising: a processor; and a memory storing computer-readable instructions, which, when executed by the processor, implement any of the methods described above for evaluating the combined detection results of biomarkers for early warning of heatstroke.

[0017] On the other hand, a computer-readable storage medium is provided, wherein at least one instruction is stored therein, the at least one instruction being loaded and executed by a processor to implement any of the above-described methods for evaluating the combined detection results of biomarkers for early warning of heatstroke.

[0018] The beneficial effects of the technical solutions provided in the embodiments of the present invention include at least the following: The biomarker combination detection result evaluation scheme for early warning of heatstroke provided in this invention, firstly, by dynamically optimizing sampling time points and based on a strategy of evaluating biomarker concentration change rate and information content, ensures that data with maximum information content is collected within different time windows, reducing data redundancy or omissions that may occur when manually selecting sampling time points. This enhances the correlation between biomarker concentration changes and clinical symptoms, providing more accurate and timely data support for early heatstroke data analysis, thereby improving the reliability of early warning results. Secondly, by establishing an early warning-driven index, the relationship between the concentration changes of the first biomarker combination and the physiological response to heatstroke was accurately quantified. This effectively integrates the changing trends of multiple indicators, avoiding the false positive or false negative problems that may exist with single biomarkers. Furthermore, it can identify potential physiological abnormalities in the early stages of heatstroke through the index, reducing the lag in relying on traditional biomarkers for judgment. This improves the accuracy and sensitivity of early warning for heatstroke and provides accurate and timely decision support for clinical practice. Attached Figure Description

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 A flowchart illustrating the steps of a method for evaluating the results of a combination of biomarkers for early warning of heatstroke, as provided in an embodiment of this application. Figure 2 A flowchart illustrating the sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples provided in this application embodiment. Figure 3 This application provides a reference graph showing the time-varying concentration changes of different biological samples at multiple time points, based on the embodiments of this application. Figure 4 is a smoothed time curve of the concentration change of different biomarker samples provided in the embodiments of this application; Figure 5 A flowchart of a method for optimizing an inflammation-driven index scoring model using machine learning, provided in an embodiment of this application; Figure 6 This is a block diagram of a biomarker combination detection result evaluation device for early warning of heatstroke provided in an embodiment of this application; Figure 7This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0021] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0022] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0023] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0024] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0025] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0026] This invention provides a method for evaluating the results of combined biomarker detection for early warning and diagnosis of heatstroke. This method can be implemented using an electronic device, which can be a terminal or a server. Figure 1 The flowchart shown is for evaluating the results of a combination of biomarker detections for early warning and diagnosis of heatstroke. The processing flow of this method may include the following steps: Step 101: Based on the sampling time point optimization method of multi-time point monitoring data of in vitro biological samples, obtain the concentration data of the first biomarker combination at the optimized sampling time point set.

[0027] Among them, the optimized sampling time point set is the set of time points generated by the sampling time point optimization process.

[0028] The first biomarker combination is a combination of multiple biomarkers used to characterize inflammation and cell damage-related states; the second biomarker is at least one biomarker used to characterize organ damage-related states; and the third biomarker is at least one biomarker used to characterize inflammatory responses and immune activity.

[0029] In a preferred embodiment, the first biomarker combination includes CitH3, NE, and HSP90; the second biomarker includes at least one of ALT and AST; and the third biomarker includes IL-6.

[0030] In one optional embodiment, the sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples, in order to obtain the concentration data of the first biomarker combination at the optimized sampling time point set, may include the following sub-steps: Sub-step 1: Obtain the concentration data of the first biomarker combination at multiple candidate sampling time points of the same in vitro biological sample, and perform batch alignment and quality control on the first biomarker combination concentration data to obtain comparable time series data; Sub-step 2: Construct a smoothed time curve of the first biomarker combination based on time series data, calculate the rate of change of each candidate sampling time point, and calculate the information content index of each candidate sampling time point; Sub-step 3: Under the premise of satisfying the preset sampling constraints, select and output the optimized sampling time point set from the candidate sampling time point set based on the information content index.

[0031] In actual implementation, sampling constraints may include at least: upper limit constraint on the number of samplings, minimum time interval constraint between adjacent sampling time points, and phased coverage constraint; the phased coverage constraint is used to limit the set of sampling time points to cover at least two of the early, middle and late stages.

[0032] This method of optionally acquiring concentration data of the first biomarker combination at an optimized set of sampling time points yields more reliable concentration data.

[0033] Step 102: Calculate the early inflammation-driven index based on the concentration data of the first biomarker combination at the optimized sampling time point set.

[0034] In one alternative embodiment, the first biomarker combination is used to optimize the sampling time point set.

[0035] The method for calculating the early inflammation-driven index from the combined concentration data is as follows: After standardizing the concentration data of the first biomarker combination at the optimized sampling time point set, the data is input into a mapping function to obtain the early inflammation-driven index. The mapping function can be any one of a weighted summation function, a nonlinear function, or a piecewise function.

[0036] Step 103: Obtain the concentration data of the second and third biomarkers for in vitro samples and construct an organ damage surrogate model.

[0037] An organ damage surrogate model was used to perform hierarchical regression decomposition of changes in the concentration of a third biomarker. These changes were quantitatively characterized using third biomarker concentration data.

[0038] Step 104: Based on the concentration data of the second biomarker and the organ damage surrogate model, perform hierarchical regression decomposition on the concentration data of the third biomarker to obtain the residual components of the third biomarker that cannot be explained by the organ damage surrogate.

[0039] In its specific implementation, step 104 may include the following sub-steps: S1: The organ injury surrogate model constructs an injury surrogate quantity using the concentration data of the second biomarker as input, and uses the injury surrogate quantity as an explanatory variable to perform hierarchical regression decomposition on the concentration change represented by the concentration data of the third biomarker to obtain the regression fit value. More specifically, the implementation method is as follows: hierarchical regression decomposition includes: a first-level regression and a second-level regression; the first-level regression uses the damage proxy to explain the concentration change of the third biomarker to obtain the intermediate residual; the second-level regression uses the early inflammation driving index to explain the intermediate residual to obtain the contribution index of the early inflammation driving index to the change of the third biomarker residual.

[0040] S2: The difference between the observed value of the third biomarker and the regression fit value is determined as the residual component of the third biomarker that cannot be explained by organ damage proxy.

[0041] Step 105: Combine the early inflammation-driven index with the residual components to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score.

[0042] In an optional embodiment, the method for merging the early inflammation-driven index and the residual component to determine the target portion of the third biomarker being affected by the first biomarker and the inflammation-driven index score can be as follows: Gated fusion of the early inflammation-driven index and the residual component is performed to determine the target portion of the third biomarker being affected by the first biomarker and the inflammation-driven index score. The target portion information and the inflammation-driven index score are then output.

[0043] The gating conditions used in gating fusion are determined by the integrity of the sampling time point set coverage and / or the validity determination results of the second biomarker.

[0044] Step 106: Classify the risk level based on the inflammation driving index and the target component.

[0045] In practice, the inflammation-driven index score can be compared with the preset values ​​corresponding to each risk level to classify patients into the corresponding risk level. The determined risk level is used to assist in early warning of heatstroke.

[0046] The biomarker combination detection result evaluation method for early warning of heatstroke provided in this invention firstly optimizes the sampling time point by using a strategy based on the biomarker concentration change rate and information content evaluation. This ensures that data with the maximum information content is collected within different time windows, reduces data redundancy or omissions that may occur when manually selecting sampling time points, enhances the correlation between biomarker concentration changes and clinical symptoms, and provides more accurate and timely data support for early data analysis of heatstroke, thereby improving the reliability of early warning results.

[0047] Secondly, by establishing an early warning-driven index, the relationship between the concentration changes of the first biomarker combination and the physiological response to heatstroke was accurately quantified. This effectively integrates the changing trends of multiple indicators, avoiding the false positive or false negative problems that may exist with single biomarkers. Furthermore, it can identify potential physiological abnormalities in the early stages of heatstroke through the index, reducing the lag in relying on traditional biomarkers for judgment. This improves the accuracy and sensitivity of early warning for heatstroke and provides accurate and timely decision support for clinical practice.

[0048] The following example illustrates the evaluation method for combined biomarker detection results for early warning of heatstroke. This method includes the following steps: S1: Establish a comparative experiment to verify that the combination of biomarkers provided by this invention has a significant effect on early warning and diagnosis of heatstroke.

[0049] This specific embodiment uses the classic heat stress chamber method to establish a mouse model of heatstroke. Healthy adult mice, weighing between 20-25 grams, were selected and evenly distributed by sex to minimize the influence of sex on the experimental results. The mice were randomly divided into an experimental group (heatstroke model group) and a control group (normal temperature group). Control group mice were raised at normal ambient temperature, while experimental group mice were placed in a heat stress chamber for heat exposure. To simulate the conditions for heatstroke, the heat stress chamber was set at 40°C and 60% humidity for 2 hours. It is important to note that during the heat exposure process, the mice's body temperature and behavioral changes should be monitored regularly to ensure the validity of the experiment and the health status of the mice.

[0050] After the heatstroke model was established, plasma samples were collected. Samples were collected at different time points after the heat exposure ended, including 0 hours (before modeling), 1 hour, 3 hours, 6 hours, and 12 hours. Plasma samples were collected from experimental and control mice using the orbital blood sampling method and sterile blood collection needles, ensuring aseptic operation and mouse safety. The collected blood samples were immediately placed in test tubes containing anticoagulant, centrifuged to obtain plasma, and the separated plasma samples were immediately frozen for subsequent testing.

[0051] S2: After sample collection, biomarkers and traditional indicators were detected. Using commercially available enzyme-linked immunosorbent assay (ELISA) kits, strictly following the instructions, the concentrations of CitH3 (citrullinated histone H3), NE (neutrophil elastase), and HSP90 (heat shock protein) in plasma samples were measured, along with the concentrations of traditional indicators such as ALT, AST, and IL-6 as comparative data. Each sample was subjected to at least three replicate experiments to ensure the reliability of the results. The measured concentration data were uploaded to a computer via an ELISA reader for further analysis.

[0052] S3: Acquire and integrate the sample concentration data uploaded by the ELISA reader, and set the concentration at 0 hours after the end of heat exposure as the baseline.

[0053] The sample concentration data obtained at each time point in step S1 were averaged to eliminate individual differences among individual mice. Let y be the average concentration at time point t, n be the number of samples at that time point, and y be the average concentration at time point t. t,i It represents the concentration of the i-th sample at time point t. From the formula: (1)

[0054] The average concentration of the sample at each time point can be calculated.

[0055] Statistical software was used to plot the concentration change curves of CitH3, NE, and HSP90 at different time points. Line graphs were used to display the concentration change trends of different samples at each time point, and these trends were compared with the 0-hour baseline. One-way ANOVA was used, based on the formula: (2)

[0056] Calculate whether the concentration changes at different time points are significant. Here, F represents the one-way variance, and MS... between It is the component mean square, MS within This is the within-group mean square. A significance level of p < 0.01 was set, and a t-test was performed on the results. The statistical formula is as follows: (3)

[0057] in, It is the sample mean. It is the mean under the null hypothesis. It is the sample standard deviation. This is the sample size. Use this t-value to find the corresponding p-value for the t-distribution. If p < 0.01, it proves that the concentration change at different time points is statistically significant; if p ≥ 0.01, it indicates that the concentration change at that time point is not significant.

[0058] S4: Based on the historical sample concentration data uploaded by the ELISA reader, select the concentrations of CitH3, NE, and HSP90 detected at the 3-hour time point, and the concentrations of multi-organ damage indicators such as ALT, AST, and creatinine detected 24 hours later, and calculate the Pearson correlation coefficient. The formula is as follows: (4)

[0059] Where, x i and y i These are the values ​​of two biomarker variables (e.g., CitH3 and ALT). and These are the means of the two variables. With a significance level set at p < 0.001, a t-test was performed on the results. The statistical formula is as follows: (5)

[0060] in, It is the Pearson correlation coefficient. That is the sample size. Then, using degrees of freedom... The p-value is determined using a t-distribution of -2, obtained by looking up the corresponding t-statistical distribution table. If p < 0.001, it proves that the changes in biomarker concentration detected by this invention are strongly correlated with multi-organ injury indicators; if p ≥ 0.001, it indicates that the changes in biomarker concentration are not significantly related to multi-organ injury.

[0061] S5: Use statistical software to create visualizations of the correlation analysis, such as... Figure 3 As shown in Figures 4(a), 4(b), and 4(c), the concentration changes of CitH3, NE, HSP90, ALT, AST, and IL_6 at different time points are displayed. The concentration changes of different biomarkers at the same time point are also displayed, thus intuitively reflecting the changes of biomarker concentrations over time after the end of heat exposure.

[0062] Since the sampling time point is manually set, and the comparative experimental results above show that the concentration of the biomarker combination changes significantly in the short period after heat exposure, manually setting the sampling time point may cause some significant concentration changes to be missed, which could lead to inaccurate diagnosis in future clinical applications. Therefore, an optimization algorithm is proposed to optimize the sample sampling time point, such as... Figure 2 The diagram shown is a flowchart of a sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples provided in this application embodiment. The specific implementation includes the following steps: Step 201: Obtain the concentration of the first biomarker uploaded by the detection device.

[0063] Step 202: Construct curves and calculate the slope of the curves at each sampling time point.

[0064] Based on the data obtained in step S2, for the concentration of each biomarker in the first biomarker combination, a smooth time curve is fitted using a spline function, and f is used to calculate the corresponding curve. CitH3 (t), f NE (t), f HSP90 (t) represents the overall trend of changes in the concentration of each biomarker, which allows us to roughly divide the time period after the end of heat exposure into early (0-2h), middle (2-6h), and late (6-12h).

[0065] Step 203: Calculate the information content index.

[0066] Within the 0-12h interval, set a set of time points T={t1,...t}. k For each candidate time point t, define the information content for |k∈[4,6]}. : (6)

[0067] Among them, f' CitH3 (t), f' NE (t), f' HSP90 I(t) represents the derivatives of the smoothed time curves for CitH3, NE, and HSP90, respectively, indicating the rate of change of these three biomarkers' concentrations at time point t. ω1, ω2, and ω3 represent the weights of CitH3, NE, and HSP90 in the information content, typically in a 1:1:1 ratio. Note that to emphasize early inflammation, CitH3 and NE can be assigned slightly higher weights than HSP90. Selecting 4-6 time points per set ensures no data omissions or redundancy. Furthermore, phased coverage constraints should be considered; the selected time points should cover at least two of the early, middle, and late stages to ensure the final results clearly show the concentration change trend. A larger I(t) indicates a greater change in biomarker concentration at that time point t, resulting in a greater amount of information, and this time point is more likely to be considered the optimal sampling point.

[0068] Step 204: Based on the calculated information content index and the condition constraints, use a greedy algorithm to iteratively obtain the optimized sampling time points.

[0069] In addition, the following constraints should be considered when selecting the sampling time point for operational feasibility: Constraint 1: 0-hour data must be retained as a reference baseline to compare changes in biomarker concentrations in samples at subsequent time points.

[0070] Constraint 2: The time interval between two consecutive blood collections must be ≥0.5 hours. This ensures data independence and reduces physiological interference caused by short-term stress or other inflammatory responses, thus minimizing the impact on judgment. Furthermore, since both blood collection and sample processing require time in actual operation, the interval cannot be set too short, otherwise it may cause operational delays.

[0071] Constraint 3: The distribution of time points should cover at least two of the three time intervals (early, middle and late) of the entire experimental time interval (0-12h) to avoid insufficient analysis due to data gaps within the time interval.

[0072] By combining the formula and constraints, and using a simple greedy algorithm to iterate, we finally find the optimal set of data points with the largest total information, which are then used as the optimized sampling time points.

[0073] Biologically, heatstroke causes organ damage, leading to the production of IL-6, which in turn exacerbates organ damage. Therefore, in clinical diagnosis, it is easy to confuse changes in IL-6 concentration caused by heatstroke or organ damage. Based on this, this embodiment constructs an organ damage surrogate model. By calculating the Early Inflammatory Driver (EID) index and using hierarchical regression analysis to remove the portion of IL-6 concentration changes caused by organ damage, the accuracy of heatstroke early warning is optimized. It should be noted that this embodiment uses the EID value at 3 hours, the IL-6 concentration at 6 hours, and the ALT / AST concentration at 24 hours as examples. In actual implementation, technicians can select different time points as inputs according to the actual situation. Other steps are similar to this embodiment and will not be repeated here.

[0074] Based on the optimized sampling time points, a time point is selected, and the concentration at that time point will be used as the input for calculating the EID index. It is important to note that the early inflammation-driven index can be obtained from various mapping functions; the following section first introduces how to calculate it using a weighted function. The formula for calculating the early inflammation-driven index (EID) is as follows: (7)

[0075] Where Z(x) represents the standardized value of the concentration of each biomarker, and the formula is: (8) This represents the mean concentration of the biomarker at time t. CitH3 represents the standard deviation. 3h NE 3h HSP90 3h These represent the concentrations of CitH3, NE, and HSP90 in the sample at the 3-hour time point, respectively. Similarly, X here refers to the concentrations of CitH3, NE, and HSP90 in the sample at the 3-hour time point.

[0076] In formula (7), a1, a2, and a3 are regression coefficients to be determined, representing the contributions of CitH3, NE, and HSP90 to the EID value, respectively. They are generally determined through multiple regression analysis and require analysis based on the actual situation. In this embodiment, the coefficients can be obtained by using regression analysis based on the concentrations of CitH3, NE, and HSP90 at the 3-hour time point and the concentrations of ALT and AST at the 24-hour time point. This process is generally implemented using Python code.

[0077] In addition to using weighted functions to calculate EID values, piecewise functions can also be used.

[0078] The specific formula for implementing a piecewise function is as follows (for illustrative purposes only): ; The definitions and values ​​of Z and a1, a2, a3, b1, b2, b3, c1, c2, and c3 are the same as above, and will not be repeated here.

[0079] It should be noted that the above formulas and value ranges are for reference only. In actual operation, technicians can make further adjustments to the formulas and condition ranges according to the actual situation.

[0080] In the organ injury surrogate model, the larger the EID value, the stronger the "upstream stress drive" may exist in the heatstroke process, that is, the greater the possibility that the inflammation is caused by heatstroke, and the inflammation has not been fully amplified by organ injury.

[0081] For IL-6 concentration at 6 hours, stratified regression was used to distinguish between "initial inflammatory signals" and "organ damage amplification signals," consisting of a two-step regression: The first step is to establish an organ damage surrogate model, i.e., the first-layer regression, to obtain the residual components. This step constructs the damage surrogate quantity, the purpose of which is to extract the concentration "not caused by organ damage" from the currently known IL-6 concentrations. According to the formula: (9)

[0082] Substituting the historical data obtained in step S2, the residual component ε is calculated. A This refers to the portion of "IL-6" concentration changes that cannot be explained by organ damage.

[0083] The second step, based on the regression in the first step, involves incorporating early EID values ​​to calculate a second-level regression, obtaining the contribution of the early inflammation-driving index to changes in IL-6 concentration. According to the formula: (10)

[0084] Here, γ1' represents the contribution of the early inflammation-driven index to the explanation of IL-6 concentration. The significance of γ1' is determined by a t-test, and the specific steps are the same as in step S3 above. If γ1' is significant, it indicates that EID can still explain part of the changes in IL-6 concentration beyond the effects of organ damage.

[0085] β0 is the intercept term, representing the baseline IL-6 concentration when all independent variables (ALT, AST, and EID) are zero, and can be calculated from historical data; β1 and β2 represent ALT and EID, respectively. 24h and AST 24hThe effect on IL-6 concentration, specifically the expected change in IL-6 concentration for every one-unit increase in ALT and AST concentrations after 24 hours, was calculated using historical data through least squares regression; ALT 24h AST 24h These represent the concentrations of ALT and AST in the sample at 24h time points, respectively. The error term represents the part of formula (10) that cannot be explained using computer language in practical applications.

[0086] When the model is applied to real-world applications, the error term may include the following aspects: the influence of other biomarkers: In addition to the biomarkers highlighted in this invention, other biomarkers, such as C-reactive protein, play a role in organ damage; the influence of external environmental pollution: External environmental pollution may also lead to inflammation, which in turn causes changes in the concentration of IL-6 in the sample; experimental errors during sample collection: In the actual sampling process, if the equipment, experimental methods, and sample collection conditions are inconsistent, it may also lead to data deviations. Based on the above and more, it is necessary to add a residual component to the model to weaken the influence of other factors on the model, thereby improving the accuracy of the model.

[0087] Based on the above steps, a gated fusion mechanism is used, with the following two gate conditions: first, whether the sampling time points are fully covered, i.e., whether the necessary time point data are complete; and second, whether the concentration of the biomarker (IL-6 in this example) is valid. The EID index and residual index are then determined based on the gate conditions. Merging. If all gating conditions are met, a weighted summation method is used to sum the EID value and the residual exponent, and a fused value is output. This fused value represents the early inflammatory information unrelated to organ damage in the IL-6 concentration change, that is, the part of IL-6 affected by the concentrations of CitH3, NE, and HSP90. The fusion can be performed using the following formula: (11)

[0088] Where F represents the fused EID and ε A μ1 and μ2 are EID and ε, respectively. A The specific values ​​for the weights need to be further optimized through machine learning combined with historical data.

[0089] Based on the fused EID and ε A To establish an inflammation-driven index scoring model, you can refer to the following formula: (12)

[0090] Here, Score represents the inflammation-driven index score, w1, w2, w3, and w4 are the weights of each biomarker, obtained through training by a machine learning module, and IL-6, ALT, and AST represent the input of real-time detection concentration data in clinical diagnosis.

[0091] It should be noted that this risk level scoring model is for reference only. The weights need to be adjusted and optimized through machine learning training when actually building the model. Figure 5 The flowchart of the method for optimizing the inflammation-driven index scoring model through machine learning provided in the embodiments of this application is as follows: All collected data, including EID values, IL-6 concentration, ALT, and AST concentrations, are packaged into a dataset and used as the input set for the machine learning model. The health status dataset of the sampled subjects is used as the output set for training the machine learning model. During training, the model uses gradient descent or least squares to minimize the error by adjusting the weights (w1, w2, w3, and w4 in the above formula). After training, the model's performance is evaluated. If the model performs well, it indicates that the weights are optimal and can be used in practice; if the model exhibits significant errors, further training is required.

[0092] Ultimately, in practical applications, the trained model outputs an inflammation-driven index score based on the input dataset. Then, by pre-setting a score threshold, it determines the risk of disease. For example, when the pre-set score threshold is 50: a score ≤ 50 indicates low risk; a score > 50 indicates high risk. High risk indicates potentially more severe inflammatory responses and organ damage, requiring more urgent treatment; low risk indicates milder inflammatory responses and organ damage, with a more stable condition.

[0093] Figure 6 This is a block diagram of a biomarker combination detection result evaluation device for early warning of heatstroke provided by an embodiment of the present invention. The device is used to perform a biomarker combination detection result evaluation method for early warning of heatstroke.

[0094] Reference Figure 6 The device includes: a first acquisition module 601, a calculation module 602, a second acquisition module 603, a residual component determination module 604, a merging module 605, and a level classification module 606. The first acquisition module 601 is used to acquire the concentration data of the first biomarker combination on the optimized sampling time point set based on the sampling time point optimization method of multi-time point monitoring data of in vitro biological samples, wherein the optimized sampling time point set is a time point set generated by the sampling time point optimization process. Calculation module 602 is used to calculate the early inflammation-driven index based on the concentration data of the first biomarker combination at the optimized sampling time point set; The second acquisition module 603 is used to acquire the concentration data of the second biomarker and the concentration data of the third biomarker for in vitro samples, and to construct an organ damage proxy model. The residual component determination module 604 is used to perform hierarchical regression decomposition on the third biomarker concentration data based on the second biomarker concentration data and the organ damage proxy model to obtain the residual components in the third biomarker that cannot be explained by the organ damage proxy. The merging module 605 is used to merge the early inflammation-driven index with the residual component to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score. The risk classification module 606 is used to classify the risk level based on the inflammation-driven index score and the target portion.

[0095] In an optional embodiment, the first biomarker combination is a combination of multiple biomarkers for characterizing inflammation and cell damage-related states; the second biomarker is at least one biomarker for characterizing organ damage-related states; and the third biomarker is at least one biomarker for characterizing inflammatory responses and immune activity.

[0096] In an optional embodiment, the first biomarker combination includes CitH3, NE, and HSP90;

[0097] The second biomarker includes at least one of ALT and AST; The third biomarker includes IL-6.

[0098] In an optional embodiment, the first acquisition module is specifically used for: The concentration data of the first biomarker combination at multiple candidate sampling time points of the same in vitro biological sample are obtained, and batch alignment and quality control are performed on the first biomarker combination concentration data to obtain comparable time series data. Based on the time series data, a smoothed time curve of the first biomarker combination is constructed, the rate of change of each candidate sampling time point is calculated, and the information content index of each candidate sampling time point is calculated. Under the premise of meeting the preset sampling constraints, the optimized sampling time point set is selected from the candidate sampling time point set and output according to the information content index.

[0099] In one optional embodiment, the sampling constraints include at least an upper limit constraint on the number of samplings, a minimum time interval constraint between adjacent sampling time points, and a phased coverage constraint. The phased coverage constraint is used to limit the set of sampling time points to cover at least two of the early, middle and late phases.

[0100] In one alternative embodiment, the computing module is specifically used for: The concentration data of the first biomarker combined on the optimized sampling time point set are standardized and then input into a mapping function to obtain the early inflammation driving index, wherein the mapping function is any one of a weighted summation function, a nonlinear function, or a piecewise function.

[0101] In one optional embodiment, the residual component determination module is specifically used for: The organ damage proxy model is invoked to construct a damage proxy quantity using the concentration data of the second biomarker as input, and the damage proxy quantity is used as an explanatory variable to perform hierarchical regression decomposition on the concentration change represented by the concentration data of the third biomarker to obtain the regression fit value. The difference between the observed value of the third biomarker and the regression fitted value is determined as the residual component of the third biomarker that cannot be explained by organ damage proxy.

[0102] In one optional embodiment, the hierarchical regression decomposition includes a first-level regression and a second-level regression; The first layer regression was used to explain the concentration changes of the third biomarker using the damage proxy to obtain intermediate residuals; The second-level regression interprets the intermediate residuals using the early inflammation-driven index to obtain the contribution index of the early inflammation-driven index to the changes in the third biomarker residuals.

[0103] In an optional embodiment, the merging module is specifically used to: perform gated fusion of the early inflammation-driving index and the residual component to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driving index score, wherein the gating conditions used during the gating fusion are determined by the coverage integrity of the sampling time point set and / or the validity determination result of the second biomarker.

[0104] The biomarker combination detection result evaluation device for early warning of heatstroke provided in this invention firstly optimizes the sampling time point by using a strategy based on the rate of change of biomarker concentration and information content evaluation. This ensures that data with the maximum information content is collected within different time windows, reduces data redundancy or omissions that may occur when manually selecting sampling time points, enhances the correlation between changes in biomarker concentration and clinical symptoms, and provides more accurate and timely data support for early data analysis of heatstroke, thereby improving the reliability of early warning results.

[0105] Secondly, by establishing an early warning-driven index, the relationship between the concentration changes of the first biomarker combination and the physiological response to heatstroke was accurately quantified. This effectively integrates the changing trends of multiple indicators, avoiding the false positive or false negative problems that may exist with single biomarkers. Furthermore, it can identify potential physiological abnormalities in the early stages of heatstroke through the index, reducing the lag in relying on traditional biomarkers for judgment. This improves the accuracy and sensitivity of early warning for heatstroke and provides accurate and timely decision support for clinical practice.

[0106] This invention also provides a biomarker combination detection result evaluation system for early warning of heatstroke, including: an optimized sampling time optimization module, used to acquire concentration data of CitH3, NE, and HSP90 at an optimized sampling time point set, wherein the optimized sampling time point set is a set of time points generated by the sampling time point optimization process.

[0107] The organ damage surrogate modeling and residual decomposition module is used to acquire ALT, AST and IL-6 concentration data for in vitro samples, construct organ damage surrogate models, and perform hierarchical regression decomposition on changes in IL-6 concentration.

[0108] The inflammation-driven index score output module is used to merge the early inflammation-driven index with the residual components, output the portion of IL-6 affected by CitH3, NE, and HSP90, and output the inflammation-driven index score as an auxiliary means for inflammation early warning.

[0109] Figure 7 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. The electronic device includes... Figure 6 The illustrated device is a biomarker combination detection result evaluation apparatus for early warning of heatstroke. Optionally, the SSS device may include a first processor 701.

[0110] Optionally, the electronic device may also include a memory 702 and a transceiver 703.

[0111] The first processor 701, memory 2002, and transceiver 2003 can be connected via a communication bus.

[0112] The following is combined Figure 7 A detailed introduction to each component of the electronic device: The first processor 701 is the control center of the electronic device and can be a single processor or a collective term for multiple processing elements. For example, the first processor 701 can be one or more central processing units (CPUs), application-specific integrated circuits (ASICs), or one or more integrated circuits configured to implement embodiments of the present invention, such as one or more digital signal processors (DSPs), or one or more field-programmable gate arrays (FPGAs).

[0113] Optionally, the first processor 701 can perform various functions of the electronic device by running or executing software programs stored in the memory 2002 and calling data stored in the memory 702.

[0114] In a specific implementation, as one example, the first processor 2001 may include one or more CPUs, such as CPU0 and CPU1 shown in Figure 1.

[0115] In a specific implementation, as one example, the electronic device may also include multiple processors, for example... Figure 7 The first processor 701 and the second processor 704 are shown. Each of these processors can be a single-core processor (single-CPU) or a multi-core processor (multi-CPU). Here, a processor can refer to one or more devices, circuits, and / or processing cores used to process data (e.g., computer program instructions).

[0116] The memory 702 is used to store the software program that executes the solution of the present invention, and is controlled by the first processor 701 to execute it. The specific implementation method can be referred to the above method embodiment, and will not be repeated here.

[0117] Optionally, the memory 702 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, or electrically erasable programmable read-only memory (EEPROM), compact disc read-only memory (CD-ROM) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto. The memory 702 may be integrated with the first processor 701 or may exist independently and be accessed through the interface circuit of the electronic device (…). Figure 7 (Not shown in the image) is coupled to the first processor 701, and this embodiment of the invention does not specifically limit this.

[0118] The transceiver 703 is used to communicate with network devices or with terminal devices.

[0119] Optionally, transceiver 703 may include a receiver and a transmitter (not shown separately in Figure @). The receiver is used to implement the receiving function, and the transmitter is used to implement the transmitting function.

[0120] Optionally, the transceiver 703 can be integrated with the first processor 701, or it can exist independently and be connected via the interface circuit of the electronic device. Figure 7 (Not shown in the image) is coupled to the first processor 701, and this embodiment of the invention does not specifically limit this.

[0121] It should be noted that, Figure 7 The structure of the electronic device shown does not constitute a limitation on the router. The actual electronic device may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0122] Furthermore, the technical effectiveness of the electronic device can be referenced from the technical effectiveness of the biomarker combination detection result evaluation method for early warning of heatstroke described in the above method embodiments, and will not be repeated here.

[0123] It should be understood that the first processor 701 in this embodiment of the invention may be a central processing unit (CPU), or it may be other general-purpose processors, digital signal processors (DSPs), application-specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor, or it may be any conventional processor.

[0124] It should also be understood that the memory in the embodiments of the present invention can be volatile memory or non-volatile memory, or may include both volatile and non-volatile memory. The non-volatile memory can be read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. The volatile memory can be random access memory (RAM), which is used as an external cache. By way of example, but not limitation, many forms of random access memory (RAM) are available, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate synchronous DRAM (DDR SDRAM), enhanced synchronous DRAM (ESDRAM), synchronous linked DRAM (SLDRAM), and direct rambus RAM (DR RAM).

[0125] The above embodiments can be implemented, in whole or in part, by software, hardware (such as circuits), firmware, or any other combination thereof. When implemented using software, the above embodiments can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, all or part of the processes or functions described in the embodiments of the present invention are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more sets of available media. The available medium can be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium. A semiconductor medium can be a solid-state drive.

[0126] It should be understood that the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. A and B can be singular or plural. Additionally, the character " / " in this article generally indicates an "or" relationship between the preceding and following related objects, but it can also represent an "and / or" relationship. Please refer to the context for a more accurate understanding.

[0127] In this invention, "at least one" means one or more, and "more than one" means two or more. "At least one of the following" or similar expressions refer to any combination of these items, including any combination of a single item or a plurality of items. For example, at least one of a, b, or c can represent: a, b, c, ab, ac, bc, or abc, where a, b, and c can be a single item or multiple items.

[0128] It should be understood that, in various embodiments of the present invention, the order of the above-mentioned process numbers does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0129] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0130] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the devices, apparatuses, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0131] In the several embodiments provided by this invention, it should be understood that the disclosed devices, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical, mechanical, or other forms.

[0132] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0133] In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0134] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0135] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for evaluating the results of combined biomarker detection for early warning of heatstroke, characterized in that, The method includes: A sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples is used to obtain the concentration data of the first biomarker combination at the optimized sampling time point set, wherein the optimized sampling time point set is a time point set generated by the sampling time point optimization process. The early inflammation-driving index was calculated based on the concentration data of the first biomarker combination at the optimized sampling time point set. Obtain the concentration data of the second and third biomarkers for in vitro samples, and construct an organ damage surrogate model; Based on the concentration data of the second biomarker and the organ damage proxy model, the concentration data of the third biomarker is subjected to hierarchical regression decomposition to obtain the residual components of the third biomarker that cannot be explained by the organ damage proxy. The early inflammation-driven index is combined with the residual component to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score; Risk levels are determined based on the inflammation-driven index score and the target component.

2. The method for evaluating the results of combined biomarker detection for early warning of heatstroke as described in claim 1, characterized in that: The first biomarker combination is a combination of multiple biomarkers used to characterize inflammation and cell damage-related states; the second biomarker is at least one biomarker used to characterize organ damage-related states; and the third biomarker is at least one biomarker used to characterize inflammatory responses and immune activity.

3. The method for evaluating the results of combined biomarker detection for early warning of heatstroke as described in claim 1 or 2, characterized in that: The first biomarker combination includes CitH3, NE, and HSP90; The second biomarker includes at least one of ALT and AST; The third biomarker includes IL-6.

4. The method for evaluating the results of combined biomarker detection for early warning of heatstroke as described in claim 1, characterized in that, The sampling time point optimization method based on multi-time point monitoring data of in vitro biological samples includes the following steps for obtaining the concentration data of the first biomarker combination at the optimized sampling time point set: The concentration data of the first biomarker combination at multiple candidate sampling time points of the same in vitro biological sample are obtained, and batch alignment and quality control are performed on the first biomarker combination concentration data to obtain comparable time series data. Based on the time series data, a smoothed time curve of the first biomarker combination is constructed, the rate of change of each candidate sampling time point is calculated, and the information content index of each candidate sampling time point is calculated. Under the premise of meeting the preset sampling constraints, the optimized sampling time point set is selected from the candidate sampling time point set and output according to the information content index.

5. The method as described in claim 4, characterized in that: The sampling constraints include at least the upper limit constraint on the number of sampling times, the minimum time interval constraint between adjacent sampling time points, and the phased coverage constraint. The phased coverage constraint is used to limit the set of sampling time points to cover at least two of the early, middle and late phases.

6. The method for evaluating the results of combined biomarker detection for early warning of heatstroke as described in claim 1, characterized in that, The steps for calculating the early inflammation-driving index based on the concentration data of the first biomarker combination at the optimized sampling time point set include: The concentration data of the first biomarker combined on the optimized sampling time point set are standardized and then input into a mapping function to obtain the early inflammation driving index, wherein the mapping function is any one of a weighted summation function, a nonlinear function, or a piecewise function.

7. The method for evaluating the results of combined biomarker detection for early warning of heatstroke as described in claim 1, characterized in that, The step of performing hierarchical regression decomposition on the third biomarker concentration data based on the second biomarker concentration data and the organ damage surrogate model to obtain the residual components of the third biomarker that cannot be explained by the organ damage surrogate model includes: The organ damage surrogate model constructs a damage surrogate quantity using the concentration data of the second biomarker as input, and uses the damage surrogate quantity as an explanatory variable to perform hierarchical regression decomposition on the concentration change represented by the concentration data of the third biomarker to obtain the regression fit value. The difference between the observed value of the third biomarker and the regression fitted value is determined as the residual component of the third biomarker that cannot be explained by organ damage proxy.

8. The method as described in claim 7, characterized in that: The hierarchical regression decomposition includes a first-level regression and a second-level regression; The first layer regression was used to explain the concentration changes of the third biomarker using the damage proxy to obtain intermediate residuals; The second-level regression interprets the intermediate residuals using the early inflammation-driven index to obtain the contribution index of the early inflammation-driven index to the changes in the third biomarker residuals.

9. The method for evaluating the results of combined biomarker detection for early warning of heatstroke as described in claim 1, characterized in that, The step of combining the early inflammation-driven index with the residual component to determine the target portion of the third biomarker affected by the first biomarker includes: The early inflammation-driving index and residual components are gated and fused to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driving index score. The gating conditions used in the gating fusion are determined by the coverage integrity of the sampling time point set and / or the validity determination result of the second biomarker.

10. A device for evaluating the results of a combination of biomarkers for early warning of heatstroke, characterized in that, The device includes: The first acquisition module is used to acquire the concentration data of the first biomarker combination on the optimized sampling time point set based on the sampling time point optimization method of multi-time point monitoring data of in vitro biological samples, wherein the optimized sampling time point set is a time point set generated by the sampling time point optimization process. The calculation module is used to calculate the early inflammation-driven index based on the concentration data of the first biomarker combination at the optimized sampling time point set; The second acquisition module is used to acquire the concentration data of the second biomarker and the concentration data of the third biomarker for in vitro samples, and to construct an organ damage proxy model. The residual component determination module is used to perform hierarchical regression decomposition on the third biomarker concentration data based on the second biomarker concentration data and the organ damage proxy model to obtain the residual components of the third biomarker that cannot be explained by the organ damage proxy. The merging module is used to merge the early inflammation-driven index with the residual components to determine the target portion of the third biomarker affected by the first biomarker and the inflammation-driven index score. The risk classification module is used to classify the risk level based on the inflammation-driven index score and the target component.

Images