CN118173253B

CN118173253B - System and method for analyzing and managing based on patient data

Info

Publication number: CN118173253B
Application number: CN202410601563.XA
Authority: CN
Inventors: 刘俊; 武鹏佳; 田婷; 刘江; 耿建忠
Original assignee: Guizhou Ankang Clinical Laboratories Inc
Current assignee: Guizhou Ankang Clinical Laboratories Inc
Priority date: 2024-05-15
Filing date: 2024-05-15
Publication date: 2024-07-16
Anticipated expiration: 2044-05-15
Also published as: CN118173253A

Abstract

The invention discloses a system and a method for analyzing and managing based on patient data, which relate to the field of data management, wherein the system for analyzing and managing based on patient data comprises the following components: the system comprises a data acquisition module, a data analysis module, a data influence module, a data management module and a data feedback module; the data acquisition module is used for acquiring patient data and hospital information; the data analysis module is used for classifying and giving weight to the patient data; the data influence module is used for analyzing the change trend of the patient data and calculating the influence value of the patient; the data management module is used for acquiring a patient data management scheme and patient monitoring diagnosis and treatment requirements; and the data feedback module is used for adjusting a patient data management scheme and patient monitoring diagnosis and treatment requirements. According to the invention, the modularized design concept is adopted by the system, and the acquisition, analysis, management and feedback of the patient data are respectively and independently formed into the modules, so that the accuracy and usability of the patient data are improved.

Description

System and method for analyzing and managing based on patient data

Technical Field

The invention relates to the field of data management, in particular to a system and a method for analyzing and managing data based on patients.

Background

Patient data refers to all medical and health information associated with individual patients, and generally covers a wide range of fields including, but not limited to, basic information of patients, medical history, diagnosis, treatment records, drug use, laboratory test results, and imaging results, whereas in modern medical systems, effective management and analysis of patient data plays a vital role in improving medical quality of service, reducing costs, enabling personalized medicine, enabling extraction of deeper insight from patient data, and further improving medical results and patient experience.

The patient data analysis and management, through deep analysis of the patient data, a doctor accurately diagnoses the disease, makes a treatment scheme more suitable for the patient, reveals trend and potential risk of disease progress, helps the doctor make a more intelligent decision, and simultaneously, the patient data analysis identifies unique requirements and response modes of each patient, thereby customizing a personalized treatment plan, improving the pertinence and effectiveness of treatment, particularly playing a vital role in modern medical system when treating complex or chronic diseases, and improving the quality and efficiency of medical services.

However, the existing analysis and management system based on patient data does not classify and partition various factor data of patients, so that the processing of the patient data cannot be accurately realized when the analysis and management system based on patient data is used, and meanwhile, the specific requirements and potential safety risks of the patients are not considered, so that the existing analysis and management system based on patient data cannot accurately analyze the patient data when the analysis and management system based on patient data is used, and the effect when the management of the patient data is performed is not ideal.

For the problems in the related art, no effective solution has been proposed at present.

Disclosure of Invention

Aiming at the problems in the related art, the invention provides a system and a method for analyzing and managing data based on patients, which are used for overcoming the technical problems existing in the prior related art.

For this purpose, the invention adopts the following specific technical scheme:

According to an aspect of the present invention, there is provided a patient data analysis management system and method, the patient data analysis management system comprising: the system comprises a data acquisition module, a data analysis module, a data influence module, a data management module and a data feedback module;

the data acquisition module is used for acquiring patient data and hospital information and cleaning the patient data;

the data analysis module is used for classifying the cleaned data and giving weight to the data classification result;

The data influence module is used for analyzing the data change trend according to the data classification weighting result and calculating a data influence value based on the data change trend and hospital information;

the data management module is used for generating a data management scheme and monitoring requirements according to the data influence value;

and the data feedback module is used for collecting the data management scheme and the monitoring requirement use feedback and optimally adjusting the data management scheme and the monitoring requirement based on the use feedback.

Preferably, the data analysis module includes: the system comprises a data factor module, a factor ratio module, a factor classification module and a data weighting module;

The data factor module is used for extracting characteristic factors in the cleaned data;

The factor ratio module is used for carrying out influence analysis on the extracted characteristic factors and sequencing the characteristic factors based on influence analysis results;

the factor classification module is used for classifying the data according to the characteristic factor sequencing result;

And the data weighting module is used for weighting the data according to the category division result and the characteristic factors.

As a preferred solution, the data weighting module includes: the system comprises a weighting calculation module, a result verification module, a weighting influence module and a weighting optimization module;

the weighting calculation module is used for calculating a class division result weight value and a characteristic factor weight value according to a preset class weighting rule and a characteristic weighting rule;

the result verification module is used for verifying the classification result weight value and the characteristic factor weight value;

and the weighting influence module is used for calculating a weighting comprehensive influence value according to the verified category division result weight value and the characteristic factor weight value.

As a preferred scheme, calculating the weighted comprehensive influence value according to the verified classification result weight value and the characteristic factor weight value comprises:

normalizing the class division result weight value and the characteristic factor weight value;

Presetting a comprehensive influence rule, and dividing the duty ratio of a class division result weight value and a characteristic factor weight value based on the comprehensive influence rule;

Calculating a weighted comprehensive influence value through a weighted average algorithm according to the class division result weight value and the duty ratio result of the characteristic factor weight value;

Setting a weighting comprehensive influence threshold, verifying the weighting comprehensive influence value, and comparing the verified weighting comprehensive influence value with the weighting comprehensive influence threshold;

and adding an influence grade label to the weighting comprehensive influence value based on the comparison result of the validated weighting comprehensive influence value and the weighting comprehensive influence threshold.

Preferably, the data influencing module comprises: the system comprises a weighting analysis module, a trend prediction module, a patient demand module, a parameter comparison module and an influence calculation module;

The weighting analysis module is used for judging the data change trend according to the data classification weighting result;

the trend prediction module is used for predicting data evolution parameters according to the data change trend judgment result and the data classification weighting result;

the patient demand module is used for extracting demand data in the data evolution parameters;

The parameter comparison module is used for comparing the demand data with the hospital information and generating a hospital matching value based on the comparison result;

and the influence calculation module is used for calculating a data influence value according to the hospital matching value and the data evolution parameter.

As a preferred scheme, predicting data evolution parameters according to the data change trend judgment result and the data classification weighting result comprises;

Preprocessing the data classification weighting result, and verifying the data change trend judgment result;

Presetting an influence factor extraction rule, and extracting a data change value in a preprocessed data classification weighting result based on the influence factor extraction rule;

setting a trend evolution rule set, and matching the result with the trend evolution rule set trend evolution rule according to the verified data change trend judgment result;

and calculating data evolution parameters according to the data change values and the trend evolution rules, and verifying predicted data evolution parameters.

Preferably, calculating the data evolution parameter according to the data change value and the trend evolution rule, and verifying the predicted data evolution parameter includes:

constructing a data evolution model through a time sequence analysis algorithm;

dividing the data change value into a training set and a testing set, and adjusting data evolution model parameters based on trend evolution rules;

Training the adjusted data evolution model through a training set, and verifying the trained data evolution model through a testing set;

substituting the data change value into the verified data evolution model, and calculating a data evolution parameter;

And cross-verifying the data evolution parameters, and evaluating the accuracy of the verified predicted data evolution parameters.

As a preferred scheme, the parameter comparison module includes: the system comprises a data docking module, a demand matching module and a verification optimizing module;

the data docking module is used for extracting characteristic parameters of the demand data and characteristic parameters of the hospital information;

The demand matching module is used for matching the demand data characteristic parameters with the hospital information characteristic parameters and calculating demand data matching values;

And the verification optimization module is used for verifying and optimizing the demand data matching value and outputting the demand data matching value after verification and optimization as a hospital matching value.

As a preferred scheme, matching the required data characteristic parameters with the hospital information characteristic parameters, and calculating a calculation formula of the required data matching value is as follows:；

Wherein W is a demand data matching value;

N is the total number of the required data characteristic parameters and the hospital information characteristic parameters;

i is an index of the characteristic parameters of the demand data and the characteristic parameters of the hospital information;

component values of the ith feature in the feature parameters of the demand data;

Is the component value of the ith feature in the hospital information feature parameters.

According to another aspect of the present invention, there is provided a patient data based analysis and management method, comprising the steps of:

s1, acquiring patient data and hospital information, and cleaning the patient data;

s2, classifying the cleaned data, and giving weight to the data classification result;

S3, analyzing a data change trend according to the data classification weighting result, and calculating a data influence value based on the data change trend and hospital information;

s4, generating a data management scheme and a monitoring requirement according to the data influence value;

s5, collecting data management scheme and monitoring requirement use feedback, and optimally adjusting the data management scheme and the monitoring requirement based on the use feedback.

The beneficial effects of the invention are as follows:

1. According to the invention, the modularized design concept is adopted by the system, the acquisition, analysis, management and feedback of the patient data are respectively and independently formed into the modules, the patient data are carefully processed, the accuracy and the usability of the patient data are improved, a solid foundation is provided for subsequent analysis, and meanwhile, the specific requirements and potential health risks of the patient are identified by classifying, weighting and trend analysis of the patient data, so that a doctor can conveniently formulate a more personalized diagnosis and treatment scheme.

2. According to the invention, the integrated data management module is used for generating and adjusting the patient monitoring diagnosis and treatment plan, so that the utilization efficiency of medical resources is improved, the medical resource waste is reduced, the use feedback is collected, the diagnosis and treatment and management scheme is optimized based on the feedback, meanwhile, the trend prediction module and the time sequence analysis algorithm are utilized for simulating and predicting the evolution trend of patient data, and the health management and preventive measure formulation of patients are realized according to the patient data.

3. According to the invention, through the design of the parameter comparison module, the patient requirements and hospital resources are accurately matched, the individuation and satisfaction degree of medical service are improved, and through comprehensively considering a plurality of data factors and classifying and weighting the patient data, the accuracy of patient data processing is improved, and the actual conditions and requirements of patients are accurately reflected.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the embodiments will be briefly described below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and other drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

FIG. 1 is a system block diagram of a patient data based analysis management system in accordance with an embodiment of the present invention;

fig. 2 is a flow chart of a method for analyzing and managing patient data according to an embodiment of the present invention.

In the figure:

1. A data acquisition module; 2. a data analysis module; 3. a data influencing module; 4. a data management module; 5. and the data feedback module.

Detailed Description

The following describes in further detail the embodiments of the present invention with reference to the drawings and examples. The following examples are illustrative of the invention and are not intended to limit the scope of the invention.

Thus, the following detailed description of the embodiments of the invention, as presented in the figures, is not intended to limit the scope of the invention, as claimed, but is merely representative of selected embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

According to an embodiment of the invention, a system and a method for analyzing and managing based on patient data are provided.

The present invention will now be further described with reference to the accompanying drawings and detailed description, as shown in fig. 1, according to one embodiment of the present invention, there is provided a patient data analysis-based management system including: the system comprises a data acquisition module 1, a data analysis module 2, a data influence module 3, a data management module 4 and a data feedback module 5;

the data acquisition module 1 is used for acquiring patient data and hospital information and cleaning the patient data;

Specifically, the information such as medical history, laboratory results, image data and the like of a patient are obtained from a hospital using system, data directly generated by equipment such as an electrocardiograph, a monitor and the like and patient self-describing health information collected through questionnaires or digital application are automatically synchronized to a data analysis platform from an electronic health record system and medical equipment, and under the condition that the automatic synchronization cannot be achieved, the data are supplemented in a manual input mode, the data are extracted from other medical systems or databases by utilizing a programming interface, and meanwhile, the related data of the patient are obtained, so that the patient needs to agree or authorize the patient.

Checking repeated items in the data, only keeping necessary records, ensuring all the data, such as date and time stamp, and the like, conforming to the same format standard, taking proper measures on the missing data, such as filling, deleting or estimating missing values by using a statistical method, identifying unreasonable or abnormal data values, correcting or deleting, converting texts into numerical values, date and the like according to the needs, and anonymizing sensitive information for protecting privacy of patients.

The data analysis module 2 is used for classifying the cleaned data and giving weight to the classification result of the data;

Specifically, the data analysis module 2 includes: the system comprises a data factor module, a factor ratio module, a factor classification module and a data weighting module;

Specifically, understanding the source, type, and information contained in the data clearly analyzes the objectives, such as whether disease prediction, patient classification, or treatment outcome assessment is to be performed, which will instruct which features are relevant, selecting features from the dataset that may affect the analysis outcome based on the objectives, including patient essential information, such as age, sex, weight, height, etc., clinical parameters, such as blood pressure, heart rate, blood test outcome, etc., treatment history, such as past accepted treatment and drug use, pathological information, such as diagnosis type, disease course progression, pathological reporting outcome, lifestyle factors, such as eating habits, exercise frequency, tobacco and wine use, etc.

The selected features are suitably transformed to fit for analysis, including transforming category data to values, such as gender male and female to 1 and 0, adjusting the numerical data to the same scale, calculating new features based on the original data, such as age from birthday, or average from a plurality of blood parameters, identifying the most influential features using feature selection techniques, such as determining an analysis target, i.e., a target variable to be predicted or classified, e.g., a disease severity of a patient, the target variable may be a numerical or classified variable, analyzing the relationship between continuous variables such as age, weight, etc., of the patient and the disease severity using pearson correlation coefficients to evaluate the correlation or degree between each feature and the target variable, selecting the features most influential to the target variable based on the results obtained by the statistical method, and sorting the features most influential according to the degree of correlation, as needed.

Specifically, a reference model is established, including all possible features, and the model may be a simple linear regression, a logistic regression, or a more complex machine learning model, depending on the analysis purpose and the data type, and the influence of each feature on the model is evaluated using a statistical or machine learning method, such as a coefficient analysis method, in which the larger the absolute value of the coefficient is checked, the larger the influence of the feature on the model is, and for each feature, the further analysis of the relationship between the feature and the target variable is performed by visualization such as a scatter diagram, a box diagram, etc. or calculation of a conditional average, and the features are ranked according to the evaluated result, and the ranking of the features is based on their contribution size to the model performance or their statistical significance, the features with high contribution degree or statistical significance are ranked in front, the stability of the feature importance is tested using cross-validation, and the above analysis process is repeated on different subsets of training sets, and the consistency of the feature importance ranking is checked.

Specifically, the most influential features are selected as the basis for classifying the patient category, while the features show higher importance in the predictive model and are obviously associated with clinical results of the patient, and then a suitable category classification method is determined according to the type and data characteristics of the selected features, including setting thresholds for continuous variables according to specific clinical significance or statistical analysis results, such as classifying blood pressure into normal, pre-hypertension and hypertension, using an unsupervised learning method such as hierarchical clustering, and the like, automatically classifying the patient into different groups according to the features, and creating rules based on logic or expert knowledge, such as classifying the patient according to combinations of age, sex and certain biomarkers.

Specifically, the most influential features are selected as the basis for classifying the patient categories, the features show higher importance in the prediction model and are obviously associated with clinical results of the patient, and suitable category classification methods are determined according to the types and data characteristics of the selected features, such as classifying blood pressure into normal, pre-hypertension and hypertension, and for other types of features, unsupervised learning methods such as hierarchical clustering and the like are adopted to automatically classify the patient into different groups according to the features, and then rules based on logic or expert knowledge are created according to the combination of age, sex and certain biomarkers to further refine and verify the classification of the patient.

The selected method is applied to categorize patients in the dataset, such as by optimizing the internal similarity of the clusters and the differences between clusters when using a clustering algorithm to determine the optimal number of clusters, and verifying the validity and practicality of the categorization by clinical results or other independent data sets, applying the category information in clinical practice, such as custom treatment plans, patient management strategies, or further study analysis.

The classification is based on the following aspects, namely selecting the characteristics which have obvious influence on the prediction model and the clinical result, considering the clinical significance of the characteristics, such as the normal range of blood pressure, the abnormal level of the biomarker and the like, and using statistical methods, such as threshold setting, cluster analysis and the like, to assist classification, and combining logic rules with expert knowledge, such as age, gender and other factors, to create finer classification rules.

The classification of patients in the data set is performed by applying a selected method, for example, when a clustering algorithm is used, the optimal number of groups is determined by optimizing the internal similarity of the groups and the difference between the groups, checking the homogeneity in the groups and the heterogeneity between the groups, ensuring that the patients in the same group are similar in key characteristics, while the patients in different groups are obviously different, verifying the effectiveness and practicality of the classification by clinical results or other independent data sets, and applying the classification information to clinical practice, for example, customizing treatment plans, patient management strategies or further research analysis, for example, different treatment schemes or prognosis management may be required for patients in different groups.

Specifically, the data weighting module includes: the system comprises a weighting calculation module, a result verification module, a weighting influence module and a weighting optimization module;

Specifically, the weighting rules of each category and each feature are defined, for example, some features may obtain higher weights due to greater contributions to disease prediction, specific numerical values or calculation methods of weights are determined, including directly assigning a fixed weight value to each category or feature, and dynamically determining the weight value according to a specific calculation formula or model output, for example, the weights are calculated according to the statistical importance, influence or expert opinion of the features, for example, the greater the gain of the features to the target variable information, the higher the weights, whereas in the regression model, the absolute values of coefficients of the features may be used as weights, the weights may be assigned according to factors such as the clinical importance of each category, the number of patients or the severity of the disease, for example, the greater the number of patients in the category, the higher the category weights may be, and the higher the categories with higher urgency or severity may be assigned higher weights.

The weight of each category or feature is multiplied by its corresponding score or value, and then all categories or features are summed to obtain a weighted total score, e.g., total score = (feature value x feature weight), whereas in practical applications, the weight setting is adjusted according to the performance of the result, which needs to be performed based on feedback and additional data analysis, ensuring that the weight distribution is fair and meets the intended goals and effects.

Specifically, a classification weighting rule and a feature weighting rule are set for diabetes, such as a mild diabetes patient, such a patient does not usually need urgent treatment, and thus this category is given a lower weight, for example, a weight of 1, a moderate diabetes patient, such a patient needs regular treatment and monitoring, and thus this category is given a weight of 2, a severe diabetes patient, such a patient's condition may be life threatening, and urgent and aggressive intervention is required, and thus this category is given a highest weight, for example, the feature weighting rule includes an age, such as an age that is older, a risk of diabetic complications is usually higher, and thus a weight is given according to an age range, such as 50-65 years = 1, 50-65 years = 2, 65 years = 3, and a body weight index is usually associated with an increase in diabetes risk, a weight distribution is a body weight index <25 = 1, 25 +.ltoreq.body weight index <30 = 2, a body weight index is a direct indicator for evaluating diabetes control, a weight is set such as normal = 1, a mild elevation = 2, a high elevation = 3, a risk of diabetic is easily controlled, and a good response is given a low patient = 1 = no response.

Assuming a patient with an age of 65 years, a body mass index of 32, a high blood glucose level rise, and a general therapeutic response, the total feature weight=3 (age) +3 (body mass index) +3 (blood glucose level) +2 (therapeutic response) =11 is calculated to evaluate the overall disease management difficulty and priority of the patient for more targeted treatment and resource allocation.

Specifically, checking consistency and logic of weight assignment, for example, ensuring that high risk category and feature having a larger influence on disease are given higher weight, verifying the effect of weight assignment by using statistical method, including checking correlation between weighted category and feature weight and target variable, such as disease severity and treatment effect, testing robustness of weight assignment by cross verification, dividing data into multiple subsets, repeating training and verification process, checking whether performance of weight assignment on different subsets is consistent, and testing weight assignment of category and feature in actual application scene by applying weight to new data set and monitoring influence on prediction performance, decision support or resource assignment optimization.

Feedback from actual applications, including treatment responses of patients and comments of other relevant stakeholders, is collected, sensitivity analysis is performed, the degree of influence of different weight settings on analysis results is evaluated, and weights requiring key verification and possible adjustment are determined.

Specifically, calculating the weighted comprehensive influence value according to the verified category classification result weight value and the characteristic factor weight value comprises the following steps:

Specifically, the goal and purpose of the explicit rules, such as whether to predict disease risk, improve treatment or optimize resource allocation, formulate weight allocation rules based on each category and the contribution of the feature to the final goal, including assigning a fixed weight based on the importance of each feature and category, and dynamically adjusting the weights based on the performance of the feature or category in the model, such as accuracy, sensitivity, while the basic weight of each feature and category is determined based on its performance in past data analysis, e.g., a feature that would be relatively high if strongly correlated with disease prognosis, then integrate all the correlated weights to form a comprehensive impact score, calculate the weight of each weight in the total weight, and verify the validity of these rules via actual data and prediction results.

Specifically, the distribution of the weighted comprehensive influence values is determined through historical data analysis, so that a threshold value, such as a mean value, a median value, or a specific percentile, is set, for example, a threshold value for distinguishing high-risk patients from low-risk patients is set according to the requirements of a prediction model, the performance of the weighted comprehensive influence values in the same data set is set by using a weighted comprehensive influence value calculation method, consistency and stability of the weighted comprehensive influence values are evaluated through cross verification or leave verification, the weighted comprehensive influence values are tested on different data sets by using a calculation method of the weighted comprehensive influence values, universality and accuracy are ensured, the calculated weighted comprehensive influence values are compared with actual clinical results or prognosis data, prediction capability of the weighted comprehensive influence values is evaluated, the weighted comprehensive influence values are classified into different risk levels or treatment categories according to the set threshold value, for example, the threshold value is possibly lower than the threshold value, the threshold value is higher than the threshold value, in actual operation, the decision is guided by using the classification, such as patient layered management, resource allocation or treatment strategy is carried out, the suitability of the actual application is evaluated according to feedback and new data, and the suitability of the threshold value is adjusted when necessary, if the weighted comprehensive influence management is required to be less than the threshold value 50, if the comprehensive influence management is required is more easily determined, and the comprehensive influence management 50 is required to be less, and if the comprehensive influence management of the comprehensive influence is required is less than 50 is determined.

Specifically, different influence levels are determined based on the weighting comprehensive influence threshold and specific requirements, the levels divide the weighting comprehensive influence value into low risk, medium risk and high risk levels according to the threshold, or finer classification is set according to specific conditions, and specific level thresholds are set according to data analysis and targets. For example, a low risk, a first quantile with a weighted composite impact value below a set threshold, a medium risk, a weighted composite impact value between the first and second quantiles, a high risk, a weighted composite impact value exceeding a threshold of the second quantile, such as assuming a set composite impact threshold of 50, the weighted composite impact value being classified as low risk (< 30), medium risk (30-60), high risk (> 60).

The weighted composite impact value of each individual or data point is classified, corresponding risk levels are assigned according to the comparison result of the value and the threshold value, different levels are displayed by using charts or color codes for enabling the impact level to be more visual, for example, low, medium and high risk levels are distinguished by using marks with different colors on a data dashboard, the level marks are used for assisting decision making processes, for example, in the priority setting of resource allocation, patient management or preventive measures, and the threshold value and classification of the impact level are periodically reevaluated and adjusted according to new data and feedback.

The data influence module 3 is used for analyzing the data change trend according to the data classification weighting result and calculating a data influence value based on the data change trend and the hospital information;

Specifically, the data influencing module 3 includes: the system comprises a weighting analysis module, a trend prediction module, a patient demand module, a parameter comparison module and an influence calculation module;

Specifically, ensuring that the collected patient data is complete and up-to-date, including various health indicators and related features of weighting classifications, such as blood pressure, blood glucose, body mass index, etc., classifying the newly collected data using existing classification weighting rules, such as assigning different weights according to severity of the condition, sorting the patient data in time series based on the weights of particular health indicators, observing the trend of the patient data using time series analysis methods, including applying moving average or exponential smoothing methods to observe the smooth trend of the data, calculating the rate of change of key health indicators for each patient or patient population, e.g., using logarithmic differences to calculate the percentage change between consecutive time points.

A statistical model is built according to historical data, such as a linear regression model and the like, the change trend of patient data in a future period is predicted, a chart is used for visualizing time series data and a prediction result, the accuracy of trend judgment and the effectiveness of the prediction model are periodically evaluated, a classification weighting rule and the prediction model are adjusted according to the latest data and clinical feedback, the health state of a patient can be accurately reflected, for example, if the blood glucose history of a diabetic patient shows month-by-month increase, the linear regression analysis is applied for predicting the blood glucose trend of a plurality of months in the future, and if the prediction result shows that blood glucose will continue to rise, a treatment scheme is required to be adjusted or monitoring is enhanced.

Specifically, predicting data evolution parameters according to the data change trend judgment result and the data classification weighting result comprises;

Specifically, the objective of the extraction rule is determined, for example, in order to evaluate the overall health condition of the patient, predict the disease progression or assist in clinical decision making, the extraction rule is formulated by determining which factors are most valuable to the objective and how the objective should be extracted based on the information available in the preprocessed patient data, selecting the health indicators related to the objective, such as blood glucose, blood pressure, cholesterol, etc., from the dataset, assigning weights to the features according to their importance and relevance, such as giving higher weights to factors such as severity of the disease, past history, etc., setting a threshold or condition, such as the range of specific indicators or the presence of specific features, and determining which data needs to be extracted.

And (3) screening the characteristics in the preprocessed patient data by applying an extraction rule, filtering out unconditional data according to the screening standard in the extraction rule, sorting the eligible data according to the characteristic weights, extracting the characteristics with the largest influence, and combining the extracted characteristics to calculate a comprehensive influence value.

Specifically, a trend evolution rule set needs to be defined according to clinical requirements and historical data analysis, and rules should describe various possible change trends of patient data and corresponding clinical meanings thereof, for example, a stable trend, patient data keeps stable for a long period, a trend is gradually improved, patient data shows a gradual improvement trend, a gradual deterioration trend, patient data shows a gradual deterioration trend, a rapid change trend, patient data shows rapid changes, indicating sudden events or acute deterioration, a specific quantization index and a specific threshold are set for each trend evolution rule, for example, a rule is specified by using a statistical index of data such as a mean value, a standard deviation, a change rate or other applicable measurement standard, for example, a stable trend, a month change rate is less than 5%, a gradual improvement trend, a continuous three month change rate is positive and increases month by month, a gradual deterioration trend, a continuous three month change rate is negative and decreases month by month, a rapid change trend is more than 20%, and a trend evolution index is calculated by using a proper data processing technology such as a time sequence analysis, a moving average or an index smoothing method.

And applying a trend evolution rule set on actual patient data, verifying the prediction accuracy and the practicability of the trend evolution rule set, ensuring that the rule set can effectively reflect the change of the health condition of the patient, matching corresponding trend evolution rules according to the data change trend judgment result of the patient, applying the matching result to clinical decision support, such as adjusting a treatment scheme, planning follow-up time or changing a patient management strategy, and providing prediction information about the future health condition change of the patient by the matched trend evolution rules.

Specifically, calculating the data evolution parameters according to the data change values and the trend evolution rules, and verifying the predicted data evolution parameters includes:

constructing a data evolution model through a time sequence analysis algorithm;

Specifically, historical health data of a patient is collected and arranged, including but not limited to blood pressure, heart rate, body weight, laboratory detection results and the like, integrity and accuracy of the data are ensured, missing values are removed or filled, a standardized data format assumes that a current value is linearly related to a historical value thereof, statistical characteristics of the data, such as stationarity, seasonality and trend, are analyzed, a graphical tool, such as a time sequence chart, an autocorrelation chart and the like, is used for assisting in identifying data characteristics, statistical software or a programming tool is used for estimating and fitting a selected time sequence model, model parameters, such as the number of hysteresis terms, differential order and the like, an optimal fitting effect is achieved, whether residual errors of the model are white noise is checked, prediction capability of the model is evaluated by using a reserved data or cross-validation method, good performance of the model on the unseen data is ensured, uncertainty of a predicted value is predicted by using the fitted model, uncertainty of a predicted result is evaluated according to confidence interval provided by the model, accuracy and relativity of the model is ensured by periodically updating the model, structural changes possibly introduced by the new data are evaluated, and the model parameters are adjusted.

specifically, the parameter comparison module includes: the system comprises a data docking module, a demand matching module and a verification optimizing module;

specifically, the calculation formula for matching the characteristic parameters of the demand data with the characteristic parameters of the hospital information and calculating the matching values of the demand data is as follows: ；

Wherein W is a demand data matching value;

Specifically, based on the degree of matching between the specificity and urgency of the treatment required by the patient and the ability of the hospital to provide services, for example, the matching value is assessed according to factors such as the equipment of the hospital, the availability of specialists, and the effect of the treatment, these parameters include the trend of change in the health index of the patient, such as the result of analysis of time-series data of blood pressure, blood sugar, body weight, etc., the influence value is calculated using a weighted sum model or a more complex machine learning model, in combination with the hospital matching value and the data evolution parameters, for example: patient impact value = K x hospital matching value + L x data evolution parameter, where K and L are weights determined from model training results.

The weights K and L are determined from historical data and clinical studies and which factors are more important for the prediction of patient outcome, the weights are estimated using a data fitting method, ensuring that all input data, hospital matching values and data evolution parameters are up-to-date and have been suitably pre-processed, such as normalized or standardized, that the outcome of the analysis using time series data, such as evolution parameters may be future values predicted based on a time series model, that the model and parameters are applied to calculate the impact value for each patient on the actual data and ensure that the application of the model reflects the original design intent, such as correct handling of outliers and adherence to data confidentiality, verifying whether the calculated impact values are consistent with the actual patient outcome, such as by subsequent health outcome tracking, adjusting model parameters based on feedback and new data, improving prediction accuracy and relevance.

The data management module 4 is used for generating a data management scheme and monitoring requirements according to the data influence values;

Specifically, patients are classified into different risk classes or management categories according to their impact values, e.g., low risk, need regular monitoring and preventive medical measures, medium risk, need more frequent examinations and possible early interventions, high risk, need intensive care and possible urgent medical interventions, and for different classes of patients, corresponding data management schemes are formulated, including data collection, processing and update frequency, low risk patients, only need regular updates of health data, such as a full physical examination every half year, stroke risk patients, need more frequent collection of critical health indicators, such as detailed examinations every quarter, high risk patients, need continuous health monitoring and real-time data updates, and heart rate and blood pressure are monitored using wearable devices.

Based on classification and data management schemes of patients, a specific monitoring diagnosis and treatment plan is formulated, low-risk patients mainly pay attention to life style guidance and conventional disease prevention, stroke risk patients need customized treatment schemes, curative effects and side effects are evaluated regularly, high-risk patients need emergency intervention plans, cooperative treatment of multi-disciplinary teams and close medical monitoring, a medical information system and a data analysis tool are applied to support implementation of data management and monitoring diagnosis and treatment, an electronic health record system is used for tracking and managing patient data, and data analysis and an artificial intelligence tool are used for predicting disease progress and responding to treatment, effects of the management scheme and diagnosis and treatment requirements are evaluated regularly, adjustment is performed according to health data changes of patients, treatment effects and health states of the patients are reviewed regularly, and treatment schemes and management measures are adjusted to better adapt to needs of the patients.

If a diabetic patient is assumed to have a high risk of impact, their data management scheme may include blood glucose monitoring and monthly physician visits at least once a week, while the monitored medical needs may include rapid response measures in emergency situations, as well as periodic nutritional and medication adjustments.

Specifically, the development of the care diagnosis and treatment plan includes that low-risk patients are focused on guidance of living modes and conventional disease prevention, comprehensive health management plans of reasonable diet, regular exercise and regular physical examination are developed, individualized treatment schemes including more frequent medical examination, disease management education and early intervention measures are developed for stroke patients, the treatment plan is adjusted according to the change of illness, emergency intervention plans and intensive care are implemented for high-risk patients, multidisciplinary team cooperation including doctors, nurses, nutritionists and other professionals are required, and tight health monitoring is performed regularly, key vital signs are monitored by using wearable equipment, and rapid response is required when necessary, and the health state and treatment progress of the patients are tracked by using an electronic health record system.

The method comprises the steps of predicting the progress of a disease using data analysis and artificial intelligence tools, adjusting a treatment plan based on the prediction results, updating health data in real time, so as to quickly respond to changes in the disease and set up a periodic review conference, evaluating the treatment effect and health status of patients, such as by a periodic conference and patient interview by a medical team, analyzing the effect of treatment regimens and patient satisfaction, identifying aspects of success and places where improvement is needed, adjusting treatment plans and management measures based on changes in the patient's health data, for example, if a diabetic patient has poor glycemic control, adjusting their medication dosage or changing treatment methods, and maintaining open and continuous communication with the patient and their family, explaining the cause and expected effect of the treatment plan and any adjustments, while providing disease-related education, helping the patient to better understand their condition and how to manage daily life, enhancing the patient's self-management ability.

The data feedback module 5 is used for collecting the data management scheme and the monitoring requirement use feedback and optimally adjusting the data management scheme and the monitoring requirement based on the use feedback.

In particular, feedback in determining which aspects are most important for assessment and improvement, including patient satisfaction, treatment outcome, ease of use, etc., appropriate tools and methods are selected to collect feedback, such as designing questionnaires for patients and healthcare workers, collecting satisfaction and advice for management of the plan and need for diagnosis, conducting one-to-one interview or focus team discussions with patients and healthcare workers, in-depth knowledge of their experiences and needs, and using existing health data and usage records to analyze the actual performance of the plan and the health outcome of the patient.

Specifically, questionnaires and surveys are designed, questionnaires for different target groups are created, such as patients and medical staff, satisfaction, treatment effect, operation convenience and other aspects are covered, one-to-one interviews or group of tissue focus discussions are regularly carried out with the patients and medical staff, deeper feedback and suggestions are obtained, meanwhile, key data and use conditions in the treatment process are automatically collected and recorded by using an electronic health record system and other digital tools, all feedback is ensured to be collected by a standardized method, and unified questionnaires and investigation tools and standardized data input formats are used.

The method comprises the steps of periodically checking the integrity and accuracy of data, ensuring the reliability and consistency of the data, utilizing methods such as time series analysis and the like, tracking the change trend of key data points, such as the change of patient satisfaction, identifying long-term effects and potential problems, analyzing the collected data, identifying the advantages and potential defects of a scheme and the trend of the change along with time, formulating specific improvement strategies such as adjusting treatment plans, optimizing data management processes or adding medical resources based on data analysis results, reducing the improvement measures to be implemented, setting clear evaluation standards and time points, and continuously monitoring the effects of the measures.

The method comprises the steps of ensuring that feedback is systematically and standardized, comparing and long-term tracking data, storing and managing the collected data by using an electronic health record system or special feedback management software, analyzing the collected feedback data in detail, identifying the advantages and potential defects of a scheme, checking the trend of time change, such as improvement or decline of patient satisfaction, identifying and analyzing specific factors causing dissatisfaction or problems, identifying which aspects or measures achieve good effects, considering whether the application can be expanded or not, formulating specific improvement measures according to feedback analysis results, including adjusting treatment plans or management measures according to specific requirements of patients, improving data management and monitoring procedures, improving efficiency and response speed, adding medical resources, such as equipment, medicines or manpower resources according to needs, and performing additional training on medical staff to improve the execution capacity and efficiency of the scheme.

The improvement is put into practice and the criteria and points in time for detecting and evaluating these improvements are set, the effect of the monitoring regimen is continued, further adjustments are made as necessary, and the patient and healthcare worker are left open and in continuous communication informing them of the performance and effectiveness of the improvement, encouraging them to continue providing feedback, forming a positive feedback loop.

As shown in fig. 2, according to another embodiment of the present invention, there is provided a patient data analysis and management method, which includes the steps of:

In summary, by means of the above technical solution of the present invention, the system adopts a modularized design concept to make the acquisition, analysis, management and feedback of the patient data independent into modules, carefully process the patient data, improve the accuracy and usability of the patient data, provide a solid foundation for subsequent analysis, and simultaneously identify the specific needs and potential health risks of the patient by classifying, weighting and trend analysis of the patient data, so as to facilitate the doctor to formulate a more personalized diagnosis and treatment scheme.

In addition, the integrated data management module is used for generating and adjusting the patient monitoring diagnosis and treatment plan, improving the utilization efficiency of medical resources, reducing the medical resource waste, collecting the use feedback, optimizing the diagnosis and treatment and management scheme based on the feedback, and simultaneously simulating and predicting the evolution trend of patient data by utilizing the trend prediction module and the time sequence analysis algorithm, so that the health management and prevention measure formulation of patients are realized according to the patient data.

In addition, the invention accurately matches the patient demands and hospital resources through the design of the parameter comparison module, improves individuation and satisfaction of medical services, and improves the accuracy of patient data processing and accurately reflects the actual conditions and demands of patients by comprehensively considering a plurality of data factors and classifying and weighting the patient data.

The foregoing description of the preferred embodiments of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, alternatives, and improvements that fall within the spirit and scope of the invention.

Claims

1. A patient data based analysis management system, comprising: the system comprises a data acquisition module, a data analysis module, a data influence module, a data management module and a data feedback module;

the data analysis module is used for classifying the cleaned data and giving weight to the classification result of the data;

The data feedback module is used for collecting data management scheme and monitoring requirement use feedback and optimizing and adjusting the data management scheme and the monitoring requirement based on the use feedback;

The data analysis module comprises: the system comprises a data factor module, a factor ratio module, a factor classification module and a data weighting module;

the data weighting module is used for weighting the data according to the classification result and the characteristic factors;

the data weighting module comprises: the system comprises a weighting calculation module, a result verification module, a weighting influence module and a weighting optimization module;

the weighting influence module is used for calculating a weighting comprehensive influence value according to the verified category division result weight value and the characteristic factor weight value;

the data influencing module comprises: the system comprises a weighting analysis module, a trend prediction module, a patient demand module, a parameter comparison module and an influence calculation module;

the influence calculation module is used for calculating a data influence value according to the hospital matching value and the data evolution parameter;

the calculating the data influence value according to the hospital matching value and the data evolution parameter comprises:

collecting and sorting influence data related to the health indexes of the patient, and preprocessing the influence data;

weighting the influence data, and calculating a data influence value of the influence data weight by adopting a weighted summation model and a data fitting method;

Verifying whether the calculated data influence value is consistent with the actual patient result, and adjusting the data influence value based on the verification result;

the generating the data management scheme and the monitoring requirement according to the data influence value comprises the following steps:

Classifying the data influence values in a management monitoring level, and making a monitoring diagnosis and treatment plan according to the classification result of the management monitoring level;

Carrying out data management and implementation of a monitoring diagnosis and treatment plan by using a preset medical information system and a data analysis tool, and collecting and managing data change of the monitoring diagnosis and treatment plan;

periodically evaluating the effect of the monitoring diagnosis and treatment plan according to the data change result, and adjusting the monitoring diagnosis and treatment plan based on the evaluation result;

the collecting the data management scheme and the monitoring requirement using feedback, and optimizing the adjusting the data management scheme and the monitoring requirement based on the using feedback comprises:

Acquiring data management scheme and monitoring demand use feedback, and extracting characteristic parameters of the data management scheme and the monitoring demand use feedback;

Analyzing the extracted characteristic parameters of the feedback, and acquiring a data management scheme and a requirement factor of the monitoring requirement using the feedback based on an analysis result;

Formulating improvement measures based on demand factors, acquiring improvement parameters of the improvement measures in real time, and adjusting a data management scheme and monitoring demands based on the improvement parameters;

the parameter comparison module comprises: the system comprises a data docking module, a demand matching module and a verification optimizing module;

the verification optimization module is used for verifying and optimizing the demand data matching value and outputting the demand data matching value after verification and optimization as a hospital matching value;

the calculation formula for matching the required data characteristic parameters with the hospital information characteristic parameters and calculating the required data matching value is as follows: ；

Wherein W is a demand data matching value;

The component value of the ith feature in the hospital information feature parameters;

The calculating the weighted comprehensive influence value according to the verified class classification result weight value and the characteristic factor weight value comprises the following steps:

Adding an influence grade label to the weighting comprehensive influence value based on the comparison result of the validated weighting comprehensive influence value and the weighting comprehensive influence threshold;

the predicting the data evolution parameters according to the data change trend judging result and the data classification weighting result comprises the following steps:

calculating data evolution parameters according to the data change values and the trend evolution rules, and verifying predicted data evolution parameters;

The calculating the data evolution parameters according to the data change values and the trend evolution rules, and verifying the predicted data evolution parameters comprises:

constructing a data evolution model through a time sequence analysis algorithm;

2. A patient data based analysis and management method for implementing the patient data based analysis and management system of claim 1, comprising the steps of: