Personalized medical-seeking path planning method and platform fused with data analysis

Abstract

The invention provides a personalized medical-seeking path planning method and platform fused with data analysis, and relates to the technical field of medical-seeking path planning, and the method comprises the steps: collecting and obtaining a multi-platform medical data flow; obtaining a medical attribute index set, and carrying out attribute classification identification on the available multi-platform medical data flow; performing data fusion analysis according to the multi-platform data attribute information set; constructing a medical summary knowledge base, and performing feature clustering analysis on the medical summary knowledge base to obtain a general medical knowledge base and a personalized medical knowledge base; and performing medical treatment path planning analysis on the target medical treatment fusion feature to obtain target medical treatment path information, and performing personalized medical treatment management on the target user. According to the method and the device, the technical problem of poor medical path planning accuracy caused by non-uniform platform data and difficulty in information intercommunication in the prior art can be solved, and the medical path planning accuracy and personalization are improved by integrating the data of a plurality of medical platforms.

Description

Technical Field

[0001] The present application relates to the technical field of medical route planning, and in particular to a personalized medical route planning method and platform integrating data analysis. Background Art

[0002] With the gradual enrichment of medical resources and the increasing complexity of patients' medical needs, providing patients with efficient, convenient and accurate medical services has become increasingly important. Medical route planning refers to generating the best medical route based on multiple factors such as the patient's condition, geographical location, medical resources, and medical service capabilities. Existing medical route planning is mostly based on the patient's basic health information and the hospital's geographical location information to optimize the patient's medical process from home to hospital and avoid excessive waiting or unnecessary referrals. Personalized route planning is not only about the patient's route from home to hospital, but also involves how to choose the most appropriate medical service, the best treatment path, the most suitable medical institution, etc. At present, medical data is usually distributed in different medical platforms and institutions. Due to the diversity of medical data and data barriers between platforms, patients' medical data is often missing or inconsistent, resulting in the inability to obtain patients' complete medical information during route planning, thereby affecting the accuracy and comprehensiveness of route planning.

[0003] In summary, the existing technology has technical problems such as the inconsistency of data on various medical platforms, inability to effectively share and integrate data, resulting in information being unable to communicate with each other, further affecting the accuracy of medical route planning. Summary of the invention

[0004] The purpose of this application is to provide a personalized medical route planning method and platform that integrates data analysis, so as to solve the technical problems in the prior art that the data of various medical platforms are not unified, cannot be effectively shared and integrated, resulting in the inability to communicate information, further affecting the accuracy of medical route planning.

[0005] In view of the above problems, this application provides a personalized medical path planning method and platform that integrates data analysis.

[0006] In the first aspect, the present application provides a personalized medical path planning method based on fusion data analysis, which is implemented through a personalized medical path planning platform based on fusion data analysis, wherein the personalized medical path planning method based on fusion data analysis includes: collecting and acquiring multi-platform medical data streams of target users, transmitting the multi-platform medical data streams to an information interaction center for integrated preprocessing, and obtaining available multi-platform medical data streams; acquiring a medical attribute indicator set, and classifying and identifying the attributes of the available multi-platform medical data streams based on the medical attribute indicator set to obtain a multi-platform data attribute information set; performing data fusion analysis on the available multi-platform medical data streams according to the multi-platform data attribute information set to obtain target medical fusion features; constructing a medical summary knowledge base, and performing feature clustering analysis on the medical summary knowledge base to obtain a general medical knowledge base and a personalized medical knowledge base; performing medical path planning analysis on the target medical fusion features based on the general medical knowledge base and the personalized medical knowledge base to obtain target medical path information, and performing personalized medical management on the target user through the target medical path information.

[0007] In the second aspect, the present application also provides a personalized medical path planning platform for integrated data analysis, which is used to execute the personalized medical path planning method for integrated data analysis as described in the first aspect, wherein the personalized medical path planning platform for integrated data analysis includes: a data acquisition module, the data acquisition module is used to collect and obtain the multi-platform medical data stream of the target user, and transmit the multi-platform medical data stream to the information interaction center for integrated preprocessing to obtain a usable multi-platform medical data stream; an attribute classification module, the attribute classification module is used to obtain a medical attribute indicator set, and attribute classification and identification of the available multi-platform medical data stream based on the medical attribute indicator set to obtain multi-platform data attributes. Information set; a data fusion module, the data fusion module is used to perform data fusion analysis on the available multi-platform medical data stream according to the multi-platform data attribute information set to obtain target medical fusion characteristics; a clustering analysis module, the clustering analysis module is used to build a medical summary knowledge base, perform feature clustering analysis on the medical summary knowledge base, and obtain a general medical knowledge base and a personalized medical knowledge base; a path planning module, the path planning module is used to perform medical path planning analysis on the target medical fusion characteristics based on the general medical knowledge base and the personalized medical knowledge base to obtain target medical path information, and perform personalized medical management on the target user through the target medical path information.

[0008] One or more technical solutions provided in this application have at least the following technical effects or advantages: By collecting and acquiring the multi-platform medical data stream of the target user, the multi-platform medical data stream is transmitted to the information exchange center for integrated preprocessing to obtain an available multi-platform medical data stream; a medical attribute indicator set is obtained, and the available multi-platform medical data stream is attributed and identified based on the medical attribute indicator set to obtain a multi-platform data attribute information set; data fusion analysis is performed on the available multi-platform medical data stream according to the multi-platform data attribute information set to obtain a target medical fusion feature; a medical summary knowledge base is constructed, and feature clustering analysis is performed on the medical summary knowledge base to obtain a general medical knowledge base and a personalized medical knowledge base; medical path planning analysis is performed on the target medical fusion feature based on the general medical knowledge base and the personalized medical knowledge base to obtain target medical path information, and personalized medical management is performed on the target user through the target medical path information. That is to say, by collecting multi-platform medical data streams, and performing attribute classification and data fusion on them, a medical summary knowledge base is constructed and feature clustering analysis is performed, and medical path planning is performed for the target user through the target medical fusion feature and knowledge base, thereby improving the accuracy and personalization of medical path planning.

[0009] The above description is only an overview of the technical solution of the present application. In order to more clearly understand the technical means of the present application, it can be implemented according to the contents of the specification, and in order to make the above and other purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are specifically cited below. It should be understood that the content described in this section is not intended to identify the key or important features of the embodiments of the present application, nor is it intended to limit the scope of the present application. Other features of the present application will become easy to understand through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0010] In order to more clearly illustrate the technical solutions in the present application or the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only exemplary, and for ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0011] Figure 1 A flowchart of a personalized medical treatment pathway planning method integrating data analysis for this application; Figure 2 Schematic diagram of the structure of the personalized medical path planning platform that integrates data analysis for this application.

[0012] Explanation of reference numerals: data acquisition module 11 , attribute classification module 12 , data fusion module 13 , cluster analysis module 14 , path planning module 15 . DETAILED DESCRIPTION

[0013] This application solves the technical problem in the prior art that the data of various medical platforms are not unified, cannot be effectively shared and integrated, and cannot be interoperable, which further affects the accuracy of path planning by providing a personalized medical path planning method and platform for integrated data analysis. By collecting medical data streams from multiple platforms, classifying their attributes and fusing their data, building a medical summary knowledge base and performing feature clustering analysis, and planning medical paths for target users through target medical fusion features and knowledge bases, the accuracy and personalization of medical path planning are improved.

[0014] Below, the technical solutions in the present application will be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments of the present application. It should be understood that the present application is not limited to the example embodiments described herein. Based on the embodiments of the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application. It should also be noted that, for the convenience of description, only the parts related to the present application are shown in the accompanying drawings, rather than all of them.

[0015] For example, please refer to the attached Figure 1 The present application provides a personalized medical path planning method based on fusion data analysis, wherein the personalized medical path planning method based on fusion data analysis is applied to a personalized medical path planning platform based on fusion data analysis, and the personalized medical path planning method based on fusion data analysis specifically includes the following steps: Step 1: Collect and obtain the multi-platform medical data stream of the target user, and transmit the multi-platform medical data stream to the information exchange center for integrated preprocessing to obtain a usable multi-platform medical data stream.

[0016] Specifically, the target user's medical data is collected from multiple medical platforms (such as hospital information systems, clinics, health monitoring devices, mobile health applications, electronic medical record platforms, etc.), including the target user's medical records, test results, historical cases, drug allergy records, disease diagnosis, imaging data, etc. The target user refers to the patient who needs to plan the medical path, and usually needs to provide personalized medical services based on their health records and medical data. Multi-platform medical data flow refers to data obtained from multiple medical platforms, including patients' medical information obtained from the hospital's electronic health record (EHR) system, clinics, health management applications, smart medical devices, etc.

[0017] Transmit multi-platform medical data streams to the information interaction center for integrated preprocessing. The information interaction center is a center that centrally manages and processes multi-source medical data. It receives, stores, and processes data from different platforms, and ensures that data can interact and share between different systems. Integrated preprocessing refers to the first step of data processing after the multi-platform medical data streams are transmitted to the information interaction center. The data from different platforms are uniformly formatted, cleaned, deduplicated, and processed to ensure data consistency, accuracy, and completeness. The data formats of different platforms may be different, so format conversion is required. After that, remove errors, missing values, or inconsistent parts of the data. In addition, different platforms may record information about the same patient, and data duplication may lead to errors in path planning and decision-making processes, which require deduplication.

[0018] After integrated preprocessing, all data are converted to a unified format, redundancy, errors, and inconsistencies are removed, and all data conform to relevant medical data standards. Sensitive data is identified for the processed multi-platform medical data stream, and each data is desensitized according to its corresponding sensitivity level to obtain a usable multi-platform medical data stream. The privacy and security of medical data during transmission and processing are ensured. Sensitive data identification and desensitization effectively reduce the risk of data leakage while retaining data availability.

[0019] Step 2: Obtain a medical attribute indicator set, classify and identify the available multi-platform medical data stream based on the medical attribute indicator set, and obtain a multi-platform data attribute information set.

[0020] Specifically, a standard set of medical attribute indicators is defined and collected, covering various aspects of patient health-related data, including clinical data, diagnostic data, laboratory test results, and patients' personal information. A medical attribute indicator set refers to a set of characteristic indicators used to describe and analyze medical data. Medical attribute indicator sets are usually defined according to standards in the medical field and can be adjusted according to specific needs to ensure the comprehensiveness and standardization of the indicator set.

[0021] The available multi-platform medical data streams are classified and identified according to the medical attribute indicator set. In the multi-platform medical data streams, each piece of data needs to be classified according to its specific type. The data identification method usually adopts a labeling method to mark each data item. In the medical data stream, data from different platforms may use different coding systems or units. In order to ensure the uniformity of the data, these data need to be standardized, usually using an internationally standardized coding system for unification (such as ICD-10 for disease diagnosis codes and LOINC for laboratory test data).

[0022] After the attribute classification and identification is completed, a multi-platform data attribute information set is obtained, which includes medical data transmitted on different platforms and is classified through unified standards and identification. By classifying and identifying medical data, medical data is made more structured, providing more accurate basic data for personalized medical path planning.

[0023] Step three: Perform data fusion analysis on the available multi-platform medical data stream according to the multi-platform data attribute information set to obtain target medical fusion features.

[0024] Specifically, the available multi-platform medical data streams are diverted and integrated according to the multi-platform data attribute information set, and the medical data streams are divided into several different categories according to the type and attributes of the data to obtain medical attribute classification data streams, that is, data streams classified according to different attributes. Data from different platforms may have missing values, duplicate records, inconsistent formats, etc., and it is necessary to perform abnormal identification and data preprocessing on the medical attribute classification data streams, that is, to process the abnormal values, missing values, duplicate values, inconsistent formats, etc. in the medical attribute classification data streams so that the data can be in a unified format and meet the analysis requirements. Then the data is standardized to convert data from different sources into the same metric.

[0025] After data integration, the data is fused and analyzed, and data from multiple sources are integrated and analyzed to extract valuable information. According to the multi-platform data attribute information set, the extraction algorithm for each data feature is determined, and a suitable data feature extraction algorithm model set is determined. Different medical data streams require different algorithms to extract valuable features. Use the corresponding algorithm to extract features and perform data fusion analysis on the standardized medical attribute classification data stream, assign corresponding weights to different types of medical data according to the importance, reliability, credibility, etc. of each attribute information, and fuse them, that is, weighted merge the features to obtain a comprehensive patient portrait and obtain the target medical fusion features. By integrating data from multiple platforms, the target medical fusion features obtained can more comprehensively and accurately assess the patient's health status, cover multi-dimensional information, and formulate tailored path planning based on the patient's specific health status and disease risk.

[0026] Step 4: Construct a medical summary knowledge base, perform feature clustering analysis on the medical summary knowledge base, and obtain a general medical knowledge base and a personalized medical knowledge base.

[0027] Specifically, by collecting data from different medical platforms (such as electronic medical records, medical research, clinical diagnosis records, etc.), a medical summary knowledge base is constructed, which contains a summary knowledge base of information such as various medical conditions, department locations, attending physicians, drug use, and test results. Collect a variety of data from medical systems, health management platforms, scientific research databases, etc., such as disease type, diagnosis information, department location, attending physician, patient medical history, etc. The collected data is processed by deduplication, missing value filling, and outlier detection to ensure the accuracy and completeness of the data. Convert data from different sources into a unified format for subsequent processing and analysis.

[0028] The purpose of clustering analysis on the features in the medical summary knowledge base is to identify and classify similar disease types, symptom characteristics, etc. Feature clustering analysis is to analyze the features of the data, find the similarities or differences between the data, and cluster the data with similar features into the same cluster to achieve the purpose of classifying and understanding the data. Common clustering analysis algorithms include K-means clustering and hierarchical clustering. Among them, K-means clustering refers to dividing the data into K clusters to optimize the uniformity of samples within the cluster and the difference of samples between clusters; hierarchical clustering refers to gradually merging or splitting data clusters according to the distance relationship between data points to form a tree structure. Through feature clustering analysis, the corresponding feature data of each medical disease type are clustered according to the knowledge attributes of medical diseases. Each medical disease corresponds to a cluster cluster, and multiple cluster clusters are obtained.

[0029] According to the preset correlation threshold, multiple clusters are analyzed for correlation, and multiple in-cluster correlation data and multiple out-cluster discrete data are obtained. The in-cluster data with strong correlation is used as a general medical knowledge base, that is, it contains common diseases; the out-cluster data with low correlation is used as a personalized medical knowledge base, that is, the personalized medical knowledge base focuses on the unique situation of individual patients, including personal medical history, genetic information, allergy information, etc. By constructing a medical summary knowledge base and performing feature clustering analysis, a general medical knowledge base and a personalized medical knowledge base are finally obtained, which helps to optimize personalized path planning.

[0030] Step 5: Perform medical path planning analysis on the target medical fusion features based on the general medical knowledge base and the personalized medical knowledge base to obtain target medical path information, and perform personalized medical management for the target user through the target medical path information.

[0031] Specifically, path planning analysis is performed based on the patient's target medical integration characteristics (including medical history, examination data, medical needs, etc.), combined with the general medical knowledge base and personalized medical knowledge base. The general medical knowledge base provides a standardized medical path based on extensive patient data, which is applicable to most common diseases and health conditions. The personalized medical knowledge base is based on the specific situation of each patient. By analyzing and summarizing their historical medical data, diagnosis and treatment needs, disease characteristics and other information, a personalized medical data set dedicated to the patient is constructed, which includes the specific patient's medical needs, medical history, family history, etc.

[0032] The target medical fusion features are traversed and compared in the general medical knowledge base, and the general medical path information with the greatest matching degree is screened out. The target medical fusion features are traversed and compared in the personalized medical knowledge base to obtain personalized medical path information. Combine the general medical path information and the personalized medical path information to obtain the target medical path information. After path planning analysis, the target medical path information obtained will include detailed medical service steps, including the patient's registration arrangement, examination and diagnosis process, drug use, allergy translation, rehabilitation plan, etc.

[0033] According to the target medical path information, personalized medical management is carried out to ensure that the patient's medical process can proceed smoothly, and the path is adjusted and optimized in a timely manner. As the patient's medical process progresses, feedback information (such as the smoothness of appointment registration, waiting time, patient experience feedback, etc.) is collected in real time, and the path is adjusted when necessary. Through the patient's electronic health record and mobile health management system, the patient's medical progress is monitored in real time to ensure that it proceeds as planned. Keep communicating with the patient to ensure that the patient understands and cooperates with the treatment path, and also obtain patient feedback in a timely manner to facilitate path adjustment. Through the integration of the general medical knowledge base and the personalized medical knowledge base, it is ensured that the medical path can be optimized according to the patient's personalized needs, reducing problems such as waiting in line and missed examinations, greatly improving the patient's medical experience, and improving the accuracy of path planning.

[0034] Furthermore, step one of this application includes: The information exchange center constructs a sensitive data identifier according to the medical data application standard; performs sensitive data identification on the multi-platform medical data stream based on the sensitive data identifier to obtain a medical sensitive data set; performs sensitivity level assessment on the medical sensitive data set according to the medical sensitivity level assessment system to obtain sensitive data level information; constructs a data desensitization grading strategy, and performs integrated and graded desensitization on the multi-platform medical data stream according to the sensitive data level information based on the data desensitization grading strategy to obtain the usable multi-platform medical data stream.

[0035] Specifically, in the information exchange center, according to the medical data application standards, a sensitive data identifier is built to detect sensitive information in medical data through automated tools. Medical data application standards refer to the data format, data structure and data exchange rules specified in the medical industry to ensure that data can be shared effectively and accurately between different platforms, including HL7, FHIR, DICOM, etc. The sensitive data identifier is used to identify the part of medical data that contains sensitive information. Sensitive information usually refers to data that threatens patient privacy or security, such as personal identity information, medical records, medical diagnosis, drug allergy records, etc.

[0036] Analyze the application standards of medical data, understand the specific format and content of the data, and formulate rules for identifying sensitive data. Define which data is considered sensitive data according to the data standards, such as ID number, disease diagnosis, drug use, etc. This information usually requires special protection to avoid unauthorized access and use. Automatically analyze and identify sensitive data based on predefined rules. For unstructured text data (such as medical records, doctor's diagnosis, etc.), use NLP technology to identify sensitive data. NLP can extract sensitive information (such as disease name, drug name, specific diagnosis and treatment plan, etc.) from the text through vocabulary matching, entity recognition and other technologies. After identifying sensitive data, the sensitive data identifier will mark it and distinguish it from non-sensitive data. Add identification to sensitive data fields, such as sensitive labels. For data marked as sensitive, use encryption, desensitization and other technologies to process it to ensure that even if the data is leaked, sensitive information cannot be recovered.

[0037] The multi-platform medical data stream is input into the sensitive data identifier for sensitive data identification, and the sensitive data contained therein is identified to obtain a medical sensitive data set. The medical sensitivity level assessment system is a standardized assessment framework used to grade medical data according to its sensitivity. Medical data is usually divided into different sensitivity levels, such as high sensitivity, medium sensitivity, and low sensitivity, which helps to determine data protection measures. The medical sensitive data set is sent to the sensitivity level assessment system, and different types of sensitive data are graded according to preset standards to obtain sensitive data level information. The data factors considered by the assessment system include the privacy, importance, and possible harm of the data after leakage. For example, personal information such as a user's ID number is usually rated as highly sensitive, while the user's blood pressure data, weight, etc. may be rated as low sensitive, which helps medical institutions decide how to handle this data and also helps to choose appropriate desensitization technology when desensitizing the data.

[0038] Construct a data desensitization classification strategy and use different desensitization methods for data of different sensitivity levels. For highly sensitive data, such as patients' personal information and medical records, encryption (such as AES encryption) or pseudo-anonymization (using pseudo-information as a substitute) and other technologies are used, so that even if these data are intercepted during transmission, the original information cannot be restored; for medium-sensitive data, the specific information is blurred; and for low-sensitive data, field truncation and other technologies are used. Data desensitization is the process of processing sensitive data, removing or changing its sensitive parts so that it cannot directly identify the patient's identity, thereby protecting privacy, including data encryption, data anonymization, and falsified data. According to the level information of sensitive data, determine the desensitization method of data of different sensitivity levels to form a data desensitization classification strategy.

[0039] According to the data desensitization classification strategy and sensitive data level information, the multi-platform medical data stream is integrated and graded for desensitization. Integrated graded desensitization refers to desensitizing data to different degrees according to the sensitivity level. According to the sensitive data level information, each data in the data stream is classified, and its sensitivity level is determined, and the corresponding desensitization strategy is used for desensitization. The desensitized data is integrated together to form a unified multi-platform medical data stream, that is, a usable multi-platform data stream. By performing sensitive identification and graded protection on the multi-platform medical data stream, it is ensured that the user's personal identity information and sensitive health data will not be leaked during the data exchange and sharing process.

[0040] Further, step three of this application includes: The available multi-platform medical data streams are split and integrated according to the multi-platform data attribute information set to obtain a medical attribute classification data stream; a data preprocessing program is obtained, and abnormal data identification and data preprocessing are performed on the medical attribute classification data stream based on the data preprocessing program to obtain a standard medical attribute classification data stream; a data feature extraction algorithm model set is determined according to the multi-platform data attribute information set; feature extraction and data fusion analysis are performed on the standard medical attribute classification data stream based on the data feature extraction algorithm model set to obtain target medical fusion features.

[0041] Specifically, the multi-platform data attribute information set refers to a structured data set containing various medical attributes formed by classifying and identifying data from different medical platforms, which includes all medical data that have been identified, classified and standardized. Use the multi-platform data attribute information set to divert and integrate the available medical data streams, and divide the available multi-platform medical data streams into several different categories according to the type and attributes of the data, so that each type of data can be processed and analyzed separately to obtain a medical attribute classification data stream. The medical attribute classification data stream is to divert and integrate the medical data stream that has been classified and identified according to its attributes, that is, to divide and merge different categories of data (such as basic patient information, patient diagnosis results, patient precautions, etc.) according to different logics.

[0042] The data preprocessor is a tool or algorithm used to perform preliminary cleaning, conversion and processing of medical data, including outlier detection, missing value filling, data normalization, etc., to improve data quality and eliminate noise and erroneous data. The data preprocessor is used to identify and preprocess the medical attribute classification data stream through abnormal data identification and data preprocessing to identify abnormal data. Abnormal data identification is abnormal values ​​or unreasonable data caused by data collection errors, data transmission errors or extreme situations, including extremely abnormal body temperature (such as 40℃), unreasonable age, etc. Abnormal data is detected by setting standard thresholds or using statistical methods (such as box plots, standard deviation methods). For example, a body temperature higher than 42℃ may be an abnormal value caused by data collection errors or extreme conditions, which needs to be identified and corrected. For processing missing data, mean filling, interpolation or deletion of missing data can be used to standardize or normalize the data to ensure that the units of all data are consistent. After preprocessing, a standard medical attribute classification data stream is obtained. The data stream is processed by outlier identification, data cleaning, formatting and standardization based on classification to ensure data quality and consistency.

[0043] According to the multi-platform data attribute information set, determine the appropriate data feature extraction algorithm model set. Different medical data streams may require different algorithms to extract valuable features. For example, for clinical data (such as weight, blood sugar, etc.), use statistical analysis methods (such as mean, standard deviation, correlation analysis) to extract useful features; for patients' medical history, diagnostic information, etc., use text analysis methods (such as natural language processing technology, TF-IDF, etc.) to extract disease-related features; for laboratory test data, use clustering algorithms (such as K-means) to discover the rules of different patient groups.

[0044] According to the data feature extraction algorithm model set, feature extraction is performed on the standard medical attribute classification data stream, and different weights are assigned to each data feature according to the importance, reliability, credibility, etc. of each attribute. The final target medical fusion feature is obtained by weighted fusion (such as weighted average) of the multi-attribute medical key feature set. The purpose of feature extraction is to extract valuable key information from various types of medical data of patients. Data fusion analysis is the process of integrating and analyzing data from different sources or platforms, with the goal of obtaining a more comprehensive and accurate feature representation by merging multiple data sources. In medical data analysis, data fusion usually involves combining different types of medical data (such as clinical data, examination data, medical history data, etc.) to obtain a comprehensive patient portrait. The target medical fusion feature is the final feature set obtained through data fusion analysis, which is the extraction result of key information in the medical data stream. Through data preprocessing and feature extraction, the data format of different platforms is ensured to be consistent. Through feature extraction and data fusion analysis, the comprehensive characteristics of the patient's health status are obtained, which is helpful to formulate a targeted and personalized medical path plan.

[0045] Furthermore, the present application also includes the following steps: Based on the data feature extraction algorithm model set, feature extraction is performed on the standard medical attribute classification data stream to obtain a multi-attribute medical related feature set; correlation analysis and feature dimension reduction are performed on the multi-attribute medical related feature set to obtain a multi-attribute medical key feature set; based on the multi-platform data attribute information set, a data attribute fusion decision factor set is determined; according to the data attribute fusion decision factor set, data weighted fusion is performed on the multi-attribute medical key feature set to obtain target medical fusion features.

[0046] Specifically, the data feature extraction algorithm model set includes feature extraction of different types of data streams, and extracting valuable key information from various types of medical data of patients. The multi-attribute medical association feature set refers to a feature set containing the relationship between multiple medical attributes obtained by extracting features from standard medical attribute classification data streams, revealing the intrinsic connection between different variables in medical data. For example, the body mass index (BMI) is extracted from blood sugar and weight data, and the diabetes risk level is extracted from blood sugar and eating habits data.

[0047] Perform correlation analysis (such as Pearson correlation coefficient, association rule mining, etc.) on multi-attribute medical association feature sets to reveal the process of mutual influence and relationship between different features. For example, through Pearson correlation analysis, calculate the correlation between features and find highly correlated feature pairs. Feature dimensionality reduction, through dimensionality reduction methods such as PCA or LDA, extract several principal components from features of multiple dimensions (such as weight, blood sugar, blood pressure, etc.), reduce the number of features, retain the main variability of the data, and reduce the complexity of the data. In the process of dimensionality reduction, unimportant or redundant features are eliminated to retain the most critical features for disease prediction and path planning.

[0048] Based on the multi-platform data attribute information set, the importance, reliability and credibility of each data attribute are analyzed to determine the data attribute fusion decision factor set. The data attribute fusion decision factor set refers to the set of key decision factors selected according to the importance, reliability and credibility of different data attributes when conducting medical pathway planning or health management. By comprehensively considering the information of different attributes, a suitable medical planning path is generated. Each attribute will be given a different weight according to its influence on the decision. A weight is assigned to each data attribute, and the weight reflects the influence of the attribute in medical decision-making. According to the needs of health management and treatment pathway planning, the most decision-making factor is selected.

[0049] According to the determined data attribute fusion decision factor set, the multi-attribute medical key feature set is weighted and fused to obtain the target medical fusion feature. In other words, according to the weight of each feature, the weighted values ​​of each feature are fused to obtain the target medical fusion feature, which fully reflects the patient's health status. Through multi-attribute medical feature extraction and weighted fusion, the influence of different data sources can be comprehensively considered, so as to more comprehensively evaluate the patient's health status.

[0050] Further, step 4 of this application includes: According to the medical summary knowledge base, a patient disease diagnosis data set is obtained; a feature clustering analysis is performed on the patient disease diagnosis data set according to the medical disease knowledge attribute to obtain a plurality of diagnosis data clustering clusters; based on a preset correlation threshold, data association analysis is performed on the plurality of diagnosis data clustering clusters to obtain a plurality of in-cluster association data and a plurality of out-cluster discrete data; based on the plurality of in-cluster association data and the plurality of out-cluster discrete data, the medical summary knowledge base is divided and integrated to obtain the general medical knowledge base and the personalized medical knowledge base.

[0051] Specifically, based on the medical summary knowledge base, the patient's disease diagnosis data set is extracted, including symptoms, signs, medical history, test results, etc. For example, the blood sugar level, weight, genetic factors and other data of diabetic patients. Medical disease knowledge attributes are the knowledge information corresponding to each medical disease (such as diabetes, heart disease, etc.). For example, the medical disease knowledge attributes of diabetes include blood sugar level, weight, eating habits, genetic factors, etc., and the medical disease knowledge attributes of heart disease include high blood pressure, family history, exercise habits, etc. Each disease has a specific set of knowledge attributes, which are used to identify, classify and analyze the disease.

[0052] According to the specific knowledge attributes of each disease, the patient disease diagnosis data set is clustered, and data with similar characteristics are classified into the same category to form multiple diagnosis data clusters. The medical disease knowledge attributes provide the corresponding diagnostic knowledge information for each medical disease type. According to the medical disease knowledge attributes, the corresponding feature data of each medical disease type are clustered. Each medical disease corresponds to a cluster. The data attributes within each cluster are similar, and the differences between clusters are large.

[0053] Through statistical analysis or the experience of domain experts, a correlation threshold is set. The preset correlation threshold is a standard value used to determine whether there is a significant correlation between different diagnostic data. By setting a threshold, the degree of correlation between different data can be distinguished, and then it can be decided whether to classify certain data into one category or one group. Intra-cluster correlation data refers to data with strong correlation in the same cluster, which usually represents the common characteristics of a certain disease type or patient group. For example, the blood sugar, weight, diet and other characteristics of diabetic patients may have a high correlation in the same cluster. Out-of-cluster discrete data refers to data that has a weak relationship with other in-cluster data, representing patient-specific disease characteristics or allergic reactions, etc., and is usually manifested as outliers or special cases far away from other data points in the feature space.

[0054] The medical summary knowledge base is divided and integrated according to multiple in-cluster associated data and multiple out-cluster discrete data. The in-cluster data with strong correlation is classified as common medical data, and the out-cluster data with low correlation is classified as unique individual medical data. The general medical knowledge base and the personalized medical knowledge base are obtained. The general medical knowledge base contains the common diseases of most patients, such as the treatment process and precautions of common diseases such as diabetes and hypertension. The personalized medical knowledge base focuses on the unique situation of individual patients, including personal medical history, genetic information, special disease characteristics, allergy information, etc. Through cluster analysis and association analysis, common diseases and individual differences are identified and distinguished, providing patients with personalized medical planning paths.

[0055] Further, step five of this application includes: Based on the target medical fusion feature, a traversal matching analysis is performed in the general medical knowledge base to obtain a medical knowledge matching set; a preset matching threshold is set, and the medical knowledge matching set is compared and optimized based on the preset matching threshold to obtain general medical path information with the greatest matching degree; if the medical knowledge matching set does not meet the preset matching threshold, the personalized medical knowledge base is called to perform matching analysis and comparison on the target medical fusion feature to obtain personalized medical path information; based on the general medical path information and the personalized medical path information, the target medical path information is determined.

[0056] Specifically, the target medical fusion feature is traversed and matched in the general medical knowledge base, and similarity measurement methods such as cosine similarity and Euclidean distance are used to evaluate the similarity between the target medical fusion feature and each disease feature in the general medical knowledge base. Each matching score represents the similarity between the target medical fusion feature and a disease feature and treatment path in the knowledge base. A high matching result means that the target feature is highly consistent with the information in the knowledge base, indicating that this is a possible applicable path. The preset matching threshold is a set threshold used to determine the degree of matching between the target medical fusion feature and the information in the knowledge base. If the matching degree is lower than the matching threshold, it means that the target feature is not consistent with the information in the general medical knowledge base, and it is necessary to further find a more suitable matching path from the personalized medical knowledge base. If the matching degree is higher than the matching threshold, it means that the information in the general medical knowledge base is sufficiently matched to meet the patient's diagnosis and treatment needs.

[0057] All matching results are compared, and disease diagnosis and treatment pathways with matching degrees greater than or equal to the preset matching degree threshold are selected as general medical path information. General medical path information is a medical path planned for the patient's illness and symptoms based on the standard medical path provided in the general medical knowledge base. It is suitable for most patient groups and can meet the diagnosis and treatment needs of common diseases.

[0058] If the medical knowledge matching degree set does not meet the preset matching degree threshold, a matching analysis and comparison is performed in the personalized medical knowledge base, and matching is performed according to the patient's personalized information (such as special medical history, family inheritance, living habits, etc.) to obtain personalized medical path information, that is, to provide customized medical path planning for specific patients' medical records, special diseases and other information. For example, if the target patient has a special disease history or allergy history (such as allergy to insulin), the standard treatment path in the general knowledge base may not meet the needs. Combine the general medical path information with the personalized medical path information to obtain the target medical path information, that is, the medical path planning provided to the patient. By performing matching analysis of the general and personalized medical knowledge bases based on the target medical fusion characteristics, combined with threshold screening and path optimization, the target medical path information is finally obtained, and the medical path that best matches the patient's health status is accurately matched to avoid delays or unnecessary referrals.

[0059] Furthermore, the present application also includes the following steps: Based on the target medical path information, personalized medical monitoring is performed on the target user to obtain medical effect feedback data; when the medical effect feedback data fails to achieve the expected medical effect, the medical effect feedback data is optimized and analyzed to obtain the medical path optimization direction; the target medical path information is adjusted and optimized according to the medical path optimization direction to obtain a medical optimization path information set; each optimized path in the medical optimization path information set is evaluated and screened to obtain target medical optimization path information, and personalized medical management is performed based on the target medical optimization path information.

[0060] Specifically, the target user seeks medical treatment according to the target medical path information, and at the same time, the target user is monitored for personalized medical treatment, and the patient's feedback effect is tracked, including but not limited to the patient's self-perceived symptoms, treatment side effects, etc. The patient's symptom data is collected through the patient's self-reported symptom changes using mobile health applications, or the patient's self-perception is understood through questionnaires. Medical effect feedback data refers to data collected from the patient's treatment process. By analyzing the patient's feedback, the bottleneck links in the medical path are identified, such as long waiting time, cumbersome registration process, etc.

[0061] Analyze the differences in patients' experience at different stages using statistical and data analysis methods to identify possible pain points. Perform optimization analysis when the feedback data on medical treatment effects does not meet the expected medical treatment effects. By comparing the expected effects with the actual effects, analyzing the deviations and specific influencing factors, we can obtain the optimization analysis results. Based on the optimization analysis results, determine the specific optimization direction, adjust the medical treatment path, obtain multiple backup optimization paths, and combine the multiple backup paths to form a medical treatment optimization path information set.

[0062] For the generated set of optimized path information, evaluate the effect of the optimized path, including the patient's health, the patient's acceptance, possible adverse reactions, etc., and select the most suitable medical path for the patient. For example, simplify the patient's medical process, such as reducing unnecessary examinations and optimizing the flow of patients between departments. Perform personalized medical management for patients based on the screened target medical optimization path information and use new medical paths. Generate personalized optimized path information based on each patient's treatment feedback to ensure that each target user can receive timely medical treatment, eliminate unnecessary links, shorten patient waiting time, and improve overall medical efficiency.

[0063] In summary, the personalized medical path planning method based on fusion data analysis provided in this application has the following technical effects: By collecting and acquiring the multi-platform medical data stream of the target user, the multi-platform medical data stream is transmitted to the information exchange center for integrated preprocessing to obtain an available multi-platform medical data stream; a medical attribute indicator set is obtained, and the available multi-platform medical data stream is attributed and identified based on the medical attribute indicator set to obtain a multi-platform data attribute information set; data fusion analysis is performed on the available multi-platform medical data stream according to the multi-platform data attribute information set to obtain a target medical fusion feature; a medical summary knowledge base is constructed, and feature clustering analysis is performed on the medical summary knowledge base to obtain a general medical knowledge base and a personalized medical knowledge base; medical path planning analysis is performed on the target medical fusion feature based on the general medical knowledge base and the personalized medical knowledge base to obtain target medical path information, and personalized medical management is performed on the target user through the target medical path information. That is to say, by collecting multi-platform medical data streams, and performing attribute classification and data fusion on them, a medical summary knowledge base is constructed and feature clustering analysis is performed, and medical path planning is performed for the target user through the target medical fusion feature and knowledge base, thereby improving the accuracy and personalization of medical path planning.

[0064] Embodiment 2, based on the same inventive concept as the personalized medical path planning method integrating data analysis in the aforementioned embodiment 1, this application also provides a personalized medical path planning platform integrating data analysis, please refer to the attached Figure 2 The personalized medical treatment path planning platform integrating data analysis includes: The data acquisition module 11 is used to collect and acquire the multi-platform medical data stream of the target user, and transmit the multi-platform medical data stream to the information exchange center for integration preprocessing to obtain the available multi-platform medical data stream.

[0065] The attribute classification module 12 is used to obtain a medical attribute indicator set, classify and identify the available multi-platform medical data stream based on the medical attribute indicator set, and obtain a multi-platform data attribute information set.

[0066] The data fusion module 13 is used to perform data fusion analysis on the available multi-platform medical data stream according to the multi-platform data attribute information set to obtain target medical fusion features.

[0067] The cluster analysis module 14 is used to construct a medical summary knowledge base, perform feature cluster analysis on the medical summary knowledge base, and obtain a general medical knowledge base and a personalized medical knowledge base.

[0068] The path planning module 15 is used to perform medical path planning analysis on the target medical fusion characteristics based on the general medical knowledge base and the personalized medical knowledge base, obtain target medical path information, and perform personalized medical management for the target user through the target medical path information.

[0069] Furthermore, the data acquisition module 11 in the personalized medical treatment path planning platform integrating data analysis is also used for: The information exchange center constructs a sensitive data identifier according to the medical data application standard; performs sensitive data identification on the multi-platform medical data stream based on the sensitive data identifier to obtain a medical sensitive data set; performs sensitivity level assessment on the medical sensitive data set according to the medical sensitivity level assessment system to obtain sensitive data level information; constructs a data desensitization grading strategy, and performs integrated and graded desensitization on the multi-platform medical data stream according to the sensitive data level information based on the data desensitization grading strategy to obtain the usable multi-platform medical data stream.

[0070] Furthermore, the data fusion module 13 in the personalized medical treatment path planning platform of fusion data analysis is also used for: The available multi-platform medical data streams are split and integrated according to the multi-platform data attribute information set to obtain a medical attribute classification data stream; a data preprocessing program is obtained, and abnormal data identification and data preprocessing are performed on the medical attribute classification data stream based on the data preprocessing program to obtain a standard medical attribute classification data stream; a data feature extraction algorithm model set is determined according to the multi-platform data attribute information set; feature extraction and data fusion analysis are performed on the standard medical attribute classification data stream based on the data feature extraction algorithm model set to obtain target medical fusion features.

[0071] Furthermore, the data fusion module 13 in the personalized medical treatment path planning platform of fusion data analysis is also used for: Based on the data feature extraction algorithm model set, feature extraction is performed on the standard medical attribute classification data stream to obtain a multi-attribute medical related feature set; correlation analysis and feature dimension reduction are performed on the multi-attribute medical related feature set to obtain a multi-attribute medical key feature set; based on the multi-platform data attribute information set, a data attribute fusion decision factor set is determined; according to the data attribute fusion decision factor set, data weighted fusion is performed on the multi-attribute medical key feature set to obtain target medical fusion features.

[0072] Furthermore, the cluster analysis module 14 in the personalized medical treatment path planning platform of fusion data analysis is also used for: According to the medical summary knowledge base, a patient disease diagnosis data set is obtained; a feature clustering analysis is performed on the patient disease diagnosis data set according to the medical disease knowledge attribute to obtain a plurality of diagnosis data clustering clusters; based on a preset correlation threshold, data association analysis is performed on the plurality of diagnosis data clustering clusters to obtain a plurality of in-cluster association data and a plurality of out-cluster discrete data; based on the plurality of in-cluster association data and the plurality of out-cluster discrete data, the medical summary knowledge base is divided and integrated to obtain the general medical knowledge base and the personalized medical knowledge base.

[0073] Furthermore, the path planning module 15 in the personalized medical path planning platform for fusion data analysis is also used for: Based on the target medical fusion feature, a traversal matching analysis is performed in the general medical knowledge base to obtain a medical knowledge matching set; a preset matching threshold is set, and the medical knowledge matching set is compared and optimized based on the preset matching threshold to obtain general medical path information with the greatest matching degree; if the medical knowledge matching set does not meet the preset matching threshold, the personalized medical knowledge base is called to perform matching analysis and comparison on the target medical fusion feature to obtain personalized medical path information; based on the general medical path information and the personalized medical path information, the target medical path information is determined.

[0074] Furthermore, the path planning module 15 in the personalized medical path planning platform for fusion data analysis is also used for: Based on the target medical path information, personalized medical monitoring is performed on the target user to obtain medical effect feedback data; when the medical effect feedback data fails to achieve the expected medical effect, the medical effect feedback data is optimized and analyzed to obtain the medical path optimization direction; the target medical path information is adjusted and optimized according to the medical path optimization direction to obtain a medical optimization path information set; each optimized path in the medical optimization path information set is evaluated and screened to obtain target medical optimization path information, and personalized medical management is performed based on the target medical optimization path information.

[0075] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. Figure 1 The personalized medical path planning method and specific examples of the fusion data analysis in the first embodiment are also applicable to the personalized medical path planning platform of the fusion data analysis in this embodiment. Through the detailed description of the personalized medical path planning method of the fusion data analysis, those skilled in the art can clearly know the personalized medical path planning platform of the fusion data analysis in this embodiment, so for the sake of brevity of the specification, it will not be described in detail here. For the platform disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part description.

[0076] The above description of the disclosed embodiments enables those skilled in the art to implement or use the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application will not be limited to the embodiments shown herein, but will conform to the widest scope consistent with the principles and novel features disclosed herein.

[0077] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application belong to the scope of the present application and its equivalent technology, the present application is also intended to include these modifications and variations.

Claims

1. A personalized medical path planning method integrating data analysis, characterized in that: include: Collect and obtain the multi-platform medical data stream of the target user, and transmit the multi-platform medical data stream to the information exchange center for integration preprocessing to obtain a usable multi-platform medical data stream; Acquire a medical attribute indicator set, and classify and identify the available multi-platform medical data stream based on the medical attribute indicator set to obtain a multi-platform data attribute information set; Determine a data feature extraction algorithm model set according to the type characteristic information of the multi-platform data attribute information set, and perform feature extraction and data weighted fusion on the available multi-platform medical data streams based on the data feature extraction algorithm model set to obtain target medical fusion features; Constructing a medical summary knowledge base, performing feature clustering analysis on the medical summary knowledge base, and obtaining a general medical knowledge base and a personalized medical knowledge base; Based on the general medical knowledge base and the personalized medical knowledge base, a medical path planning analysis is performed on the target medical fusion feature to obtain target medical path information, and personalized medical management is performed on the target user through the target medical path information; The obtained general medical knowledge base and personalized medical knowledge base include: Acquire a patient disease diagnosis data set according to the medical summary knowledge base; Performing feature clustering analysis on the patient disease diagnosis data set according to medical disease knowledge attributes to obtain multiple diagnosis data clusters; Based on a preset correlation threshold, data association analysis is performed on the plurality of diagnostic data clusters to obtain a plurality of in-cluster association data and a plurality of out-cluster discrete data; Based on multiple in-cluster associated data and multiple out-cluster discrete data, the medical summary knowledge base is divided and integrated to obtain the general medical knowledge base and the personalized medical knowledge base; The obtaining of target medical treatment path information includes: Performing traversal matching analysis in the general medical knowledge base based on the target medical fusion feature to obtain a medical knowledge matching degree set; Setting a preset matching threshold, comparing and optimizing the medical knowledge matching set based on the preset matching threshold, and obtaining universal medical path information with the greatest matching degree; If the medical knowledge matching degree set does not meet the preset matching degree threshold, the personalized medical knowledge base is called to perform matching analysis and comparison on the target medical fusion feature to obtain personalized medical path information; Based on the general medical path information and the personalized medical path information, the target medical path information is determined.

2. The personalized medical path planning method based on fusion data analysis as claimed in claim 1, characterized in that: The available multi-platform medical data streams include: The information exchange center constructs a sensitive data identifier according to medical data application standards; Performing sensitive data identification on the multi-platform medical data stream based on the sensitive data identifier to obtain a medical sensitive data set; Performing a sensitivity level assessment on the medical sensitive data set according to a medical sensitivity level assessment system to obtain sensitive data level information; A data desensitization classification strategy is constructed, and based on the data desensitization classification strategy, the multi-platform medical data stream is integrated and graded and desensitized according to the sensitive data level information to obtain the usable multi-platform medical data stream.

3. The personalized medical path planning method based on fusion data analysis as claimed in claim 1, characterized in that: The obtaining of the target medical fusion feature comprises: Divide and integrate the available multi-platform medical data streams according to the multi-platform data attribute information set to obtain a medical attribute classification data stream; Obtain a data preprocessing program, and perform abnormal data identification and data preprocessing on the medical attribute classification data stream based on the data preprocessing program to obtain a standard medical attribute classification data stream; Determining a data feature extraction algorithm model set according to type characteristic information of the multi-platform data attribute information set; Based on the data feature extraction algorithm model set, feature extraction and data fusion analysis are performed on the standard medical attribute classification data stream to obtain target medical fusion features.

4. The personalized medical path planning method based on fusion data analysis as claimed in claim 3, characterized in that: The obtaining of the target medical fusion feature comprises: Based on the data feature extraction algorithm model set, feature extraction is performed on the standard medical attribute classification data stream respectively to obtain a multi-attribute medical association feature set; Performing correlation analysis and feature dimension reduction on the multi-attribute medical related feature set to obtain a multi-attribute medical key feature set; Determining a data attribute fusion decision factor set according to the multi-platform data attribute information set; The multi-attribute medical key feature set is subjected to data weighted fusion according to the data attribute fusion decision factor set to obtain target medical fusion features.

5. The personalized medical path planning method based on fusion data analysis as claimed in claim 1, characterized in that: The performing personalized medical management on the target user through the target medical path information includes: Based on the target medical path information, personalized medical monitoring is performed on the target user to obtain medical effect feedback data; When the medical treatment effect feedback data does not achieve the expected medical treatment effect, optimizing and analyzing the medical treatment effect feedback data to obtain the optimization direction of the medical treatment path; Adjust and optimize the target medical path information according to the medical path optimization direction to obtain a medical path optimization information set; Each optimization path in the medical optimization path information set is evaluated and screened to obtain target medical optimization path information, and personalized medical management is performed based on the target medical optimization path information.

6. A personalized medical path planning platform integrating data analysis, characterized by: The steps for implementing the personalized medical treatment path planning method of fusion data analysis as described in any one of claims 1 to 5, wherein the personalized medical treatment path planning platform of fusion data analysis comprises: A data acquisition module, which is used to collect and acquire a multi-platform medical data stream of a target user, and transmit the multi-platform medical data stream to an information exchange center for integration preprocessing to obtain a usable multi-platform medical data stream; An attribute classification module, the attribute classification module is used to obtain a medical attribute indicator set, and to classify and identify the available multi-platform medical data stream based on the medical attribute indicator set to obtain a multi-platform data attribute information set; A data fusion module, the data fusion module is used to perform data fusion analysis on the available multi-platform medical data stream according to the multi-platform data attribute information set to obtain a target medical fusion feature; A cluster analysis module, which is used to construct a medical summary knowledge base, perform feature cluster analysis on the medical summary knowledge base, and obtain a general medical knowledge base and a personalized medical knowledge base; A path planning module is used to perform medical path planning analysis on the target medical fusion characteristics based on the general medical knowledge base and the personalized medical knowledge base, obtain target medical path information, and perform personalized medical management for the target user through the target medical path information.

Images