Death risk prediction model for severe burn and abnormal liver enzyme, construction method and storage medium

Abstract

The invention relates to a death risk prediction model for severe burns and abnormal liver enzymes, a construction method and a storage medium. The construction method of the death risk prediction model comprises the following steps: collecting clinical feature data of patients with severe burns and abnormal liver enzymes; using an RF-RFE combined regression model to analyze and screen the clinical feature data of the training set and whether death occurs, and screening out clinical feature data related to serious burn and liver enzyme abnormality and death; and constructing a death risk prediction model based on the screened clinical feature data. The invention also provides a death risk prediction model constructed by the construction method. The invention further provides a storage medium. According to the method, risk factors causing death of patients suffering from serious burns and liver enzyme abnormality are identified based on clinical basic data and disease related indexes of the patients, so that the death probability of the patients is conveniently, quickly and accurately predicted in advance, and doctors are helped to take corresponding prevention and treatment measures for the patients.

Description

Technical Field

[0001] The present invention relates to the field of medical diagnosis technology, and in particular to a death risk prediction model for severe burns combined with abnormal liver enzymes, a construction method and a storage medium. Background Art

[0002] Burns are a very common type of trauma in daily life, which poses a great threat to human life and health. Burns may not only cause long-term physical and mental problems, but also a large number of patients lose their precious lives due to burns every year around the world. The classification of burns mainly includes the depth of the burn and the area of ​​the burn. The depth of burns ranges from first to fourth degree, while the burn area is divided into mild and severe. Mild burns are characterized by a burn area less than 10% of the total body surface area (TBSA), mainly manifested as surface burns. In general, patients with a total surface area burn of >20% or full-thickness burns of >10% are defined as severe burns.

[0003] Although advances in resuscitation techniques, antibiotic therapy, surgery, and nutritional support have reduced the mortality rate of patients with severe burns in recent years, their mortality rate remains high compared to other critically ill patients. One of the reasons for this phenomenon may be multiple organ dysfunction syndrome (MODS) caused by organ damage, especially in patients with a total burn area (TBSA) of 20% or more, the incidence of MODS can be as high as 63.4%. In MODS, liver damage is extremely common, but research on the liver's response to burns is relatively limited.

[0004] After burns, a variety of factors may cause liver damage, including hypoperfusion caused by hypotension, hepatocyte death caused by proinflammatory cytokines, liver edema and fat changes. Studies have shown that the mechanism of liver damage may be related to ischemic hepatitis caused by early effective blood volume reduction after burns, as well as hepatocyte apoptosis and abnormal liver enzymes caused by ischemia-reperfusion injury. Similarly, there is evidence that liver-related factors affect the prognosis of burn patients, because persistent disorders of liver enzymes predict sepsis and poor prognosis in burn wounds. Common serum liver enzymes include aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP) and γ-glutamyl transferase (GGT). Among them, AST and ALT are the most common liver enzymes, and liver damage is usually diagnosed by measuring their levels. The content of liver enzymes reflects the degree and type of liver damage.

[0005] The occurrence of liver dysfunction after burns has been reported since the late 1930s. In the 1980s, a series of autopsies revealed an increase in liver volume after burns. Subsequent studies have shown that the immediate increase in liver volume after burns is due to the formation of edema. The increase in edema may cause cell damage and release liver enzymes, which return to normal levels during liver regeneration, thus affecting prognosis. The traumatic response caused by severe burns is both general and specific, and is crucial for the epidemiological analysis and mortality risk assessment of patients with abnormal liver enzymes. However, there is no existing technology to predict whether patients with severe burns and abnormal liver enzymes will die in the future, resulting in the inability to prevent and treat them in a timely manner. Therefore, timely and accurate identification of risk factors for death is crucial to assessing the patient's clinical prognosis and guiding individualized treatment. Summary of the invention

[0006] In view of this, the purpose of the present invention is to provide a death risk prediction model, construction method and storage medium for severe burns combined with abnormal liver enzymes, so as to identify the risk factors leading to death of patients with severe burns combined with abnormal liver enzymes based on the patient's clinical basic data and disease-related indicators, and use these factors to establish a fully verified death risk prediction model to predict the patient's probability of death in advance, conveniently, quickly and accurately, thereby helping doctors take corresponding preventive and treatment measures for patients.

[0007] In order to achieve the above object, the technical solution adopted by the present invention is as follows:

[0008] A method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes, characterized by comprising the following steps:

[0009] S1. Data collection and preprocessing: Clinical characteristics data of patients with severe burns and abnormal liver enzymes were collected and preprocessed;

[0010] S2. Feature data screening: The preprocessed clinical feature data were divided into a training set and a validation set. The RF-RFE combined with LASSO regression model was used to analyze and screen the clinical feature data of the training set with a p value less than 0.05 and whether the patient died, and the clinical feature data related to severe burns combined with abnormal liver enzymes and death were screened out;

[0011] S3. Model construction: Based on the screened clinical characteristic data related to death, a mortality risk prediction model for severe burns combined with abnormal liver enzymes was constructed.

[0012] Among them, abnormal liver enzymes were defined as at least one of the indicators of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP) and γ-glutamyl transferase (GGT) being higher or lower than the normal value.

[0013] Preferably, the clinical characteristic data include: age, gender, admission date, injury date, BMI, cause of burns, burn area, burn location, inhalation injury, length of hospital stay, admission ALT, admission AST, liver enzyme abnormality, duration, death and length of hospital stay;

[0014] Preferably, the clinical characteristic data related to death of severe burns combined with abnormal liver enzymes include: age, total burn area (TBSA) and direct bilirubin value at admission.

[0015] Preferably, the mortality risk prediction model is a nomogram.

[0016] Preferably, in the nomogram:

[0017] The first column is the score scale (b(Xm)terms), with a score range of -2 to 9; the second column is the direct bilirubin on admission (DBIL), ranging from 0 to 280; the third column is the total burn area (TBSA), ranging from 20 to 100; the fourth column is the patient's age (age), ranging from 10 to 90; the fifth column is the total score, ranging from -4 to 10; the sixth column is the probability of death risk, ranging from 0.01 to 0.998.

[0018] In the nomogram, the second to fourth rows are related factors, and different situations of different factors correspond to different scores on the scale; the total score of the fifth row is the sum of the scores of each factor, and there is a corresponding relationship between the sixth row and the total score of the fifth row. According to the different scores, they are projected to the corresponding positions, which are the corresponding mortality risk probabilities of severe burns combined with abnormal liver enzymes.

[0019] Preferably, the average AUROC value of the nomogram model is 0.806.

[0020] Preferably, the death risk prediction model also includes a death risk probability calculation formula:

[0021]

[0022] In formula (I), X1 represents age; X2 represents total burn area (TBSA); X3 represents direct bilirubin (DBIL); and P represents the mortality risk probability of severe burns combined with hyperbilirubinemia.

[0023] Preferably, the p-value is obtained by a t-test or a chi-square test.

[0024] Preferably, it also includes:

[0025] The clinical characteristic data are randomly divided into a 70% training set and a 30% validation set; the model is trained by the training data set to obtain a death risk prediction model;

[0026] The death risk prediction model was validated using the validation set.

[0027] The present invention also provides a mortality risk prediction model for severe burns combined with abnormal liver enzymes obtained by the construction method of the present invention.

[0028] Preferably, the mortality risk prediction model for severe burns combined with abnormal liver enzymes includes:

[0029] The data collection and preprocessing module is used to collect clinical characteristic data of patients who died and survived from severe burns with abnormal liver enzymes, and to preprocess the clinical characteristic data;

[0030] The feature data screening module is used to divide the preprocessed clinical feature data into a training set and a validation set, and use the RF-RFE combined with the LASSO regression model to analyze and screen the clinical feature data of the training set with a p value less than 0.05 and whether or not the patient died, and screen out the clinical feature data related to death in patients with severe burns combined with abnormal liver enzymes;

[0031] The prediction model building module is used to construct a death risk prediction model for severe burns combined with abnormal liver enzymes based on the screened clinical feature data related to death.

[0032] The present invention also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps of the construction method described in the present invention are implemented.

[0033] Beneficial effects of the present invention:

[0034] The method for constructing a death risk prediction model for severe burns with abnormal liver enzymes of the present invention ensures that the basic data for constructing the prediction model are comprehensive and representative by collecting clinical characteristic data of patients with severe burns with abnormal liver enzymes; in the data preprocessing stage, the collected clinical characteristic data are cleaned and standardized to improve the data quality and provide accurate and reliable data support for subsequent analysis; the RF-RFE combined with the LASSO regression model is used for feature screening, which can effectively process high-dimensional data and screen out clinical characteristics with a high correlation with death, thereby improving the prediction accuracy of the model; the model construction is not only based on the clinical manifestations of the clinical characteristic data, but also combines experimental detection indicators, so that the prediction model is more comprehensive and can comprehensively reflect the severity of the patient's condition and the risk of death; the death risk prediction model constructed based on the screened indicators can provide clinicians with an accurate and practical tool, realize result visualization, help evaluate the patient's risk of death, and thus formulate a more personalized treatment plan, and has promotion and application value in the field of medical diagnosis technology. BRIEF DESCRIPTION OF THE DRAWINGS

[0035] Figure 1Flow chart of data collection and preprocessing for clinical characteristics of patients with severe burns and abnormal liver enzymes;

[0036] Figure 2 RF-RFE combined with LASSO is used to screen risk predictors of death, where 2(a) is feature variable screening based on RF-RFE; 2(b) is feature variable screening based on LASSO regression (lambda: 1SE); 2(c) is RF-RFE combined with LASSO; LASSO: least absolute shrinkage and selection operator; RFE: recursive feature elimination RF: random forest;

[0037] Figure 3 A nomogram model for predicting mortality risk in patients with severe burns and abnormal liver enzymes;

[0038] Figure 4 This is the receiver operating characteristic curve of the validation cohort for the logistic regression model;

[0039] Figure 5 This is a graph showing the results of the calibration curve analysis of the validation set cohort of the nomogram model;

[0040] Figure 6 This is a graph showing the results of the calibration curve analysis of the validation set cohort of the nomogram model;

[0041] Figure 7 Receiver operating characteristic curve for the training set cohort of the logistic regression model;

[0042] Figure 8 This is a graph showing the results of the calibration curve analysis of the training set cohort of the nomogram model;

[0043] Fig. 9 This is a graph showing the calibration curve analysis results for the training set cohort of the nomogram model. DETAILED DESCRIPTION

[0044] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.

[0045] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention, and thus the drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.

[0046] The present invention intends to disclose a death risk prediction model, construction method and storage medium for severe burns combined with abnormal liver enzymes, so as to identify the risk factors leading to death of patients with severe burns combined with abnormal liver enzymes based on the patients' clinical basic data and disease-related indicators, and use these factors to establish a fully verified death risk prediction model to predict the patient's probability of death in advance, conveniently, quickly and accurately, thereby helping doctors take corresponding preventive and treatment measures for patients.

[0047] Among them, the method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes includes the following steps:

[0048] S1. Data collection: Clinical characteristic data of patients with severe burns and abnormal liver enzymes who died and survived were collected, and the clinical characteristic data were preprocessed to remove patients with incomplete clinical characteristic data;

[0049] S2. Feature data screening: The preprocessed clinical feature data were divided into a training set and a validation set. The clinical feature data of the training set with a p-value of less than 0.05 in the t-test or chi-square test and whether the patient died were analyzed and screened using the RF-RFE combined with LASSO regression model. The clinical feature data that met the screening results of both RF-RFE and LASSO regression were screened out to obtain the clinical feature data related to death in patients with severe burns combined with abnormal liver enzymes.

[0050] S3. Model construction: Based on the screened clinical characteristic data related to death, a mortality risk prediction model for severe burns combined with abnormal liver enzymes was constructed.

[0051] Among them, abnormal liver enzymes were defined as at least one of the indicators of serum aspartate aminotransferase (AST), alanine aminotransferase (ALT), alkaline phosphatase (ALP) and γ-glutamyl transferase (GGT) being higher or lower than the normal value.

[0052] The diagnostic classification and evaluation of the injury pattern was based on the R value (R = (ALT / ULN) / (ALP / ULN)), with a compliance rate of 96%. R ≥ 5 indicates liver enzyme abnormalities of hepatocellular injury type; 2 < R < 2 indicates mixed liver enzyme abnormalities; R ≤ 2 indicates cholestatic liver enzyme abnormalities; ULN represents the normal value.

[0053] Among them, abnormal liver enzymes are defined as at least one of serum ALT, AST, ALP, GGT being higher or lower than normal. Mild liver enzyme elevation means at least one of serum ALT, AST, ALP, GGT is 1-5 times ULN, moderate liver enzyme elevation means at least one of serum ALT, AST, ALP, GGT is 5-15 times ULN, and severe liver enzyme elevation means at least one of serum ALT, AST, ALP, GGT is greater than 15 times ULN.

[0054] In some embodiments, clinical characteristic data include: age, gender, admission date, injury date, BMI, cause of burns, burn area, burn location, inhalation injury, length of hospital stay, admission ALT, admission AST, liver enzyme abnormality, duration, etc., death and hospital stay.

[0055] In some embodiments, clinical characteristic data associated with death in patients with severe burns combined with abnormal liver enzymes include: age, total burned area (TBSA), and admission direct bilirubin value.

[0056] In some embodiments, the mortality risk prediction model is a nomogram model.

[0057] Exemplarily, in S1, data collection and preprocessing include:

[0058] like Figure 1 As shown, the data of 655 patients who were continuously admitted to the Burn Department of Southwest Hospital from 2017 to 2022 were collected and preprocessed to remove the data of patients with missing information, remove the data of patients with TBSA ≤ 20%, remove the data of patients whose biochemical indicators (blood AST, ALT, bilirubin, etc.) were not tested within 72 hours after admission, remove the data of patients with extremely severe chest and abdomen or craniocerebral injuries; remove the data of patients who were hospitalized for less than 24 hours; remove the data of patients who were hospitalized more than 7 days after the burn, and remove the patients who were less than or equal to 18 years old, and 569 patients with severe burns were obtained. Among them, 479 patients had abnormal liver enzymes, and the 479 patients with abnormal liver enzymes were included in the risk prediction analysis of death during hospitalization in patients with severe burns and abnormal liver enzymes.

[0059] Exemplarily, S2, feature data screening, includes:

[0060] Patients with abnormal liver enzymes were further divided into two groups (survival group (N=394) and non-survival group (N=85). Variables that were consistent with RF-RFE and LASSO screening were selected. The analysis results showed that demographic factors included age, laboratory test indicators included admission direct bilirubin value, and burn severity factors included total burn area (TBSA).

[0061] Using logistics analysis, it was found that mixed injury (OR, 4.43; 95% CI, 2.70-7.27; P < 0.001), hepatocellular injury (OR, 49.75; 95% CI, 16.25-152.31; P < 0.001) and AST / ALT (OR, 1.61; 95% CI, 1.35-1.92; P < 0.001) were considered to be risk factors for the development of moderate to severe liver enzymes during hospitalization.

[0062] Among them, 479 patients had abnormal liver enzymes, and 85 of them died, with a mortality rate of 17.75%. As shown in Table 1, the age of the deceased patients was older than that of the surviving group (52.00 (43.00, 67.00) vs 46.00 (35.00, 53.00), p < 0.01). Furthermore, the severity of burns was significantly higher in patients who died, including total burn area (54.00 (33.00, 85.00) vs 37.00 (28.00, 52.00), p<0.001), Baux score (118.00 (101.00, 141.00) vs 90.00 (76.00, 108.00), p<0.001), and burn index (61.50 (34.00) vs 33.50 (22.50, 53.00), 92.00), p<0.001). It is worth noting that the AST of the deceased patients on admission was significantly higher (67.40 (47.90, 128.10) vs 37.00 (28.00, 52.00), p < 0.01), and the total bilirubin was significantly higher (25.07 (14.00, 43.67) vs 17.40 (12.51, 25.63), p < 0.001).

[0063] Table 1 Comparison of clinical characteristics between the survival group and the death group in patients with abnormal liver enzymes

[0064]

[0065]

[0066]

[0067] RF-RFE combined with LASSO is used to screen risk predictors of death, such as Figure 2 As shown in the figure, a total of three non-zero characteristic variables (including total burn area, age, and direct bilirubin) were screened as constructed predictive variables. SPSS27.0 software was used for regression analysis, nomogram construction, and evaluation.

[0068] Exemplarily, S3, model construction, includes: based on the above-screened clinical characteristic data related to death, based on the LASSO regression model, constructing a nomogram for predicting mortality, such as Figure 3 shown.

[0069] Among them, in the nomogram model:

[0070] The first column is the score scale (b(Xm)terms), with a score range of -2 to 9; the second column is the direct bilirubin on admission (DBIL), ranging from 0 to 280; the third column is the total burn area (TBSA), ranging from 20 to 100; the fourth column is the patient's age (age), ranging from 10 to 90; the fifth column is the total score, ranging from -2 to 3; the sixth column is the probability of death risk, ranging from 0.01 to 0.998.

[0071] In the nomogram, the second to fourth rows are related factors, and different situations of different factors correspond to different scores on the scale; the total score of the fifth row is the sum of the scores of each factor, and there is a corresponding relationship between the sixth row and the total score of the fifth row. According to the different scores, they are projected to the corresponding positions, which are the corresponding mortality risk probabilities of severe burns combined with abnormal liver enzymes.

[0072] The average AUROC value of the nomogram model was 0.806, the calibration slope was 1.245, and the Brier score was 0.131, indicating a good fit.

[0073] In some embodiments, the death risk prediction model further includes a death risk value formula:

[0074] The formula for calculating the probability of death risk is:

[0075]

[0076] In formula (I), X1 represents age; X2 represents total burn area (TBSA); X3 represents direct bilirubin (DBIL); and P represents the mortality risk probability of severe burns combined with hyperbilirubinemia.

[0077] Patients with abnormal liver enzymes were analyzed according to the type and degree of abnormal liver enzymes. Variables with more than 30% missing were deleted, and the remaining variables were all missing less than 4%, so mean interpolation was used. After interpolation, all clinical characteristic data of patients with abnormal liver enzymes were randomly divided into a training set (70%) and a validation set (30%). Logistic regression was used to fit the prediction model, and LASSO regression and RF-RFE were performed on the estimated training cohort to screen risk factors, fit the model, and construct a dynamic nomogram. Finally, the model was evaluated on the validation set. The model was evaluated using the area under the receiver operating characteristic curve (AUC), calibration curve, and decision curve analysis. The results are shown in the figure. Figures 4 to 7 Shown

[0078] in, Figure 4 This is the operating characteristic curve of the validation set cohort of the logistic regression model; Figure 5 Results of calibration curve analysis for the validation set cohort of the nomogram model; Figure 6 Results of calibration curve analysis for the validation set cohort of the nomogram model; Figure 7 The training set subject cohort operating characteristic curve for the logistic regression model; Figure 8 Results of calibration curve analysis for the training set cohort of the nomogram model; Fig. 9 This is the result of the calibration curve analysis of the training set cohort of the nomogram model. The decision curve analysis showed that the prediction model has good clinical benefits.

[0079] In some embodiments, a mortality risk prediction model for severe burns combined with abnormal liver enzymes obtained by the construction method in any of the above embodiments is also provided, comprising:

[0080] The data collection and preprocessing module is used to collect clinical characteristic data of patients who died and survived from severe burns with abnormal liver enzymes, and to preprocess the clinical characteristic data;

[0081] The feature data screening module is used to divide the preprocessed clinical feature data into a training set and a validation set, and use the RF-RFE combined with the LASSO regression model to analyze and screen the clinical feature data of the training set with a p value less than 0.05 and whether or not the patient died, and screen out the clinical feature data related to death in patients with severe burns combined with abnormal liver enzymes;

[0082] The prediction model building module is used to construct a death risk prediction model for severe burns combined with abnormal liver enzymes based on the screened clinical feature data related to death.

[0083] In some embodiments, a computer-readable storage medium is also provided, on which a computer program is stored. When the computer program is executed by a processor, the steps of the construction method in any of the above embodiments are implemented.

[0084] In summary, the nomogram model for predicting death due to severe burns combined with abnormal liver enzymes of the present invention can comprehensively evaluate the clinical characteristic indicators and laboratory data of patients, and reduce misjudgment caused by one-sided information. When a severely burned patient has abnormal liver enzymes, a nomogram model based on clinical characteristic indicators and experimental test data characteristics can be used to assist in predicting whether the severely burned patient with abnormal liver enzymes will die. The model of the present invention is simple, intuitive, and easy to promote and apply. It can effectively allow clinicians to conduct accurate individualized assessments of severely burned patients when they are receiving treatment, bringing better survival benefits to patients. It has important value for the effective application of prognosis and monitoring of severe burns combined with abnormal liver enzymes, and provides personalized and professional health management plans for different stratifications. It has promotion and application value in the field of medical diagnostic technology.

[0085] The above embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Any equivalent substitution or change made by a person skilled in the art based on the present invention is within the protection scope of the present invention.

Claims

1. A method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes, characterized in that: The following steps are involved: S1. Data collection and preprocessing: Clinical characteristics data of patients with severe burns and abnormal liver enzymes were collected and preprocessed; S2. Feature data screening: The preprocessed clinical feature data were divided into a training set and a validation set. The RF-RFE combined with LASSO regression model was used to analyze and screen the clinical feature data of the training set with a p value less than 0.05 and whether the patient died, and the clinical feature data related to severe burns combined with abnormal liver enzymes and death were screened out; S3. Model construction: Based on the screened clinical characteristic data related to death, a mortality risk prediction model for severe burns combined with abnormal liver enzymes was constructed.

2. The method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 1, characterized in that: The clinical characteristic data include: age, gender, admission date, injury date, BMI, burn etiology, burn area, burn location, inhalation injury, hospital stay, admission ALT, admission AST, admission bilirubin, liver enzyme abnormality, duration, death and hospital stay; And / or, clinical characteristic data related to death of severe burns combined with abnormal liver enzymes, including: age, total burn area (TBSA) and direct bilirubin value on admission.

3. The method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 2, characterized in that: The death risk prediction model is a nomogram model.

4. The method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 3, characterized in that: In the nomogram model: The first column is the score scale (b(Xm)terms), with a range of -2 to 9; the second column is the admission direct bilirubin (DBIL), with a range of 0 to 280; the third column is the total burn area (TBSA), with a range of 20 to 100; the fourth column is the patient's age (age), with a range of 10 to 90; the fifth column is the total score, with a range of -4 to 10; The sixth column is the probability of death risk, ranging from 0.01 to 0.

998.

5. The method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 4, characterized in that: The average AUROC value of the nomogram model was 0.

806.

6. The method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 3, characterized in that: The death risk prediction model also includes a death risk probability calculation formula: In formula (I), X1 represents age; X2 represents total burn area (TBSA); X3 represents direct bilirubin (DBIL); and P represents the mortality risk probability of severe burns combined with abnormal liver enzymes.

7. The method for constructing a mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 1, characterized in that: The p value is obtained by t test or chi-square test; And / or, the construction method further comprises: The clinical characteristic data were randomly divided into 70% training set and 30% validation set; The model is trained through the training data set to obtain the death risk prediction model; The death risk prediction model was validated using the validation set.

8. A mortality risk prediction model for severe burns combined with abnormal liver enzymes constructed by the construction method according to any one of claims 1 to 7.

9. The mortality risk prediction model for severe burns combined with abnormal liver enzymes according to claim 8, characterized in that: include: Data collection and preprocessing module, used to collect and preprocess clinical characteristic data of severe burns combined with abnormal liver enzymes; The feature data screening module is used to divide the preprocessed clinical feature data into a training set and a validation set, and use the RF-RFE combined with the LASSO regression model to analyze and screen the clinical feature data of the training set with a p value less than 0.05 and whether or not the patient died, and screen out the clinical feature data related to death in patients with severe burns combined with abnormal liver enzymes; The prediction model building module is used to construct a death risk prediction model for severe burns combined with abnormal liver enzymes based on the screened clinical feature data related to death.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the construction method according to any one of claims 1 to 7 are implemented.

Images