System and method for customizing nursing measures based on evaluation result of patient subjected to lung cancer chemotherapy

Abstract

The invention discloses a system and a method for customizing nursing measures based on evaluation results of patients subjected to lung cancer chemotherapy, which are used in the field of data analysis, and the system for customizing nursing measures based on evaluation results of patients subjected to lung cancer chemotherapy comprises a data acquisition module, an evaluation and analysis module, a nursing measure generation module, a nursing monitoring module and an interaction and display module. According to the method, unimportant variable coefficients are reduced to zero through penalty coefficients, so that variables significantly related to life quality and health conditions of patients are identified; an optimal penalty parameter is found by using a cross validation method, so that the prediction precision and generalization ability of the model can be maximized, and the risk of overfitting is reduced; through the multiple linear regression model, the specific influence of each characteristic variable on the life quality and health condition of the patient can be quantified, and possible side effect risks can be identified, so that more personalized and safer treatment suggestions are provided for the patient.

Description

Technical Field

[0001] The present invention relates to the field of data analysis, and in particular, to a system and method for customizing nursing measures based on evaluation results of lung cancer chemotherapy patients. Background Art

[0002] Lung cancer is a malignant tumor that originates from lung tissue. It is one of the most common types of cancer worldwide and the cause of the most cancer deaths. Lung cancer is mainly divided into two categories: small cell lung cancer (SCLC) and non-small cell lung cancer (NSCLC); Small cell lung cancer (SCLC) is an aggressive type of lung cancer that is often strongly associated with smoking. It grows quickly and tends to spread.

[0003] Non-small cell lung cancer (NSCLC) is the most common type of lung cancer, accounting for approximately 85% of all lung cancer cases. It includes several different lung cancer types, such as adenocarcinoma, squamous cell carcinoma, and large cell carcinoma.

[0004] Chemotherapy is a treatment method that uses chemical drugs to kill or inhibit the growth of cancer cells. For lung cancer patients, chemotherapy can be used alone or in combination with surgery or radiotherapy to achieve the treatment goal. The care of chemotherapy patients is a multifaceted process that focuses on reducing the side effects of chemotherapy and improving the patient's quality of life.

[0005] Traditional analysis methods may have difficulty in effectively processing high-dimensional data and can easily lead to model overfitting. In complex medical data analysis, many models are easily limited by the characteristics of the training data set, resulting in poor performance on new data. Among the multi-factor influences on patient treatment, traditional methods often find it difficult to clearly quantify the specific impact of each factor on the patient's health status. Traditional nursing measure evaluations often lack systematic verification methods and cannot comprehensively evaluate treatment effects. In multi-professional medical teams, information island problems often affect team collaboration and decision-making efficiency.

[0006] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0007] In order to overcome the above problems, the present invention aims to propose a system and method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients. The purpose is to solve the problem that traditional methods often find it difficult to clearly quantify the specific impact of various factors on the patient's health status. Traditional nursing measure evaluations often lack systematic verification methods and cannot comprehensively evaluate the treatment effects. In multi-professional medical teams, information island problems often affect team collaboration and decision-making efficiency.

[0008] To this end, the specific technical solution adopted by the present invention is as follows: According to one aspect of the present invention, a system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients is provided. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients comprises: a data acquisition module, an evaluation and analysis module, a nursing measure generation module, a nursing monitoring module, and an interaction and display module; The data collection module, the evaluation and analysis module, the nursing measures generation module, the nursing monitoring module and the interaction and display module are connected in sequence; The data collection module is used to integrate the patient's basic information, clinical data, health indicators and individual response to chemotherapy, and to create an electronic health record for patients undergoing chemotherapy for lung cancer; The evaluation and analysis module analyzes the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identifies the side effects caused by chemotherapy and provides corresponding management suggestions; Nursing measures generation module, which is used to generate personalized nursing measures based on the identification results and share the internal information of the nursing team through the collaboration platform; Nursing monitoring module, which is used to track the implementation of nursing measures and collect feedback from lung cancer chemotherapy patients, and adaptively adjust nursing measures based on the feedback; The interactive and display module is used to provide a visual interface for interactive communication and use visual tools to display health indicators and nursing progress; The evaluation and analysis module analyzes the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identifies the side effects caused by chemotherapy and provides corresponding management suggestions, including: The Lasso regression model was used to screen variables related to the patient's quality of life and health status from the electronic health records, and the coefficient changes of different variables were observed by adjusting the penalty parameter value to identify the characteristic variables; The characteristic variables were used to construct a multivariate linear regression model, and the regression coefficients of each variable were analyzed to evaluate the specific impact of the regression coefficients on the patient's physical condition and quality of life; Perform F-test in multiple linear regression models and t-test for different variables; calculate the coefficient of determination and adjusted coefficient of determination at the same time; The multivariate linear regression model was tested using cross-validation method; Based on the F test, t test, coefficient of determination and adjusted coefficient of determination and the multivariate linear regression model after testing, the impact of each variable on the quality of life and health status of patients was analyzed, and the patient groups with side effects were identified and marked.

[0009] Optionally, the Lasso regression model is used to screen variables related to the patient's quality of life and health status from the electronic health record, and by adjusting the penalty parameter value, the coefficient changes of different variables are observed, and the characteristic variables are identified, including: Extract relevant feature data from electronic health records and organize the extracted feature data into a feature matrix; According to the analysis requirements, determine the preset target and construct the corresponding target vector; Instantiate the Lasso regression model and set the initial penalty parameter value; The feature matrix and target vector are input into the Lasso regression model, and each penalty parameter value is tested by cross-validation method. The change of variable coefficient under each penalty parameter value is observed to identify the characteristic variables of the target variable.

[0010] Optionally, the feature matrix and target vector are input into the Lasso regression model, each penalty parameter value is tested by cross-validation method, and the change of variable coefficient under each penalty parameter value is observed. The characteristic variables of the target variable are identified as follows: Input the prepared feature matrix and target vector into the Lasso regression model and configure the cross-validation parameters; Test a series of preset penalty parameter values ​​in the Lasso regression model and evaluate the performance of the Lasso regression model and the changes in the characteristic coefficients under each parameter; For each penalty parameter value obtained from the penalty parameter value test, retrain the model through each fold of cross-validation, record the characteristic coefficient of each fold, and observe the change of the characteristic coefficient with the penalty parameter value; The influence of each variable on the target vector is evaluated based on the stability and significance of the characteristic coefficient, and the variables with significant and stable coefficients under multiple penalty parameter value settings are identified and marked as key characteristic variables.

[0011] Optionally, a multivariate linear regression model is constructed using characteristic variables, and the regression coefficient of each variable is analyzed to evaluate the specific impact of the regression coefficient on the patient's physical condition and quality of life, including: Multiple linear regression models were constructed using the identified characteristic variables; Input the feature matrix and target vector into the multivariate linear regression model for training, and calculate the regression coefficient of each feature variable; The obtained regression coefficients were analyzed to obtain the positive or negative impact of each characteristic variable on the patient's physical condition and quality of life, and to evaluate the specific impact of each characteristic variable on the patient's condition.

[0012] Optionally, perform an F test in the multiple linear regression model and perform a t test on different variables; calculate the coefficient of determination and the adjusted coefficient of determination simultaneously including: The multivariate linear regression model was fitted using the least squares method to obtain the estimated values ​​of the model parameters; Based on the total sum of squares, regression sum of squares, and residual sum of squares, the F statistic is calculated, and the P value of the F statistic is obtained based on the degrees of freedom; Perform a t-test on the coefficient of each explanatory variable in the model, calculate the t-statistic for each coefficient, and find the P value of each t-statistic based on the degrees of freedom; The coefficient of determination and the adjusted coefficient of determination were calculated.

[0013] Optionally, the expression for calculating the F statistic is: ; The expression for calculating the t-statistic for each coefficient is: ; The expressions for calculating the coefficient of determination and the adjusted coefficient of determination are: ; ; In the formula, p represents the number of explanatory variables in the model; n represents the total number of samples; F express F test; t express t test; represents the estimated value of the coefficient; represents the standard error of the coefficient; represents the coefficient of determination; represents the adjusted coefficient of determination; represents the total sum of squares; represents the residual sum of squares; represents the regression sum of squares.

[0014] Optionally, based on the F test, t test, coefficient of determination and adjusted coefficient of determination and the tested multiple linear regression model, the influence of each variable on the patient's quality of life and health status is analyzed, and the patient groups with side effects are identified and marked, including: The F test was used to confirm the overall significance of the multivariate linear regression model, and the t test was used to identify the independent variables that had an impact on the dependent variable; The explanatory power of the multiple linear regression model was measured by the coefficient of determination, and the validity of the multiple linear regression model was determined by the adjusted coefficient of determination; Based on the output of the multivariate linear regression model, factors associated with reduced quality of life or worsening health status were identified, and the sources of side effects were located; cross-validation was used to test the stability and predictive ability of the model.

[0015] Optionally, the nursing measure generation module includes, when generating personalized nursing measures according to the recognition results and sharing the internal information of the nursing team through the collaboration platform: Design a rule engine based on pre-defined clinical indicators and outputs of multivariate regression models; The random forest algorithm was used to analyze the electronic health records of lung cancer chemotherapy patients to predict their response to different nursing measures. Combine the identification results of the rule engine and the prediction results of the random forest model to generate personalized care measures for each patient; Through the collaborative platform, the generated nursing action plan is shared in real time to the nursing team’s internal information database.

[0016] Optionally, the random forest algorithm is used to analyze the electronic health records of lung cancer chemotherapy patients to predict the response of lung cancer chemotherapy patients to different nursing measures including: Obtain electronic health records of patients undergoing chemotherapy for lung cancer and identify characteristic data related to chemotherapy response based on the electronic health records; A prediction model was constructed using the random forest algorithm to use the identified features to predict the response of lung cancer chemotherapy patients to different care interventions; Evaluate the performance of the model and use accuracy metrics to measure the effectiveness of the prediction model in predicting the response of lung cancer patients to chemotherapy; Predictive models are used to predict patient responses and generate personalized care recommendations for each lung cancer chemotherapy patient.

[0017] According to another aspect of the present invention, a method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients is also provided. The method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients comprises the following steps: S1. Integrate patients’ basic information, clinical data, health indicators and individual responses to chemotherapy, and create electronic health records for patients undergoing chemotherapy for lung cancer; S2. Analyze the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identify the side effects caused by chemotherapy and provide corresponding management suggestions; S3, Generate personalized care measures based on the identification results and share the internal information of the care team through the collaborative platform; S4. Track the implementation of nursing measures and collect feedback from patients undergoing lung cancer chemotherapy, and adaptively adjust nursing measures based on the feedback; S5. Provide a visual interface for interactive communication and use visual tools to display health indicators and care progress.

[0018] Compared with the prior art, this application has the following beneficial effects: 1. The Lasso regression of the present invention can effectively process high-dimensional data, and reduce the coefficients of unimportant variables to zero through the penalty coefficient, thereby helping to identify variables that are significantly related to the patient's quality of life and health status; using the cross-validation method to find the optimal penalty parameter can maximize the prediction accuracy and generalization ability of the model and reduce the risk of overfitting; through the multivariate linear regression model, the specific impact of each characteristic variable on the patient's quality of life and health status can be quantified, such as understanding the positive and negative impact and impact intensity of each variable through the regression coefficient; performing F test and t test, as well as calculating the determination coefficient and the adjusted determination coefficient, can verify the statistical significance and explanatory power of the model and ensure the reliability of the analysis results; by analyzing the significant coefficients in the model, especially those variables with negative effects, possible side effect risks can be identified, thereby providing patients with more personalized and safe treatment recommendations.

[0019] 2. The present invention generates personalized nursing recommendations based on the patient's specific health data by utilizing the random forest algorithm and rule engine, which can more accurately meet the needs of each patient and improve the adaptability and effectiveness of treatment. The random forest algorithm improves the accuracy and stability of predictions by constructing multiple decision trees and integrating their results, which helps to reduce prediction errors and provide more reliable nursing recommendations. Personalized nursing plans are shared in real time through a collaborative platform to ensure that all relevant nursing team members have instant access to the latest nursing information, thereby speeding up the decision-making process and improving nursing efficiency. By collecting feedback from the implemented nursing measures, the model and nursing methods can be continuously optimized, which helps to continuously improve the patient's treatment experience and outcomes.

[0020] 3. The present invention can monitor the implementation of nursing measures in real time by using an electronic health record system or a specialized nursing management software, ensure that nursing activities are carried out promptly and correctly according to established guidelines, and provide a visual interface for interactive communication, so that patients can intuitively see their health indicators and nursing progress, thereby promoting communication and understanding between patients and the medical team. Through a shared interface, medical personnel of different professions can more easily access and understand the work of other team members and the status of patients, thereby promoting collaborative work within the team. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above characteristics, features and advantages of the present invention and their implementation methods and methods will become more clearly understood in conjunction with the following description of the embodiments, which are described in detail in conjunction with the accompanying drawings. Herein, a schematic diagram is shown: Figure 1 is a principle block diagram of a system for customizing nursing measures based on evaluation results of lung cancer chemotherapy patients according to an embodiment of the present invention; Figure 24 is a flowchart of a method for customizing nursing measures based on evaluation results of lung cancer chemotherapy patients according to an embodiment of the present invention.

[0022] In the figure: 1. Data collection module; 2. Evaluation and analysis module; 3. Nursing measures generation module; 4. Nursing monitoring module; 5. Interaction and display module. DETAILED DESCRIPTION

[0023] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0024] According to an embodiment of the present invention, a system and method for customizing nursing measures based on evaluation results of lung cancer chemotherapy patients are provided.

[0025] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to one embodiment of the present invention, a system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients is provided, and the system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients includes: a data acquisition module 1, an evaluation and analysis module 2, a nursing measure generation module 3, a nursing monitoring module 4 and an interaction and display module 5; The data collection module 1, the evaluation and analysis module 2, the nursing measures generation module 3, the nursing monitoring module 4 and the interaction and display module 5 are connected in sequence; Data collection module 1 is used to integrate the patient's basic information, clinical data, health indicators and individual response to chemotherapy, and to create an electronic health record for patients undergoing chemotherapy for lung cancer.

[0026] It needs to be explained that the patient's basic information, such as name, gender, age, contact information, etc., is collected; it is usually completed through the front desk reception or through the online pre-registration system when the patient is first admitted to the hospital, using a form or electronic input system, and the patient or medical staff enters basic information. The information will be entered into the hospital's information management system as part of the patient's electronic health record; the patient's clinical data, including medical records, surgical records, diagnostic reports, etc., are integrated. Clinical data is collected from different medical departments, such as outpatient clinics, operating rooms, radiology departments, etc., and data integration is performed through the hospital's clinical information system (CIS). Doctors and nurses will update the patient's clinical information, including the diagnosis and treatment process, in the clinical information system.

[0027] Evaluation and analysis module 2 analyzes the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identifies the side effects caused by chemotherapy and provides corresponding management recommendations.

[0028] Among them, the evaluation and analysis module 2 analyzes the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identifies the side effects caused by chemotherapy and provides corresponding management suggestions, including: The Lasso regression model was used to screen variables related to patients' quality of life and health status from electronic health records, and the coefficient changes of different variables were observed by adjusting the penalty parameter value to identify characteristic variables.

[0029] Preferably, the Lasso regression model is used to screen variables related to the patient's quality of life and health status from the electronic health record, and by adjusting the penalty parameter value, the coefficient changes of different variables are observed, and the identified characteristic variables include: Extract relevant feature data from electronic health records and organize the extracted feature data into a feature matrix; According to the analysis requirements, determine the preset target and construct the corresponding target vector; Instantiate the Lasso regression model and set the initial penalty parameter value; The feature matrix and target vector are input into the Lasso regression model, and each penalty parameter value is tested by cross-validation method. The change of variable coefficient under each penalty parameter value is observed to identify the characteristic variables of the target variable.

[0030] Preferably, the feature matrix and the target vector are input into the Lasso regression model, each penalty parameter value is tested by cross-validation method, and the change of the variable coefficient under each penalty parameter value is observed. The characteristic variables of the target variable are identified as follows: Input the prepared feature matrix and target vector into the Lasso regression model and configure the cross-validation parameters; Test a series of preset penalty parameter values ​​in the Lasso regression model and evaluate the performance of the Lasso regression model and the changes in the characteristic coefficients under each parameter; For each penalty parameter value obtained from the penalty parameter value test, retrain the model through each fold of cross-validation, record the characteristic coefficient of each fold, and observe the change of the characteristic coefficient with the penalty parameter value; The influence of each variable on the target vector is evaluated based on the stability and significance of the characteristic coefficient, and the variables with significant and stable coefficients under multiple penalty parameter value settings are identified and marked as key characteristic variables.

[0031] It should be explained that data directly related to the patient's health status, such as physiological measurements, treatment history, disease status, etc., are screened out, and the required data are extracted from the electronic health record database using data extraction tools or programming languages ​​(such as Python). The data are screened according to the specific needs of the study, for example, variables such as age, gender, weight, medication information, and treatment response are selected, and the extracted data are organized into a format that can be processed by the mathematical model, that is, the feature matrix, and the corresponding target vector. The data are cleaned, standardized, and converted into a feature matrix using data processing software or programming libraries (such as Pandas in Python); the target vector is set according to the purpose of the study, such as the patient's quality of life score or health status indicator. The Lasso model is used to select and shrink variables in order to identify the most important predictor variables, and machine learning libraries (such as scikit-learnin Python) instantiates the Lasso model and sets an initial penalty parameter (alpha), which controls the strength of regularization. A larger alpha value leads to more features being excluded (the coefficient is reduced to zero). The optimal alpha value is determined through cross-validation to find the optimal model complexity and achieve effective variable selection. Cross-validation (such as k-fold cross-validation) is used to test the impact of different alpha values ​​on model performance. In scikit-learn, LassoCV can be used, which automatically finds the optimal solution among multiple alpha values ​​and observes the changes in coefficients under different alpha values. According to the Lasso regression results, variables with non-zero coefficients are selected as key features. The Lasso regression results are analyzed and variables with non-zero coefficients are extracted. Variables are considered to have the greatest impact on the model, that is, features that have a significant impact on the patient's quality of life and health status.

[0032] In the process of feature selection and model optimization using Lasso regression, systematic testing of different values ​​of the penalty parameter (alpha) is a crucial step; through cross-validation, the model is retrained for each alpha value and the feature coefficients of each fold are recorded; by observing the changes in feature coefficients under different penalty strengths; in each fold of cross-validation, the model is trained with data other than the current fold and the feature coefficients on the current fold are calculated to ensure the independence and accuracy of the evaluation; finally, by comparing the changes in coefficients under different alpha values, it is possible to determine which features remain stable under regularization pressure, thereby identifying the features that are most critical to the model's predictive ability.

[0033] A multivariate linear regression model was constructed using characteristic variables, and the regression coefficient of each variable was analyzed to evaluate the specific impact of the regression coefficient on the patient's physical condition and quality of life.

[0034] Preferably, a multivariate linear regression model is constructed using characteristic variables, and the regression coefficient of each variable is analyzed to evaluate the specific impact of the regression coefficient on the patient's physical condition and quality of life, including: Multiple linear regression models were constructed using the identified characteristic variables; Input the feature matrix and target vector into the multivariate linear regression model for training, and calculate the regression coefficient of each feature variable; The obtained regression coefficients were analyzed to obtain the positive or negative impact of each characteristic variable on the patient's physical condition and quality of life, and to evaluate the specific impact of each characteristic variable on the patient's condition.

[0035] It should be explained that to construct a multiple linear regression model to analyze the relationship between multiple independent variables (identified feature variables) and a dependent variable (such as the patient's quality of life), first ensure that the selected feature data has been properly preprocessed, such as standardized or normalized, to eliminate the impact of different dimensions; use statistical software or programming libraries (such as statsmodels or scikit-learn in Python) to initialize a multiple linear regression model, prepare to input data into the model, and input the preprocessed feature matrix and target vector (such as the patient's quality of life score) into the model; use the least squares method as the estimation method, aiming to minimize the difference between the predicted value and the actual value. The sum of squared errors between the independent variables is calculated by training the model to estimate the influence of each independent variable on the dependent variable (i.e., the regression coefficient). The regression coefficient of each characteristic variable is obtained through model training. These coefficients indicate the importance and influence direction of the characteristic variable in the prediction model. After the model training is completed, the regression coefficients are analyzed. The positive and negative signs of each coefficient indicate whether the variable has a positive or negative impact on the dependent variable, while the absolute value of the coefficient indicates the intensity of the impact. The significance of each regression coefficient is explained to determine the specific impact of each characteristic variable on the patient's physical condition and quality of life. The coefficient of each variable is evaluated to determine its positive or negative impact on the patient's quality of life. In addition, the overall fit of the model is evaluated, such as through R 2 The p-value is used to measure the proportion of variation explained by the model, and the p-value is used to test the statistical significance of each coefficient.

[0036] F-tests were performed in the multiple linear regression models, and t-tests were performed for different variables; the coefficient of determination and adjusted coefficient of determination were calculated at the same time.

[0037] Preferably, performing an F test in a multiple linear regression model and performing a t test on different variables; and calculating the coefficient of determination and the adjusted coefficient of determination simultaneously comprises: The multivariate linear regression model was fitted using the least squares method to obtain the estimated values ​​of the model parameters; Based on the total sum of squares, regression sum of squares, and residual sum of squares, the F statistic is calculated, and the P value of the F statistic is obtained based on the degrees of freedom; Perform a t-test on the coefficient of each explanatory variable in the model, calculate the t-statistic for each coefficient, and find the P value of each t-statistic based on the degrees of freedom; The coefficient of determination and the adjusted coefficient of determination were calculated.

[0038] Preferably, the expression for calculating the F statistic is: ; The expression for calculating the t-statistic for each coefficient is: ; The expressions for calculating the coefficient of determination and the adjusted coefficient of determination are: ; ; In the formula, p represents the number of explanatory variables in the model; n represents the total number of samples; F express F test; t express t test; represents the estimated value of the coefficient; represents the standard error of the coefficient; represents the coefficient of determination; represents the adjusted coefficient of determination; represents the total sum of squares; represents the residual sum of squares; represents the regression sum of squares.

[0039] It should be explained that the purpose of fitting a model by least squares is to estimate the linear effect of each explanatory variable in the model on the dependent variable, that is, to calculate the estimated value of the model parameters. Using least squares in software or programming libraries (such as Python's statsmodels library), inputting the preprocessed feature matrix and target vector for model training will obtain the coefficient estimate and intercept of each variable. The F test is used to test whether at least one explanatory variable in the model has a statistically significant effect on the dependent variable. The calculation of the F statistic requires the regression sum of squares (variation explained by the model), the residual sum of squares (variation not explained by the model), and the residual sum of squares (variation not explained by the model). The F statistic is calculated by comparing the goodness of fit of the model with that of a simple average model without explanatory variables. The F statistic and its corresponding P value are usually calculated and reported automatically. The t-test is used to assess whether the coefficient of each variable in the model is significantly different from zero, that is, whether each variable has a significant effect on the dependent variable. To calculate the t-statistic for each coefficient, the coefficient estimate and standard error (SE) are required. The t-statistic is obtained by dividing the coefficient estimate by its standard error. The corresponding P value can be obtained automatically by looking up a table or using statistical software, and is determined based on the degrees of freedom and t-distribution. The coefficient of determination ( R 2 ) measures the proportion of the total variation explained by the model, while the adjusted coefficient of determination adjusts the degrees of freedom to account for the number of variables in the model and is calculated as the regression sum of squares divided by the total sum of squares. The adjusted coefficient of determination subtracts the bias that may be introduced by the introduction of more variables from the coefficient of determination.

[0040] The multivariate linear regression model was tested using cross-validation.

[0041] Based on the F test, t test, coefficient of determination and adjusted coefficient of determination and the multivariate linear regression model after testing, the impact of each variable on the quality of life and health status of patients was analyzed, and the patient groups with side effects were identified and marked.

[0042] Preferably, based on the F test, t test, determination coefficient and adjusted determination coefficient and the tested multiple linear regression model, the influence of each variable on the patient's quality of life and health status is analyzed, and the patient groups with side effects are identified and marked, including: The F test was used to confirm the overall significance of the multivariate linear regression model, and the t test was used to identify the independent variables that had an impact on the dependent variable; The explanatory power of the multiple linear regression model was measured by the coefficient of determination, and the validity of the multiple linear regression model was determined by the adjusted coefficient of determination; Based on the output of the multivariate linear regression model, factors associated with reduced quality of life or worsening health status were identified, and the sources of side effects were located; cross-validation was used to test the stability and predictive ability of the model.

[0043] It should be explained that the F test is used to determine whether at least one explanatory variable in the model has a significant statistical effect on the dependent variable, thereby verifying the overall significance of the model. After the model is trained, statistical software (such as statsmodels) is used to calculate the F statistic and the corresponding P value; if the P value is lower than the set significance level (usually 0.05), the model is considered statistically significant, indicating that at least one of the explanatory variables included in the model has an effect on the dependent variable. The t test is used to evaluate whether the coefficient of each explanatory variable in the model is significantly different from zero, thereby determining the significance of each variable. A t test is performed on the coefficient of each variable, and the t value and P value of each coefficient are calculated; if the P value of a variable is lower than the significance level, it indicates that the variable has a significant effect on the dependent variable, and the determination coefficient ( R 2 ) is used to measure the proportion of the total variation explained by the model, reflecting the explanatory power of the model, and is calculated R 2 Value, evaluate the model's fit to the data, a higher R 2 The value indicates that the model can better explain the variation of the dependent variable. The adjusted coefficient of determination adjusts the number of variables in the model and can more accurately reflect the effectiveness of the model. R 2 It provides a more rigorous evaluation of the model effect by subtracting the errors that may be caused by introducing more explanatory variables. The stability and prediction accuracy of the model are evaluated by cross-validation, which is an important test of the generalization ability of the model. The data set is divided into several subsets, and one part is used as the test set and the rest as the training set in turn. This process is repeated to ensure that each data point has the opportunity to be used as test data. The model performance indicators of each round are calculated, such as R 2 By analyzing the significant model coefficients, we can identify the variables that may lead to a decline in the patient's quality of life or worsening of their health status. For each significant negative coefficient, we further analyze how these variables are associated with the patient's quality of life and health status, paying special attention to those factors that may cause side effects.

[0044] The nursing measure generation module 3 is used to generate personalized nursing measures according to the recognition results and share the internal information of the nursing team through the collaboration platform.

[0045] Preferably, the nursing measure generation module 3 generates personalized nursing measures according to the recognition results and shares the internal information of the nursing team through the collaboration platform, including: Design a rule engine based on pre-defined clinical indicators and the output of the multivariate regression model.

[0046] The random forest algorithm was used to analyze the electronic health records of patients undergoing lung cancer chemotherapy and to predict their response to different nursing measures.

[0047] Preferably, the random forest algorithm is used to analyze the electronic health records of lung cancer chemotherapy patients to predict the response of lung cancer chemotherapy patients to different nursing measures, including: Obtain electronic health records of patients undergoing chemotherapy for lung cancer and identify characteristic data related to chemotherapy response based on the electronic health records; A prediction model was constructed using the random forest algorithm to use the identified features to predict the response of lung cancer chemotherapy patients to different care interventions; Evaluate the performance of the model and use accuracy metrics to measure the effectiveness of the prediction model in predicting the response of lung cancer patients to chemotherapy; Predictive models are used to predict patient responses and generate personalized care recommendations for each lung cancer chemotherapy patient.

[0048] It should be explained that the purpose of obtaining electronic health records is to collect sufficient data to support subsequent analysis and model building, extract the patient's detailed health records from the hospital information system, including diagnostic information, chemotherapy process, past medical records, physiological indicators, etc., ensure that the collected data comply with relevant privacy and data protection regulations, screen out key features related to chemotherapy response from a large amount of health data, use data analysis techniques such as correlation analysis and feature selection methods to determine which data are significantly correlated with chemotherapy response, features include patients' basic biomarkers, previous chemotherapy response records, genetic information, etc., use the random forest algorithm to build a predictive model to predict patients' response to different care measures, and use the scikit-learn library in the selected programming environment (such as Python) to select the key features related to chemotherapy response. Create a model using the random forest algorithm; random forest is a powerful machine learning model that enhances the accuracy and stability of predictions by building multiple decision trees and fusing their predictions. Evaluate the performance of the prediction model to ensure its accuracy and practicality. Evaluate the performance of the model through techniques such as cross-validation. Use metrics such as accuracy, recall, precision, and F1 score to measure the performance of the model in practical applications. Metrics can help understand the behavior of the model when it predicts correctly and incorrectly. Based on the model's predictions, generate personalized nursing recommendations for each lung cancer chemotherapy patient. Use the trained random forest model to make predictions for new patient data and provide targeted nursing recommendations based on the predictions. Recommendations may include recommending specific nursing measures or adjusting existing treatment plans to optimize patients' treatment outcomes and quality of life.

[0049] Combine the identification results of the rule engine and the prediction results of the random forest model to generate personalized care measures for each patient; Through the collaborative platform, the generated nursing action plan is shared in real time to the nursing team’s internal information database.

[0050] It should be explained that the identification results of the rule engine are combined with the prediction results of the random forest model to form a comprehensive judgment framework to generate more accurate and comprehensive nursing measures. First, the output of the rule engine is determined, such as the marked patient risk level or specific warning; then, the prediction results provided by the random forest model are integrated, such as the response prediction to specific treatments or nursing measures; the outputs of these two methods are integrated using logical or weighted algorithms (such as weighted average or decision matrix) to produce a comprehensive treatment and nursing recommendation, and personalized nursing measures are designed for each patient based on the comprehensive analysis results. According to the patient's specific situation and the comprehensive prediction of the model, a personalized care plan is formulated, including adjustment of drug dosage. , change treatment plans, recommend specific rehabilitation activities, etc., to ensure that each recommendation is based on the best match between the patient's unique needs and predicted outcomes, and use the collaborative platform to share the generated nursing action plan in real time to the nursing team's internal information library so that the nursing team can access and execute these measures immediately, select a suitable collaborative platform (such as a hospital information system, electronic medical record system, or specialized nursing collaboration software), and ensure that the platform can support real-time data sharing and updating; upload the generated nursing action plan to the platform and set appropriate access rights to ensure that relevant nursing team members can access this information; in addition, implement a tracking and feedback mechanism to collect data on implementation effectiveness and feedback from team members for future improvements.

[0051] Nursing monitoring module 4 is used to track the implementation of nursing measures, collect feedback from lung cancer chemotherapy patients, and adaptively adjust nursing measures based on the feedback.

[0052] It should be explained that the monitoring mechanism should ensure that nursing measures are correctly and timely implemented, and the effectiveness of each measure can be tracked. The electronic health record system (EHR) or special nursing management software should be used to record the implementation details of nursing measures in real time, such as the time of nursing operations, the executors, the patients' immediate reactions, etc. Automatic reminders and alarm functions should be set to ensure that nursing measures are not ignored or incorrectly implemented. By systematically collecting feedback from patients, the effectiveness of nursing measures and patient satisfaction can be understood to provide a basis for the optimization of nursing measures. A feedback collection mechanism should be designed and implemented, including regular questionnaires, face-to-face interviews or the use of digital tools to collect immediate feedback from patients; ensure that the feedback covers the acceptance of nursing measures, patient comfort and Any side effects or problems, based on patient feedback, make necessary adjustments to nursing measures to improve their effectiveness and patient satisfaction, create a feedback loop, which includes evaluating patient feedback and analyzing data to determine which measures need to be adjusted and how to adjust them; using data analysis and decision support systems to help identify points for improvement; regularly reviewing and updating care plans, especially for those measures that have been reported to perform poorly, through continuous monitoring and feedback mechanisms, achieving continuous improvement and personalized adjustment of nursing measures, and establishing a dynamic care management framework that can be quickly iterated and adjusted based on real-time data and patient feedback; using advanced data analysis technologies, such as machine learning algorithms, to automatically identify patterns and trends, thereby making more precise recommendations for care adjustments.

[0053] The interaction and display module 5 is used to provide a visual interface for interactive communication and use visual tools to display health indicators and nursing progress.

[0054] It should be explained that the interface should be designed and developed using front-end development technologies (such as HTML, CSS, JavaScript) and frameworks (such as React or Angular); ensure that the interface design meets user experience best practices, such as clear navigation, responsive design to adapt to different devices (mobile phones, tablets, computers) and accessibility standards, and use data visualization libraries (such as D3.js, Chart.js or Highcharts) to create dynamic charts and dashboards; data can be dynamically obtained from the back-end system through APIs and rendered in real time on the front-end; charts should include but are not limited to bar charts, line charts, pie charts and scatter charts, each of which clearly displays important health indicators and trends, and the back-end system needs to be configured with an effective data processing and transmission mechanism (such as WebSocket or polling mechanism) so that the interface can receive the latest data; the front-end should implement dynamic updates of data so that users can view the latest information without manual refresh, and integrate forms, chat windows or other interactive elements in the interface to allow users to enter data, submit feedback or request help; ensure that the interactive design is concise and clear, and the user operation is intuitive and convenient.

[0055] Specific examples are as follows: Patient information: XX Hospital, male, Mr. Zhang, undergoing first round of chemotherapy.

[0056] 1) Data collection Collect Mr. Zhang's basic information, physiological indicators before and after chemotherapy (such as blood pressure, blood cell count), treatment history and past medical records; All data were integrated into Mr. Zhang's electronic health record, including the latest chemotherapy response data.

[0057] 2) Evaluation and analysis The Lasso regression analysis model was used to screen out features significantly associated with chemotherapy response from Mr. Zhang’s health data, such as white blood cell count and liver function indicators; The analysis pointed out that Mr. Zhang has a higher risk of infection after chemotherapy and needs special monitoring.

[0058] 3) Generation of nursing measures Based on the results of the model analysis, nursing measures were proposed to increase the use of antibiotics to prevent infection and adjust the drug dosage to reduce the burden on liver function; Share care recommendations with the care team, including doctors, nurses, and pharmacists, through a collaborative platform.

[0059] 4) Nursing monitoring Real-time tracking of care delivery, such as medication timing and dosage, through electronic health record systems; Collect Mr. Zhang’s feedback regularly, including satisfaction with the nursing measures and physical feelings.

[0060] 5) Interaction and display An interactive interface was used to show Mr. Zhang his health progress, including a comparison of his physiological indicators before and after chemotherapy; The interface provides instant messaging capabilities, allowing Mr. Zhang to directly provide feedback to the nursing team or request further assistance.

[0061] It can be concluded that through real-time monitoring and feedback, it was found that Mr. Zhang responded well to the new nursing measures and his physiological indicators were stable; based on Mr. Zhang's continuous feedback and health data updates, the nursing measures were adjusted to optimize the treatment effect and quality of life.

[0062] According to another embodiment of the present invention, Figure 2 As shown, a method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients is also provided. The method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients includes the following steps: S1. Integrate patients’ basic information, clinical data, health indicators and individual responses to chemotherapy, and create electronic health records for patients undergoing chemotherapy for lung cancer; S2. Analyze the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identify the side effects caused by chemotherapy and provide corresponding management suggestions; S3, Generate personalized care measures based on the identification results and share the internal information of the care team through the collaborative platform; S4. Track the implementation of nursing measures and collect feedback from patients undergoing lung cancer chemotherapy, and adaptively adjust nursing measures based on the feedback; S5. Provide a visual interface for interactive communication and use visual tools to display health indicators and care progress.

[0063] In summary, with the help of the above technical solutions of the present invention, the Lasso regression of the present invention can effectively process high-dimensional data, and reduce the coefficients of unimportant variables to zero through the penalty coefficient, thereby helping to identify variables that are significantly related to the patient's quality of life and health status; using the cross-validation method to find the optimal penalty parameter can maximize the prediction accuracy and generalization ability of the model and reduce the risk of overfitting; through the multivariate linear regression model, the specific impact of each characteristic variable on the patient's quality of life and health status can be quantified, such as understanding the positive and negative impact and impact intensity of each variable through the regression coefficient; performing F test and t test, as well as calculating the determination coefficient and the adjusted determination coefficient, can verify the statistical significance and explanatory power of the model and ensure the reliability of the analysis results; by analyzing the significant coefficients in the model, especially those variables with negative effects, possible side effect risks can be identified, thereby providing patients with more personalized and safe treatment recommendations; the present invention generates personalized nursing recommendations based on the patient's specific health data by using the random forest algorithm and rule engine, which can more accurately meet the needs of each patient. The random forest algorithm improves the accuracy and stability of predictions by constructing multiple decision trees and integrating their results, which helps to reduce prediction errors and provide more reliable nursing recommendations. It shares personalized nursing plans in real time through a collaborative platform to ensure that all relevant nursing team members have instant access to the latest nursing information, thereby speeding up the decision-making process and improving nursing efficiency. By collecting feedback from the implemented nursing measures, the model and nursing methods can be continuously optimized, which helps to continuously improve the patient's treatment experience and results. The present invention uses an electronic health record system or a dedicated nursing management software to monitor the implementation of nursing measures in real time, ensure that nursing activities are carried out in a timely and correct manner in accordance with established guidelines, and provide a visual interface for interactive communication, so that patients can intuitively see their health indicators and nursing progress, promote communication and understanding between patients and medical teams, and through a shared interface, medical personnel of different professions can more easily access and understand the work of other team members and patient status, and promote collaborative work within the team.

[0064] Although the present invention has been disclosed as above with preferred embodiments, the embodiments are merely examples for the convenience of description and are not intended to limit the present invention. Those skilled in the art may make several changes and modifications without departing from the spirit and scope of the present invention. The scope of protection claimed by the present invention shall be based on the claims.

Claims

1. A system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients, characterized in that: The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients includes: a data acquisition module, an evaluation and analysis module, a nursing measure generation module, a nursing monitoring module, and an interaction and display module; The data collection module, the evaluation and analysis module, the nursing measures generation module, the nursing monitoring module and the interaction and display module are connected in sequence; The data collection module is used to integrate the patient's basic information, clinical data, health indicators and individual response to chemotherapy, and to create an electronic health record for patients undergoing chemotherapy for lung cancer; The evaluation and analysis module analyzes the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identifies the side effects caused by chemotherapy and provides corresponding management suggestions; Nursing measures generation module, which is used to generate personalized nursing measures based on the identification results and share the internal information of the nursing team through the collaboration platform; Nursing monitoring module, which is used to track the implementation of nursing measures and collect feedback from lung cancer chemotherapy patients, and adaptively adjust nursing measures based on the feedback; The interactive and display module is used to provide a visual interface for interactive communication and use visual tools to display health indicators and nursing progress; The evaluation and analysis module includes: analyzing the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identifying the side effects caused by chemotherapy and providing corresponding management suggestions; The Lasso regression model was used to screen variables related to the patient's quality of life and health status from the electronic health records, and the coefficient changes of different variables were observed by adjusting the penalty parameter value to identify the characteristic variables; The characteristic variables were used to construct a multivariate linear regression model, and the regression coefficients of each variable were analyzed to evaluate the specific impact of the regression coefficients on the patient's physical condition and quality of life; Perform F-test in multiple linear regression models and t-test for different variables; calculate the coefficient of determination and adjusted coefficient of determination at the same time; The multivariate linear regression model was tested using cross-validation method; Based on the F test, t test, coefficient of determination and adjusted coefficient of determination and the multivariate linear regression model after testing, the impact of each variable on the quality of life and health status of patients was analyzed, and the patient groups with side effects were identified and marked.

2. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 1, characterized in that: The Lasso regression model is used to screen variables related to the patient's quality of life and health status from the electronic health records, and by adjusting the penalty parameter value, the coefficient changes of different variables are observed, and the characteristic variables are identified, including: Extract relevant feature data from electronic health records and organize the extracted feature data into a feature matrix; According to the analysis requirements, determine the preset target and construct the corresponding target vector; Instantiate the Lasso regression model and set the initial penalty parameter value; The feature matrix and target vector are input into the Lasso regression model, and each penalty parameter value is tested by cross-validation method. The change of variable coefficient under each penalty parameter value is observed to identify the characteristic variables of the target variable.

3. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 2, characterized in that: The characteristic matrix and target vector are input into the Lasso regression model, each penalty parameter value is tested by cross-validation method, and the change of variable coefficient under each penalty parameter value is observed to identify the characteristic variables of the target variable, including: Input the prepared feature matrix and target vector into the Lasso regression model and configure the cross-validation parameters; Test a series of preset penalty parameter values ​​in the Lasso regression model and evaluate the performance of the Lasso regression model and the changes in the characteristic coefficients under each parameter; For each penalty parameter value obtained from the penalty parameter value test, retrain the model through each fold of cross-validation, record the characteristic coefficient of each fold, and observe the change of the characteristic coefficient with the penalty parameter value; The influence of each variable on the target vector is evaluated based on the stability and significance of the characteristic coefficient, and the variables with significant and stable coefficients under multiple penalty parameter value settings are identified and marked as key characteristic variables.

4. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 3, characterized in that: The method of constructing a multivariate linear regression model using characteristic variables, analyzing the regression coefficient of each variable, and evaluating the specific impact of the regression coefficient on the patient's physical condition and quality of life includes: Multiple linear regression models were constructed using the identified characteristic variables; Input the feature matrix and target vector into the multivariate linear regression model for training, and calculate the regression coefficient of each feature variable; The obtained regression coefficients were analyzed to obtain the positive or negative impact of each characteristic variable on the patient's physical condition and quality of life, and to evaluate the specific impact of each characteristic variable on the patient's condition.

5. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 4, characterized in that: The F test was performed in the multiple linear regression model, and the t test was performed on different variables; Calculating both the coefficient of determination and the adjusted coefficient of determination involves: The multivariate linear regression model was fitted using the least squares method to obtain the estimated values ​​of the model parameters; Based on the total sum of squares, regression sum of squares, and residual sum of squares, the F statistic is calculated, and the P value of the F statistic is obtained based on the degrees of freedom; Perform a t-test on the coefficient of each explanatory variable in the model, calculate the t-statistic for each coefficient, and find the P value of each t-statistic based on the degrees of freedom; The coefficient of determination and the adjusted coefficient of determination were calculated.

6. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 5, characterized in that: The expression for calculating the F statistic is: ； The expression for calculating the t statistic for each coefficient is: ； The expressions for calculating the coefficient of determination and the adjusted coefficient of determination are: ； ； In the formula, p represents the number of explanatory variables in the model; n represents the total number of samples; F express F test; t express t test; represents the estimated value of the coefficient; represents the standard error of the coefficient; represents the coefficient of determination; represents the adjusted coefficient of determination; represents the total sum of squares; represents the residual sum of squares; represents the regression sum of squares.

7. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 6, characterized in that: The F test, t test, determination coefficient and adjusted determination coefficient and the multivariate linear regression model after testing are used to analyze the impact of each variable on the patient's quality of life and health status, and identify and mark the patient groups with side effects, including: The F test was used to confirm the overall significance of the multivariate linear regression model, and the t test was used to identify the independent variables that had an impact on the dependent variable; The explanatory power of the multiple linear regression model was measured by the coefficient of determination, and the validity of the multiple linear regression model was determined by the adjusted coefficient of determination; Based on the output of the multivariate linear regression model, factors associated with reduced quality of life or worsening health status were identified, and the sources of side effects were located; cross-validation was used to test the stability and predictive ability of the model.

8. The system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to claim 1, characterized in that: The nursing measure generation module generates personalized nursing measures according to the recognition results and shares the internal information of the nursing team through the collaboration platform, including: Design a rule engine based on pre-defined clinical indicators and outputs of multivariate regression models; The random forest algorithm was used to analyze the electronic health records of lung cancer chemotherapy patients to predict their response to different nursing measures. Combine the identification results of the rule engine and the prediction results of the random forest model to generate personalized care measures for each patient; Through the collaborative platform, the generated nursing action plan is shared in real time to the nursing team’s internal information database.

9. The system for customizing nursing measures based on evaluation results of lung cancer chemotherapy patients according to claim 8, characterized in that: The random forest algorithm is used to analyze the electronic health records of lung cancer chemotherapy patients to predict the response of lung cancer chemotherapy patients to different nursing measures, including: Obtain electronic health records of patients undergoing chemotherapy for lung cancer and identify characteristic data related to chemotherapy response based on the electronic health records; A prediction model was constructed using the random forest algorithm to use the identified features to predict the response of lung cancer chemotherapy patients to different care interventions; Evaluate the performance of the model and use accuracy metrics to measure the effectiveness of the prediction model in predicting the response of lung cancer patients to chemotherapy; Predictive models are used to predict patient responses and generate personalized care recommendations for each lung cancer chemotherapy patient.

10. A method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients, used to implement the system for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients according to any one of claims 1 to 9, characterized in that: The method for customizing nursing measures based on the evaluation results of lung cancer chemotherapy patients comprises the following steps: S1. Integrate patients’ basic information, clinical data, health indicators and individual responses to chemotherapy, and create electronic health records for patients undergoing chemotherapy for lung cancer; S2. Analyze the physical condition, quality of life and disease risk of lung cancer chemotherapy patients based on electronic health records, identify the side effects caused by chemotherapy and provide corresponding management suggestions; S3, Generate personalized care measures based on the identification results and share the internal information of the care team through the collaborative platform; S4. Track the implementation of nursing measures and collect feedback from patients undergoing lung cancer chemotherapy, and adaptively adjust nursing measures based on the feedback; S5. Provide a visual interface for interactive communication and use visual tools to display health indicators and care progress.

Images