CN113936789A - Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram - Google Patents

Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram Download PDF

Info

Publication number
CN113936789A
CN113936789A CN202111195517.7A CN202111195517A CN113936789A CN 113936789 A CN113936789 A CN 113936789A CN 202111195517 A CN202111195517 A CN 202111195517A CN 113936789 A CN113936789 A CN 113936789A
Authority
CN
China
Prior art keywords
cirrhosis
hepatitis
diagnosis
liver
model
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111195517.7A
Other languages
Chinese (zh)
Inventor
周俭
胡捷
张翔宇
王征
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Zhongshan Hospital Fudan University
Original Assignee
Zhongshan Hospital Fudan University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Zhongshan Hospital Fudan University filed Critical Zhongshan Hospital Fudan University
Priority to CN202111195517.7A priority Critical patent/CN113936789A/en
Publication of CN113936789A publication Critical patent/CN113936789A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B40/00—ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16B—BIOINFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR GENETIC OR PROTEIN-RELATED DATA PROCESSING IN COMPUTATIONAL MOLECULAR BIOLOGY
    • G16B45/00—ICT specially adapted for bioinformatics-related data visualisation, e.g. displaying of maps or networks
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/40—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for data related to laboratory analysis, e.g. patient specimen analysis
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H30/00—ICT specially adapted for the handling or processing of medical images

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Public Health (AREA)
  • General Health & Medical Sciences (AREA)
  • Epidemiology (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Primary Health Care (AREA)
  • Physics & Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Evolutionary Biology (AREA)
  • Databases & Information Systems (AREA)
  • Theoretical Computer Science (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Biotechnology (AREA)
  • Radiology & Medical Imaging (AREA)
  • Pathology (AREA)
  • Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
  • Artificial Intelligence (AREA)
  • Bioethics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Software Systems (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

The invention provides a method for constructing a noninvasive hepatitis B cirrhosis diagnosis model and application of a diagnosis nomogram, wherein the method comprises the following steps: extracting a prediction factor from clinical data (such as hematology and imaging examination) of a training queue by adopting a minimum absolute contraction and selection algorithm (Lasso algorithm), and selecting parameters with nonzero coefficients by a multiple cross-validation method to perform binary Logistic regression to obtain a hepatitis B cirrhosis diagnosis model which is 0.231 multiplied by type III procollagen +0.011 multiplied by type IV collagen +0.003 multiplied by hyaluronic acid-0.013 multiplied by platelets +0.124 multiplied by liver hardness-2.387; compared with the traditional liver puncture biopsy, the method is non-invasive inspection and easy to popularize, can continuously read for multiple times to dynamically evaluate the liver cirrhosis, and can also be used for evaluating the liver cirrhosis before the operation of the hepatitis B related liver cancer patient; the model adopts a large sample independent verification queue to carry out external verification, has reliable results and is more persuasive.

Description

Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram
Technical Field
The invention belongs to the technical field of hepatitis B diagnosis, and particularly relates to a method for non-invasive diagnosis of hepatitis B-related cirrhosis and application thereof, namely a construction method of a non-invasive hepatitis B cirrhosis diagnosis model and application of a diagnosis nomogram.
Background
China is a big country with hepatitis B, and chronic hepatitis B for a long time can develop into hepatic fibrosis and further gradually develop into cirrhosis and liver cancer. Hepatitis B cirrhosis is a high-risk factor of liver cancer, and the incidence and mortality of liver cancer in China are the top of the world. Hepatitis B cirrhosis is a chronic liver disease characterized pathologically by progressive liver fibrosis and destruction of the hepatic lobular structure. Patients with hepatitis b cirrhosis are at higher risk for developing serious complications, including ascites, esophageal varices, liver failure, and primary liver cancer. For primary liver cancer patients, cirrhosis is the main cause of poor prognosis after hepatectomy. Whether cirrhosis determines the safe range of liver resection and the size of the future residual liver (FLR), and the size of the range FLR of liver resection is considered to be a determinant of liver failure after liver resection. Early diagnosis of liver cirrhosis and timely intervention can prevent further damage of liver, reduce morbidity and mortality of primary liver cancer, and improve liver transplantation success rate. Therefore, in the diagnosis and treatment process of the primary liver cancer patient, accurate diagnosis of liver cirrhosis and evaluation of the degree of liver cirrhosis are of great importance for selection of a treatment scheme and evaluation of safety.
Pathological diagnosis by liver biopsy is the gold standard for diagnosing hepatitis B-related cirrhosis, but has the following disadvantages: (1) liver biopsy is invasive operation, and complications such as abdominal bleeding and the like may occur; (2) liver puncture can only obtain a small amount of strip-shaped liver tissues, and the sampling position is closely related to the puncture direction and depth and often cannot reflect the whole cirrhosis condition of the liver; (3) the acceptance of the patient on hepatic puncture is low, and the patient is not suitable for dynamic monitoring of multiple times of puncture; (4) patients with blood coagulation dysfunction cannot perform liver biopsy, while patients with hepatitis b-related cirrhosis often have blood coagulation dysfunction of varying degrees.
In recent years, due to the defects of liver biopsy, researchers have proposed noninvasive methods for diagnosing hepatitis b-related cirrhosis by measuring liver hardness, examining serum biomarker levels, and the like. At present, the noninvasive diagnosis methods for hepatitis B related cirrhosis reported in the literature are mainly divided into two main categories. The first method is a prediction model based on serological indexes, including fibrosis-4 index (FIB-4), glutamic-oxaloacetic transaminase-platelet ratio index (APRI), Forns index, King's score and the like, but the prediction accuracy of the models is low, and the models cannot be widely used clinically. The second method is a non-invasive elastography technique, such as Two-dimensional shear-wave elastography (2D-SWE), in which Liver Stiffness (LS) is evaluated by a physical method, but the results are easily affected by inflammation, congestion, edema and other factors, and accurate results cannot be obtained. Therefore, there is an urgent clinical need for a noninvasive cirrhosis diagnosis method based on a larger sample.
Disclosure of Invention
Aiming at the defects in the prior art, the invention aims to establish a method for non-invasive diagnosis of hepatitis B related cirrhosis, assist clinicians in diagnosis of cirrhosis and assessment of the severity of hepatic fibrosis, and particularly for patients who cannot perform hepatic puncture due to combined blood coagulation dysfunction.
In order to achieve the above purpose, the solution of the invention is as follows:
in a first aspect, the invention provides a method for constructing a hepatitis B cirrhosis diagnosis model, which comprises the steps of extracting a prediction factor from clinical data (such as hematology and imaging examination) of a training queue by adopting a Lasso algorithm, and selecting a parameter with a nonzero coefficient by a multiple cross-validation method to perform binary Logistic regression to obtain the hepatitis B cirrhosis diagnosis model.
Among them, type III procollagen (PIII-NP) +0.011 × IV collagen (IV-C) +0.003 × Hyaluronic Acid (HA) -0.013 × Platelet (PLT) +0.124 × liver hardness (LS) -2.387 were diagnosed as hepatitis b cirrhosis models.
As a preferred embodiment of the present invention, the hematological examination index includes PIII-NP, IV-C, HA, and PLT; the imaging examination is the result of liver hardness measurement (LS) of two-dimensional shear wave elastography (2D-SWE).
In a second aspect, the invention provides a hepatitis B cirrhosis diagnosis model obtained by the above construction method.
In the third aspect, the invention makes the more complex hepatitis B cirrhosis diagnosis model into the hepatitis B cirrhosis diagnosis nomogram which is more convenient to use clinically.
In a fourth aspect, the invention provides the use of a hepatitis B cirrhosis diagnostic nomogram in the preparation of an assessment diagnostic for cirrhosis.
As a preferred embodiment of the invention, 5 clinical common indexes PIII-NP, IV-C, HA, PLT and LS are found out corresponding scores in a nomogram for diagnosing the hepatitis B cirrhosis, and the total score is calculated, and the total score corresponds to the probability that the patient suffers from the cirrhosis.
In a fifth aspect, the invention provides an application of a hepatitis B cirrhosis diagnosis nomogram in preparation of evaluating liver fibrosis.
In a sixth aspect, the invention provides an application of a hepatitis B cirrhosis diagnosis nomogram in preparation of a medicine for preventing and treating cirrhosis.
Due to the adoption of the scheme, the invention has the beneficial effects that:
the invention creatively combines the hematology examination index and the 2D-SWE, is non-invasive and non-invasive examination, is easy to popularize, can continuously read and dynamically monitor the development condition of the liver cirrhosis of the chronic hepatitis B patient for many times, and can also be used for the assessment of the liver cirrhosis of the hepatitis B related liver cancer patient before operation. The model adopts a large sample independent verification queue to carry out external verification, has reliable results and is more persuasive.
Secondly, the invention incorporates a large sample amount (1115 cases) of hepatitis B patients to evaluate and verify the diagnosis model of liver cirrhosis, and reduces the influence of liver fibrosis stage and individual difference of patients on the diagnosis model as much as possible.
Thirdly, the pathological diagnosis sample used by the invention is from hepatectomy, more liver tissue samples can be obtained compared with liver puncture biopsy, and the cirrhosis condition of the patient can be evaluated more objectively.
Drawings
FIG. 1 is a statistical plot of the coefficients of variation (Binomial development is the Binomial deviation) of a Lasso model with 10-fold cross validation for identifying clinical features using a Lasso-binary logistic regression model in an embodiment of the present invention.
FIG. 2 is a statistical chart of Coefficients of clinical features (i.e., Lasso coefficient chart of clinical indices, total 26 clinical indices were subjected to Lasso regression, and 5 key variables (Coefficients are ordinate Coefficients) were determined by binary logistic regression) in an example of the present invention.
Fig. 3 is an alignment chart of diagnosis of hepatitis b cirrhosis according to an embodiment of the present invention, in which the numerical values of 5 non-invasive tests of a patient are added to the score corresponding to the top line to obtain a total score, and the score below the total score is the probability that the patient has cirrhosis.
FIG. 4 is a calibration curve analysis diagram of the modeling and verification sets according to an embodiment of the present invention.
FIG. 5 is a receiver operating characteristic curve (ROC) plot of the modeling and validation set models in an embodiment of the present invention.
Detailed Description
The invention provides a method for constructing a noninvasive hepatitis B cirrhosis diagnosis model and application of a diagnosis nomogram. The construction method of the hepatitis B cirrhosis diagnosis model extracts 26 indexes from hematology examination and imaging examination results as potential factors, screens variables through a Lasso algorithm, finally selects 5 nonzero-coefficient parameters to perform binary Logistic regression, establishes the hepatitis B cirrhosis diagnosis model (HLC) by taking operation pathological diagnosis as a gold standard, and draws a hepatitis B cirrhosis diagnosis nomogram for calculating the probability of the liver cirrhosis of clinical cases and assisting clinical diagnosis and treatment decision.
The construction method of the noninvasive hepatitis B cirrhosis diagnosis model has the following advantages:
(1) solves the problem of collinearity
A traditional diagnosis model adopts a linear Logistic regression method, a clinical outcome is used as a dependent variable, clinical information is used as an independent variable, and the diagnosis model is constructed. However, in the case of clinical examination indices, some indices are correlated, such as total bilirubin and direct bilirubin, and these indices are co-linear. The traditional single-factor screening-linear Logistic regression model construction can not effectively solve the problem of index collinearity, and the invention adopts the Lasso-Logistic model construction method to establish the hepatitis B cirrhosis diagnosis model for the first time, so that the collinearity index can be effectively eliminated, and the detection efficiency of the model is higher.
(2) Variable screening method is more optimized
Clinical test indexes are numerous. In practical application, it is necessary to determine which indexes play important roles and which indexes can be discarded, so as to simplify the model to the greatest extent. The traditional linear Logistic regression adopts a single-factor check variable method, takes a P value as a reference standard, and eliminates certain indexes too urgently. In fact, hepatitis B virus may not be different in single factor tests, for example, due to sample size, but is necessarily associated with cirrhosis, based on clinical experience. The variable screening of the linear Logistic regression loses some important information. Sparse estimation can be generated by screening variables through Lasso, and each variable put in the model is ensured to influence the result, so that the reliability of the model is improved.
(3) Visualization of diagnostic results
Compared with the prior art, the method establishes the hepatitis B cirrhosis diagnosis nomogram, scores are assigned to the value level of each influence factor according to the contribution degree (the size of a regression coefficient) of each influence factor in the model to the result variable, then the scores are added to obtain a total score, and finally the predicted value of the individual result event is calculated through the function conversion relation between the total score and the occurrence probability of the result event. The hepatitis B cirrhosis diagnosis nomogram converts a complex regression equation into a visual graph, so that the result of a diagnosis model is more readable, the evaluation of a patient is facilitated, and the diagnosis nomogram is closer to clinical application.
< method for constructing diagnosis model of hepatitis B cirrhosis >
The method for constructing the hepatitis B cirrhosis diagnosis model adopts the Lasso algorithm to extract the prediction factors from clinical data (such as hematology and imaging examination) of a training queue, and selects parameters with nonzero coefficients to perform binary Logistic regression by a multiple cross-validation method to obtain the hepatitis B cirrhosis diagnosis model.
Among them, type III procollagen (PIII-NP) +0.011 × IV collagen (IV-C) +0.003 × Hyaluronic Acid (HA) -0.013 × Platelet (PLT) +0.124 × liver hardness (LS) -2.387 were diagnosed as hepatitis b cirrhosis models.
Indices for hematological examinations include PIII-NP, IV-C, HA, and PLT; the imaging examination is the result of liver hardness measurement of two-dimensional shear wave elastography (2D-SWE).
As a quantitative tool for evaluating risks and benefits, the model can provide more objective and accurate information for decisions of doctors and patients. The hepatitis B cirrhosis diagnosis model is a clinical prediction model constructed based on clinical characteristics, imaging examination results and laboratory examination results. The traditional regression analysis for constructing a prediction model is only suitable for research with few independent variables, the model contains 26 clinical characteristics, and the variable selection is very difficult. While the Lasso algorithm can limit the regression coefficients, thereby solving the problem of multivariate screening. The invention screens 5 clinical characteristics of PIII-NP, IV-C, HA, PLT and LS through a Lasso algorithm, and further constructs a hepatitis B cirrhosis diagnosis model through Logistic regression analysis.
< model of diagnosis of hepatitis B cirrhosis >
The hepatitis B cirrhosis diagnosis model is obtained by the construction method.
< nomogram for diagnosing hepatitis B cirrhosis >
The invention makes the hepatitis B cirrhosis diagnosis model into a hepatitis B cirrhosis diagnosis nomogram which is more convenient to use clinically.
< application of nomogram for diagnosing hepatitis B cirrhosis >
The invention relates to an application of a hepatitis B cirrhosis diagnosis nomogram in preparation, evaluation and diagnosis of cirrhosis.
Wherein, 5 clinical common indexes PIII-NP, IV-C, HA, PLT and LS are found out corresponding scores in a hepatitis B cirrhosis diagnosis nomogram, and a total score is calculated, wherein the total score is corresponding to the probability that the patient suffers from cirrhosis.
The hepatitis B cirrhosis diagnosis nomogram of the invention can be applied to the preparation and the evaluation of hepatic fibrosis.
The nomogram for diagnosing hepatitis B cirrhosis can also be applied to the preparation of drugs for preventing and treating the cirrhosis.
The present invention will be further described with reference to the following examples.
Example (b):
in this example, 754 clinical data of patients accompanied by Chronic viral hepatitis B (CHB) were collected during hepatectomy in the secondary zhongshan hospital of the university of countermand between 7 months in 2015 and 4 months in 2017, and these cases were used as training cohorts to establish a noninvasive diagnosis model. 421 cases of hepatitis b disease were prospectively enrolled in the validation cohort for hepatoresection at the subsidiary zhongshan hospital of the university of compound denier during the period of 5 months in 2017 to 11 months in 2017. The 421 patients are subjected to hematology examination and 2D-SWE examination before hepatectomy, the model constructed by the embodiment is adopted to diagnose liver cirrhosis, and the result of the diagnosis model is compared with the result of pathological diagnosis of liver cirrhosis after operation, so that the diagnosis accuracy is evaluated.
The construction method of the embodiment extracts 26 potential parameters from the demographic and clinical data of the training queue, and adopts a regularization Lasso algorithm to reduce the coefficients of some parameters and even directly change some coefficients with smaller absolute values into 0, so that main characteristics are found out from a large number of clinical characteristics, and the generalization capability of the model is enhanced. According to the screening result of the Lasso algorithm, a total of 5 parameters with nonzero coefficients are selected for binary Logistic regression (shown in figure 1 and figure 2), and finally, a diagnosis model of hepatitis B related cirrhosis with PIII-NP, IV-C, HA, PLT and LS as prediction factors is determined. The diagnostic model for hepatitis B cirrhosis (HLC Index) of this example is as follows: HLC ═ 0.231 × PIII-NP +0.011 × IV-C +0.003 × HA-0.013 × PLT +0.124 × LS-2.387. Meanwhile, the reported noninvasive serological cirrhosis scoring model including APRI, FIB-4, King's score and Forns index was calculated using clinical data of training cohorts and laboratory examination data. Wherein APRI ═ AST (unit/liter (U/L))/AST normal reference upper limit (set to 40U/L) × 100/PLT (× 10)9L). Forns index 7.811-3.131 xln (PLT count (A))×109/L)) +0.781 xln (gamma-GT (U/L)) +3.467 xln (age (years)) -0.014 xln (cholesterol (mg/dL)); king's score ═ AST (U/L). times.INR/PLT (× 10)9L); FIB-4 is age (years) x AST (U/L/[ ALT (U/L))1/2×PLT(×109L). The model constructed in the embodiment is compared with other models.
In which fig. 1 shows the relation between the logarithm of λ and the mean square error as well as the number of variables in the model. The two vertical dashed lines in fig. 1 represent the logarithm of the minimum mean square error λ (left dashed line) and the logarithm of the standard error of the minimum distance λ (right dashed line). Two values of lambda can be chosen, the logarithm of the minimum mean square error lambda being the optimum value, and the logarithm of the standard error of the minimum distance lambda being a more compact model within one time of the mean square error. The present embodiment selects the log lambda value of the standard error of the minimum distance.
In fig. 2, as λ decreases, the compression parameter decreases and the absolute value of the coefficient increases.
In this embodiment, a hepatitis b cirrhosis diagnosis nomogram (fig. 3) is established based on the hepatitis b cirrhosis diagnosis model, and a quantitative and visual tool is provided for a clinician to diagnose cirrhosis. Each predictor was assigned a score based on the patient's demographic and clinical characteristics, and a total score for the likelihood of cirrhosis was calculated.
In fig. 3, the numerical values of 5 non-invasive tests of the patient are added to the score corresponding to the top line to obtain a total score, and the score below the total score is the probability that the patient has cirrhosis. The probability of cirrhosis in the case of FIG. 3 is 0.0441, indicating that the probability of cirrhosis in this case is extremely low.
For patients with hepatitis B who need to be diagnosed with liver cirrhosis or hepatitis B-related liver cancer who need to be evaluated for liver cirrhosis preoperatively, venous blood was drawn for PIII-NP, IV-C, HA, and PLT examinations, and LS was determined by 2D-SWE. The obtained detection values are mapped in the diagnostic nomogram of fig. 4 and mapped to the scores of the first row in the diagnostic nomogram, and the scores of the 5 indices are obtained, respectively. And calculating the sum of the 5 scores to obtain a total score. The total score of the patient corresponds to the last row, the obtained numerical value is the probability of the patient suffering from liver cirrhosis, and the calibration curve of the probability of liver cirrhosis in the training queue shows that the prediction and the observation result are well matched (figure 4), so that the embodiment can accurately diagnose the liver cirrhosis of the hepatitis B patient and provide guidance for prevention and treatment decisions.
In fig. 4, the abscissa represents the probability of being predicted to be liver cirrhosis, and the ordinate represents the probability of actually being liver cirrhosis. If the predicted probability matches the probability of a true event, the calibration curve is a straight line with a slope of 1.
The accuracy of the different cirrhosis diagnostic methods was assessed using the receiver operating characteristic curve (ROC) (fig. 5). Compared with the noninvasive diagnosis model of cirrhosis (2D-SWE, APRI, FIB-4, King's score and Forns index) reported in the literature, the diagnosis model of the embodiment has the best discrimination capability and diagnosis performance, and the AUC is the highest (modeling queue AUC is 0.866; validation queue AUC is 0.852).
Wherein, the larger the area under the curve in fig. 5, the higher the accuracy of the model in diagnosing liver cirrhosis. As can be seen from fig. 5, the diagnostic model of the present embodiment (curved solid line) has the largest area under the curve compared to other noninvasive diagnostic models of liver cirrhosis reported in the literature.
The previous description of the disclosed embodiments is provided to enable any person skilled in the art to make or use the present invention. It will be readily apparent to those skilled in the art that various modifications to these embodiments and the generic principles defined herein may be applied to other embodiments without the use of the inventive faculty. Therefore, the present invention is not limited to the above-described embodiments. Those skilled in the art should appreciate that many modifications and variations are possible in light of the above teaching without departing from the scope of the invention.

Claims (9)

1. A method for constructing a hepatitis B cirrhosis diagnosis model is characterized by comprising the following steps: extracting a prediction factor from clinical data of a training queue by adopting a Lasso algorithm, and selecting a parameter with a nonzero coefficient by a multiple cross-validation method to perform binary Logistic regression to obtain a hepatitis B cirrhosis diagnosis model;
the diagnostic model of hepatitis B cirrhosis is 0.231X type III procollagen + 0.011X type IV collagen + 0.003X hyaluronic acid-0.013X platelet + 0.124X liver hardness-2.387.
2. The method for constructing a diagnostic model of hepatitis B cirrhosis according to claim 1, characterized in that: the clinical data includes hematological and imaging examinations.
3. The method for constructing a diagnostic model of hepatitis B cirrhosis according to claim 2, characterized in that: the examination indexes of hematology comprise type III procollagen, type IV collagen, hyaluronic acid and platelets, and the imaging examination is a liver hardness value obtained by two-dimensional shear wave elastography.
4. A diagnostic model for hepatitis b cirrhosis, characterized by: which is obtained by the method of claim 1.
5. A nomogram for diagnosing hepatitis B cirrhosis, characterized in that: the diagnostic model of hepatitis B cirrhosis according to claim 4.
6. Use of the diagnostic nomogram for hepatitis B cirrhosis according to claim 5 for the preparation of an evaluation diagnostic for cirrhosis.
7. Use according to claim 6, characterized in that: finding out corresponding scores of 5 clinical common indexes of type III procollagen, type IV collagen, hyaluronic acid, platelet and liver hardness in a hepatitis B cirrhosis diagnosis nomogram, and calculating a total score, wherein the total score corresponds to the probability that the patient has cirrhosis.
8. Use of the hepatitis B cirrhosis diagnostic nomogram of claim 5 in the preparation of a diagnostic for assessing liver fibrosis.
9. The use of the nomogram for diagnosing hepatitis B cirrhosis according to claim 5 for the preparation of a medicament for the prevention and treatment of cirrhosis.
CN202111195517.7A 2021-10-14 2021-10-14 Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram Pending CN113936789A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111195517.7A CN113936789A (en) 2021-10-14 2021-10-14 Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111195517.7A CN113936789A (en) 2021-10-14 2021-10-14 Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram

Publications (1)

Publication Number Publication Date
CN113936789A true CN113936789A (en) 2022-01-14

Family

ID=79279404

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111195517.7A Pending CN113936789A (en) 2021-10-14 2021-10-14 Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram

Country Status (1)

Country Link
CN (1) CN113936789A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113436731A (en) * 2021-07-15 2021-09-24 王新兴 Liver hemodynamic detection method and system based on multiple hepatic vein oscillograms

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113436731A (en) * 2021-07-15 2021-09-24 王新兴 Liver hemodynamic detection method and system based on multiple hepatic vein oscillograms

Similar Documents

Publication Publication Date Title
Pavlov et al. Transient elastography for diagnosis of stages of hepatic fibrosis and cirrhosis in people with alcoholic liver disease
JP6185393B2 (en) A non-invasive method for assessing the presence or severity of liver fibrosis based on a novel detailed classification
CN112420196A (en) Prediction method and system for survival rate of acute myocardial infarction patient within 5 years
CN113327679A (en) Pulmonary embolism clinical risk and prognosis scoring method and system
CN112669960A (en) Liver fibrosis prediction model construction method based on machine learning method, prediction system, device and storage medium
CN109585011A (en) The Illnesses Diagnoses method and machine readable storage medium of chest pain patients
Kelleher et al. Assessment of liver fibrosis in co-infected patients
RU2583939C1 (en) Method for laboratory diagnosis of stages of hepatic fibrosis in chronic viral hepatitis c
CN113936789A (en) Method for constructing noninvasive hepatitis B cirrhosis diagnosis model and application of diagnosis nomogram
CN112837818B (en) Model for evaluating liver fibrosis degree of hepatitis B patient
CN115602319B (en) Noninvasive hepatic fibrosis assessment device
CN117497170A (en) Construction method and application of early warning model for converting acute kidney injury into chronic kidney disease
CN117079810A (en) Cardiovascular disease unscheduled re-hospitalization risk prediction method
Li et al. Development and validation of a routine blood parameters-based model for screening the occurrence of retinal detachment in high myopia in the context of PPPM
CN115240855A (en) Gastric cancer resection postoperative severe complication nursing risk prediction model and construction method thereof
CN114898859A (en) Acute aortic dissection hospital internal prognosis prediction system
CN110070942A (en) A kind of chronic liver disease risk evaluating system promoting tree-model based on gradient
TW202030741A (en) Liver fibrosis assessment model, liver fibrosis assessment system and liver fibrosis assessment method
Germing et al. Diastolic dysfunction without abnormalities in left atrial and left ventricular geometry does not affect quality of life in elderly women
CN117577330B (en) Device and storage medium for predicting liver fibrosis degree of nonalcoholic fatty liver disease
CN111986806B (en) Prediction model of acute myocardial infarction infection risk based on clinical-biological indexes
WO2021090909A1 (en) Heart failure marker
Lin et al. Comprehensive clinical application analysis of artificial intelligence-enabled electrocardiograms for screening multiple valvular heart diseases
CN114628026A (en) Apparatus for diagnosing HCM and apparatus for predicting HCM patient prognosis
CN116844718A (en) Community senile sarcopenia screening method based on arteriosclerosis index

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination