Search

CN-121983299-A - Method for establishing post-stroke inflammation injury prediction model and evaluation method thereof

CN121983299ACN 121983299 ACN121983299 ACN 121983299ACN-121983299-A

Abstract

The invention discloses a method for establishing a post-stroke inflammation injury prediction model and an evaluation method thereof, belonging to the technical field of post-stroke inflammation injury prediction, wherein the model establishment step comprises the following steps of 1, collecting clinical index data of an AIS group and a health control group; the method comprises the steps of (1) carrying out tendency scoring matching on an AIS group and a healthy control group according to gender and age to obtain a matched AIS group with the same number of people as the healthy control group, and obtaining matching characteristics of the healthy control group and the matched AIS group, wherein (3) carrying out differentiation and correlation analysis on the matching characteristics between the healthy control group and the matched AIS group, (4) carrying out single factor analysis on fibrinogen level influence indexes in the AIS group, and (5) inputting the screened confounding factors into a fully-automatic machine learning model for model training and verification to obtain a post-stroke inflammation injury prediction Logistic regression model. The regression model established through the steps has good sensitivity and high accuracy in predicting inflammatory injury after stroke.

Inventors

  • ZHU DESHENG
  • ZHOU XIAJUN
  • SONG YAYING
  • DING JIE
  • ZHAO NAN
  • HAN LU

Assignees

  • 朱德生

Dates

Publication Date
20260505
Application Date
20240315

Claims (10)

  1. 1. The method for establishing the post-stroke inflammation injury prediction model is characterized by comprising the following steps of: step 1, collecting clinical index data of an AIS group and a health control group; Step 2, after tendency scoring matching is carried out on the AIS group and the healthy control group according to gender and age, the matched AIS group with the same number of people as the healthy control group is obtained, and the matching characteristics of the healthy control group and the matched AIS group are obtained; Step 3, performing differentiation and correlation analysis on the matching characteristics between the healthy control group and the matching AIS group; step 4, on the basis of the analysis in the step 3, carrying out single factor analysis on fibrinogen level influence indexes in the AIS group; And 5, inputting the confounding factors analyzed and screened in the step 4 into a full-automatic machine learning model for model training and verification to obtain a Logistic regression model for predicting inflammatory injury after stroke by fibrinogen level.
  2. 2. The method of claim 1, wherein the clinical index data collected in step 1 comprises sex, age, past medical history, blood routine index, serum biochemical index, blood lipid index, fibrinogen level, D-dimer, partial thromboplastin time, prothrombin time, NSE, tumor index, and treatment items received before administration.
  3. 3. The method of claim 2, wherein the history of history comprises hypertension, coronary heart disease, diabetes, and atrial fibrillation, and the blood normative indicators comprise erythrocyte count, leukocyte count, lymphocyte count, neutrophil count, and platelet count.
  4. 4. The method of claim 3, wherein the matched features obtained in step 2 include gender, age, white blood cell count, neutrophil count, fibrinogen level, NSE, hypertension, diabetes, coronary heart disease and atrial fibrillation.
  5. 5. The method of claim 4, wherein the specific operation of step 3 comprises the steps of: Step 301, differential expression level analysis of NSE, neutrophils and fibrinogen between a healthy control group and a matched AIS group; step 302, ROC curve analysis of NSE and fibrinogen between healthy control group and matched AIS group; Step 303, correlation analysis of fibrinogen and NSE between healthy control group and matched AIS group.
  6. 6. The method according to claim 5, wherein the correlation between fibrinogen and NSE expression levels in AIS patients is analyzed before single factor analysis of fibrinogen level indicators in AIS patients, and the dependency between fibrinogen and NSE expression levels in AIS patients is determined.
  7. 7. The method for constructing a predicted model of post-stroke inflammatory injury according to claim 6, wherein confounding factors determined by the single factor analysis of step 4 include gender, age, neutrophil, monocyte, platelet, albumin, D-D dimer, thrombin time.
  8. 8. The method for establishing a post-stroke inflammatory injury prediction model according to claim 7, wherein after the single factor analysis in the step 4, a curve fitting analysis of the relationship between fibrinogen and central nerve cell injury markers of the AIS patient is performed, specifically as follows: Confounding factors screened by the single factor analysis were included in a smooth curve fitting analysis of fibrinogen and NSE to determine that the extent of central nerve cell damage in AIS patients increased with increasing fibrinogen levels.
  9. 9. The method according to claim 8, wherein the Logistic regression model for predicting post-stroke inflammatory injury based on fibrinogen levels established in the step 5 is expressed as y= -7.00338+5.62873× (1/fibrinogen 2 ) -27.20105 × [ 1/fibrinogen 2 ×log (fibrinogen) ] +0.76015 × (male=1) +0.04274 ×age +0.07542 ×monocyte +0.00343 ×platelets +0.04120 ×albumin +0.11142 ×thrombin time.
  10. 10. A method for assessing the extent of post-stroke inflammatory injury using a predicted model of post-stroke inflammatory injury as set forth in any one of claims 1-9, comprising the steps of: (1) Collecting fibrinogen levels, monocyte counts, platelet counts, albumin counts, and thrombin time of the patient; (2) The collected fibrinogen level, monocyte count, platelet count, albumin count, thrombin time and age and sex of the patient are input into a Logistic regression model, and a predicted value of inflammatory injury after stroke is calculated.

Description

Method for establishing post-stroke inflammation injury prediction model and evaluation method thereof Technical Field The invention belongs to the technical field of post-stroke inflammation injury prediction, and particularly relates to a method for establishing a post-stroke inflammation injury prediction model and an evaluation method thereof. Background The brain parenchyma after the occurrence of acute ischemic cerebral apoplexy (acute ischemic stroke, AIS) can undergo cascade amplified inflammatory storm, complex inflammation related proteins can be generated in the process, various inflammatory cells are infiltrated and activated, and the inflammatory components and infiltrated cells mainly cause the change of microenvironment in brain tissues and cannot be detected and analyzed by detection means such as imaging in a short time. However, the expression levels of these protein molecules may change in the peripheral circulating blood or cerebrospinal fluid at the first time, and the inflammatory reaction of the dynamic response center progresses. By detecting these inflammatory indicators, or monitoring the change in the proportion of inflammatory cells, it is possible to be a fast and practical means of monitoring the level of inflammatory response in the brain parenchyma after infarction, and many studies are now focused on how to predict stroke prognosis, control inflammatory response by changes in the inflammatory indicators in the blood. In the past, many studies have confirmed that the expression level of various inflammatory factors is positively correlated with nerve function injury after infarction, for example, many clinical studies on AIS patients find that the clinical studies on patients are evaluated by using the national institute of health stroke scale (national Institute of Health stroke scale, NIHSS), and that the inflammatory indexes such as IL-6, C-reactive protein (CRP) and TNF-alpha in peripheral blood or cerebrospinal fluid are also obviously increased in a patient test group with high NIHSS score, and meanwhile, the inflammatory indexes are also correlated with poor prognosis. However, some studies have found that inflammatory factors have no correlation with lesion size and prognosis in the early stages, and other cytokines, such as IL-6, exert dual effects of inflammatory injury and neuroprotection at different stages of the inflammatory response after acute ischemia. The clinical studies have different numbers of cases, disease courses, sampling time, follow-up time and the like, and the research conclusion also shows that the expression mode and the functional effect of the cytokines are greatly heterogeneous in stroke patients. Thus, there is still a need for further research on how various inflammatory factors act as markers of the inflammatory level response. In recent years, more views are that interaction of various factors and intercellular signal cascade amplification during inflammatory reaction, and only one factor of single analysis can possibly encounter the problems of large heterogeneity, single analysis angle, and the like, so that in recent years, many researches are focused on combined analysis of various inflammatory indexes or cell counts, and an analysis model is constructed for assessment of severity of inflammatory level after ischemic cerebral apoplexy, prognosis of illness, and the like. But these attempts remain largely limited to small sample population studies, and the exploration and determination of post-stroke inflammatory-related markers remains a number of challenges. Early animal studies have demonstrated that fibrinogen is deposited into the infarcted area following acute ischemia in a Middle Cerebral Artery Occlusion (MCAO) animal model and activates neutrophils to exacerbate post-infarct inflammatory response and secondary brain injury. However, these injury reactions all occur centrally and are difficult to monitor clinically in a timely manner. It is noted that fibrinogen is also a typical index of inflammatory response in peripheral blood. Fibrinogen levels in peripheral blood have been shown by many studies to correlate with AIS pathogenesis and prognosis. Fibrinogen is gradually increased within 24h of the acute onset of AIS, and high fibrinogen levels in the peripheral blood of AIS patients at the time of admission are associated with poor prognosis. These phenomena are in part coincident with the inflammatory response that we observe in the central axis in the MCAO model. In summary, it is believed that changes in fibrinogen levels in the peripheral blood of AIS patients to some extent reflect dynamic changes in inflammatory levels in the hub and severity of inflammatory lesions in the hub, and further analytical verification is needed. Disclosure of Invention In order to solve the technical problems provided by the background, the invention provides a method for establishing a post-stroke inflammation injury prediction model and a