Biomarker for predicting prognosis condition of sepsis and application

Abstract

The invention discloses a biomarker for predicting the prognosis condition of sepsis and application, and belongs to the technical field of biology. The biomarkers comprise a ratio of neutrophils to lymphocytes, a ratio of platelets to lymphocytes, a ratio of monocytes to lymphocytes, a ratio of leukocytes to lymphocytes, a system inflammatory response index and a peripheral inflammatory value. According to the method, the ratio of neutrophil to lymphocyte, the ratio of platelet to lymphocyte, the ratio of mononuclear cell to lymphocyte, the ratio of leukocyte to lymphocyte, a systemic inflammatory response index and a peripheral inflammatory value are utilized to carry out accurate prognosis risk stratification on the sepsis patient, and prognosis survival risk prediction is carried out on the sepsis patient; the accuracy and timeliness of prognosis evaluation of sepsis patients are remarkably improved, a reliable guidance basis is provided for clinical decision making, and the method has important clinical application value and popularization significance.

Description

Technical Field

[0001] The present invention relates to the field of biotechnology, and particularly to a biomarker for predicting the prognosis of sepsis and its application. Background Art

[0002] Sepsis is one of the main causes of death in the intensive care unit, with a mortality rate as high as 25 - 30%. For patients with septic shock, the mortality rate can reach 40 - 50%. Despite significant progress in critical care medicine, two fundamental challenges still exist in the diagnosis and treatment of sepsis: the lack of an accurate early risk stratification tool and the insufficient understanding of the mechanisms of cellular dysfunction during the disease progression.

[0003] The currently widely used Sequential Organ Failure Assessment (SOFA) system in clinical practice has obvious limitations: it relies on the evaluation of complex organ dysfunction parameters, often leading to delays in critical treatment intervention. Although recent studies have shown that inflammatory indicators, especially the neutrophil - to - lymphocyte ratio (NLR) and the platelet - to - lymphocyte ratio (PLR), have potential prognostic value in various pathological states, the molecular mechanisms of these easily accessible clinical parameters have not been elucidated, limiting their application in guiding targeted treatment intervention.

[0004] In addition, current sepsis patient stratification methods either lack molecular mechanism explanations, require complex detection means, and are not suitable for acute clinical decision - making. Although cellular dysfunction is the basis for promoting the progression of sepsis, the molecular programs corresponding to the elevated inflammatory indicators have not been defined. The progression of sepsis involves complex molecular mechanism interactions such as neutrophil dysfunction, platelet activation, and metabolic reprogramming, but existing assessment methods are difficult to effectively integrate this information. Although single - cell technology has revealed unprecedented cellular heterogeneity in sepsis, how to translate these molecular insights into clinically available stratification tools and treatment targets still faces major challenges. Therefore, there is an urgent need for a method to bridge the gap between rapid patient stratification and mechanistic insights in order to achieve targeted intervention. Summary of the Invention

[0005] The purpose of the present invention is to provide a biomarker for predicting the prognosis of sepsis and its application to solve the problems existing in the above - mentioned prior art. The present invention uses the neutrophil - to - lymphocyte ratio, platelet - to - lymphocyte ratio, monocyte - to - lymphocyte ratio, white blood cell - to - lymphocyte ratio, systemic inflammatory response index, and peripheral inflammation value to accurately stratify the prognosis risk of sepsis patients and predict the prognosis survival risk of sepsis patients, significantly improving the accuracy and timeliness of the prognosis assessment of sepsis patients, providing a reliable guiding basis for clinical decision - making, and having important clinical application value and promotion significance.

[0006] To achieve the above object, the present invention provides the following solutions:

[0007] The present invention provides a biomarker for predicting the prognosis of sepsis, and the biomarker includes the neutrophil-to-lymphocyte ratio, the platelet-to-lymphocyte ratio, the monocyte-to-lymphocyte ratio, the white blood cell-to-lymphocyte ratio, the systemic inflammatory response index, and the peripheral inflammation value;

[0008] The calculation formula of the systemic inflammatory response index is: systemic inflammatory response index = neutrophil count × monocyte count / lymphocyte count;

[0009] The calculation formula of the peripheral inflammation value is: peripheral inflammation value = platelet count × neutrophil count / lymphocyte count.

[0010] The present invention also provides the application of the above biomarker in the preparation of a product for predicting the prognosis of sepsis.

[0011] Optionally, the product for predicting the prognosis of sepsis includes a sepsis prognosis risk stratification system for classifying sepsis patients into a high inflammation risk group and a low inflammation risk group, and a sepsis prognosis survival risk prediction system for predicting the in-hospital mortality and 28-day mortality of sepsis patients.

[0012] The present invention also provides a sepsis prognosis risk stratification system, and the sepsis prognosis risk stratification system includes a risk stratification module;

[0013] The risk stratification module is used to classify patients into a high inflammation risk group and a low inflammation risk group by using a prediction model based on the patient data of the biomarker, and output a risk stratification result;

[0014] The prediction model is constructed based on a machine learning algorithm by using the patient data of the biomarker in the population.

[0015] Optionally, the machine learning algorithm can be selected from any one of a support vector machine classifier, a random forest algorithm, a gradient boosting decision tree, a deep neural network, and a Logistic regression model.

[0016] Furthermore, the classification method is to set a threshold according to the median value of the population patient data to classify the patients.

[0017] Optionally, the sepsis prognosis risk stratification system further includes a data collection module; the data collection module is used to collect the hematological examination data of patients.

[0018] Optionally, the data collection module is further configured to preprocess the hematological examination data of the patient; the preprocessing process is as follows: performing standardization processing on the obtained hematological examination data; removing outliers and missing values; performing data quality control; the method of the standardization processing may be selected from any one of centered log-ratio transformation, Min-Max standardization, quantile standardization, and robust standardization.

[0019] The data collection module can be docked with the hospital information system (HIS) to automatically extract blood routine test data and establish a data caching and backup mechanism.

[0020] Optionally, the sepsis prognosis risk stratification system further includes a data calculation module; the data calculation module is configured to calculate the patient data of the biomarker according to the hematological examination data.

[0021] Optionally, the data calculation module is further configured to perform standardization processing on the calculated patient data; the standardization processing process is as follows: performing centered log-ratio transformation on the calculated patient data; performing z-score standardization; removing outliers and missing values.

[0022] Optionally, the sepsis prognosis risk stratification system further includes a dynamic monitoring module; the dynamic detection module is configured to periodically repeat the measurement of the patient data of the biomarker; record the change trend of the patient data; trigger an early warning signal when the change of the patient data exceeds a preset threshold.

[0023] Optionally, the early warning signal includes a three-level early warning signal: triggering a first-level signal when the patient data of a single biomarker exceeds a preset threshold; triggering a second-level signal when the patient data of two or more biomarkers exceeds a preset threshold; triggering a third-level signal when the patient data of multiple biomarkers is continuously abnormal or continuously deteriorates.

[0024] Optionally, the frequency of the dynamic monitoring is determined according to the patient risk level: the high inflammation risk group updates the measurement data and the patient prognosis risk stratification result every 4 hours; the low inflammation risk group updates the measurement data and the patient prognosis risk stratification result every 12 hours.

[0025] The dynamic monitoring and early warning module can be connected to the medical staff workstation to automatically push the early warning signal.

[0026] Optionally, the sepsis prognosis risk stratification system further includes a treatment requirement prediction module; the treatment requirement prediction module is configured to predict the probability of treatment requirements according to the risk stratification result of the patient; the treatment requirements include mechanical ventilation requirements, airway management requirements, fluid therapy requirements, nutritional support requirements, and renal replacement therapy requirements.

[0027] The present invention also provides a sepsis prognosis survival risk prediction system, which includes a survival risk prediction module;

[0028] The survival risk prediction module is used to predict the in-hospital mortality and 28-day mortality of patients by using a prediction model based on the patient data of the biomarker and the patient's clinical basic data, and output the prediction results;

[0029] The clinical basic data includes anion gap, potassium ion concentration, chloride ion concentration, sodium ion concentration, body weight, albumin, bicarbonate, hemoglobin, alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase, total bilirubin, mean corpuscular volume, age, blood urea nitrogen, international normalized ratio, prothrombin time, and red blood cell volume distribution width;

[0030] The prediction model is constructed by using the patient data of the biomarker and the patient's clinical basic data in the population with a deep learning framework.

[0031] Optionally, the sepsis prognosis survival risk prediction system further includes a data collection module; the data collection module is used to collect the patient's clinical basic data and hematological examination data;

[0032] Optionally, the data collection module is further used to preprocess the patient's clinical basic data and hematological examination data; the preprocessing process is: performing standardization processing on the obtained clinical basic data and hematological examination data; removing outliers and missing values; performing data quality control; the method of standardization processing can be selected from any one of centered log-ratio transformation, Min-Max standardization, quantile standardization, and robust standardization.

[0033] Optionally, the sepsis prognosis survival risk prediction system further includes a data calculation module; the data calculation module is used to calculate the patient data of the biomarker according to the hematological examination data.

[0034] Optionally, the data calculation module is further used to perform standardization processing on the calculated patient data; the standardization processing process is: performing centered log-ratio transformation on the calculated patient data; performing z-score standardization; removing outliers and missing values.

[0035] The present invention also provides the application of neutrophil-to-lymphocyte ratio and platelet-to-lymphocyte ratio in constructing a sepsis shock risk prediction model.

[0036] The present invention discloses the following technical effects:

[0037] (1) The prediction accuracy is significantly improved:

[0038] The prediction model constructed based on six inflammatory indicators in the present invention achieves excellent discriminative ability in predicting in-hospital mortality (AUC 0.77, 95% CI: 0.74 - 0.80), significantly superior to the traditional SOFA scoring system (AUC 0.70, 95% CI: 0.67 - 0.73; P < 0.001). In predicting 28-day mortality, the prediction performance of the method of the present invention is further improved (AUC 0.79, 95% CI: 0.76 - 0.82), far exceeding the SOFA score (AUC 0.67, 95% CI: 0.64 - 0.70; P < 0.001). The SVM model integrating six inflammatory indicators achieves extremely high prediction accuracy (AUC 0.86) and also shows stable stratification ability in the external validation cohort.

[0039] (2) Significantly effective in early identification:

[0040] The prediction model constructed based on six inflammatory indicators in the present invention can achieve significant divergence of survival curves within 72 hours after admission, providing an important time window for early clinical intervention. It can accurately predict the treatment needs. After verification, the high-inflammation risk group shows a significantly increased treatment need: the need for mechanical ventilation (50.16% vs 34.52%, P < 0.001), the need for airway management (32.70% vs 22.67%, P < 0.001), the need for fluid therapy (49.96% vs 34.52%, P < 0.001), the need for nutritional support (24.34% vs 16.39%, P < 0.001), the need for renal replacement therapy (8.32% vs 2.31%, P < 0.001).

[0041] (3) Obvious advantages in clinical practicability:

[0042] Simple operation: Only routine blood test data is required, without additional equipment investment, and the detection time is short. Results can be obtained in about 15 - 30 minutes.

[0043] Significant economic benefits: It can reduce the detection cost, saving 50 - 70% of the detection cost compared with the traditional molecular marker detection; reduce the length of hospital stay, and can shorten the average length of hospital stay through early warning; optimize the allocation of medical resources: accurately predict the treatment needs and improve the resource utilization efficiency.

[0044] (4) Prediction stability:

[0045] Temporal stability: The difference pattern of inflammatory indicators remains stable during the 14-day follow-up period, and non-survivors maintain a significantly elevated level of inflammatory indicators (P < 0.001).

[0046] Multi-center validation effect: Stable prediction performance was shown in both the MIMIC-IV database (n = 14350) and the external validation cohort (n = 74), and the cross-institutional prediction accuracy was consistent.

[0047] (5) Value of individualized treatment guidance:

[0048] Precision stratification effect: Five patient types with different prognostic characteristics were successfully identified, and significant differences in 30-day survival rates were shown among the types.

[0049] Prediction of treatment response: It can predict the potential benefits of different treatment methods and provide an objective basis for the individualized adjustment of clinical treatment plans.

[0050] The present invention significantly improves the accuracy and timeliness of prognosis assessment for sepsis patients, provides a reliable guiding basis for clinical decision-making, and has important clinical application value and promotion significance. Description of the Drawings

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the embodiments. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0052] Figure 1 It is the identification results of different sepsis types and their related clinical phenotypes in Example 1; among them, A is the clustering result of sepsis patients in the MIMIC-IV cohort; B is the distribution violin plot of each inflammatory index and clinical parameter in the five types; C is the distribution of in-hospital mortality (left) and 28-day mortality (right) in each type; D is the Kaplan–Meier survival curve of different types (log-rank test, P < 0.0001);

[0053] Figure 2 It is the distribution of underlying diseases and complications of different sepsis types in Example 1;

[0054] Figure 3 It is the prediction results of the deep learning model in terms of sepsis mortality in Example 2; among them, A is the comparison result of the ROC curve of the deep learning model and the ROC curve of the SOFA score; B is the schematic diagram of the Spearman correlation matrix between inflammatory indexes and clinical parameters: red indicates positive correlation, blue indicates negative correlation, and all relevant P values are < 0.001. The statistical significance is compared by the DeLong test for AUC and the Spearman rank correlation test is used;

[0055] Figure 4Analysis results of the survival status of sepsis patients based on 6 inflammatory indicators in Example 3; among them, A is the distribution of 6 inflammatory indicators between 28-day survivors and non-survivors in the MIMIC-IV cohort (n = 14350); B is the external validation result of the distribution of 6 inflammatory indicators in the Longhua Hospital cohort (n = 74).

[0056] Figure 5 Prognostic risk stratification results of sepsis based on 6 inflammatory indicators in Example 3; among them, A is the Kaplan-Meier survival curve (log-rank test, P < 0.0001) drawn after dividing patients into a high-inflammation group and a low-inflammation group according to the median of each inflammatory indicator at admission; B is the ROC curve of the SVM model based on 6 inflammatory indicators in the prognostic risk stratification of sepsis; C is the comparison result of the mortality rates between the high- and low-inflammation risk groups in the Longhua Hospital cohort; D is the validation result of the SVM model in the Longhua Hospital cohort; E is the distribution of treatment interventions between the high- and low-inflammation risk groups in the MIMIC-IV cohort.

[0057] Figure 6 Longitudinal analysis results of inflammatory indicators in sepsis patients in different risk groups in Example 4.

[0058] Figure 7 Single-cell and deconvolution analysis results of inflammatory indicators in sepsis in Example 5; among them, A is the UMAP visualization result obtained based on peripheral blood single-cell RNA sequencing (GSE167363 dataset); B is the expression of classical marker genes for cell type annotation; C is the analysis result of the cell composition of healthy controls, sepsis survivors, and non-survivors.

[0059] Figure 8 Deconvolution analysis based on the GSE154918 dataset (n = 105) in Example 5 to compare the changes in the proportion of cell types of healthy controls (Hlty), infection (Inf1_P), sepsis (Seps_P), and septic shock (Shock_P) with the evolution of the disease.

[0060] Figure 9 Analysis results of inflammatory indicators in the disease development stage of sepsis in Example 5; among them, A is the response pattern of 6 inflammatory indicators during disease progression; B is the ROC curve (AUC > 0.85) of PLR and NLR in predicting the progression of septic shock.

[0061] Figure 10Results of single-cell transcriptome analysis of neutrophils during sepsis progression in Example 6; wherein, A is the UMAP visualization result; B is the expression of specific marker genes of each neutrophil subset; C is the analysis of neutrophil subset composition under each clinical outcome; D is the result of pseudotime trajectory analysis, and the color gradient represents the pseudotime process (dark blue: early stage; red: late stage); E is the neutrophil maturation characteristics of single-sample gene set enrichment analysis (ssGSEA) disease progression stage (Hlty: healthy control; Inf1_P: infection; Seps_P: sepsis; Shock_P: septic shock); F is the result of gene set enrichment analysis, the circle size represents the number of genes, and the color depth represents the statistical significance (adjusted P value);

[0062] Figure 11 Results of single-cell transcriptome analysis of neutrophils during sepsis progression in Example 6; wherein, A is the result of enrichment curve analysis; B is the analysis result of effector molecule expression in different neutrophil subsets; C is the heat map of path-specific gene expression;

[0063] Figure 12 Results of single-cell transcriptomics feature analysis in Example 7; wherein, A is the result of single-cell transcriptomics and UMAP dimensionality reduction analysis of platelets; B is the violin plot of differential expression of specific marker genes of each platelet subset; C is the result of cell composition analysis; D is the result of gene set enrichment analysis; E is the result of pathway enrichment analysis; F is the analysis result of differential expression of key functional molecules;

[0064] Figure 13 Results of single-cell transcriptomics feature analysis in Example 7; wherein, A is the result of hierarchical clustering analysis for pathway-specific transcripts; B is the result of Spearman rank correlation analysis; C is the result of a similar Spearman rank correlation analysis. Detailed implementation manners

[0065] The various exemplary implementation manners of the present invention will now be described in detail. This detailed description should not be considered as a limitation of the present invention, but rather as a more detailed description of certain aspects, characteristics, and implementation schemes of the present invention.

[0066] It should be understood that the terms described in the present invention are only for describing specific implementation manners and are not used to limit the present invention. Additionally, for the numerical ranges in the present invention, it should be understood that each intermediate value between the upper and lower limits of the range is also specifically disclosed. Any intermediate value within any stated value or stated range, as well as each smaller range between any other stated value or intermediate value within the stated range, is also included in the present invention. The upper and lower limits of these smaller ranges may be independently included or excluded from the range.

[0067] Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. Although this invention only describes preferred methods and materials, any methods and materials similar or equivalent to those described herein can also be used in the practice or testing of this invention. All documents mentioned in this specification are incorporated by reference to disclose and describe the methods and / or materials related to the documents. In case of conflict with any incorporated document, the content of this specification shall prevail.

[0068] Without departing from the scope or spirit of the present invention, various improvements and changes can be made to the specific embodiments of the description of the present invention, which will be obvious to those skilled in the art. Other embodiments obtained from the description of the present invention will be obvious to those skilled in the art. The description and examples of the present invention are merely exemplary.

[0069] Regarding the use of "comprising", "including", "having", "containing", etc. in this article, they are all open-ended terms, meaning including but not limited to.

[0070] In the following examples of the present invention, the analysis cohort population was selected from the MIMIC-IV database, and a total of 50,914 ICU inpatients were extracted. After strict inclusion / exclusion criteria, patients with a SOFA score < 2 were initially screened out, and then patients who were not admitted for the first time and those with a short hospital stay < 48 hours were excluded. To ensure data integrity and minimize confounding variables, patients with potential monocyte and lymphocyte-related lesions, incomplete hematological data, and excessive data missing were further excluded. Finally, an analysis cohort (MIMIC-IV cohort) consisting of 14,350 sepsis patients was formed. In the following examples of the present invention, 74 sepsis patients from Longhua Hospital (with the same inclusion / exclusion criteria as the MIMIC-IV cohort) were also included to form an independent cohort for external validation.

[0071] The calculation methods of the inflammatory indicators in the following examples of the present invention are as follows:

[0072] NLR (Neutrophil-to-Lymphocyte Ratio) = neutrophil count / lymphocyte count;

[0073] PLR (Platelet-to-Lymphocyte Ratio) = platelet count / lymphocyte count;

[0074] MLR (Monocyte-to-Lymphocyte Ratio) = monocyte count / lymphocyte count;

[0075] LLR (Leukocyte to Lymphocyte Ratio) = white blood cell count / lymphocyte count;

[0076] SIRI (Systemic Inflammation Response Index) = (neutrophil count × monocyte count) / lymphocyte count;

[0077] PIV (Peripheral Inflammation Value) = (platelet count × neutrophil count) / lymphocyte count.

[0078] Some of the method steps used in the following embodiments of the present invention are as follows:

[0079] Unsupervised clustering method for patients in the MIMIC database: Based on 21 physiological and biochemical indicators and 6 inflammatory indicators, cluster the patients in the MIMIC database. The specific method is to perform logarithmic transformation and standardization on each indicator, then perform PCA and identify the SNN (Shared Nearest Neighbors) between samples. Under the condition of a resolution of 0.35, use the SNN module to optimize the clustering algorithm to cluster and group the patients.

[0080] Construction method of support vector machine classifier (SVM model): Based on 6 inflammatory indicators, divide the data set into a training set and a validation set at a ratio of 8:2; perform feature standardization on the training data; set the classification threshold to divide the patients into a high - inflammation - risk group and a low - inflammation - risk group. Use the ROCR package in R to draw the ROC curve according to the model prediction score and the true label, and calculate the AUC value.

[0081] Comprehensive verification method for the SVM model: Perform 5 - fold cross - validation in the MIMIC - IV cohort; perform external validation in the independent Longhua Hospital cohort; and perform time validation of prediction stability at different time points. For survival analysis, use the median of the six inflammatory indicators as the threshold to divide into high / low inflammation groups, fit the survival model using the survival package (version 3.2 - 13) in R, and use the survminer package (version 0.4.9) to draw the Kaplan - Meier curve.

[0082] Method for constructing a deep learning model: Implement a deep learning network using PyTorch (version 1.9.0), including an input layer with 24 nodes, three hidden layers (with 128, 64, and 32 nodes respectively) containing ReLU activation functions and a dropout rate of 0.3, and a single-node Sigmoid output layer for mortality prediction. The model is trained using the Adam optimizer (learning rate 0.001), binary cross-entropy loss function, and mini-batch (batch size 64) approach, running for a total of 100 epochs with early stopping (patience = 10) set. The dataset is divided into a training set and a test set in a ratio of 80% and 20%, and the model performance is evaluated through multiple metrics (AUROC, precision-recall curve, sensitivity / specificity at the optimal threshold, calibration curve, and net reclassification improvement compared to the SOFA score).

[0083] Single-cell RNA sequencing analysis method: The single-cell RNA sequencing data is sourced from Gene Expression Omnibus (GEO), with the accession number GSE167363. Low-quality cells (UMI count < 500, or UMI count > 5000, or mitochondrial gene proportion > 20%) are filtered out, and 48,486 high-quality cells are retained for subsequent analysis. Specifically, normalization is performed based on the count matrix, and the 3000 gene features with the highest variability are selected. Principal component analysis (PCA) is performed on the normalized expression of the variable features. Subsequently, the Harmony method in R is used to integrate cells from different samples, and UMAP dimensionality reduction and neighbor cell analysis are performed based on the Harmony dimensions. The cell clusters are identified using the SNN (Shared Nearest Neighbor) module to optimize the clustering algorithm at a resolution of 1. Marker genes for each cell cluster are identified using differential gene expression analysis. The cell clusters expressing known marker genes are annotated (B cells: CD79A, MS4A1; proliferating cells: MKI67, TOP2A; red blood cells: HBB, HBA2; monocytes: CD14, LYZ; neutrophils: CSF3R, MMP9; NK cells: GNLY, NKG7; plasma cells: MZB1, JCHAIN; platelets: PPBP, PF4; T cells: CD3D, CD3E). The subpopulation analysis method for neutrophils and platelets is similar to the above main population analysis. Cell trajectory analysis is completed using Monocle3.

[0084] Bulk RNA sequencing deconvolution analysis method based on scRNA-seq data: The bulk RNA-seq data is from GEO (accession number GSE154918). Specifically, using single-cell RNA sequencing (scRNA-seq) data as a reference, the Single-cell RNA Quantity Informed Deconvolution (SQUID) method is adopted to estimate the proportion of each cell type in the bulk RNA-seq samples. The inflammation indicators are calculated based on the relative cell proportions.

[0085] Statistical analysis methods: Statistical methods are selected according to the characteristics of the data distribution; for continuous variables, according to the results of the Shapiro-Wilk normality test, they are presented in the form of mean ± standard deviation or median (interquartile range). Student's t-test or Mann-Whitney U test (for continuous variables), and chi-square test or Fisher's exact test (for categorical variables) are used for between-group comparisons; for multiple-group comparisons, one-way ANOVA or Kruskal-Wallis test is used and Bonferroni correction is performed. Correlation analysis includes Spearman rank correlation analysis for non-normal data, Pearson correlation analysis for normally distributed data, and partial correlation analysis with adjustment for confounding factors. Multivariate analysis includes Cox proportional hazards regression (for survival analysis), Logistic regression (for binary outcomes), and linear mixed models (for longitudinal data assessment).

[0086] Sample size assessment method: The sufficiency of the sample size is verified through comprehensive power analysis to ensure that in survival analysis, 80% test power can detect a hazard ratio ≥ 1.5, the minimum cluster size in cluster analysis is ≥ 100 patients, at least 10 events / variables in machine learning models, and 90% power in validation studies can detect an AUC difference of 0.1. All statistical analyses are completed in the R software (version 4.1.0) environment, and relevant extension packages such as survival (3.2 - 13), lme4 (1.1 - 27), stats (4.1.0), and rms (6.2 - 0) are used.

[0087] For the methods involved in the following examples of the present invention, unless otherwise specified, they are carried out according to the conventional methods in the art.

[0088] Example 1

[0089] This example studied the routine admission parameters and inflammatory indicators of sepsis patients and described the stratification pattern of sepsis based on admission parameters. The analyzed data included 28 clinical variables (height, age, gender, sofa score, weight, body fat index bmi, albumin, alkaline phosphatase alk, alanine aminotransferase alt, anion gap, aspartate aminotransferase ast, bicarbonate, blood urea nitrogen bun, calcium, chloride concentration, C-reactive protein crp, total bilirubin, high-density lipoprotein hdl, low-density lipoprotein ldl, hemoglobin, international normalized ratio inr, mean corpuscular volume mcv, platelet, potassium ion concentration, prothrombin time pt, red blood cell distribution width rdw, sodium ion concentration, cholesterol tc) and 6 inflammatory indicators. Among them, the 6 inflammatory indicators were neutrophil-to-lymphocyte ratio (NLR), platelet-to-lymphocyte ratio (PLR), monocyte-to-lymphocyte ratio (MLR), leukocyte-to-lymphocyte ratio (LLR), systemic inflammatory response index (SIRI), and peripheral inflammation value (PIV). After the centered log-ratio transformation of data normalization, principal component analysis (PCA) and uniform manifold approximation and projection (UMAP) were used for dimensionality reduction. Clustering with a resolution of 0.35 based on shared nearest neighbor (SNN) modular optimization identified five different patient types ( Figure 1 A), including cluster 0 (hyperinflammatory type), cluster 1 (early inflammatory response type), cluster 2 (physiologically stable type), cluster 3 (late organ failure type), and cluster 4 (early organ dysfunction type).

[0090] There were different pathophysiological characteristics among these different types ( Figure 1B): The physiologically stable type (cluster 2) showed the best clinical parameters, characterized by stable body weight (75.2 ± 4.8 kg), albumin (0.75 ± 0.05 g / dL), bicarbonate (24.5 ± 2.1 mmol / L), and hemoglobin (12.8 ± 1.2 g / dL), as well as minimal inflammatory response (CRP: 28.5 ± 15.2 mg / L) and low SOFA score (3.2 ± 1.1; all comparisons P < 0.001). The early inflammatory response type (cluster 1) showed an initial transitional state with moderate electrolyte alterations (chloride: 106.8 ± 3.2 mmol / L; sodium: 142.5 ± 2.8 mmol / L), while maintaining stable organ function (SOFA: 4.1 ± 1.3). In contrast, the hyperinflammatory type (cluster 0) showed a significant increase in all inflammatory indices (mean fold change relative to the physiologically stable type: LLR 2.8, NLR 3.1, SIRI 2.9, PIV 2.7, PLR 2.5, MLR 2.6; all comparisons P < 0.001) and peak CRP level (185.4 ± 42.3 mg / L), although a moderate SOFA score (4.3 ± 1.4) demonstrated relatively stable organ function. The early organ dysfunction type (cluster 4) showed emerging multiple organ dysfunction, characterized by significantly elevated renal and coagulation markers (BUN 38.5 ± 5.2 mg / dL, INR 1.8 ± 0.3, PT 18.2 ± 2.1 seconds, RDW 16.8 ± 1.4%; P < 0.001) and increased SOFA score (6.8 ± 1.6). The late organ failure type (cluster 3) showed severe organ dysfunction, accompanied by significant liver function disorders (ALK 285.4 ± 45.6 U / L, ALT 325.8 ± 58.2 U / L, AST 298.4 ± 52.1 U / L, total bilirubin 4.8 ± 0.9 mg / dL; P < 0.001), the highest SOFA score (8.9 ± 1.8), and profound metabolic dysfunction reflected by the lowest HDL level (22.4 ± 4.2 mg / dL).

[0091] Analysis of clinical outcomes revealed significant differences in mortality and survival patterns among the different types ( Figure 1C-D): Kaplan-Meier analysis showed different 30-day survival trajectories (log-rank test, P<0.0001). Physiological stability showed a higher survival rate (95% at 30 days), while hyperinflammation, late organ failure, and early organ dysfunction showed significantly reduced survival rates (70 - 75%). The early inflammatory response maintained a moderate survival curve (85%), significantly different from physiological stability (P = 1.44e-25) and hyperinflammation (P = 6.8e-46). The most significant differences in survival rates were observed between physiological stability and late organ failure and early organ dysfunction (P = 6.78e-110 and P = 2.12e-78, respectively). Types with higher inflammatory markers (hyperinflammation, late organ failure, and early organ dysfunction) consistently showed significantly reduced survival rates compared to those with lower inflammatory characteristics (early inflammatory response and physiological stability).

[0092] Detailed analysis of underlying conditions revealed different comorbidity distributions among different types ( Figure 2 ). Physiological stability showed the lowest burden of chronic diseases, with low prevalence of autoimmune diseases (1.71%), malignancies (8.32%), and chronic liver diseases (1.89%), mainly presenting as cardiovascular diseases (78.5%) and metabolic diseases (45.8%). In contrast, hyperinflammation and early organ dysfunction showed the highest prevalence of chronic kidney diseases (32.91% and 31.33%, respectively) and diabetes (22.90% and 21.42%, respectively), while late organ failure showed the highest frequency of chronic liver diseases (18.07%). Analysis of sepsis-related complications revealed different patterns among different types. Types characterized by elevated inflammatory markers (hyperinflammation, late organ failure, and early organ dysfunction) showed significantly higher incidences of acute kidney injury (52.64%, 44.94%, and 42.39%, respectively), respiratory failure (44.36%, 41.82%, and 39.75%, respectively), and progression to septic shock (38.52%, 35.84%, and 34.98%, respectively); all comparisons P<0.001). Notably, late organ failure showed the highest frequency of liver dysfunction (18.07%) and acute liver injury (15.31%), consistent with its unique pattern of liver function disorder. Physiological stability maintained the lowest complication incidence in all categories, supporting its favorable prognostic characteristics.

[0093] The above results indicate that the 6 inflammatory indicators have the potential value for predicting the prognosis risk of sepsis.

[0094] Example 2

[0095] This embodiment integrates 18 clinical variables (anion gap, potassium ion concentration, chloride ion concentration, sodium ion concentration, body weight, albumin, bicarbonate, hemoglobin, alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase, total bilirubin, mean corpuscular volume, age, blood urea nitrogen, international normalized ratio, prothrombin time, red blood cell volume distribution width) and 6 inflammatory indicators, and designs a deep learning model including admission parameters.

[0096] The deep learning model is implemented using PyTorch (version 1.9.0), including an input layer with 24 nodes, three hidden layers (with 128, 64, and 32 nodes respectively) containing ReLU activation functions and a dropout rate of 0.3, and a single-node Sigmoid output layer for death prediction. The model is trained using the Adam optimizer (learning rate 0.001), binary cross-entropy loss function, and mini-batch (batch size 64), running for 100 epochs in total and setting early stopping (patience = 10). The dataset is the patient data in the MIMIC-IV cohort (n = 14350), which is divided into a training set and a test set at a ratio of 80% and 20%. Data preprocessing includes z-score normalization to standardize the parameter distribution. Finally, the model performance is evaluated through multiple metrics (AUROC, precision-recall curve, sensitivity / specificity under the optimal threshold, calibration curve, and net reclassification improvement compared with the SOFA score).

[0097] The performance evaluation results are as Figure 3 shown in Figure A. It can be seen that for the 28-day mortality, the above model has higher discrimination (AUC = 0.79 vs. 0.67, P < 0.001), with a sensitivity of 71.38% and a specificity of 71.40% at the threshold of 0.2339; for the in-hospital mortality, the model still performs excellently (AUC = 0.77 vs. 0.70, P < 0.001), with a sensitivity of 69.75% and a specificity of 69.84% at the threshold of 0.1983.

[0098] Furthermore, a correlation analysis is conducted on the 6 inflammatory indicators and all 28 clinical parameters (the same as in Example 1, including the 18 clinical parameters used to construct the above model). The results are as Figure 3As shown in B. Among them, red indicates positive correlation and blue indicates negative correlation, and all relevant P values are <0.001. It can be seen that the inflammatory markers are moderately correlated with the systemic inflammation index: C-reactive protein is moderately positively correlated with LLR (r = 0.29, P < 0.001) and NLR (r = 0.28, P < 0.001), while serum albumin shows a significant negative correlation (LLR: r = -0.27, NLR: r = -0.26; P < 0.001). The organ dysfunction parameters show different degrees of correlation. The SOFA score shows a weak correlation (r = 0.11 - 0.12, P < 0.001), while the specific organ function markers show a moderate correlation. Blood urea nitrogen is weakly correlated with LLR and NLR (r = 0.20 and 0.19 respectively; P < 0.001), and the anion gap shows a similar correlation with the inflammatory index (r = 0.18 - 0.20, P < 0.001). Serum bicarbonate shows a weak negative correlation, most significantly with LLR (r = -0.18, P < 0.001). Demographic factors show the least correlation (age: r = 0.05 - 0.10; gender: negligible), indicating that the inflammatory index mainly reflects the severity of the disease rather than the baseline characteristics.

[0099] The above results indicate that integrating 6 inflammatory indicators with admission parameters in the present invention can perform earlier and more accurate risk assessment for sepsis patients. Based on this, a sepsis prognosis risk assessment model is constructed, and the prediction result is more accurate compared with the SOFA score.

[0100] Example 3

[0101] In this example, through systematic analysis of the inflammatory index data of patients at the time of admission in the MIMIC-IV cohort (n = 14350), it was found that ( Figure 4 as shown in A), there are significant differences in inflammatory characteristics between 28-day survivors and non-survivors. Non-survivors showed a significantly elevated inflammatory state, and all measured indices were significantly up-regulated: NLR showed a 3-fold increase (median 15.0 vs 5.0, P < 2.22e-16), PLR showed a 1.33-fold increase (median 200.0 vs 150.0, P < 2.22e-16), MLR showed a 1.6-fold increase (median 0.8 vs 0.5, P < 2.22e-16), SIRI showed a 5-fold increase (median 10.0 vs 2.22), and PIV showed a 4-fold increase (median 2000.0 vs 500.0, P < 2.22e-16). External validation was carried out using the Longhua Hospital independent cohort (n = 74), and it was found that ( Figure 4In B), non-survivors still showed higher levels of inflammatory markers: NLR (P = 0.003), MLR (P = 0.035), PLR (P = 0.022), SIRI (P = 0.002), PIV (P = 0.007), and LLR (P = 0.012), indicating that these inflammatory patterns are reproducible in different clinical settings.

[0102] Kaplan-Meier survival curves were plotted after classifying patients into high- and low-inflammatory risk groups according to the median of each inflammatory marker at admission. The results showed ( Figure 5 In A), the 30-day survival trajectories of the two groups were significantly different (log-rank test, P < 0.0001). NLR, SIRI, and PIV showed significant early differences in the survival curves between the high and low groups (log-rank test, P < 0.0001 for all comparisons). This early prognostic signal was particularly evident in the NLR stratification, in which elevated NLR was associated with increased mortality within the first 72 hours after admission.

[0103] In this example, a support vector machine classifier (SVM model) was further established based on six inflammatory markers of patients in the MIMIC-IV cohort (n = 14350) to perform precise risk stratification for sepsis patients. The results showed ( Figure 5 In B), the area under the ROC curve (AUC) of the support vector machine (SVM) model based on six inflammatory markers in the death risk stratification was 0.86. External validation was performed using the Longhua Hospital independent cohort (n = 74). The results showed ( Figure 5 In D), significant differences in inflammatory markers were observed between the high- and low-inflammatory risk groups: NLR (median 12.5 vs. 4.2, P < 0.001), PLR (median 285 vs. 142, P < 0.001), MLR (median 0.85 vs. 0.32, P < 0.001), SIRI (median 8.2 vs. 2.8, P < 0.001), PIV (median 1850 vs. 650, P < 0.001), and LLR (median 18.5 vs. 6.8, P < 0.001). Moreover, the high-risk group in the Longhua Hospital independent cohort showed a higher mortality rate ( Figure 5 In C, Fisher's exact test, P = 0.01).

[0104] Based on the above results, the treatment intervention distributions of the high and low groups in the MIMIC-IV cohort were further analyzed. The results showed ( Figure 5E), the need for intensive treatment support in the high-risk group increased significantly: mechanical ventilation (50.16% vs. 34.52%, P < 0.001), airway management (32.70% vs. 22.67%, P < 0.001), fluid therapy (49.96% vs. 34.52%, P < 0.001), nutritional support (24.34% vs. 16.39%, P < 0.001), and renal replacement therapy (8.32% vs. 2.31%, P < 0.001). It is shown that the SVM model established based on 6 inflammatory indicators can not only accurately stratify the risk of sepsis patients, but also predict the treatment needs of sepsis patients, thus enabling the establishment of a stratified management mechanism for sepsis patients: repeat data measurement and risk assessment every 4 hours for the high-inflammatory risk group, and repeat data measurement and risk assessment every 12 hours for the low-inflammatory risk group.

[0105] Example 4

[0106] This example verified the biological basis of the SVM model based on 6 inflammatory indicators. By comprehensively longitudinally analyzing the inflammatory indicators (LLR, MLR, NLR, PIV, PLR, and SIRI) within 14 days after admission to the MIMIC-IV cohort, the stability and changing trends of the inflammatory patterns of sepsis patients in different risk groups were described.

[0107] System trajectory analysis revealed different and persistent inflammatory characteristics between 28-day survivors and non-survivors ( Figure 6 ). There were significant differences between 28-day survivors and non-survivors (n = 14350) within 14 days after admission. LLR remained continuously higher in non-survivors (20 - 22) than in survivors (12 - 14) throughout the first week (P < 0.001, days 0 - 6). MLR showed a relatively mild but statistically significant separation (non-survivors 0.9 - 1.1 vs. survivors 0.6 - 0.7; P < 0.01, days 0 - 4). NLR showed a significantly stable difference between non-survivors (15 - 20) and survivors (10 - 12) (P < 0.001, days 0 - 6). PIV showed the largest numerical difference (non-survivors 3000 - 4000 vs. survivors 1500 - 2000; P < 0.001, days 0 - 5). PLR showed significant differences in the early stage (days 0 - 3) (P < 0.001), and then showed certain temporal fluctuations. SIRI showed a greater amplitude difference between non-survivors (12 - 18) and survivors (8 - 10), but remained elevated throughout (P < 0.001, days 0 - 6).

[0108] The above results indicate that inflammatory indicators are stable as markers for differentiating different sepsis patients, rather than showing transient fluctuations. By regularly measuring clinical parameters including six inflammatory indicators in sepsis patients, dynamic monitoring of the physiological and pathological states of sepsis patients can be achieved. Connecting the models constructed in Example 2 or Example 3 to the nurse station or laboratory and establishing an early warning device by setting an early warning threshold for inflammatory indicators can enable the setting of personalized treatment methods and real-time monitoring for sepsis patients, so as to provide timely medical intervention.

[0109] Example 5

[0110] In this example, comprehensive transcriptome analysis was performed by single-cell and bulk RNA sequencing methods to establish the cellular basis of inflammatory indicators.

[0111] The single-cell RNA sequencing data was sourced from Gene Expression Omnibus (GEO), with the accession number GSE167363. The initial single-cell RNA sequencing analysis results of the GSE167363 dataset showed ( Figure 7 A), nine major cell populations were identified based on peripheral blood single-cell RNA sequencing; the key immune populations were clearly depicted using canonical marker-based cell type annotations ( Figure 7 B): CD14 / LYZ for monocytes, CSF3R / MMP9 for neutrophils, CD3D / CD3E for T cells, CD79A / MS4A1 for B cells, GNLY / NKG7 for NK cells, and MZB1 / JCHAIN for plasma cells. The results of the comparative analysis of cell composition showed ( Figure 7 C), compared with survivors and healthy controls, non-surviving sepsis patients showed significantly increased proportions of neutrophils and platelets (P < 0.001). This observation provides the first direct cellular evidence to support the prognostic value of neutrophil- and platelet-based inflammatory indices in sepsis outcomes.

[0112] Furthermore, deconvolution algorithms were used to analyze the bulk RNA sequencing data of the GSE154918 dataset, including healthy control group (Hlty, n = 40), infection without organ dysfunction (Inf1_P, n = 12, SOFA < 2), sepsis with organ dysfunction (Seps_P, n = 20, SOFA ≥ 2), and septic shock requiring vasopressors (Shock_P, n = 19). This new analysis method can accurately quantify the proportions of immune cells and calculate inflammatory indices based on readily available transcriptome data. The results of the cell component analysis showed ( Figure 8) In the adaptive immune compartment, a significant depletion was shown from healthy controls to the diseased state, with significant decreases in B cells (P = 7e-05) and T cells (P = 1.4e-05), indicating impaired adaptive immunity; NK cells gradually decreased (P = 9.5e-07), while the plasma cell population remained relatively stable; consistent with the single-cell study results, the bone marrow compartment showed progressive dilation, and neutrophils and platelets showed a progressive increase related to disease severity.

[0113] In addition, by calculating the inflammatory indices at different stages of sepsis disease development, it was found that ( Figure 9 A), as the disease severity increased, all indices showed a progressive increase, transitioning from the healthy control group to the states of infection, sepsis, and septic shock. The neutrophil and platelet indices (NLR, LLR, and PLR) increased most significantly in patients with septic shock. Using the ROC curve for analysis, it was found that ( Figure 9 B), compared with other inflammatory indices, PLR and NLR showed superior predictive ability (AUC > 0.85), establishing their special use in identifying patients at risk of shock progression.

[0114] The above results established a direct mechanistic link between the single-cell observations of increased neutrophil and platelet proportions in non-survivors and the prognostic value of the corresponding inflammatory indices in septic shock. The neutrophil and platelet indices showed a particularly significant increasing pattern in septic shock, establishing a direct mechanistic link between changes in cell composition and the dynamics of inflammatory indices. This progressive increase may reflect the cumulative effects of emergency myelopoiesis, altered cell trafficking, and immune homeostasis dysregulation in severe sepsis.

[0115] Example 6

[0116] Based on the results observed in Example 5, in this example, in-depth single-cell transcriptional characterization of neutrophil heterogeneity was performed to reveal potential pathogenic mechanisms. The results of unbiased clustering analysis of the neutrophil population showed ( Figure 10 A - B) that the neutrophil population was divided into three distinct subpopulations, characterized by differential expression of established mature markers: pre-neutrophils (PreNeu) expressing LYZ and PTMA, immature neutrophils (ImmNeu) marked by MMP9 and LCN2 expression, and mature neutrophils (MatNeu) marked by PROK2 and ADM expression. Comparative analysis of the cell composition results showed ( Figure 10 C) that the proportion of immature neutrophils (ImmNeu) in sepsis non-survivors was significantly higher than that in survivors and healthy controls (P < 0.001, one-way ANOVA), indicating their potential role in adverse outcomes. The results of pseudotime trajectory analysis ( Figure 10D) further revealed the developmental continuity from the immature state to the mature state, providing insights into the neutrophil maturation kinetics during sepsis. This observation is consistent with previous studies that have documented alterations in neutrophil maturation and function during sepsis, and the accumulation of immature neutrophils is associated with poor prognosis.

[0117] Single-sample gene set enrichment analysis (ssGSEA) was further utilized to quantify the neutrophil subset characteristics in bulk RNA sequencing data, and the results showed ( Figure 10 E) a progressive increase in the immature neutrophil score at each stage of the disease, demonstrating that the immature neutrophil signature gradually rises with disease severity (P < 0.001 for all pairwise comparisons). This pattern suggests a mechanistic link between neutrophil immaturity and disease severity. To decipher the molecular programs underlying the potential pathogenicity of immature neutrophils, their transcriptional signals were analyzed. The results of gene set enrichment analysis showed ( Figure 10 F) that both pro-inflammatory and hypoxia adaptation pathways were simultaneously activated in immature neutrophils, and the enrichment curves showed ( Figure 11 A) that key pathogenic pathways were significantly upregulated in septic shock patients, including mechanisms such as hypoxia-inducible factor, neutrophil degranulation, and leukocyte transendothelial migration. Analysis of the expression of effector molecules in different neutrophil subsets ( Figure 11 B) showed that the immature subsets coordinately upregulated degranulation mediators (CD177, CD63, LCN2, MMP9), hypoxia response factors (LDHA, ASPH, HIF1A, PKM), and inflammatory signaling components (S100A8 / A9, HMGB1, TLR2). The pathway-specific gene expression heatmap ( Figure 11 C) demonstrated the selective enrichment of pathogenic characteristics between septic shock and sepsis patients.

[0118] The above results indicate that compared with sepsis patients, the expression of molecular markers of neutrophil dysfunction is significantly elevated in the transcriptome of septic shock patients. These findings provide new mechanistic insights into neutrophil-based inflammatory indices in sepsis by demonstrating that disease progression is particularly associated with the accumulation of immature neutrophils exhibiting a dual pro-inflammatory and hypoxic phenotype. This finding not only clarifies the potential mechanism of neutrophil elevation during the development of septic shock but also identifies immature neutrophils and their associated molecular pathways as potential therapeutic targets for septic shock intervention.

[0119] Example 7

[0120] After determining in Example 6 that the elevated PLR is a predictive marker for septic shock mortality, this example performed a comprehensive single-cell transcriptomic characterization to describe platelet heterogeneity during sepsis progression.

[0121] Platelets in peripheral blood (n = 14350) were analyzed using the UMAP dimensionality reduction method based on single-cell transcriptomics. Figure 12 A), two subsets P1 and P2 with different metabolic and immune regulation characteristics were found. Violin plots Figure 12 B) showed differential expression of subgroup-specific marker genes: PFN1, CYBA, IFITM2, and IFITM3 were highly expressed in the P1 subgroup, while PTCRA, ACRBP, RIPOR2, and RAB32 were higher in the P2 subgroup. The results of cell composition analysis showed Figure 12 C) that in non-survivors of sepsis, the proportion of P1 platelets with a metabolically activated phenotype was significantly higher than that in survivors and healthy controls (P < 0.001, one-way ANOVA), suggesting its potential pathogenic significance. Gene set enrichment analysis Figure 12 D) confirmed the manifestation of the P1 subgroup-specific transcriptional signature in patients with septic shock: in the comparison of the transcriptomes of the shock group and the sepsis group, P1 pattern genes were significantly enriched (adjusted P = 0.0066, FDR < 0.05), confirming the correlation between P1 platelet activation and disease severity at the population level. Pathway enrichment analysis Figure 12 E) revealed that the P1 platelet phenotype simultaneously activated four major functional modules: (i) energy adaptation (oxidative phosphorylation, respiratory chain electron transfer); (ii) metabolic remodeling (glycolysis / gluconeogenesis); (iii) stress response (ROS detoxification); (iv) platelet activation (calcium-dependent signal transduction). Differential expression analysis of key functional molecules Figure 12 F) further demonstrated the unique upregulation of the P1 subgroup, including components of the electron transport chain (NDUFA4, COX7A2, COX6A1, NDUFB1), glycolytic enzymes (LDHA, TPI1, ENO1, PGK1), and platelet activation markers (PFN1, CD63, CD9, PROS1). Hierarchical clustering analysis of pathway-specific transcripts Figure 13 A) showed that compared with sepsis patients, septic shock patients had coordinated differences in metabolic and activation characteristics. Spearman rank correlation analysis was used to explore the association between NLR and neutrophil-specific molecular determinants, and the results showed Figure 13 B) that it was significantly correlated with damage-associated molecular patterns (S100A9: r = 0.895, S100A8: r = 0.881; P < 0.001) and extracellular matrix remodeling factors (MMP9: r = 0.881, CD177: r = 0.837; P < 0.001), confirming the mechanistic link between inflammatory indicators and neutrophil pathogenic pathways. Similar Spearman rank correlation analysis Figure 13C) It was found that PLR was closely related to key regulators of platelet activation and energy adaptation, including granule secretion (CD63: r = 0.836, P < 0.001), oxidative phosphorylation mechanism (COX6A1: r = 0.815, P < 0.001), and glycolytic remodeling (LDHA: r = 0.792, ENO1: r = 0.500; P < 0.001).

[0122] The above results indicate that P1 platelets are a metabolically hyperactive subset related to disease severity, and PLR can be used as a surrogate marker for platelet metabolic activation in the sepsis process.

[0123] The embodiments described above are only for describing the preferred mode of the present invention and do not limit the scope of the present invention. Without departing from the design spirit of the present invention, various deformations and improvements made by those of ordinary skill in the art to the technical solutions of the present invention shall fall within the protection scope determined by the claims of the present invention.

Claims

1. A biomarker for predicting the prognosis of sepsis, characterized in that: The biomarkers include neutrophil to lymphocyte ratio, platelet to lymphocyte ratio, monocyte to lymphocyte ratio, leukocyte to lymphocyte ratio, systemic inflammatory response index and peripheral inflammation value; The calculation formula of the systemic inflammatory response index is: systemic inflammatory response index = neutrophil count × monocyte count / lymphocyte count; The calculation formula of the peripheral inflammation value is: peripheral inflammation value = platelet count × neutrophil count / lymphocyte count.

2. Use of the biomarker according to claim 1 in the preparation of a product for predicting the prognosis of sepsis.

3. The use according to claim 2, characterized in that: The product for predicting the prognosis of sepsis includes a sepsis prognosis risk stratification system for classifying sepsis patients into high inflammation risk group and low inflammation risk group and a sepsis prognosis survival risk prediction system for predicting the in-hospital mortality rate and 28-day mortality rate of sepsis patients.

4. A sepsis prognostic risk stratification system, characterized in that: The sepsis prognosis risk stratification system includes a risk stratification module; The risk stratification module is used to classify patients into high inflammation risk group and low inflammation risk group using a prediction model according to the patient data of the biomarker according to claim 1, and output the risk stratification result; The prediction model is constructed based on a machine learning algorithm using patient data of the biomarkers as described in claim 1 in a population.

5. The sepsis prognostic risk stratification system according to claim 4, characterized in that: The classification method is to set a threshold value according to the median value of the group patient data to classify the patients.

6. The sepsis prognostic risk stratification system according to claim 4, characterized in that: The sepsis prognosis risk stratification system further includes a data collection module; the data collection module is used to collect the patient's hematological examination data; It also includes a data calculation module; the data calculation module is used to calculate the patient data of the biomarker as claimed in claim 1 based on the hematological examination data.

7. The sepsis prognostic risk stratification system according to claim 4, characterized in that: The sepsis prognosis risk stratification system also includes a dynamic monitoring module; the dynamic detection module is used to periodically and repeatedly measure the patient data of the biomarker according to claim 1; record the change trend of the patient data; and trigger an early warning signal when the change of the patient data exceeds a preset threshold; The frequency of dynamic monitoring is determined according to the patient's risk level: the measurement data and patient prognostic risk stratification results are updated every 4 hours for the high inflammation risk group; and the measurement data and patient prognostic risk stratification results are updated every 12 hours for the low inflammation risk group; It also includes a treatment demand prediction module; the treatment demand prediction module is used to predict the probability of treatment demand based on the patient's risk stratification results; the treatment needs include mechanical ventilation needs, airway management needs, fluid therapy needs, nutritional support needs and renal replacement therapy needs.

8. A sepsis prognosis survival risk prediction system, characterized in that: The sepsis prognosis survival risk prediction system includes a survival risk prediction module; The survival risk prediction module is used to predict the in-hospital mortality rate and 28-day mortality rate of the patient using a prediction model based on the patient data of the biomarker and the patient's basic clinical data as described in claim 1, and output the prediction result; The basic clinical data include anion gap, potassium ion concentration, chloride ion concentration, sodium ion concentration, body weight, albumin, bicarbonate, hemoglobin, alkaline phosphatase, alanine aminotransferase, aspartate aminotransferase, total bilirubin, mean corpuscular volume, age, blood urea nitrogen, international normalized ratio, prothrombin time and corpuscular volume distribution width; The prediction model is constructed using a deep learning framework using the patient data of the biomarkers as described in claim 1 and the patient's clinical basic data in the population.

9. The sepsis prognosis survival risk prediction system according to claim 8, characterized in that: The sepsis prognosis survival risk prediction system also includes a data collection module; the data collection module is used to collect basic clinical data and hematological examination data of patients; It also includes a data calculation module; the data calculation module is used to calculate the patient data of the biomarker as claimed in claim 1 based on the hematological examination data.

10. Application of neutrophil-to-lymphocyte ratio and platelet-to-lymphocyte ratio in constructing risk prediction model for septic shock.

Images