Skin temperature monitoring system based on tumor thermal therapy patient

By designing a skin temperature monitoring system that includes preference data collection, index system construction and machine learning models in tumor thermal therapy, the problem of lack of real-time monitoring and personalized solutions in traditional thermal therapy is solved, and the personalization and accuracy of treatment is improved. , improve patient satisfaction and treatment effect.

CN119004274AInactive Publication Date: 2024-11-22NANTONG TUMOR HOSPITAL

Patent Information

Application Number
CN202411486192.1
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-10-23
Publication Date
2024-11-22
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

The lack of real-time skin temperature monitoring and personalized treatment plans in traditional tumor thermal therapy leads to poor treatment results and patient discomfort, and a lack of effective solutions.

Method used

Design a skin temperature monitoring system based on tumor thermal therapy patients, including preference data collection module, index system construction module, timeliness determination module, model construction module, real-time monitoring module, prediction module and personalized recommendation module, and real-time through machine learning models Monitor skin temperature and provide personalized treatment recommendations.

Benefits of technology

It improves the personalization and accuracy of the treatment plan, enhances patient satisfaction and treatment compliance, and achieves more precise treatment results by predicting skin temperature risks in real time and adjusting treatment parameters.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119004274A_ABST
    Figure CN119004274A_ABST
Patent Text Reader

Abstract

The invention relates to the field of tumor thermal therapy, and discloses a skin temperature monitoring system based on a tumor thermal therapy patient, and the skin temperature monitoring system based on the tumor thermal therapy patient comprises a preference data collection module, an index system construction module, an aging determination module, a model construction module, a real-time monitoring module, a prediction module and a personalized recommendation module. The preference data collection module is connected with the index system construction module, the index system construction module is connected with the timeliness determination module and the model construction module, the model construction module is connected with the prediction module, the prediction module is connected with the real-time monitoring module, and the real-time monitoring module is connected with the personalized recommendation module. According to the invention, individual demands and preferences of the patient can be considered in the treatment scheme, and the satisfaction degree and treatment compliance of the patient are improved; by analyzing preference data, a comprehensive index system is created, and the treatment effect and patient satisfaction can be evaluated more accurately.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the field of tumor hyperthermia, and in particular to a skin temperature monitoring system for patients undergoing tumor hyperthermia. Background Art

[0002] Tumor hyperthermia is a treatment method that uses high temperature to directly kill tumor cells or enhance the effects of other treatments (such as chemotherapy and radiotherapy). The effectiveness of hyperthermia depends largely on the ability to accurately control and maintain the temperature in the treatment area within an appropriate treatment window. Too high a temperature may cause damage to normal tissues, while too low a temperature may lead to poor treatment effects. Therefore, it is particularly important to monitor the patient's skin temperature during tumor hyperthermia.

[0003] In traditional hyperthermia treatment, doctors usually rely on experience and regularly measured skin temperature data to adjust treatment parameters. This method has certain limitations. First, the intermittent nature of the measurement may result in insufficient continuity of the data and fail to reflect the actual changes in skin temperature during the hyperthermia process in real time. Second, doctors need to use experience to judge when to adjust the output of the treatment device. This method that relies on personal experience may be subjective, leading to uncertainty in the treatment effect. In addition, due to the lack of personalized treatment plans, traditional methods may not fully take into account the specific circumstances of each patient.

[0004] For example, Chinese patent 201610533332.5 discloses a skin temperature monitoring system for tumor hyperthermia, which includes multiple bags, heat exchange chambers, circulation pipes, etc., which can better control the skin temperature of patients during tumor hyperthermia. However, the above method still has the following shortcomings: not considering the patient's personal preferences and specific circumstances may result in the treatment plan being unable to be adjusted to the unique needs of each patient. This may cause some patients to feel uncomfortable or the treatment effect may be poor. In addition, not considering the patient's preferences may lead to low patient participation, thereby affecting their compliance with the treatment plan and overall satisfaction.

[0005] Currently, no effective solution has been proposed for the problems in the related technologies. Summary of the invention

[0006] In view of this, the purpose of the embodiments of the present application is to provide a skin temperature monitoring system for patients undergoing tumor hyperthermia therapy to overcome the above-mentioned technical problems existing in the existing related technologies.

[0007] To this end, the specific technical solution adopted by the present invention is as follows: A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy, comprising a preference data collection module, an indicator system construction module, a timeliness determination module, a model construction module, a real-time monitoring module, a prediction module and a personalized recommendation module; wherein the preference data collection module is connected to the indicator system construction module, the indicator system construction module is connected to the timeliness determination module and the model construction module, the model construction module is connected to the prediction module, the prediction module is connected to the real-time monitoring module, and the real-time monitoring module is connected to the personalized recommendation module; A preference data collection module is used to collect patient information and preference information for hyperthermia therapy to obtain a preference data set; The indicator system construction module is used to build a personalized treatment indicator system based on the patient's preference data; The time determination module is used to determine the optimal duration and temperature setting of each thermal treatment, and record the actual time information of the treatment to obtain historical treatment data; Model building module, used to build machine learning models based on historical treatment data and patient-specific treatment indicators; Real-time monitoring module, used to monitor the patient's skin temperature in real time during the thermal therapy process, and collect treatment time information and instant feedback from the patient; The prediction module is used to use the built machine learning model to predict the risk of the patient's skin temperature in real time. If a high risk is predicted, the treatment parameters are adjusted immediately; The personalized recommendation module is used to provide personalized treatment suggestions based on the prediction results of the machine learning model and the patient's personalized treatment indicator system.

[0008] Furthermore, the indicator system construction module includes a data analysis module and a system establishment module; The data analysis module is used to analyze the potential relationships in the preference data set, and to group patients using clustering algorithms. At the same time, it uses classification and regression algorithms to obtain the results of patients' satisfaction with the treatment plan. The system establishment module is used to select several indicators that are most relevant to treatment effects and patient satisfaction, classify the indicators, and assign weights to each indicator.

[0009] Furthermore, in the selected feature set, when analyzing the potential relationship in the data, each variable in the selected feature set is classified and determined to be quantitative or qualitative data, and the quantitative data is subjected to a distribution test to determine whether it obeys a normal distribution, and a report on the data type and distribution characteristics is obtained; According to the data type and distribution characteristics in the distribution characteristic report, a corresponding correlation analysis algorithm is selected, and the corresponding correlation analysis algorithm is applied to calculate the correlation coefficients between the variables to obtain the correlation coefficient matrix between each variable pair; Conduct hypothesis tests on each correlation coefficient to obtain significance test results, and based on the significance test results, identify and eliminate several variables with the lowest correlation with the treatment effect to obtain a streamlined variable set; Integrate the streamlined variable set and its correlation coefficients and significance test results to write a data relationship report.

[0010] Furthermore, when selecting the corresponding correlation analysis algorithm based on the data type and distribution characteristics in the distribution characteristics report, the Pearson correlation coefficient is selected for quantitative and normally distributed data; For nonnormal or rank data, choose the Spearman correlation coefficient; For qualitative data, choose the Chi-Square test.

[0011] Furthermore, when obtaining the results of patient satisfaction with the treatment plan through classification and regression algorithms, the correlation between each feature and the patient's satisfaction with the treatment plan is analyzed in the selected feature set, and several features that are most correlated with the treatment satisfaction are selected to obtain a screening feature set; Extract additional derived features based on the features in the filtered feature set, process the time series data, extract trend and periodic features, add the derived features to the filtered feature set, obtain an extended feature set, and convert the categorical features into numerical features using encoding technology to obtain an encoded feature set; Obtain patient satisfaction data on treatment plans from historical records and use classification and regression algorithms to build a satisfaction prediction model.

[0012] Furthermore, when using the classification regression algorithm to construct a satisfaction prediction model, if satisfaction is categorical data, a classification model is constructed; If satisfaction is continuous rating data, a regression model is constructed.

[0013] Furthermore, after constructing the satisfaction prediction model, a training data set is constructed, which includes the coded features and satisfaction of the patients, and the parameters of the satisfaction prediction model are trained and tuned using cross-validation technology, and the performance of the satisfaction prediction model on the validation set is evaluated; A test data set was constructed, which contained the encoded features and satisfaction of patients not used for training. The satisfaction prediction model was evaluated using the test data set, and the performance indicators of the satisfaction prediction model were calculated. At the same time, the satisfaction of patients with the treatment plan was predicted by the tested satisfaction prediction model.

[0014] Furthermore, when predicting the patient's satisfaction with the treatment plan through the tested satisfaction prediction model, characteristic data of new patients are obtained, and the format of the characteristic data of new patients is consistent with the data format used in the training of the satisfaction prediction model; The tested satisfaction prediction model is used to predict the satisfaction of the characteristic data of the new patient. If the satisfaction prediction model is a classification model, the category label of the satisfaction is output; if the satisfaction prediction model is a regression model, a continuous satisfaction score is output.

[0015] Furthermore, the model building module includes a feature selection module, a model training module and a model verification module; Among them, the feature selection module is used to pre-process the historical treatment data and use statistical methods to select features related to skin temperature risk to obtain a model input feature set, and divide the model input feature set into a training set and a validation set; Model training module, used to train the machine learning model using the training set; The model validation module is used to test the performance of the machine learning model on the validation set and adjust the machine learning model parameters to optimize the performance.

[0016] Furthermore, when using statistical methods to select features related to skin temperature risk, if the features are continuous, the point bi-order correlation coefficient is calculated, where the formula for calculating the point bi-order correlation coefficient is: ; represents the point bi-order correlation coefficient; X i and Y i Respectively represent i The first and second continuous eigenvalues ​​of samples; and Represent the sample means of the first and second continuous features respectively; If it is a categorical feature, the chi-square test is used to obtain the correlation measure between each feature and skin temperature risk, and multivariate statistical methods are used to evaluate the interaction between features, analyze the overall contribution of the features to skin temperature risk, and obtain multivariate analysis results; Based on the results of multivariate analysis, features are selected according to statistical significance and practical significance to determine the model input feature set.

[0017] Embodiments of the present invention include the following benefits: (1) The skin temperature monitoring system for patients undergoing tumor hyperthermia provided by the present invention ensures that the treatment plan can take into account the individual needs and preferences of the patient, thereby improving patient satisfaction and treatment compliance. By analyzing preference data and creating a comprehensive indicator system, the treatment effect and patient satisfaction can be evaluated more accurately.

[0018] (2) The present invention analyzes the potential relationships of data in the preference data set, uses clustering algorithms to group patients, identifies different preference patterns, and predicts patient satisfaction with treatment plans through classification and regression algorithms. Clustering algorithms can help identify preference patterns of different patient groups, which is crucial for personalized treatment. The potential relationships in the data are analyzed to obtain several indicators that are most relevant to the treatment effect; the use of classification and regression algorithms can predict the patient's satisfaction with the treatment plan based on the patient's characteristic data, thereby providing an important reference dimension for constructing an indicator system. At the same time, the analysis determines the indicators that are significantly related to the treatment effect and satisfaction, and these indicators will constitute the core of the personalized treatment indicator system. The classification and weight allocation of indicators reflect the importance of different indicators in treatment, which helps to formulate more accurate and personalized treatment plans.

[0019] (3) Use historical treatment data and personalized indicators to build machine learning models to make treatment decisions more data-driven and accurate. Use machine learning models to predict the risk of patient skin temperature in real time, enabling earlier intervention and treatment adjustments. Provide personalized treatment recommendations based on prediction results and treatment indicator systems to further improve the personalization and accuracy of treatment. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0021] Figure 1 The schematic diagram of a skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to an embodiment of the present invention is shown in FIG.

[0022] In the figure: 1. Preference data collection module; 2. Index system construction module; 3. Timeliness determination module; 4. Model construction module; 5. Real-time monitoring module; 6. Prediction module; 7. Personalized recommendation module. DETAILED DESCRIPTION

[0023] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0024] To further illustrate each embodiment, the present invention provides drawings, which are part of the disclosure of the present invention and are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these contents, ordinary technicians in the field should be able to understand other possible implementations and advantages of the present invention. The components in the figures are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0025] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0026] According to an embodiment of the present invention, a skin temperature monitoring system for patients undergoing tumor hyperthermia therapy is provided.

[0027] The present invention is further described with reference to the accompanying drawings and specific embodiments. Figure 1 As shown, according to an embodiment of the present invention, the skin temperature monitoring system for patients with tumor hyperthermia therapy includes a preference data collection module 1, an indicator system construction module 2, a timeliness determination module 3, a model construction module 4, a real-time monitoring module 5, a prediction module 6 and a personalized recommendation module 7; Among them, the preference data collection module 1 is connected to the indicator system construction module 2, the indicator system construction module 2 is connected to the timeliness determination module 3 and the model construction module 4, the model construction module 4 is connected to the prediction module 6, the prediction module 6 is connected to the real-time monitoring module 5, and the real-time monitoring module 5 is connected to the personalized recommendation module 7; The preference data collection module 1 is used to collect the basic information of the patient and the preference information for thermal therapy to obtain a preference data set; Specifically, the patient's personal information (age, gender, weight, height), medical background (condition, treatment experience, drug use), lifestyle (diet, activity level, sleep), preference for heat therapy (treatment time, frequency, temperature) and psychological condition (treatment attitude, anxiety level) are collected, and a structured patient data set is output. The structured patient data set is cleaned and feature selection is performed at the same time to screen out several features that are most relevant to the construction of the treatment indicator system, and the selected feature set, i.e., the preference data set, is obtained.

[0028] Indicator system construction module 2, used to construct a personalized treatment indicator system based on the patient's preference data; In a further embodiment, the indicator system construction module includes a data analysis module and a system establishment module; The data analysis module is used to analyze the potential relationships in the preference data set, and use clustering algorithms to group patients and identify different preference patterns. At the same time, classification and regression algorithms are used to obtain the patient's satisfaction with the treatment plan. It should be noted that clustering algorithm is an unsupervised learning algorithm that groups objects in a data set so that objects in the same group (i.e., a cluster) have a high degree of similarity, while objects in different groups have a low degree of similarity. The key to clustering algorithm is to define a measure of "similarity", which is usually achieved through some distance metric, such as Euclidean distance, Manhattan distance, or cosine similarity. In medical data analysis, clustering algorithms are used to identify groups of patients with similar medical characteristics, so as to provide more targeted treatment plans or medical services for these groups.

[0029] The system establishment module is used to select several indicators that are most relevant to treatment effectiveness and patient satisfaction, classify the indicators (physiological, therapeutic, psychological, etc.), and assign weights to each indicator to reflect its importance in personalized treatment.

[0030] In a further embodiment, when analyzing the potential relationships in the data in a selected feature set, each variable in the selected feature set (such as age, condition, treatment history, thermal therapy temperature preference, etc.) is classified and determined to be quantitative or qualitative data, and a distribution test is performed on the quantitative data to determine whether it obeys a normal distribution, and a data type and distribution characteristic report is obtained.

[0031] According to the data type and distribution characteristics in the distribution characteristics report, the corresponding correlation analysis algorithm is selected, and the corresponding correlation analysis algorithm is applied to calculate the correlation coefficient between the variables, such as calculating the correlation between the hyperthermia temperature preference and the treatment effect, to obtain the correlation coefficient matrix between each pair of variables.

[0032] A hypothesis test is performed on each correlation coefficient to determine whether it is statistically significant, that is, whether the relationship between the variables reflects the overall trend, and the significance test results are obtained. Based on the significance test results, several variables with the lowest correlation with the treatment effect (target variable), such as certain lifestyle factors, are identified and eliminated to obtain a streamlined set of variables, that is, several indicators most relevant to the treatment effect, which only include variables that are significantly related to the treatment effect.

[0033] It should be noted that the calculated correlation coefficient itself does not directly indicate whether the correlation is statistically significant, that is, whether the relationship is strong enough that it is unlikely to be caused by random factors. In order to determine this, a hypothesis test is required for the correlation coefficient. For the Pearson coefficient, a t-test is usually used to assess its significance. The basic idea of ​​hypothesis testing is to set a null hypothesis (H0), that is, there is no correlation between the two variables (the correlation coefficient is 0), and then calculate the probability of observing the current or more extreme correlation coefficient (ie, p-value) when this null hypothesis is true.

[0034] Assume that the result of the test is a p-value, which indicates the probability of obtaining the current observation result (or more extreme) if the null hypothesis is true. If this p-value is less than the pre-set significance level (usually 0.05 or 0.01), the null hypothesis is rejected and the correlation between the two variables is considered significant. Through hypothesis testing, determine which variables are statistically significantly correlated with the treatment effect (target variable). Then select those variables with significant correlation and eliminate those with insignificant correlation. For example, if certain lifestyle factors are not significantly correlated with the treatment effect, these factors are excluded from the model.

[0035] Obtain a simplified variable set: Through the above process, a simplified variable set containing variables significantly related to the treatment effect is finally obtained. These variables will be used to build a satisfaction prediction model or further analysis because they are more likely to reflect the real factors that affect the treatment effect.

[0036] At the same time, the streamlined variable set and its correlation coefficients and significance test results were integrated to write a data relationship report, which described the correlation between the variables in detail, and especially emphasized the variables that had an important impact on the therapeutic effect of hyperthermia.

[0037] In a further embodiment, when selecting a corresponding correlation analysis algorithm based on the data type and distribution characteristics in the distribution characteristic report, for quantitative and normally distributed data, the Pearson correlation coefficient is selected; for non-normal or rank data, the Spearman correlation coefficient is selected; for qualitative data, the chi-square test is selected.

[0038] It should be noted that the Pearson correlation coefficient is applicable to both variables when they are quantitative data and when these data are normally distributed, the Pearson correlation coefficient is used to measure the linear relationship between them. The Pearson correlation coefficient reflects the strength and direction of the linear relationship between the two variables, and the coefficient value ranges from -1 (completely negative linear relationship) to +1 (completely positive linear relationship), and 0 indicates no linear relationship.

[0039] The Spearman correlation coefficient is applicable to quantitative data when at least one variable is non-normally distributed, or when the data is rank (order) data, the Spearman correlation coefficient is used to measure the monotonic relationship between them. The Spearman correlation coefficient is based on the rank of the variable value rather than the actual value, and is applicable to the measurement of nonlinear relationships. The coefficient value range is also from -1 to +1.

[0040] The chi-square test is used to determine whether there is an association between two variables when the data is qualitative (categorical) data. The chi-square test compares the difference between the observed frequency and the expected frequency to assess the independence between the variables. The test results include the chi-square statistic and the corresponding p-value, which are used to determine the significance of the association.

[0041] In a further embodiment, when obtaining the patient satisfaction result with the treatment plan through the classification and regression algorithm, the correlation between each feature and the patient satisfaction with the treatment plan is analyzed in the selected feature set, and several features most correlated with the treatment satisfaction are selected to obtain a screening feature set; Extract additional derived features based on the features in the screening feature set, such as the time point after the start of treatment, changes in the patient's psychological condition, etc., and process the time series data to extract trend and periodic features. At the same time, add the derived features to the screening feature set to obtain an extended feature set, and convert categorical features (such as gender, treatment type) into numerical features using encoding technology to obtain an encoded feature set, such as using One-Hot Encoding or Label Encoding; Obtain patient satisfaction data on treatment options from historical records, which are usually categorical data (such as satisfied, dissatisfied) or continuous rating data, and use classification and regression algorithms (such as random forest, logistic regression, support vector machine, etc.) to build a satisfaction prediction model, depending on the type of satisfaction data.

[0042] In a further embodiment, when a classification regression algorithm is used to construct a satisfaction prediction model, if the satisfaction is categorical data, a classification model is constructed; if the satisfaction is continuous rating data, a regression model is constructed.

[0043] In a further embodiment, after constructing the satisfaction prediction model, a training data set is constructed, the training data set includes the coded features and satisfaction of the patient, the parameters of the satisfaction prediction model are trained and tuned using a cross-validation technique, and the performance of the satisfaction prediction model on the validation set is evaluated; A test dataset is constructed, which contains the encoded features and satisfaction of patients that are not used for training. The satisfaction prediction model is evaluated using the test dataset, and the performance indicators of the satisfaction prediction model, such as accuracy, recall, F1 score, etc., are calculated, depending on whether the problem is classification or regression. At the same time, the satisfaction prediction model after the test is used to predict the patient's satisfaction with the treatment plan.

[0044] In a further embodiment, when predicting the patient's satisfaction with the treatment plan through the satisfaction prediction model after the test, the characteristic data of the new patient is obtained, and the format of the characteristic data of the new patient is consistent with the data format used when training the satisfaction prediction model; The tested satisfaction prediction model is used to predict the satisfaction of the characteristic data of the new patient. If the satisfaction prediction model is a classification model, the category label of the satisfaction is output; if the satisfaction prediction model is a regression model, a continuous satisfaction score is output.

[0045] The time-effect determination module 3 is used to determine the optimal duration and temperature setting of each thermal treatment to maximize the treatment effect and patient comfort, and to record the actual time-effect information of the treatment to obtain historical treatment data for subsequent analysis; In addition, detailed treatment records are generated, including treatment parameter settings, patient feedback, treatment effect evaluation, etc., for future reference and regulatory compliance.

[0046] Model building module 4, used to build a machine learning model based on historical treatment data and patient-personalized treatment indicators; In a further embodiment, the model building module includes a feature selection module, a model training module and a model validation module; Among them, the feature selection module is used to pre-process the historical treatment data and use statistical methods to select features related to skin temperature risk to obtain a model input feature set, and divide the model input feature set into a training set and a validation set; Among them, when preprocessing the historical treatment data, data cleaning is performed on the historical treatment data, including processing missing values, outliers and noise.

[0047] Model training module, used to train machine learning models (such as logistic regression, random forest, gradient boosting tree, etc.) using training sets; The model validation module is used to test the performance of the machine learning model on the validation set and adjust the machine learning model parameters to optimize the performance.

[0048] In a further embodiment, when a feature related to skin temperature risk is selected using a statistical method, if it is a continuous feature, the point bi-order correlation coefficient is calculated, wherein the formula for calculating the point bi-order correlation coefficient is: ; It represents the point bi-order correlation coefficient, which measures the strength and direction of the linear relationship between two variables; X i and Y i Respectively represent i The first and second continuous eigenvalues ​​of samples; and Represent the sample means of the first and second continuous features respectively; If it is a categorical feature, the chi-square test is used to obtain the correlation measure between each feature and skin temperature risk, and multivariate statistical methods (such as logistic regression analysis) are used to evaluate the interaction between features, analyze the overall contribution of the features to skin temperature risk, and obtain multivariate analysis results; Based on the results of multivariate analysis, features are selected according to statistical significance and practical significance, and the model input feature set is determined using methods such as forward selection, backward elimination, or stepwise regression.

[0049] Among them, forward selection starts with a model without any features and gradually adds features. At each step, consider adding each feature that is not already included in the model and select the feature that improves the model performance the most. Repeat this process until there are no more features that can significantly improve the model performance.

[0050] Backward elimination starts with a model that contains all candidate features and removes features step by step. At each step, consider removing each feature from the model and select the feature that improves the model performance the least after removing it. Repeat this process until removing any one feature significantly degrades the model performance.

[0051] Stepwise regression is a method that combines forward selection and backward elimination. At each step, features are considered for addition or removal, and based on some criteria (such as the Akaike Information Criterion AIC or the Bayesian Information Criterion BIC), a decision is made as to which operation provides the best model improvement. This process is repeated until model performance no longer improves.

[0052] Real-time monitoring module 5, used to monitor the patient's skin temperature and related indicators, such as heart rate, blood pressure, blood oxygen saturation, etc., in real time during the thermal treatment process, so as to provide more comprehensive patient status information and collect treatment time information and instant feedback from the patient; In addition, the integrated abnormal warning mechanism can remind medical staff in real time when the monitored parameters exceed the normal range, ensuring that the monitoring data can be synchronized to the central database or electronic medical record system in real time for historical data analysis and remote monitoring.

[0053] Prediction module 6, used to use the constructed machine learning model to predict the risk of the patient's skin temperature in real time. If a high risk is predicted, the treatment parameters are immediately adjusted, such as lowering the temperature or shortening the treatment time; As new data is continuously collected, the machine learning model is regularly updated to maintain its prediction accuracy, ensuring that the prediction results can be quickly fed back to the real-time monitoring module 5 to form a closed-loop control.

[0054] The personalized recommendation module 7 is used to provide personalized treatment recommendations based on the prediction results of the machine learning model and the patient's personalized treatment indicator system. These recommendations include treatment duration, temperature settings, and possible food and activity recommendations. At the same time, an intuitive user interface is designed so that patients and medical staff can easily understand and use it.

[0055] In summary, the skin temperature monitoring system based on tumor hyperthermia provided by the present invention ensures that the treatment plan can take into account the personalized needs and preferences of patients, and improves patient satisfaction and treatment compliance. By analyzing preference data and creating a comprehensive indicator system, the treatment effect and patient satisfaction can be more accurately evaluated. The present invention analyzes the potential relationship of data in the preference data set, uses a clustering algorithm to group patients, identifies different preference patterns, and predicts patient satisfaction with the treatment plan through classification and regression algorithms, so that the clustering algorithm can help identify the preference patterns of different patient groups, which is crucial for personalized treatment. The potential relationship in the data is analyzed to obtain several indicators that are most relevant to the treatment effect; the use of classification and regression algorithms can predict the patient's satisfaction with the treatment plan based on the patient's characteristic data, thereby providing an important reference dimension for constructing an indicator system. At the same time, the indicators that are significantly related to the treatment effect and satisfaction are analyzed and determined, and these indicators will constitute the core of the personalized treatment indicator system. The classification and weight allocation of indicators reflect the importance of different indicators in treatment, which helps to formulate more accurate and personalized treatment plans. The machine learning model is constructed using historical treatment data and personalized indicators to make treatment decisions more data-driven and accurate. Use machine learning models to predict the risk of patient skin temperature in real time, enabling earlier intervention and treatment adjustments. Provide personalized treatment recommendations based on prediction results and treatment indicator systems to further improve the personalization and accuracy of treatment.

[0056] Those skilled in the art will appreciate that all or some of the steps in the methods disclosed above, and the functional charging modules / units in the systems and devices may be implemented as software, firmware, hardware, or appropriate combinations thereof.

[0057] The terms "first", "second", "third", "fourth", etc. (if any) in the specification of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0058] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy, characterized in that: The skin temperature monitoring system based on tumor hyperthermia patients includes a preference data collection module, an indicator system construction module, a timeliness determination module, a model construction module, a real-time monitoring module, a prediction module and a personalized recommendation module; The preference data collection module is connected to the indicator system construction module, the indicator system construction module is connected to the timeliness determination module and the model construction module, the model construction module is connected to the prediction module, the prediction module is connected to the real-time monitoring module, and the real-time monitoring module is connected to the personalized recommendation module; The preference data collection module is used to collect patient information and preference information for thermal therapy to obtain a preference data set; The indicator system construction module is used to construct a personalized treatment indicator system based on the patient's preference data; The time-effect determination module is used to determine the optimal duration and temperature setting of each thermal treatment, and record the actual time-effect information of the treatment to obtain historical treatment data; The model building module is used to build a machine learning model based on historical treatment data and patient-specific treatment indicators; The real-time monitoring module is used to monitor the patient's skin temperature in real time during the thermal treatment process and collect treatment time information and instant feedback from the patient; The prediction module is used to use the constructed machine learning model to predict the risk of the patient's skin temperature in real time, and if a high risk is predicted, the treatment parameters are adjusted immediately; The personalized recommendation module is used to provide personalized treatment suggestions based on the prediction results of the machine learning model and the patient's personalized treatment indicator system.

2. A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 1, characterized in that: The indicator system construction module includes a data analysis module and a system establishment module; The data analysis module is used to analyze the potential relationships in the preference data set, and to group the patients using a clustering algorithm, and to obtain the patient's satisfaction with the treatment plan through classification and regression algorithms; The system establishment module is used to select several indicators that are most relevant to treatment effects and patient satisfaction, classify the indicators, and assign weights to each indicator.

3. A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 2, characterized in that: When analyzing the potential relationship in the data in the selected feature set, each variable in the selected feature set is classified and determined to be quantitative or qualitative data, and a distribution test is performed on the quantitative data to determine whether it obeys a normal distribution, thereby obtaining a data type and distribution characteristic report; According to the data type and distribution characteristics in the distribution characteristic report, a corresponding correlation analysis algorithm is selected, and the corresponding correlation analysis algorithm is applied to calculate the correlation coefficients between the variables to obtain the correlation coefficient matrix between each variable pair; Conduct hypothesis tests on each correlation coefficient to obtain significance test results, and based on the significance test results, identify and eliminate several variables with the lowest correlation with the treatment effect to obtain a streamlined variable set; Integrate the streamlined variable set and its correlation coefficients and significance test results to write a data relationship report.

4. A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 3, characterized in that: When selecting the corresponding correlation analysis algorithm according to the data type and distribution characteristics in the distribution characteristic report, for quantitative and normally distributed data, the Pearson correlation coefficient is selected; For nonnormal or rank data, choose the Spearman correlation coefficient; For qualitative data, choose the Chi-Square test.

5. A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 4, characterized in that: When obtaining the patient's satisfaction with the treatment plan through the classification and regression algorithm, the correlation between each feature and the patient's satisfaction with the treatment plan is analyzed in the selected feature set, and several features most correlated with the treatment satisfaction are selected to obtain a screening feature set; Extract additional derived features based on the features in the filtered feature set, process the time series data, extract trend and periodic features, add the derived features to the filtered feature set, obtain an extended feature set, and convert the categorical features into numerical features using encoding technology to obtain an encoded feature set; Obtain patient satisfaction data on treatment plans from historical records and use classification and regression algorithms to build a satisfaction prediction model.

6. A skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 5, characterized in that: When the classification regression algorithm is used to construct the satisfaction prediction model, if the satisfaction is classified data, a classification model is constructed; If satisfaction is continuous rating data, a regression model is constructed.

7. The skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 6, characterized in that: After the satisfaction prediction model is constructed, a training data set is constructed, the training data set includes the coded features and satisfaction of the patients, the parameters of the satisfaction prediction model are trained and tuned using a cross-validation technique, and the performance of the satisfaction prediction model on the validation set is evaluated; A test data set was constructed, which contained the encoded features and satisfaction of patients not used for training. The satisfaction prediction model was evaluated using the test data set, and the performance indicators of the satisfaction prediction model were calculated. At the same time, the satisfaction of patients with the treatment plan was predicted by the tested satisfaction prediction model.

8. The skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 7, characterized in that: When predicting the patient's satisfaction with the treatment plan through the satisfaction prediction model after the test, the characteristic data of the new patient is obtained, and the format of the characteristic data of the new patient is consistent with the data format used when training the satisfaction prediction model; The tested satisfaction prediction model is used to predict the satisfaction of the characteristic data of the new patient. If the satisfaction prediction model is a classification model, the category label of the satisfaction is output; if the satisfaction prediction model is a regression model, a continuous satisfaction score is output.

9. The skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 8, characterized in that: The model building module includes a feature selection module, a model training module and a model verification module; The feature selection module is used to pre-process the historical treatment data and select features related to skin temperature risk using statistical methods to obtain a model input feature set, and divide the model input feature set into a training set and a validation set; The model training module is used to train the machine learning model using the training set; The model validation module is used to test the performance of the machine learning model on the validation set and adjust the machine learning model parameters to optimize the performance.

10. The skin temperature monitoring system for patients undergoing tumor hyperthermia therapy according to claim 9, characterized in that: When the statistical method is used to select the feature related to the skin temperature risk, if it is a continuous feature, the point bi-order correlation coefficient is calculated, wherein the formula for calculating the point bi-order correlation coefficient is: ; represents the point bi-order correlation coefficient; X i and Y i Respectively represent i The first and second continuous eigenvalues ​​of samples; and Represent the sample means of the first and second continuous features respectively; If it is a categorical feature, the chi-square test is used to obtain the correlation measure between each feature and skin temperature risk, and multivariate statistical methods are used to evaluate the interaction between features, analyze the overall contribution of the features to skin temperature risk, and obtain multivariate analysis results; Based on the results of multivariate analysis, features are selected according to statistical significance and practical significance to determine the model input feature set.

Citation Information

Patent Citations

  • Skin temperature monitoring system used for tumor thermotherapy

    CN106038038A

  • User satisfaction prediction model establishment method, device and system and medium

    CN116776169A

  • Method based on correlation analysis

    CN117932549A

  • Knowledge graph-driven medical large model diagnosis method

    CN118280562A

  • Personalized treatment scheme recommendation system based on machine learning

    CN118538356A

Cited By

  • Method and system for controlling infusion heating tube of integrated multi-channel medical heating blanket

    CN120616892A