Inpatient child CRT risk early warning system and method
By collecting and processing health information of hospitalized children, using artificial intelligence and logistic regression models to automatically evaluate their CRT risk levels, the problems of low evaluation efficiency and high missed diagnosis in the existing technology are solved, and more accurate and efficient risk assessment is achieved.
Patent Information
- Application Number
- CN202411946451.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-27
- Publication Date
- 2025-04-18
AI Technical Summary
The prior art is difficult to effectively evaluate and manage the risk of central venous catheter-associated thrombosis (CRT) in hospitalized children, resulting in low evaluation efficiency and high missed diagnosis rate.
By collecting and processing health information of hospitalized children, using artificial intelligence semantic analysis and logistic regression models, their CRT risk levels are automatically evaluated, and visual risk levels and comprehensive prevention information are output.
A more comprehensive and accurate assessment of CRT risks for hospitalized children has been achieved, significantly reducing the time required for medical staff to evaluate and improving the evaluation efficiency.
Smart Images
Figure CN119851946A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of medical risk management, and in particular to a CRT risk early warning system and method for hospitalized children. Background Art
[0002] The current status of the use of central venous access devices (CVADs) in hospitalized children and the challenges they bring are as follows: About 25% of hospitalized children have central venous access devices implanted in their bodies, such as central venous catheters, peripherally inserted central venous catheters, and infusion tubes, which can provide safe and reliable vascular access for children's treatment. Although the use of CVAD has greatly improved the quality of medical care for hospitalized children, it can lead to central venous access device-related thrombosis (CRT), which is the presence of occlusive or non-occlusive thrombosis in at least one blood vessel caused by CVAD placement, as diagnosed by venography or vascular venous ultrasound.
[0003] Studies using prospective screening techniques (i.e., in the absence of clinical symptoms) have shown that the rate of CRT formation in hospitalized pediatric patients is as high as 66.67%, and in CRT, asymptomatic DVT is more common. Specifically, CRT in children can lead to serious acute and chronic complications. In terms of acute complications, CRT can cause sepsis and loss of important venous access in children. In severe cases, life-threatening pulmonary embolism or superior vena cava syndrome may occur; in terms of chronic complications, it can lead to recurrent thrombosis, post-phlebitis syndrome, complications of anticoagulant therapy, post-thrombotic syndrome, and even loss of limb function in children. It can be seen that it is particularly important to strengthen the management of high-risk children with CRT.
[0004] Currently, medical staff mainly rely on their own medical experience to analyze the physical examination results of hospitalized children to assess their CRT risk. However, if the evaluation is based solely on the experience of medical staff, as the number of hospitalized children increases, they will spend more time on CRT risk assessment, resulting in slower evaluation results for hospitalized children. Therefore, relying entirely on the experience of medical staff not only increases their workload, but also reduces the efficiency of evaluation. In addition, clinical medical staff often only conduct relevant examinations when the child shows clinical symptoms; due to lack of experience, even if the child shows symptoms, they may not be able to make accurate judgments and take measures in time, resulting in a very high rate of missed diagnosis of CRT in children.
[0005] Therefore, it is crucial to develop a CRT risk warning system for hospitalized children. Summary of the invention
[0006] The technical problem to be solved by the present invention is how to automatically evaluate the CRT risk of a child to be evaluated using the health information of the child to be evaluated.
[0007] In order to solve the above technical problems, the present invention provides a CRT risk warning method for hospitalized children, the method comprising the following steps: S1. Data collection: collecting identification information and health data of hospitalized children; establishing a first corresponding relationship between the collected identification information and the health data; storing the identification information and the health data in a CRT database according to the first corresponding relationship; S2, data processing: acquiring health data corresponding to the identification information from the CRT database, extracting original CRT feature data matching the CRT keyword from the health data, and converting the data into structured target CRT feature data; establishing a second corresponding relationship between the identification information and the target CRT feature data; storing the identification information and the target CRT feature data in the CRT database according to the second corresponding relationship; S3, data analysis: acquiring the target CRT feature data corresponding to the identification information from the CRT database, and acquiring the evaluation score of each feature data in the target CRT feature data according to the CRT risk assessment strategy, superimposing the evaluation scores to obtain the total evaluation score of the target CRT feature data, and acquiring the corresponding CRT risk level according to the total evaluation score; S4. Data display: Output a visualized CRT risk level and comprehensive prevention information of the child to be evaluated.
[0008] Furthermore, the health data in S1 includes data of on-site health assessment of the child to be assessed, existing physical examination result data and / or past medical record data; wherein the existing physical examination result data and / or past medical record data come from at least one of the HIS system, hand numbness system, medical order system, testing and examination system, nursing management system and electronic medical record system.
[0009] Furthermore, the extracting of original CRT feature data matching the CRT keyword from the health data in S2 and converting the data into structured target CRT feature data includes: Extracting all feature data matching the CRT keyword from the health data; Using an artificial intelligence semantic analysis system to perform semantic analysis on all the feature data; Feature data whose semantic analysis results meet the CRT feature requirements are screened out to serve as the structured target CRT feature data that matches the CRT keyword.
[0010] Furthermore, in S3, according to the CRT risk assessment strategy, an assessment score of each feature data in the target CRT feature data is obtained, including: S31. Construct a logistic regression model for the risk of CRT in hospitalized children, and use univariate logistic regression analysis to screen the target CRT characteristic data; S31, bringing the target CRT characteristic data into multivariate logistic regression analysis to obtain independent risk factors for CRT risk and corresponding scoring models; S32. Assign the assessment score to each of the independent risk factors based on the regression coefficient β in the logistic regression model for the risk of CRT in hospitalized children.
[0011] Furthermore, the independent risk factors include patient-related risk factors, treatment-related risk factors and CVAD-related risk factors.
[0012] Furthermore, the step of allocating the evaluation score to each of the independent risk factors in S32 includes: The distance between the grouping of the independent risk factors and the set basic risk reference value is recorded as D, which satisfies the formula:
[0013] in, is the reference value for each group of independent risk factors; is the basic risk reference value; is the regression coefficient; The score corresponding to each group of the independent risk factors Satisfy the formula:
[0014] in, A constant corresponding to the set evaluation score unit score.
[0015] Furthermore, the CRT risk levels include high risk, medium risk and low risk, where: The comprehensive prevention information for the low-risk and medium-risk patients includes health education, basic prevention and flushing / sealing; The comprehensive prevention information for the high risk includes bleeding risk assessment, which includes low risk and high risk.
[0016] Furthermore, the bleeding risk assessment includes: Extracting original bleeding risk feature data from the health data and converting it into structured target bleeding risk feature data; Establishing a third corresponding relationship between the identification information and the target bleeding risk characteristic data; storing the identification information and the target bleeding risk characteristic data in the CRT database according to the third corresponding relationship; According to the bleeding risk assessment strategy, obtaining an assessment result of each feature data in the target bleeding risk feature data; The bleeding risk level of the person to be evaluated is obtained according to the evaluation result of each characteristic data in the target bleeding risk characteristic data.
[0017] Furthermore, the bleeding risk assessment strategy includes: Constructing a judgment rule for each characteristic data of the target bleeding risk characteristic data; It is determined whether the evaluation rule corresponding to each characteristic data of the target bleeding risk characteristic data is triggered to obtain the corresponding bleeding risk assessment result.
[0018] A CRT risk warning system for hospitalized children, using the CRT risk warning method for hospitalized children, comprises: A data collection module, used to collect identification information and health data of the person to be evaluated; establish a first corresponding relationship between the identification information and the health data; and store the identification information and the health data in a CRT database according to the first corresponding relationship; A data processing module, used for extracting target CRT feature data matching the CRT keyword from the health data; A data analysis module, according to a CRT risk assessment strategy, obtains an assessment score of each feature data in the target CRT feature data, and superimposes the total assessment score of the target CRT feature data to obtain a corresponding CRT risk level; and A data display module is connected to the output end of the data analysis module and is used to output a visualized CRT risk level and comprehensive prevention information of the person to be evaluated.
[0019] Compared with the prior art, the present invention has the following beneficial effects: The present invention extracts CRT characteristic data from the health information of the child to be evaluated, and evaluates the risk level of CRT in hospitalized children based on these data. Since the data analyzed by the system comes from comprehensive health information rather than relying solely on physical examination results, it can more comprehensively evaluate the risk of CRT in hospitalized children; in addition, the evaluation is performed using a computer system, which can significantly reduce the time required for medical staff to evaluate CRT risks, making the evaluation process faster and more accurate. BRIEF DESCRIPTION OF THE DRAWINGS
[0020] Figure 1 The figure is a schematic diagram of the overall process disclosed in the embodiment of the present invention. DETAILED DESCRIPTION
[0021] In order to make the technical scheme and technical effect of the present invention clearer, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments.
[0022] The first aspect of the present invention is to provide a CRT risk warning method for hospitalized children, so as to automatically assess the CRT risk of the child to be assessed using the health information of the child to be assessed.
[0023] See also Figure 1 , the early warning method specifically includes the following steps: 1. Data Collection Collect identification information and health data of hospitalized children; establish a first corresponding relationship between the collected identification information and health data; and store the identification information and health data in a CRT database according to the first corresponding relationship.
[0024] In a further scheme, the health data includes the data of on-site health assessment of the child to be assessed, existing physical examination result data and / or past medical record data; wherein the existing physical examination result data and / or past medical record data come from at least one of the HIS system, hand numbness system, medical order system, testing and examination system, nursing management system and electronic medical record system.
[0025] 2. Data processing Acquire health data corresponding to the identification information from the CRT database, extract original CRT feature data matching the CRT keywords from the health data, and convert them into structured target CRT feature data; establish a second corresponding relationship between the identification information and the target CRT feature data; store the identification information and the target CRT feature data into the CRT database according to the second corresponding relationship.
[0026] In a further solution, the CRT keywords are as follows:
[0027] In a further solution, all feature data matching the CRT keywords are extracted from the health data; all feature data are semantically analyzed using an artificial intelligence semantic analysis system; and feature data whose semantic analysis results meet the CRT feature requirements are screened out as structured target CRT feature data matching the CRT keywords.
[0028] 3. Data Analysis Obtain target CRT feature data corresponding to the identification information from the CRT database, and according to the CRT risk assessment strategy, obtain the assessment score of each feature data in the target CRT feature data, superimpose the assessment scores to obtain the total assessment score of the target CRT feature data, and obtain the corresponding CRT risk level according to the total assessment score.
[0029] In a further solution, the step of obtaining the evaluation score of each feature data in the target CRT feature data further includes: 3.1. Construct a logistic regression model for the risk of CRT in hospitalized children and use univariate logistic regression analysis to screen target CRT characteristic data.
[0030] 3.2. The target CRT characteristic data were brought into multivariate logistic regression analysis to obtain independent risk factors for CRT risk and the corresponding scoring model; among them, independent risk factors included patient-related risk factors, treatment-related risk factors, and CVAD-related risk factors.
[0031] Specifically, the distance between the grouping of independent risk factors and the set basic risk reference value is recorded as D, satisfying the formula:
[0032] in, is the reference value for each group of independent risk factors; is the basic risk reference value; is the regression coefficient.
[0033] The score corresponding to each independent risk factor group Satisfy the formula:
[0034] in, A constant corresponding to the unit score of the set evaluation score.
[0035] 3.3. According to the regression coefficient β in the logistic regression model of the risk of CRT in hospitalized children, the assessment score is assigned to each independent risk factor.
[0036] In the further scheme, the CRT risk levels include high risk, intermediate risk and low risk, among which: Comprehensive prevention information for moderate risk includes health education, basic prevention, and flushing / sealing.
[0037] Comprehensive prevention information for high-risk individuals includes bleeding risk assessment, which includes low risk and high risk.
[0038] In a further approach, the bleeding risk assessment includes the following steps: Extracting raw bleeding risk feature data from health data and converting it into structured target bleeding risk feature data; establishing a third corresponding relationship between the identified information and the target bleeding risk characteristic data; storing the identification information and the target bleeding risk characteristic data in the CRT database according to the third corresponding relationship; According to the bleeding risk assessment strategy, obtaining an assessment result of each feature data in the target bleeding risk feature data; The bleeding risk level of the person to be evaluated is obtained according to the evaluation result of each feature data in the target bleeding risk feature data.
[0039] In further planning, the bleeding risk assessment strategy includes: Constructing a judgment rule for each characteristic data of target bleeding risk characteristic data; Determine whether the evaluation rule corresponding to each characteristic data of the target bleeding risk characteristic data is triggered to obtain the corresponding bleeding risk assessment result.
[0040] In a second aspect, the present invention further provides a CRT risk warning system for hospitalized children using the above method, the system mainly comprising a CRT database, a data acquisition module, a data processing module, a data analysis module and a data display module, wherein: The data collection module is used to collect the identification information and health data of the person to be evaluated; establish a first corresponding relationship between the identification information and the health data; and store the identification information and the health data in the CRT database according to the first corresponding relationship.
[0041] The data processing module is used to extract target CRT feature data matching the CRT keywords from the health data.
[0042] The data analysis module obtains the evaluation score of each feature data in the target CRT feature data according to the CRT risk assessment strategy, and superimposes the total evaluation score of the target CRT feature data to obtain the corresponding CRT risk level.
[0043] The data display module is connected to the output end of the data analysis module and is used to output a visualized CRT risk level and comprehensive prevention information of the person to be evaluated.
[0044] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.
Claims
1. A CRT risk warning method for hospitalized children, characterized in that: The method comprises the following steps: S1. Data collection: collecting identification information and health data of hospitalized children; establishing a first corresponding relationship between the collected identification information and the health data; storing the identification information and the health data in a CRT database according to the first corresponding relationship; S2, data processing: acquiring health data corresponding to the identification information from the CRT database, extracting original CRT feature data matching the CRT keyword from the health data, and converting the data into structured target CRT feature data; establishing a second corresponding relationship between the identification information and the target CRT feature data; storing the identification information and the target CRT feature data in the CRT database according to the second corresponding relationship; S3, data analysis: acquiring the target CRT feature data corresponding to the identification information from the CRT database, and acquiring the evaluation score of each feature data in the target CRT feature data according to the CRT risk assessment strategy, superimposing the evaluation scores to obtain the total evaluation score of the target CRT feature data, and acquiring the corresponding CRT risk level according to the total evaluation score; S4. Data display: Output a visualized CRT risk level and comprehensive prevention information of the child to be evaluated.
2. The CRT risk warning method for hospitalized children according to claim 1, characterized in that: The health data in S1 includes the data of on-site health assessment of the child to be assessed, existing physical examination result data and / or past medical record data; wherein the existing physical examination result data and / or past medical record data come from at least one of the HIS system, hand numbness system, doctor's order system, testing and examination system, nursing management system and electronic medical record system.
3. The CRT risk warning method for hospitalized children according to claim 1, characterized in that: The S2 extracts original CRT feature data matching the CRT keyword from the health data and converts it into structured target CRT feature data, including: Extracting all feature data matching the CRT keyword from the health data; Using an artificial intelligence semantic analysis system to perform semantic analysis on all the feature data; Feature data whose semantic analysis results meet the CRT feature requirements are screened out to serve as the structured target CRT feature data that matches the CRT keyword.
4. The CRT risk warning method for hospitalized children according to claim 3, characterized in that: In S3, according to the CRT risk assessment strategy, an assessment score of each feature data in the target CRT feature data is obtained, including: S31. Construct a logistic regression model for the risk of CRT in hospitalized children, and use univariate logistic regression analysis to screen the target CRT characteristic data; S31, bringing the target CRT characteristic data into multivariate logistic regression analysis to obtain independent risk factors for CRT risk and corresponding scoring models; S32. Assign the assessment score to each of the independent risk factors based on the regression coefficient β in the logistic regression model for the risk of CRT in hospitalized children.
5. The CRT risk warning method for hospitalized children according to claim 4, characterized in that: The independent risk factors include patient-related risk factors, treatment-related risk factors and CVAD-related risk factors.
6. The CRT risk warning method for hospitalized children according to claim 4, characterized in that: The step S32 assigning the evaluation score to each of the independent risk factors includes: The distance between the grouping of the independent risk factors and the set basic risk reference value is recorded as D, which satisfies the formula: in, is the reference value for each group of independent risk factors; is the basic risk reference value; is the regression coefficient; The score corresponding to each group of the independent risk factors Satisfy the formula: in, A constant corresponding to the set evaluation score unit score.
7. The CRT risk warning method for hospitalized children according to claim 1, characterized in that: The CRT risk levels include high risk, medium risk and low risk, among which: The comprehensive prevention information for the low-risk and medium-risk patients includes health education, basic prevention and flushing / sealing; The comprehensive prevention information for the high risk includes bleeding risk assessment, which includes low risk and high risk.
8. The CRT risk warning method for hospitalized children according to claim 7, characterized in that: The bleeding risk assessment includes: Extracting original bleeding risk feature data from the health data and converting it into structured target bleeding risk feature data; Establishing a third corresponding relationship between the identification information and the target bleeding risk characteristic data; storing the identification information and the target bleeding risk characteristic data in the CRT database according to the third corresponding relationship; According to the bleeding risk assessment strategy, obtaining an assessment result of each feature data in the target bleeding risk feature data; The bleeding risk level of the person to be evaluated is obtained according to the evaluation result of each characteristic data in the target bleeding risk characteristic data.
9. The CRT risk warning method for hospitalized children according to claim 8, characterized in that: The bleeding risk assessment strategy includes: Constructing a judgment rule for each characteristic data of the target bleeding risk characteristic data; It is determined whether the evaluation rule corresponding to each characteristic data of the target bleeding risk characteristic data is triggered to obtain the corresponding bleeding risk assessment result.
10. A CRT risk warning system for hospitalized children, using the CRT risk warning method for hospitalized children as described in any one of claims 1 to 9, characterized in that: include: A data collection module is used to collect identification information and health data of the person to be evaluated; Establishing a first corresponding relationship between the identification information and the health data; storing the identification information and the health data in a CRT database according to the first corresponding relationship; A data processing module, used for extracting target CRT feature data matching the CRT keyword from the health data; A data analysis module, according to a CRT risk assessment strategy, obtains an assessment score of each feature data in the target CRT feature data, and superimposes the total assessment score of the target CRT feature data to obtain a corresponding CRT risk level; as well as A data display module is connected to the output end of the data analysis module and is used to output a visualized CRT risk level and comprehensive prevention information of the person to be evaluated.