CN114141363B - Machine learning method-based severe pancreatitis prediction model construction method - Google Patents
Machine learning method-based severe pancreatitis prediction model construction method Download PDFInfo
- Publication number
- CN114141363B CN114141363B CN202111484319.2A CN202111484319A CN114141363B CN 114141363 B CN114141363 B CN 114141363B CN 202111484319 A CN202111484319 A CN 202111484319A CN 114141363 B CN114141363 B CN 114141363B
- Authority
- CN
- China
- Prior art keywords
- model
- data
- severe pancreatitis
- training
- prediction
- 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.)
- Active
Links
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
- 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
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
- Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
- Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Databases & Information Systems (AREA)
- Pathology (AREA)
- Epidemiology (AREA)
- General Health & Medical Sciences (AREA)
- Primary Health Care (AREA)
- Investigating Or Analysing Biological Materials (AREA)
Abstract
The invention discloses a machine learning method-based severe pancreatitis prediction model construction method, which comprises the following steps: acquiring clinical and laboratory index related data of a predicted object, and preprocessing the acquired data; removing the index of missing or little change; data is divided into boxes; screening a severe pancreatitis data prediction model; pre-training the selected model; screening and determining characteristic indexes associated with 10 severe pancreatitis according to the pre-training result, and training again; and determining a threshold value applicable to a prediction algorithm according to the working characteristic curve of the subject, and further obtaining a final severe pancreatitis data prediction model. The noninvasive diagnosis of severe pancreatitis model based on a plurality of clinical indexes solves the problem of diagnosis hysteresis, has high prediction speed and high accuracy, and can provide reference for clinical diagnosis of severe pancreatitis.
Description
Technical Field
The invention relates to the field of disease diagnosis, in particular to a machine learning method-based severe pancreatitis prediction model construction method.
Background
The existing severe pancreatitis comprises a scoring system for acute physiology and chronic health assessment or a scoring system for predicting the severity of acute pancreatitis (acute pancreatitis, AP) based on a single or a small number of laboratory indexes, and the existing laboratory technology scoring system has complex evaluation indexes such as acute physiology and chronic health assessment (Acute Physiology and Chronic Health Evaluation, APACHE II), bedside index of AP severity (Bedside Index for Severity in Acute Pancreatitis, BISAP) and the like, and has high working pressure of doctors and subjective evaluation. In addition, the severity of the AP can be further determined only by more than 48 hours or even 72 hours after the disease, so that the disease treatment diagnosis is seriously delayed, and the optimal time of treatment is influenced; moreover, obtaining multiple pieces of laboratory diagnostic technical data from different medical centers and different laboratories requires interpretation by a clinician, and has the defects of general repeatability and subjective bias.
Disclosure of Invention
Aiming at the problems, the invention provides a machine learning method-based severe pancreatitis prediction model construction method, which has the advantages of shortening diagnosis time and obtaining precious time for early stage grading diagnosis prediction.
The technical scheme of the invention is as follows:
a method for constructing a severe pancreatitis prediction model based on a machine learning method comprises the following construction steps:
s1, acquiring a prediction object discharge diagnosis result and clinically relevant data, and preprocessing the acquired data;
s2, screening the preprocessed data, carrying out data binning, and combining the binned data with the data which are not binned;
s3, establishing a plurality of severe pancreatitis prediction models through the data after the box division and the data after the box not division, comparing and selecting excellent models, and simultaneously constructing a predicted subject working characteristic curve;
s4, pre-training the selected model, screening 10 characteristic indexes with the largest model contribution degree, and re-training the 10 characteristic indexes with the largest model contribution degree screened by the pre-training model;
s5, confirming a threshold range for evaluating the severe pancreatitis and non-severe pancreatitis models according to the maximum index sensitivity and the specificity of the working characteristic curve of the test subject, obtaining a parameter range of a final model according to the confirmed threshold, and accordingly determining a model with fixed final parameters and predicting.
In S1, the discharge diagnosis result and the clinically relevant data include basic information of the patient, the discharge diagnosis result of the patient, and laboratory examination data.
In the step S1, the data preprocessing method is as follows;
1) And (3) carrying out numerical variable normalization processing, wherein the normalization formula is as follows:
Xnnormalized＝(Xn–Xmin)/(Xmax–Xmin)；
xn represents any numerical variable, xnndormalized represents the normalized value of the numerical variable Xn, xmax represents the maximum value of the numerical variable, xmin represents the minimum value of the numerical variable;
4) Recording category type variables, including urine color and clarity, encoded by severity;
5) Indicators of deletion or subtle changes include (1) deletion of features with deletion values exceeding 90%; (2) and removing the characteristic that the value is unchanged or 95% of the value is unchanged.
In the step S2, the data is divided into boxes by taking variables with the number of numerical values or class values not exceeding 10 as non-box-dividing data, and the other variables are equally divided into boxes by data, wherein the number of the boxes is 12.
In the step S3, the method for screening the severe pancreatitis data prediction model is as follows:
(1) Combining the binned data and the non-binned data to form a dataset 1;
(2) Filling missing values in the data set 1 by adopting a KNN algorithm to form a data set 2;
(3) The indexes screened by the data set 2 are further finely screened by lasso regression to obtain a data set 3;
(4) Dividing a training set and a testing set by using the data set 3;
(5) Inputting the training sample set into 8 types of to-be-selected prediction models to finish training of the to-be-selected models;
(6) Inputting the test sample set into 8 types of to-be-selected prediction models, and outputting probability values predicted by the to-be-selected models;
(7) Establishing 8 prediction subject work characteristic curves ROC according to probability values predicted by the model to be selected;
(8) Comparing the area AUC under the ROC curve, and selecting the optimal model as a distributed gradient lifting classifier (LightGBM).
The 8 prediction models to be selected are constructed by the following methods of (1) a logistic regression classifier LR, (2) a Gaussian distribution naive Bayesian classifier GNB, (3) a polynomial distribution naive Bayesian classifier MNB, (4) a support vector machine classifier SVC, (5)K) a neighbor classifier KNN, (6) a decision tree classifier DTC, (7) a random forest classifier RF, and (8) a distributed gradient lifting classifier LightGBM.
In the step S4, the 10 feature indexes are specifically:
(1) Dividing a training set and a testing set by using the data set 1;
(2) Establishing a severe pancreatitis prediction model by using a LightGBM method, and setting model parameters, wherein the parameters comprise a learning rate, iteration times, a maximum depth of a tree model, the number of leaf nodes, a sub-sampling ratio, a characteristic sampling ratio, an L1 regularization parameter and an L2 regularization parameter;
(3) Inputting a training sample set into a model, and pre-training the model;
(4) Inputting the test sample set into the model, and outputting a probability value predicted by the model;
(5) Adjusting the model parameters in the step (2) with the aim of obtaining the maximum AUC value;
(6) And screening out 10 characteristic indexes with the largest contribution degree to the model at the maximum AUC value.
The retraining method comprises the following steps:
(1) The 10 columns with the greatest contribution in the data set 1 are taken as the data set 4
(2) Dividing a training set and a testing set by using the data set 4;
(3) Establishing a severe pancreatitis prediction model by using a LightGBM method, and setting model parameters, wherein the parameters comprise a learning rate, iteration times, a maximum depth of a tree model, the number of leaf nodes, a sub-sampling ratio, a characteristic sampling ratio, an L1 regularization parameter and an L2 regularization parameter;
(4) Inputting a training sample set into the model, and training the model;
(5) Inputting the test sample set into the model, and outputting a probability value predicted by the model;
and (3) aiming at obtaining the maximum AUC value, adjusting the model parameters in the step (2) until the AUC value is not increased any more and modeling is completed.
The beneficial effects of the invention are as follows:
acquiring clinical and laboratory index related data of a predicted object, and preprocessing the acquired data; removing the index of missing or little change; data is divided into boxes; screening a severe pancreatitis data prediction model; pre-training the selected model; screening and determining characteristic indexes associated with 10 severe pancreatitis according to the pre-training result, and training again; and determining a threshold value applicable to a prediction algorithm according to the working characteristic curve of the subject, and further obtaining a final severe pancreatitis data prediction model. The noninvasive diagnosis of severe pancreatitis model based on a plurality of clinical indexes solves the problem of diagnosis hysteresis, has high prediction speed and high accuracy, and can provide reference for clinical diagnosis of severe pancreatitis.
Drawings
FIG. 1 is a schematic overall flow chart of a prediction model construction method according to an embodiment of the present invention;
fig. 2 is a schematic diagram of screening a model to be selected among 8 types of prediction models to be selected according to an embodiment of the present invention.
FIG. 3 shows the results obtained after the training method of the present invention for screening out 10 feature indexes with the largest model contribution.
Detailed Description
Embodiments of the present invention are further described below with reference to the accompanying drawings.
Examples:
as shown in fig. 1-2, a machine learning method-based severe pancreatitis prediction model construction method comprises the following construction steps:
s1, acquiring a prediction object discharge diagnosis result and clinically relevant data, and preprocessing the acquired data;
s2, screening the preprocessed data, carrying out data binning, and combining the binned data with the data which are not binned;
s3, establishing a plurality of severe pancreatitis prediction models through the data after the box division and the data after the box not division, comparing and selecting excellent models, and simultaneously constructing a predicted subject working characteristic curve;
s4, pre-training the selected model, screening 10 characteristic indexes with the largest model contribution degree, and re-training the 10 characteristic indexes with the largest model contribution degree screened by the pre-training model;
s5, confirming a threshold range for evaluating the severe pancreatitis and non-severe pancreatitis models according to the maximum index sensitivity and the specificity of the working characteristic curve of the test subject, obtaining a parameter range of a final model according to the confirmed threshold, and accordingly determining a model with fixed final parameters and predicting.
In S1, the discharge diagnosis result and the clinically relevant data include basic information of the patient, the discharge diagnosis result of the patient, and laboratory examination data.
In S1, the data preprocessing method is as follows;
1) And (3) carrying out numerical variable normalization processing, wherein the normalization formula is as follows:
Xnnormalized＝(Xn–Xmin)/(Xmax–Xmin)；
xn represents any numerical variable, xnndormalized represents the normalized value of the numerical variable Xn, xmax represents the maximum value of the numerical variable, xmin represents the minimum value of the numerical variable;
6) Recording category type variables, including urine color and transparency, encoded by severity, for example, urine color: [ 'colorless': 0, 'pale yellow': 1, 'pale yellow': 2, 'yellow': 3, 'deep yellow': 4, 'orange': 5, 'deep orange': 6, 'brown': 7, 'brown': 8], the basic information of the patient includes age, sex; the patient discharge diagnosis results comprise whether severe pancreatitis exists or not, and malignant tumors are excluded from the past history of the following disease cases (1) in the report; (2) During the recent period (3 months before image examination), the pelvic cavity and abdominal cavity are subjected to major surgery; (3) infectious diseases; laboratory test data include aspartate aminotransferase, alanine aminotransferase, alkaline phosphatase, glutamyl transpeptidase, adenosine deaminase, a-L-fucosidase, 5-nucleotidase, cholinesterase, homoamino aminopeptidase, prealbumin, total protein, albumin, globulin, albumin: globulin ratio, total bile acid, total bilirubin, direct bilirubin, indirect bilirubin, triglycerides, total cholesterol, high density lipoprotein cholesterol, low density lipoprotein cholesterol, very low density lipoprotein cholesterol, potassium, sodium, chlorine, total calcium, total carbon dioxide, magnesium, apolipoprotein A1, apolipoprotein B100, urea, creatinine, glomerular filtration rate estimates, uric acid, (urine) glucose, inorganic phosphorus, lactic acid, cystatin C, lactate deaminase, lipoprotein (a), hypersensitive C-reactive protein, homocysteine, pancreatic amylase, glycosylated hemoglobin, (blood) glucose, average blood glucose, total glycosylated hemoglobin, procalcitonin, leucocytes, neutrophil absolute, lymphocyte absolute, monocyte absolute, eosinophil absolute, basophil absolute, neutrophil percentage, lymphocyte percentage, monocyte percentage, eosinophil percentage, basophil percentage, erythrocytes, hemoglobin, hematocrit, mean erythrocytes volume, mean erythrocytes hemoglobin content, mean erythrocytes hemoglobin concentration, erythrocytes distribution width variation, platelets, mean thrombocyte volume, thrombocyte hematocrit, thrombocyte distribution width, urine body fluid appearance, urine color, blood color, clarity, urine chemistry analysis, alkalinity, specific gravity, nitrite, protein, bilirubin, urobilinogen, occult blood, ketone body, leukocyte esterase, quantitative analysis of urine sediment, leukocyte count, erythrocyte count, upper branch cell count, crystallization count, budding yeast, transparent tube count, pathological tube count, microscopic examination of urine sediment;
7) Indicators of deletion or subtle changes include (1) deletion of features with deletion values exceeding 90%; (2) and removing the characteristic that the value is unchanged or 95% of the value is unchanged.
In S2, the data is divided into the data without dividing the data into boxes by using the variables with the number of numerical values or class values not more than 10, and the number of the data divided into boxes is 12 by using other variables at equal intervals.
In S3, the method for screening the severe pancreatitis data prediction model is as follows:
(1) Combining the binned data and the non-binned data to form a dataset 1;
(2) Filling missing values in the data set 1 by adopting a KNN algorithm to form a data set 2;
(3) The indexes screened by the data set 2 are further finely screened by lasso regression to obtain a data set 3;
(4) Dividing a training set and a testing set by using the data set 3;
(5) Inputting the training sample set into 8 types of to-be-selected prediction models to finish training of the to-be-selected models;
(6) Inputting the test sample set into 8 types of to-be-selected prediction models, and outputting probability values predicted by the to-be-selected models;
(7) Establishing 8 prediction subject work characteristic curves ROC according to probability values predicted by the model to be selected;
(8) Comparing the area AUC under the ROC curve, and selecting the optimal model as a distributed gradient lifting classifier (LightGBM).
The 8 prediction models to be selected are constructed by the following methods of (1) a logistic regression classifier LR, (2) a Gaussian distribution naive Bayesian classifier GNB, (3) a polynomial distribution naive Bayesian classifier MNB, (4) a support vector machine classifier SVC, (5)K) a neighbor classifier KNN, (6) a decision tree classifier DTC, (7) a random forest classifier RF, and (8) a distributed gradient lifting classifier LightGBM.
In S4, the 10 feature indexes specifically are:
(1) Dividing a training set and a testing set by using the data set 1;
(2) Establishing a severe pancreatitis prediction model by using a LightGBM method, and setting model parameters, wherein the parameters comprise a learning rate, iteration times, a maximum depth of a tree model, the number of leaf nodes, a sub-sampling ratio, a characteristic sampling ratio, an L1 regularization parameter and an L2 regularization parameter;
(3) Inputting a training sample set into a model, and pre-training the model;
(4) Inputting the test sample set into the model, and outputting a probability value predicted by the model;
(5) Adjusting the model parameters in the step (2) with the aim of obtaining the maximum AUC value;
(6) Screening out 10 characteristic indexes with the largest contribution degree to the model at the maximum AUC value, wherein the indexes are as follows:
the retraining method comprises the following steps:
(1) Taking out 10 columns with the largest contribution degree in the data set 1 as a data set 4;
(2) Dividing a training set and a testing set by using the data set 4;
(3) Establishing a severe pancreatitis prediction model by using a LightGBM method, and setting model parameters, wherein the parameters comprise a learning rate, iteration times, a maximum depth of a tree model, the number of leaf nodes, a sub-sampling ratio, a characteristic sampling ratio, an L1 regularization parameter and an L2 regularization parameter;
(4) Inputting a training sample set into the model, and training the model;
(5) Inputting the test sample set into the model, and outputting a probability value predicted by the model;
and (3) aiming at obtaining the maximum AUC value, adjusting the model parameters in the step (2) until the AUC value is not increased any more and modeling is completed.
The foregoing examples merely illustrate specific embodiments of the invention, which are described in greater detail and are not to be construed as limiting the scope of the invention. It should be noted that it will be apparent to those skilled in the art that several variations and modifications can be made without departing from the spirit of the invention, which are all within the scope of the invention.
Claims (8)
1. The method for constructing the severe pancreatitis prediction model based on the machine learning method is characterized by comprising the following steps of:
s1, acquiring a prediction object discharge diagnosis result and clinically relevant data, and preprocessing the acquired data;
s2, screening the preprocessed data, carrying out data binning, and combining the binned data with the data which are not binned; the data box division is carried out by taking variables with the number of numerical values or class values not exceeding 10 as non-box division data, carrying out data equidistant box division on other variables, and the number of the box division is 12;
s3, establishing a plurality of severe pancreatitis prediction models through the data after the box division and the data after the box not division, comparing and selecting excellent models, and simultaneously constructing a predicted subject working characteristic curve;
s4, pre-training the selected model, screening 10 characteristic indexes with the largest model contribution degree, and re-training the 10 characteristic indexes with the largest model contribution degree screened by the pre-training model;
s5, confirming a threshold range for evaluating the severe pancreatitis and non-severe pancreatitis models according to the maximum index sensitivity and the specificity of the working characteristic curve of the test subject, obtaining a parameter range of a final model according to the confirmed threshold, and accordingly determining a model with fixed final parameters and predicting.
2. The method for constructing a machine learning method-based severe pancreatitis prediction model according to claim 1, wherein in S1, the discharge diagnosis result and the clinically relevant data include basic information of the patient, the discharge diagnosis result of the patient, and laboratory test data.
3. The method for constructing a machine learning method-based severe pancreatitis prediction model according to claim 2, wherein in S1, the data preprocessing method is as follows;
1) And (3) carrying out numerical variable normalization processing, wherein the normalization formula is as follows:
Xnnormalized＝(Xn–Xmin)/(Xmax–Xmin)；
xn represents any numerical variable, xnndormalized represents the normalized value of the numerical variable Xn, xmax represents the maximum value of the numerical variable, xmin represents the minimum value of the numerical variable;
2) Recording category type variables, including urine color and clarity, encoded by severity;
3) Indicators of deletion or subtle changes include (1) deletion of features with deletion values exceeding 90%; (2) and removing the characteristic that the value is unchanged or 95% of the value is unchanged.
4. The method for constructing the severe pancreatitis prediction model based on the machine learning method according to claim 1, wherein in the step S2, the data is divided into boxes by equal distance, the number of the boxes is 12, and the number of the boxes is not divided by using variables with the number of the occurrences of no more than 10 of numerical values or class values as non-box-dividing data.
5. The method for constructing a prediction model of severe pancreatitis based on machine learning method according to claim 1, wherein in S3, the method for screening the prediction model of severe pancreatitis data is as follows:
(1) Combining the binned data and the non-binned data to form a dataset 1;
(2) Filling missing values in the data set 1 by adopting a KNN algorithm to form a data set 2;
(3) The indexes screened by the data set 2 are further finely screened by lasso regression to obtain a data set 3;
(4) Dividing a training set and a testing set by using the data set 3;
(5) Inputting the training sample set into 8 types of to-be-selected prediction models to finish training of the to-be-selected models;
(6) Inputting the test sample set into 8 types of to-be-selected prediction models, and outputting probability values predicted by the to-be-selected models;
(7) Establishing 8 prediction subject work characteristic curves ROC according to probability values predicted by the model to be selected;
(8) Comparing the area AUC under the ROC curve, and selecting the optimal model as a distributed gradient lifting classifier (LightGBM).
6. The machine learning method-based severe pancreatitis prediction model construction method according to claim 5, wherein the 8 candidate prediction models are constructed by the following methods of (1) a logistic regression classifier LR, (2) a gaussian distribution naive bayes classifier GNB, (3) a polynomial distribution naive bayes classifier MNB, (4) a support vector machine classifier SVC, (5)K neighbor classifier KNN, (6) a decision tree classifier DTC, (7) a random forest classifier RF, and (8) a distributed gradient boost classifier LightGBM.
7. The machine learning method-based severe pancreatitis prediction model construction method according to claim 1, wherein in S4, 10 characteristic indexes are specifically:
(1) Dividing a training set and a testing set by using the data set 1;
(2) Establishing a severe pancreatitis prediction model by using a LightGBM method, and setting model parameters, wherein the parameters comprise a learning rate, iteration times, a maximum depth of a tree model, the number of leaf nodes, a sub-sampling ratio, a characteristic sampling ratio, an L1 regularization parameter and an L2 regularization parameter;
(3) Inputting a training sample set into a model, and pre-training the model;
(4) Inputting the test sample set into the model, and outputting a probability value predicted by the model;
(5) Adjusting the model parameters in the step (2) with the aim of obtaining the maximum AUC value;
(6) And screening out 10 characteristic indexes with the largest contribution degree to the model at the maximum AUC value.
8. The machine learning method-based severe pancreatitis prediction model construction method according to claim 7, wherein the retraining method is as follows:
(1) The 10 columns with the greatest contribution in the data set 1 are taken as the data set 4
(2) Dividing a training set and a testing set by using the data set 4;
(3) Establishing a severe pancreatitis prediction model by using a LightGBM method, and setting model parameters, wherein the parameters comprise a learning rate, iteration times, a maximum depth of a tree model, the number of leaf nodes, a sub-sampling ratio, a characteristic sampling ratio, an L1 regularization parameter and an L2 regularization parameter;
(4) Inputting a training sample set into the model, and training the model;
(5) Inputting the test sample set into the model, and outputting a probability value predicted by the model;
and (3) aiming at obtaining the maximum AUC value, adjusting the model parameters in the step (2) until the AUC value is not increased any more and modeling is completed.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111484319.2A CN114141363B (en) | 2021-12-07 | 2021-12-07 | Machine learning method-based severe pancreatitis prediction model construction method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202111484319.2A CN114141363B (en) | 2021-12-07 | 2021-12-07 | Machine learning method-based severe pancreatitis prediction model construction method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114141363A CN114141363A (en) | 2022-03-04 |
| CN114141363B true CN114141363B (en) | 2023-09-12 |
Family
ID=80384395
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202111484319.2A Active CN114141363B (en) | 2021-12-07 | 2021-12-07 | Machine learning method-based severe pancreatitis prediction model construction method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114141363B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115240854B (en) * | 2022-07-29 | 2023-10-03 | 中国医学科学院北京协和医院 | Pancreatitis prognosis data processing method and system |
| CN115331819A (en) * | 2022-07-29 | 2022-11-11 | 中国医学科学院北京协和医院 | Pancreatitis prognosis data processing method and system based on artificial intelligence |
| CN116580847B (en) * | 2023-07-14 | 2023-11-28 | 天津医科大学总医院 | Method and system for predicting prognosis of septic shock |
| CN116741384B (en) * | 2023-08-14 | 2023-11-21 | 惠民县人民医院 | Bedside care-based severe acute pancreatitis clinical data management method |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096632A (en) * | 2016-06-02 | 2016-11-09 | 哈尔滨工业大学 | Based on degree of depth study and the ventricular function index prediction method of MRI image |
| US10599984B1 (en) * | 2018-03-20 | 2020-03-24 | Verily Life Sciences Llc | Validating a machine learning model after deployment |
| CN112669960A (en) * | 2020-12-31 | 2021-04-16 | 鲁小杰 | Liver fibrosis prediction model construction method based on machine learning method, prediction system, device and storage medium |
| CN113160994A (en) * | 2020-12-31 | 2021-07-23 | 南京鼓楼医院 | Construction method, prediction system, device and storage medium for noninvasive screening of non-alcoholic steatohepatitis model |
| CN113555122A (en) * | 2021-07-21 | 2021-10-26 | 中南大学湘雅二医院 | Prediction model of transient and persistent AKI of severe sepsis patient and construction method thereof |
-
2021
- 2021-12-07 CN CN202111484319.2A patent/CN114141363B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106096632A (en) * | 2016-06-02 | 2016-11-09 | 哈尔滨工业大学 | Based on degree of depth study and the ventricular function index prediction method of MRI image |
| US10599984B1 (en) * | 2018-03-20 | 2020-03-24 | Verily Life Sciences Llc | Validating a machine learning model after deployment |
| CN112669960A (en) * | 2020-12-31 | 2021-04-16 | 鲁小杰 | Liver fibrosis prediction model construction method based on machine learning method, prediction system, device and storage medium |
| CN113160994A (en) * | 2020-12-31 | 2021-07-23 | 南京鼓楼医院 | Construction method, prediction system, device and storage medium for noninvasive screening of non-alcoholic steatohepatitis model |
| CN113555122A (en) * | 2021-07-21 | 2021-10-26 | 中南大学湘雅二医院 | Prediction model of transient and persistent AKI of severe sepsis patient and construction method thereof |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114141363A (en) | 2022-03-04 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN114141363B (en) | Machine learning method-based severe pancreatitis prediction model construction method | |
| CN111261282A (en) | Sepsis early prediction method based on machine learning | |
| CN106202968B (en) | Cancer data analysis method and device | |
| CN112991320B (en) | System and method for predicting hematoma expansion risk of cerebral hemorrhage patient | |
| CN114023441A (en) | Severe AKI early risk assessment model and device based on interpretable machine learning model and development method thereof | |
| CN111081381B (en) | Intelligent screening method for critical indexes of prediction of hospital fatal alimentary canal re-bleeding | |
| CN114220540A (en) | Construction method and application of diabetic nephropathy risk prediction model | |
| CN111312401A (en) | After-physical-examination chronic disease prognosis system based on multi-label learning | |
| Farajollahi et al. | Diabetes diagnosis using machine learning | |
| CN115099331A (en) | Auxiliary diagnosis system for malignant pleural effusion based on interpretable machine learning algorithm | |
| CN109585011A (en) | The Illnesses Diagnoses method and machine readable storage medium of chest pain patients | |
| CN115410716A (en) | Method and system for establishing neural function prognosis prediction model of cerebral hemorrhage patient | |
| CN113128654A (en) | Improved random forest model for coronary heart disease pre-diagnosis and pre-diagnosis system thereof | |
| CN113571176A (en) | Depression identification method based on blood routine test data | |
| CN117612703A (en) | Diabetes retinopathy classification method based on medical test indexes | |
| CN111175480A (en) | Method for calculating gender and age by blood biochemical indexes | |
| CN115910360A (en) | Smog disease risk prediction model construction method | |
| CN113782197B (en) | New coronary pneumonia patient outcome prediction method based on interpretable machine learning algorithm | |
| CN116091484A (en) | Method and system for predicting Ki-67 expression level of lung cancer | |
| CN115240852A (en) | Method and system for establishing 90-day mortality risk prediction model of cerebral hemorrhage patient | |
| Umut et al. | Prediction of sepsis disease by Artificial Neural Networks | |
| CA3193886A1 (en) | System and method for generating augmented complete blood count reports | |
| CN115064267B (en) | Biliary tract occlusion risk assessment system and establishment method thereof | |
| Meyer et al. | A Supervised Machine Learning Approach with Feature Selection for Sex-Specific Biomarker Prediction | |
| Silahtaroğlu et al. | An early prediction and diagnosis of sepsis in intensive care units: An unsupervised machine learning model |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |