Risk early warning method and system for postoperative malnutrition of gastric cancer

Abstract

The invention provides a risk early warning method and system for stomach cancer postoperative malnutrition, and relates to the technical field of medical treatment, and the method comprises the steps: obtaining gene expression data and clinical indexes of a stomach cancer patient based on a cancer genome map database; gastric cancer methylation data is collected, and a methylation risk scoring model is constructed to determine a methylation risk score; performing quantitative evaluation on the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a nutritional risk level; determining risk characteristics related to stomach cancer postoperative dystrophy according to the gene expression data, the clinical indexes, the methylation risk score and the nutrition risk grade, and establishing a risk early warning model based on a support vector machine; and acquiring current risk data of the patient side, inputting the current risk data into the risk early warning model, outputting a risk score and a risk level, generating a risk intervention measure according to the risk level, and feeding back the risk intervention measure to the medical care side, so that accurate and comprehensive early warning can be performed on the risk of postoperative malnutrition of the gastric cancer patient, and a timely and effective nutrition intervention scheme is formulated.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a risk warning method and system for malnutrition after gastric cancer surgery. Background Art

[0002] Gastric cancer is one of the most common malignant tumors in the world, with high morbidity and mortality. Surgery is the main treatment for gastric cancer, however, postoperative malnutrition is a common complication after gastric cancer surgery, which seriously affects the patient's postoperative recovery and quality of life.

[0003] At present, although the Patient Subjective Global Assessment Scale (PG-SGA) has good reliability and validity, its evaluation results often rely on the patient's own feelings and the subjective judgment of medical staff, which may affect the accuracy of the evaluation results. In addition, most of the existing risk warning models are based on single-center, small-sample research data, and their generalization ability and prediction accuracy are poor. In the process of model establishment, some important variables may be ignored due to missing data or limitations of statistical methods, which in turn limits the predictive ability of the model. In addition, most of the existing warning models are established after surgery, and high-risk patients cannot be identified in time before or during surgery, which may lead to a delay in the implementation of nutritional intervention measures and miss the best time for nutritional support. In addition, the current nutritional support plans are often formulated based on unified nutritional guidelines and lack personalized intervention plans for patients, which may lead to poor nutritional support effects and even waste of resources.

[0004] Therefore, it is necessary to provide a risk warning method and system for malnutrition after gastric cancer surgery to solve the above technical problems. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a risk warning method and system for malnutrition after gastric cancer surgery, which are used to solve the problems that the existing risk warning methods cannot accurately and comprehensively warn the risk of malnutrition after gastric cancer surgery, and cannot formulate timely and effective nutritional intervention plans.

[0006] The present invention provides a method for early warning of the risk of malnutrition after gastric cancer surgery, the method comprising:

[0007] Based on the Cancer Genome Atlas database, we obtained gene expression data and clinical indicators of gastric cancer patients;

[0008] Constructing a methylation risk scoring model, collecting gastric cancer methylation data of the gastric cancer patient, and determining a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model;

[0009] quantitatively assessing the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a corresponding nutritional risk level;

[0010] Determine the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and construct a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine;

[0011] The current risk data of the patient is obtained and input into the risk warning model, the risk score and risk level of malnutrition of the patient are output, corresponding risk intervention measures are generated according to the risk level and fed back to the medical care end.

[0012] Preferably, the constructing of the methylation risk scoring model, collecting gastric cancer methylation data of the gastric cancer patient, and determining the corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model specifically includes:

[0013] Based on the gastric cancer methylation spectrum data set, collecting gastric cancer methylation data of the gastric cancer patient, and performing a preprocessing operation on the gastric cancer methylation data;

[0014] Identifying differentially methylated sites in the gastric cancer methylation data based on a bioinformatics method, wherein the differentially methylated sites include hypermethylated sites and hypomethylated sites;

[0015] The methylation risk score model is constructed, and the corresponding methylation risk score is calculated according to the identified differential methylation sites. The calculation formula of the methylation risk score is:

[0016] J=∑(L m *T m )

[0017] In the formula, J represents the methylation risk score; L m represents the weight coefficient of the mth differentially methylated site; T m Represents the methylation level value of the mth differentially methylated site.

[0018] Preferably, the nutritional status of the gastric cancer patient is quantitatively assessed based on the nutritional risk screening tool to obtain the corresponding nutritional risk level, specifically including:

[0019] The nutritional risk screening tool is selected according to clinical needs;

[0020] Using the nutritional risk screening tool to quantitatively evaluate the nutritional status of the gastric cancer patient to generate corresponding quantitative evaluation results;

[0021] Based on the quantitative assessment results, combined with the age, gender, weight, dietary intake and digestive tract symptoms of the gastric cancer patient, the nutritional risk level is comprehensively determined;

[0022] The generated nutritional risk level is adjusted and optimized accordingly according to the clinical indicators and methylation risk scores.

[0023] Preferably, the risk characteristics associated with malnutrition after gastric cancer surgery are determined according to the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and a risk warning model for malnutrition after gastric cancer surgery is constructed according to the risk characteristics based on a support vector machine, specifically comprising:

[0024] Based on the Pearson correlation coefficient, the correlation between the risk data corresponding to the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels and the risk scores was calculated. The calculation formula for the correlation was:

[0025]

[0026] Where r represents the correlation between the risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels and the risk scores; X d represents the risk data corresponding to the dth gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; represents the mean of the risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; Y d represents the corresponding d-th risk score; represents the mean of the corresponding risk score; D represents the total number of risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels;

[0027] Obtain a preset number of risk features, sort the risk data corresponding to the gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels based on the degree of association, and select the risk data whose sequence number is less than or equal to the preset number of risk features as the risk feature.

[0028] Preferably, based on a support vector machine, the objective function corresponding to the risk warning model is determined according to the risk characteristics, and the expression of the objective function is:

[0029]

[0030] In the formula, min represents the minimum operation; w represents the normal vector of the hyperplane; b represents the intercept of the hyperplane;

[0031] The constraints corresponding to the objective function are:

[0032]

[0033] In the formula, y i represents the risk category label corresponding to the ith risk feature; w represents the normal vector of the hyperplane; x i represents the i-th risk feature; b represents the intercept of the hyperplane; n represents the total number of risk features.

[0034] Preferably, a radial basis function kernel is introduced to map the risk features in the original space to a high-dimensional feature space. The calculation formula of the radial basis function kernel is:

[0035] K(x i , x j ) = exp(-γ||x i -x j || 2 )

[0036] Where K represents the radial basis function kernel; exp() represents the exponential function with the natural constant e as the base; γ represents the parameter of the radial basis function kernel; ||x i -x j || represents the i-th risk feature x i and the j-th risk feature x j The Euclidean distance between

[0037] Introducing the Lagrange multiplier α * , determine the Lagrange multiplier α * The corresponding support vector x * , generate the decision function corresponding to the risk warning model, the calculation formula of the decision function is:

[0038]

[0039] In the formula, FXPF(x) represents the calculation result of the decision function, that is, the risk score; x represents the input risk feature; α * represents the Lagrange multiplier; x * represents the support vector; y * represents the risk category label corresponding to the support vector; K represents the radial basis function kernel; b represents the intercept of the hyperplane; N represents the number of support vectors.

[0040] Preferably, the current risk data of the patient is obtained and input into the risk warning model, the risk score and risk level of malnutrition of the patient are output, and corresponding risk intervention measures are generated according to the risk level and fed back to the medical care end, specifically including:

[0041] Acquiring current risk data of the patient, performing preprocessing operations on the current risk data and inputting the data into the risk warning model, and calculating a risk score of malnutrition of the patient;

[0042] Get the minimum risk score T min and the highest risk score T max , based on the risk score, the patient is classified into risk levels:

[0043] If the risk score FXPF is less than or equal to the minimum risk score T min , the patient is at a low risk level;

[0044] If the risk score FXPF is greater than the minimum risk score T min , and is less than the highest risk score T max , then the patient is at a medium risk level;

[0045] If the risk score FXPF is greater than or equal to the highest risk score T max , the patient is at a high risk level;

[0046] Corresponding risk intervention measures are generated according to the risk level, and the risk score, risk level and risk intervention measures are displayed to the medical staff through a visual interface.

[0047] The present invention provides a risk warning system for malnutrition after gastric cancer surgery, the system comprising:

[0048] An acquisition module, used to obtain gene expression data and clinical indicators of gastric cancer patients based on the Cancer Genome Atlas database;

[0049] A collection module, used to construct a methylation risk scoring model, collect gastric cancer methylation data of the gastric cancer patient, and determine a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model;

[0050] An evaluation module, used to quantitatively evaluate the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a corresponding nutritional risk level;

[0051] A construction module is used to determine the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and to construct a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine;

[0052] The output module is used to obtain the current risk data of the patient and input it into the risk warning model, output the risk score and risk level of malnutrition of the patient, generate corresponding risk intervention measures according to the risk level and feed it back to the medical care end.

[0053] Compared with the related art, the risk warning method and system for malnutrition after gastric cancer surgery provided by the present invention have the following beneficial effects:

[0054] The present invention is based on the cancer genome atlas database, obtains the gene expression data and clinical indicators of gastric cancer patients; constructs a methylation risk scoring model, collects gastric cancer methylation data of gastric cancer patients, and determines the corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model; quantitatively evaluates the nutritional status of gastric cancer patients based on the nutritional risk screening tool to obtain the corresponding nutritional risk level; determines the risk characteristics related to malnutrition after gastric cancer surgery according to gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and constructs a risk warning model for malnutrition after gastric cancer surgery based on risk characteristics based on a support vector machine; obtains the current risk data of the patient end and inputs it into the risk warning model, outputs the risk score and risk level of malnutrition on the patient end, generates corresponding risk intervention measures according to the risk level and feeds back to the medical care end. The present invention can accurately warn the risk of malnutrition after gastric cancer surgery by comprehensively considering the patient's gene expression data, clinical indicators, methylation risk scores and nutritional risk levels; and can formulate personalized nutritional intervention plans, provide scientific intervention basis for medical staff, and reduce the incidence of postoperative malnutrition; at the same time, the present invention can improve the nutritional status of patients, improve surgical effects and patient prognosis through timely and effective nutritional intervention. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is a flow chart of a risk warning method for malnutrition after gastric cancer surgery of the present invention;

[0056] Figure 2 A schematic diagram of the patient-side risk level classification of the present invention;

[0057] Figure 3 A schematic diagram showing a visualization interface of the present invention;

[0058] Figure 4 The present invention is a system block diagram of a risk warning system for malnutrition after gastric cancer surgery. DETAILED DESCRIPTION

[0059] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.

[0060] Embodiment 1

[0061] like Figure 1 As shown, a risk warning method for malnutrition after gastric cancer surgery, the method comprising:

[0062] S1, based on the Cancer Genome Atlas database, obtain gene expression data and clinical indicators of gastric cancer patients;

[0063] S2, constructing a methylation risk scoring model, collecting gastric cancer methylation data of the gastric cancer patient, and determining a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model;

[0064] S3, quantitatively assessing the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a corresponding nutritional risk level;

[0065] S4, determining risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and constructing a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine;

[0066] S5, obtaining the current risk data of the patient and inputting it into the risk warning model, outputting the risk score and risk level of malnutrition of the patient, generating corresponding risk intervention measures according to the risk level and feeding back to the medical care end.

[0067] Among them, first of all, the gene expression data and clinical indicators of gastric cancer patients can be systematically obtained from the Cancer Genome Atlas (TCGA) database. These data include extensive molecular level information as well as the patient's treatment history and physiological status.

[0068] It is understandable that methylation, as an important modification of DNA, is closely related to the occurrence and development of tumors. In order to deeply understand the epigenetic characteristics of gastric cancer, a methylation risk scoring model can be further constructed. According to the gastric cancer methylation data of each patient, the corresponding methylation risk score can be accurately calculated. Based on this score, the degree of abnormal methylation in the patient's tumor tissue can be determined, providing a scientific basis for the subsequent evaluation of the malignant potential of the tumor.

[0069] Furthermore, professional nutritional risk screening tools, such as NRS-2002, can be used to conduct a comprehensive quantitative assessment of the nutritional status of gastric cancer patients and derive the nutritional risk level of each patient. This step not only takes into account the patient's weight changes and dietary intake, but also comprehensively considers the impact of the disease on their nutritional needs, providing a basis for the subsequent formulation of personalized nutritional intervention strategies.

[0070] By integrating gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels, and using bioinformatics methods, we can identify risk features that are closely related to malnutrition after gastric cancer surgery. Using the support vector machine (SVM) machine learning algorithm, we can build a risk warning model for malnutrition after gastric cancer surgery based on risk features.

[0071] Finally, the risk warning model can be applied to actual clinical scenarios. First, the current risk data on the patient side can be obtained, including the latest genetic test results, clinical examination indicators, methylation status and nutritional assessment results, and input into the risk warning model, which can then output the risk score and risk level of malnutrition on the patient side. According to these risk levels, targeted risk intervention measures can be automatically generated, such as adjusting diet plans, increasing nutritional supplements and optimizing treatment plans, and timely feedback to the medical side through a visual interface to guide clinical decision-making and promote rapid recovery of patients after surgery.

[0072] In a specific implementation process, the methylation risk scoring model is constructed, gastric cancer methylation data of the gastric cancer patient is collected, and the corresponding methylation risk score is determined according to the gastric cancer methylation data based on the methylation risk scoring model, specifically including:

[0073] Based on the gastric cancer methylation spectrum data set, collecting gastric cancer methylation data of the gastric cancer patient, and performing a preprocessing operation on the gastric cancer methylation data;

[0074] Identifying differentially methylated sites in the gastric cancer methylation data based on a bioinformatics method, wherein the differentially methylated sites include hypermethylated sites and hypomethylated sites;

[0075] The methylation risk score model is constructed, and the corresponding methylation risk score is calculated according to the identified differential methylation sites. The calculation formula of the methylation risk score is:

[0076] J=∑(L m *T m )

[0077] In the formula, J represents the methylation risk score; L m represents the weight coefficient of the mth differentially methylated site; T m Represents the methylation level value of the mth differentially methylated site.

[0078] It is understandable that gastric cancer methylation data of gastric cancer patients can be systematically collected from high-quality gastric cancer methylation profile datasets. The gastric cancer methylation profile datasets are usually derived from high-throughput sequencing technologies, such as whole genome bisulfite sequencing (WGBS) or reduced bisulfite sequencing (RRBS), which can accurately measure DNA methylation status.

[0079] Furthermore, the collected gastric cancer methylation data can be preprocessed to ensure the accuracy and reliability of the data. The specific preprocessing steps include data cleaning, missing value processing, batch effect correction, and standardization, thereby reducing the impact of technical variation and noise on subsequent analysis.

[0080] In practical applications, bioinformatics methods can be used to deeply analyze the pre-processed gastric cancer methylation data and identify differentially methylated sites (DMPs). Differentially methylated sites refer to DNA sites with significant differences in methylation levels between different samples (such as tumor tissue and normal tissue), which may directly affect the occurrence and development of gastric cancer. This step usually involves statistical tests, such as t-tests, Wilcoxon rank sum tests, etc., so that sites with significant changes in methylation levels can be determined, and these sites can be further subdivided into high methylation sites (hyper-DMPs) and low methylation sites (hypo-DMPs). Based on the identified differentially methylated sites, a methylation risk scoring model can be constructed. The model can assign a weight coefficient L to each differentially methylated site. m , these coefficients reflect the contribution of each site to the overall methylation risk score. The determination of weight coefficients usually relies on machine learning algorithms, such as Lasso regression or random forest, which can screen out the most predictive variables from a large number of features.

[0081] It should be noted that the calculation formula for the methylation risk score is J = ∑(L m *T m ), this calculation formula can integrate the methylation level values ​​T of all differentially methylated sites m and its corresponding weight coefficient L m , thereby quantifying the methylation risk level of individual patients and providing a scientific basis for the early diagnosis, prognosis assessment and formulation of personalized treatment strategies for gastric cancer.

[0082] The quantitative assessment of the nutritional status of the gastric cancer patient based on the nutritional risk screening tool to obtain the corresponding nutritional risk level specifically includes:

[0083] The nutritional risk screening tool is selected according to clinical needs;

[0084] Using the nutritional risk screening tool to quantitatively evaluate the nutritional status of the gastric cancer patient to generate corresponding quantitative evaluation results;

[0085] Based on the quantitative assessment results, combined with the age, gender, weight, dietary intake and digestive tract symptoms of the gastric cancer patient, the nutritional risk level is comprehensively determined;

[0086] The generated nutritional risk level is adjusted and optimized accordingly according to the clinical indicators and methylation risk scores.

[0087] In practical applications, nutritional risk screening tools can be used to conduct a systematic and quantitative nutritional status assessment of patients to determine their nutritional risk level.

[0088] Specifically, first of all, appropriate nutritional risk screening tools can be determined based on actual clinical needs. These tools usually include Nutritional Risk Screening 2002 (NRS-2002), Malnutrition Universal Screening Tool (MUST) and Mini Nutritional Assessment (MNA), etc., which are suitable for gastric cancer patients at different stages of the disease and with different nutritional needs. By comprehensively considering the specific conditions of the patients, the sensitivity and specificity of the screening tools, and hospital resources, the required nutritional risk screening tools can be determined to ensure the effectiveness and accuracy of the assessment results.

[0089] Subsequently, the selected nutritional risk screening tool can be used to conduct a comprehensive quantitative assessment of the nutritional status of gastric cancer patients. This step involves collecting multi-dimensional data such as the patient's weight, height, body mass index (BMI), recent weight changes, dietary intake, and digestive symptoms, and calculating the final quantitative assessment results. These results are usually presented in the form of scores or grades, which can intuitively reflect the patient's nutritional status.

[0090] Since the patient's age and gender may affect their basal metabolic rate and nutritional needs, and dietary intake is directly related to the intake of energy and nutrients, and gastrointestinal symptoms reflect the patient's ability to absorb and utilize nutrients, therefore, through comprehensive analysis of these factors, the patient's nutritional risk level can be more accurately determined, providing a scientific basis for subsequent nutritional intervention.

[0091] Finally, the preliminary nutritional risk level can be adjusted and optimized in combination with the patient's clinical indicators and methylation risk scores to ensure the comprehensiveness and accuracy of the nutritional risk assessment, so as to facilitate the subsequent formulation of more precise and personalized nutritional support strategies for gastric cancer patients and promote their recovery process.

[0092] The step of determining the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and constructing a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine specifically includes:

[0093] Based on the Pearson correlation coefficient, the correlation between the risk data corresponding to the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels and the risk scores was calculated. The calculation formula for the correlation was:

[0094]

[0095] Where r represents the correlation between the risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels and the risk scores; X d represents the risk data corresponding to the dth gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; represents the mean of the risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; Y d represents the corresponding d-th risk score; represents the mean of the corresponding risk score; D represents the total number of risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels;

[0096] Obtain a preset number of risk features, sort the risk data corresponding to the gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels based on the degree of association, and select the risk data whose sequence number is less than or equal to the preset number of risk features as the risk feature.

[0097] Based on the support vector machine, the objective function corresponding to the risk warning model is determined according to the risk characteristics. The expression of the objective function is:

[0098]

[0099] In the formula, min represents the minimum operation; w represents the normal vector of the hyperplane; b represents the intercept of the hyperplane;

[0100] The constraints corresponding to the objective function are:

[0101]

[0102] In the formula, y i represents the risk category label corresponding to the ith risk feature; w represents the normal vector of the hyperplane; x i represents the i-th risk feature; b represents the intercept of the hyperplane; n represents the total number of risk features.

[0103] The radial basis function kernel is introduced to map the risk features in the original space to the high-dimensional feature space. The calculation formula of the radial basis function kernel is:

[0104] K(x i , x j ) = exp(-γ||x i -x j || 2 )

[0105] Where K represents the radial basis function kernel; exp() represents the exponential function with the natural constant e as the base; γ represents the parameter of the radial basis function kernel; ||x i -x j || represents the i-th risk feature x i and the j-th risk feature x j The Euclidean distance between

[0106] Introducing the Lagrange multiplier α * , determine the Lagrange multiplier α * The corresponding support vector x * , generate the decision function corresponding to the risk warning model, the calculation formula of the decision function is:

[0107]

[0108] In the formula, FXPF(x) represents the calculation result of the decision function, that is, the risk score; x represents the input risk feature; α * represents the Lagrange multiplier; x * represents the support vector; y * represents the risk category label corresponding to the support vector; K represents the radial basis function kernel; b represents the intercept of the hyperplane; N represents the number of support vectors.

[0109] Among them, the Pearson correlation coefficient (PCC) can be used to quantify the correlation between gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels and malnutrition risk scores.

[0110] It should be noted that the Pearson correlation coefficient is a statistic that measures the degree of linear correlation between two variables. Its value is between -1 and 1. The larger the absolute value, the stronger the correlation. The present invention uses the following formula to calculate the degree of correlation between each risk data and the risk score: Among them, r represents the degree of association; X d and Y d represent the d-th risk data and risk score respectively; and Represent the means corresponding to the risk data and risk scores respectively; D represents the total number of risk data. By determining the degree of correlation between each risk data and the risk score, the contribution of each risk data to the risk of malnutrition can be evaluated. In order to screen out the most representative risk features, the threshold of the number of risk features can be preset in advance according to actual clinical needs, and the risk data can be sorted based on the calculated degree of correlation, and the risk data with a sequence number less than or equal to the threshold can be selected as the final risk feature.

[0111] Based on the selected risk features, a risk warning model for malnutrition after gastric cancer surgery can be constructed using support vector machine (SVM). Specifically, the corresponding objective function is: Among them, w represents the normal vector of the hyperplane; b represents the intercept of the hyperplane, and the objective function aims to minimize the norm of the hyperplane. At the same time, the objective function is subject to the following constraints: Among them, y i represents the risk category label corresponding to the ith risk feature; x i represents the i-th risk feature.

[0112] In order to deal with nonlinear problems, the radial basis function kernel can be introduced to map the risk features in the original space into the high-dimensional feature space, thereby increasing the linear separability. The calculation formula of the radial basis function kernel is: K(x i , x j ) = exp(-γ||x i -x j || 2 ), where γ represents the parameter of the radial basis function kernel, which can control the width of the radial basis function kernel; ||x i -x j || represents the Euclidean distance between the i-th and j-th risk features. Finally, the Lagrange multiplier can be introduced, and its corresponding support vector can be determined to obtain the decision function corresponding to the risk warning model: Among them, FXPF(x) represents the calculation result of the decision function, that is, the risk score; α * represents the Lagrange multiplier; x * represents the support vector; N represents the number of support vectors; y * Represents the risk category label corresponding to the support vector.

[0113] By integrating multi-source data and using the Pearson correlation coefficient to screen risk features, we can construct a risk warning model for malnutrition after gastric cancer surgery, and introduce the radial basis function kernel to deal with the nonlinear problems therein, so that we can provide an accurate warning of the risk of malnutrition after gastric cancer surgery, facilitate the subsequent formulation of personalized nutritional intervention plans, provide scientific intervention basis for medical staff, and reduce the incidence of postoperative malnutrition.

[0114] The current risk data of the patient is obtained and input into the risk warning model, the risk score and risk level of malnutrition of the patient are output, and corresponding risk intervention measures are generated according to the risk level and fed back to the medical care end, specifically including:

[0115] Acquiring current risk data of the patient, performing preprocessing operations on the current risk data and inputting the data into the risk warning model, and calculating a risk score of malnutrition of the patient;

[0116] Get the minimum risk score T min and the highest risk score T max , based on the risk score, the patient is classified into risk levels:

[0117] If the risk score FXPF is less than or equal to the minimum risk score T min , the patient is at a low risk level;

[0118] If the risk score FXPF is greater than the minimum risk score T min , and is less than the highest risk score T max , then the patient is at a medium risk level;

[0119] If the risk score FXPF is greater than or equal to the highest risk score T max , the patient is at a high risk level;

[0120] Corresponding risk intervention measures are generated according to the risk level, and the risk score, risk level and risk intervention measures are displayed to the medical staff through a visual interface.

[0121] Specifically, first of all, current risk data from the patient can be systematically collected through an electronic health record system or dedicated data collection tools, including key information such as weight changes, dietary intake, biomarker levels, body mass index (BMI), and past medical history.

[0122] In order to ensure data quality and meet the model input requirements, the collected current risk data can be preprocessed, including data cleaning, data conversion, and data integration to ensure data consistency and accuracy, and data cleaning operations include removing outliers and filling missing values; data conversion operations include standardization; data integration operations include integrating information from different sources. Furthermore, the preprocessed current risk data can be input into the pre-built risk warning model to calculate the risk score of malnutrition on the patient side.

[0123] In order to intuitively display the nutritional risk status of patients, a minimum risk score and a maximum risk score can be set as a benchmark. These two thresholds are usually determined based on clinical guidelines, expert consensus, and statistical analysis results of a large amount of historical data. Figure 2 As shown, the patient side can be divided into risk levels according to the calculated risk score: if the risk score is less than or equal to the minimum risk score, the patient side is judged to be at a low risk level, that is, the patient side currently has a low risk of malnutrition; if the risk score is greater than the minimum risk score and less than the maximum risk score, the patient side is judged to be at a medium risk level; if the risk score is greater than or equal to the highest risk score, the patient side is judged to be at a high risk level, which means that the patient side faces a serious risk of malnutrition and requires immediate emergency intervention measures.

[0124] Finally, personalized risk intervention measures can be automatically generated based on the determined risk level, including adjusting diet plans, adding nutritional supplements, optimizing treatment plans, and recommending further medical evaluations. Figure 3 As shown, the risk score, risk level and risk intervention measures corresponding to the patient will be clearly displayed to the medical staff through the visual interface of the terminal device, which is convenient for medical staff to quickly understand the patient's condition and formulate or adjust the care plan, thereby effectively improving the patient's nutritional management efficiency and treatment effect.

[0125] Embodiment 2

[0126] like Figure 4 As shown, a risk warning system for malnutrition after gastric cancer surgery, the system comprises:

[0127] An acquisition module, used to obtain gene expression data and clinical indicators of gastric cancer patients based on the Cancer Genome Atlas database;

[0128] A collection module, used to construct a methylation risk scoring model, collect gastric cancer methylation data of the gastric cancer patient, and determine a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model;

[0129] An evaluation module, used to quantitatively evaluate the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a corresponding nutritional risk level;

[0130] A construction module is used to determine the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and to construct a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine;

[0131] The output module is used to obtain the current risk data of the patient and input it into the risk warning model, output the risk score and risk level of malnutrition of the patient, generate corresponding risk intervention measures according to the risk level and feed it back to the medical care end.

[0132] Through the introduction of the above embodiments, the present invention uses a risk warning method and system for malnutrition after gastric cancer surgery, and obtains gene expression data and clinical indicators of gastric cancer patients based on a cancer genome atlas database; constructs a methylation risk scoring model, collects gastric cancer methylation data of gastric cancer patients, and determines the corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model; quantitatively evaluates the nutritional status of gastric cancer patients based on a nutritional risk screening tool to obtain a corresponding nutritional risk level; determines risk characteristics related to malnutrition after gastric cancer surgery based on gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and constructs a risk warning model for malnutrition after gastric cancer surgery based on risk characteristics based on a support vector machine; obtains the current risk data on the patient side and inputs it into the risk warning model, outputs the risk score and risk level of malnutrition on the patient side, generates corresponding risk intervention measures according to the risk level and feeds back to the medical side. The present invention can provide accurate early warning of the risk of postoperative malnutrition in gastric cancer patients by comprehensively considering the patient's gene expression data, clinical indicators, methylation risk scores and nutritional risk levels; and can formulate personalized nutritional intervention plans to provide medical staff with a scientific basis for intervention and reduce the incidence of postoperative malnutrition; at the same time, the present invention can improve the patient's nutritional status, improve the surgical effect and the patient's prognosis through timely and effective nutritional intervention.

[0133] The present application is described with reference to the flowcharts and / or block diagrams of the methods, devices (systems), and computer program products according to the embodiments of the present application. It should be understood that each process and / or box in the flowchart and / or block diagram, as well as the combination of the processes and / or boxes in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0134] Those skilled in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing related hardware through a program, and the program can be stored in a computer-readable storage medium, and the storage medium includes a read-only memory (ROM), a random access memory (RAM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), a one-time programmable read-only memory (OTPROM), an electronically erasable rewritable read-only memory (EEPROM), a compact disc (CD-ROM) or other optical disc storage, magnetic disk storage, magnetic tape storage, or any other computer-readable medium that can be used to carry or store data.

[0135] It should also be noted that the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, method, commodity or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, commodity or device. In the absence of more restrictions, the elements defined by the sentence "comprises a ..." do not exclude the existence of other identical elements in the process, method, commodity or device including the elements.

Claims

1. A risk warning method for malnutrition after gastric cancer surgery, characterized in that: The method comprises: Based on the Cancer Genome Atlas database, we obtained gene expression data and clinical indicators of gastric cancer patients; Constructing a methylation risk scoring model, collecting gastric cancer methylation data of the gastric cancer patient, and determining a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model; quantitatively assessing the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a corresponding nutritional risk level; Determine the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and construct a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine; The current risk data of the patient is obtained and input into the risk warning model, the risk score and risk level of malnutrition of the patient are output, corresponding risk intervention measures are generated according to the risk level and fed back to the medical care end.

2. The method for early warning of risk of malnutrition after gastric cancer surgery according to claim 1, characterized in that: The step of constructing a methylation risk scoring model, collecting gastric cancer methylation data of the gastric cancer patient, and determining a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model specifically includes: Based on the gastric cancer methylation spectrum data set, collecting gastric cancer methylation data of the gastric cancer patient, and performing a preprocessing operation on the gastric cancer methylation data; Identifying differentially methylated sites in the gastric cancer methylation data based on a bioinformatics method, wherein the differentially methylated sites include hypermethylated sites and hypomethylated sites; The methylation risk score model is constructed, and the corresponding methylation risk score is calculated according to the identified differential methylation sites. The calculation formula of the methylation risk score is: J=Σ(L m *T m ) In the formula, J represents the methylation risk score; L m represents the weight coefficient of the mth differentially methylated site; T m Represents the methylation level value of the mth differentially methylated site.

3. The method for early warning of risk of malnutrition after gastric cancer surgery according to claim 1, characterized in that: The quantitative assessment of the nutritional status of the gastric cancer patient based on the nutritional risk screening tool to obtain the corresponding nutritional risk level specifically includes: The nutritional risk screening tool is selected according to clinical needs; Using the nutritional risk screening tool to quantitatively evaluate the nutritional status of the gastric cancer patient to generate corresponding quantitative evaluation results; Based on the quantitative assessment results, combined with the age, gender, weight, dietary intake and digestive tract symptoms of the gastric cancer patient, the nutritional risk level is comprehensively determined; The generated nutritional risk level is adjusted and optimized accordingly according to the clinical indicators and methylation risk scores.

4. The method for early warning of risk of malnutrition after gastric cancer surgery according to claim 1, characterized in that: The step of determining the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and constructing a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine specifically includes: Based on the Pearson correlation coefficient, the correlation between the risk data corresponding to the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels and the risk scores was calculated. The calculation formula for the correlation was: Where r represents the correlation between the risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels and the risk scores; X d represents the risk data corresponding to the dth gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; represents the mean of the risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; Y d represents the corresponding d-th risk score; represents the mean of the corresponding risk score; D represents the total number of risk data corresponding to gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels; Obtain a preset number of risk features, sort the risk data corresponding to the gene expression data, clinical indicators, methylation risk scores, and nutritional risk levels based on the degree of association, and select the risk data whose sequence number is less than or equal to the preset number of risk features as the risk feature.

5. The method for early warning of risk of malnutrition after gastric cancer surgery according to claim 4, characterized in that: Based on the support vector machine, the objective function corresponding to the risk warning model is determined according to the risk characteristics. The expression of the objective function is: In the formula, min represents the minimum operation; w represents the normal vector of the hyperplane; b represents the intercept of the hyperplane; The constraints corresponding to the objective function are: In the formula, y i represents the risk category label corresponding to the ith risk feature; w represents the normal vector of the hyperplane; x i represents the i-th risk feature; b represents the intercept of the hyperplane; n represents the total number of risk features.

6. The method for early warning of risk of malnutrition after gastric cancer surgery according to claim 5, characterized in that: The radial basis function kernel is introduced to map the risk features in the original space to the high-dimensional feature space. The calculation formula of the radial basis function kernel is: K(x i ,x j )=exp(-γ||x i -x j || 2 ) Where K represents the radial basis function kernel; exp() represents the exponential function with the natural constant e as the base; γ represents the parameter of the radial basis function kernel; ||x i -x j || represents the i-th risk feature x i and the j-th risk feature x j The Euclidean distance between The Lagrange multiplier α* is introduced, the support vector x* corresponding to the Lagrange multiplier a* is determined, and the decision function corresponding to the risk warning model is generated. The calculation formula of the decision function is: Where FXPF(x) represents the calculation result of the decision function, i.e., the risk score; x represents the input risk feature; α* represents the Lagrange multiplier; x* represents the support vector; y* represents the risk category label corresponding to the support vector; K represents the radial basis function kernel; b represents the intercept of the hyperplane; and N represents the number of support vectors.

7. The method for early warning of risk of malnutrition after gastric cancer surgery according to claim 1, characterized in that: The current risk data of the patient is obtained and input into the risk warning model, the risk score and risk level of malnutrition of the patient are output, and corresponding risk intervention measures are generated according to the risk level and fed back to the medical care end, specifically including: Acquiring current risk data of the patient, performing preprocessing operations on the current risk data and inputting the data into the risk warning model, and calculating a risk score of malnutrition of the patient; Get the minimum risk score T min and the highest risk score T max , based on the risk score, the patient is classified into risk levels: If the risk score FXPF is less than or equal to the minimum risk score T min , the patient is at a low risk level; If the risk score FXPF is greater than the minimum risk score T min , and is less than the highest risk score T max , then the patient is at a medium risk level; If the risk score FXPF is greater than or equal to the highest risk score T max , the patient is at a high risk level; Corresponding risk intervention measures are generated according to the risk level, and the risk score, risk level and risk intervention measures are displayed to the medical staff through a visual interface.

8. A risk warning system for malnutrition after gastric cancer surgery, applied to a risk warning method for malnutrition after gastric cancer surgery as claimed in any one of claims 1 to 7, the system comprising: An acquisition module, used to obtain gene expression data and clinical indicators of gastric cancer patients based on the Cancer Genome Atlas database; A collection module, used to construct a methylation risk scoring model, collect gastric cancer methylation data of the gastric cancer patient, and determine a corresponding methylation risk score according to the gastric cancer methylation data based on the methylation risk scoring model; An evaluation module, used to quantitatively evaluate the nutritional status of the gastric cancer patient based on a nutritional risk screening tool to obtain a corresponding nutritional risk level; A construction module is used to determine the risk characteristics associated with malnutrition after gastric cancer surgery based on the gene expression data, clinical indicators, methylation risk scores and nutritional risk levels, and to construct a risk warning model for malnutrition after gastric cancer surgery based on the risk characteristics based on a support vector machine; The output module is used to obtain the current risk data of the patient and input it into the risk warning model, output the risk score and risk level of malnutrition of the patient, generate corresponding risk intervention measures according to the risk level and feed it back to the medical care end.

Images