Hemodynamics database construction method, device, equipment and product

Abstract

The invention discloses a construction method, device, equipment and product of a haemodynamics database, and provides a standardized construction method of a multi-modal haemodynamics database, which realizes integration and standardization of data with different structures and formats, such as clinical system data and monitor data. Meanwhile, associated storage of the unstructured data and the structured data is also realized, and a query API interface with uniform data is configured for the database; therefore, direct query can be realized through the uniform interface, so that a remote data query function is realized, and on the basis, tedious data preprocessing processes of automatically installing and downloading a database, automatically integrating and standardizing data and the like in the prior art are omitted; therefore, the data use cost can be reduced while the data analysis efficiency is greatly improved; therefore, the method is very suitable for large-scale application and popularization.

Description

Technical Field

[0001] The present invention belongs to the technical field of data processing, and in particular relates to a method, device, equipment and product for constructing a hemodynamic database. Background Art

[0002] The hemodynamic database is a database that collects, stores, manages and retrieves physiological parameters and medical data related to hemodynamics, including multi-modal related physiological parameters such as patient basic information, blood pressure measurements, cardiac output, heart rate and stroke volume. Analyzing the patient's hemodynamic data based on this database is of great significance to clinical patient applications, medical device development, and physician scientific research and teaching.

[0003] However, the existing hemodynamic data basically comes from clinical system data and monitor data. The data is widely distributed, and the data structure and format are diverse, making it difficult to integrate and connect. This will cause data islands, that is, the existing hemodynamic data exists in different databases, so users need to download and install various databases by themselves. Therefore, using such island data for data analysis requires a lot of time and effort to do data parsing, cleaning and standardization. Based on this, not only the efficiency of data analysis is reduced, but also the cost of data use is increased. Therefore, based on the aforementioned deficiencies, how to provide a standardized hemodynamic database construction method has become a problem that needs to be solved urgently. Summary of the invention

[0004] The purpose of the present invention is to provide a method, device, equipment and product for constructing a hemodynamic database, so as to solve the problem that existing hemodynamic data is difficult to integrate and connect, thus causing data islands.

[0005] In order to achieve the above object, the present invention adopts the following technical solutions:

[0006] In a first aspect, a method for constructing a hemodynamic database is provided, comprising:

[0007] Acquire source database data of hemodynamics, wherein the source database data includes some basic patient data, some monitoring instrument data, some imaging data and some electronic medical record data;

[0008] Based on the medical process sequence, a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data are processed to obtain a number of initial medical data, and a number of initial medical data are processed for time synchronization and alignment to obtain a number of medical data after the time synchronization and alignment processing;

[0009] Divide each medical data into structured data and unstructured data, and perform data standardization on each unstructured data to obtain each standardized unstructured data;

[0010] Performing data association processing on the structured data in each medical data and the corresponding standardized unstructured data to obtain associated structured data and associated unstructured data;

[0011] Constructing a multimodal database based on the integration of lake and warehouse, and storing each associated structured data and each associated unstructured data in the multimodal database to obtain a hemodynamic database;

[0012] Configure the data query API interface for the hemodynamic database to complete the construction of the hemodynamic database after configuration.

[0013] Based on the above disclosed content, the present invention first obtains the source database data of hemodynamics, and then, based on the medical process sequence, performs data combing and processing on the patient basic data, monitoring instrument data, imaging data and electronic medical record data in the source database data, and after the data combing, performs time synchronization and alignment to obtain a number of medical data; in this way, the key medical process combing can be achieved and the time deviation of the collected data can be reduced; after completing the time synchronization of the data, each medical data can be divided into structured data and unstructured data, and each unstructured data (i.e., heterogeneous data) can be standardized to achieve data unification; then, each structured data is associated with the corresponding standardized unstructured data; then, a multimodal database based on the lake warehouse is constructed, and the associated structured data and unstructured data are stored in the database; finally, a data query API interface is configured for the database to complete the construction of the hemodynamic database.

[0014] Through the above design, the present invention provides a method for constructing a standardized multimodal hemodynamic database, which realizes the integration and standardization of data with different structures and formats such as clinical system data and monitor data; at the same time, it also realizes the associated storage of unstructured data and structured data, and configures a unified data query API interface for the database; in this way, it can be directly queried through the unified interface, thereby realizing the remote data query function, based on this, eliminating the need for the existing technology to install and download the database by itself, as well as the cumbersome data preprocessing process such as self-integration and standardization of data; thereby, the data analysis efficiency can be greatly improved while reducing the cost of data use; therefore, the present invention is very suitable for large-scale application and promotion.

[0015] In a possible design, each data in the source database data is stored in a database table, wherein, based on the medical process sequence, a number of patient basic data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data are sorted and processed to obtain a number of initial medical data, including:

[0016] According to the subject of the database table, data segmentation processing is performed on a number of patient basic information, a number of monitoring instrument data, a number of imaging data, and a number of electronic medical record data in the source database data to obtain a number of segmented medical data;

[0017] According to the sequence of medical procedures, each divided medical data is sorted and processed to obtain a number of initial medical data after data processing.

[0018] In a possible design, each initial medical data is subjected to time synchronization alignment processing to obtain a number of aligned initial medical data, including:

[0019] For any initial medical data, determining the calibration time of the initial medical data;

[0020] According to the calibration time, the generation time of each data in any of the initial medical data is subjected to time synchronization alignment processing, so as to obtain the aligned any of the initial medical data after the time synchronization alignment processing, wherein the alignment accuracy of different types of data in any of the initial medical data is different.

[0021] In a possible design, each associated structured data and each associated unstructured data are stored in the multimodal database to obtain an initial hemodynamic database, including:

[0022] Determine the storage method for each structured data and each unstructured data;

[0023] Constructing a data model, and using the data model to perform data processing on each associated structured data and each associated unstructured data, so as to obtain each processed structured data and each processed unstructured data;

[0024] According to the storage method of each structured data and each unstructured data, each processed structured data and each processed unstructured data are stored in the multimodal database, so as to obtain a hemodynamic database after storage.

[0025] In a possible design, each unstructured data is stored in the following manner: in the form of binary files;

[0026] Wherein, according to the storage mode of each structured data and each unstructured data, each processed structured data and each processed unstructured data are stored in the multimodal database, including:

[0027] Directly storing each processed structured data in a data warehouse in the multimodal database, and converting each processed unstructured data into binary data to obtain a binary file corresponding to each processed unstructured data;

[0028] Each binary file is stored in the data lake in the multimodal database, so as to complete the storage of each processed structured data and each processed unstructured data after storage.

[0029] In a possible design, when storing each processed structured data and each processed unstructured data in a multimodal database, ETL technology is used to perform storage optimization.

[0030] In a possible design, the hemodynamic database uses incremental storage and full storage methods to perform data update processing.

[0031] In a second aspect, a device for constructing a hemodynamic database is provided, comprising:

[0032] A data acquisition unit, used for acquiring source database data of hemodynamics, wherein the source database data includes a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data;

[0033] A data combing and correction unit is used to perform data combing processing on a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data based on the medical process sequence to obtain a number of initial medical data, and perform time synchronization and alignment processing on the number of initial medical data to obtain a number of medical data after the time synchronization and alignment processing;

[0034] A data division unit, used for dividing each medical data into structured data and unstructured data, and performing data standardization processing on each unstructured data to obtain each standardized unstructured data;

[0035] A data association unit, used for performing data association processing on the structured data in each medical data and the corresponding standardized unstructured data to obtain associated structured data and associated unstructured data;

[0036] A data storage unit is used to construct a multimodal database based on the lake-warehouse integration, and store each associated structured data and each associated unstructured data in the multimodal database to obtain a hemodynamic database;

[0037] The data storage unit is also used to configure a data query API interface for the hemodynamic database, so as to complete the construction of the hemodynamic database after configuration.

[0038] In the third aspect, another device for constructing a hemodynamic database is provided. Taking the device as an electronic device as an example, the device includes a memory, a processor and a transceiver which are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for constructing a hemodynamic database as in the first aspect or any possible design of the first aspect.

[0039] In a fourth aspect, a storage medium is provided, on which instructions are stored, and when the instructions are executed on a computer, the method for constructing a hemodynamic database as in the first aspect or any possible design in the first aspect is executed.

[0040] In a fifth aspect, a computer program product comprising instructions is provided, which, when executed on a computer, causes the computer to execute the method for constructing a hemodynamic database as described in the first aspect or any possible design of the first aspect.

[0041] Beneficial effects:

[0042] (1) The present invention provides a method for constructing a standardized multimodal hemodynamic database, which realizes the integration and standardization of data of different structures and formats such as clinical system data and monitor data; at the same time, it also realizes the associated storage of unstructured data and structured data, and configures a unified data query API interface for the database; thus, it can be directly queried through the unified interface, thereby realizing the remote data query function, based on which, the tedious data preprocessing process such as the need to install and download the database by oneself, and to perform data integration and standardization by oneself in the prior art is eliminated; thereby, the data analysis efficiency can be greatly improved while reducing the cost of data use; therefore, the present invention is very suitable for large-scale application and promotion.

[0043] (2) The present invention integrates waveform data and image data (i.e., unstructured data) with excessive amounts of data in advance with key data in the source database (i.e., associates them with structured data), and adopts a new binary storage method that is suitable for big data to perform metadata management. In this way, the convenience of associated query and the flexibility of data storage can be improved.

[0044] (3) When integrating data, the present invention organizes the data based on the sequence of medical procedures, thereby improving the accuracy and objectivity of the data.

[0045] (4) The present invention uses ETL technology to optimize storage during storage, thereby reducing the time for data loading and processing. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 A schematic flow chart of the steps of a method for constructing a hemodynamic database provided by an embodiment of the present invention;

[0047] Figure 2 A schematic diagram of the structure of a device for constructing a hemodynamic database provided by an embodiment of the present invention;

[0048] Figure 3 A schematic diagram of the structure of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the present invention will be briefly introduced below in combination with the drawings and the description of the embodiments or the prior art. Obviously, the following description of the structure of the drawings is only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work. It should be noted that the description of these embodiments is used to help understand the present invention, but does not constitute a limitation of the present invention.

[0050] It should be understood that although the terms first, second, etc. may be used herein to describe various units, these units should not be limited by these terms. These terms are only used to distinguish one unit from another unit. For example, a first unit can be referred to as a second unit, and similarly, a second unit can be referred to as a first unit without departing from the scope of the exemplary embodiments of the present invention.

[0051] It should be understood that the term "and / or" that may appear in this article is merely a description of the association relationship between associated objects, indicating that three relationships may exist. For example, A and / or B can represent three situations: A exists alone, B exists alone, and A and B exist at the same time. The term " / and" that may appear in this article describes another type of association object relationship, indicating that two relationships may exist. For example, A / and B can represent two situations: A exists alone, and A and B exist alone. In addition, the character " / " that may appear in this article generally indicates that the previous and next associated objects are in an "or" relationship.

[0052] Example:

[0053] See also Figure 1As shown, the method for constructing a hemodynamic database provided in this embodiment realizes the integration and standardization of data of different structures and formats such as clinical system data and monitor data; at the same time, it also realizes the associated storage of unstructured data and structured data, and configures a query API interface for the database; in this way, it can be directly queried through the unified interface, thereby realizing the remote data query function, based on this, it eliminates the need for the existing technology to install and download the database by itself, as well as the cumbersome data preprocessing process such as self-integration and standardization of data; thereby, it can greatly improve the efficiency of data analysis while reducing the cost of data use; therefore, this method is very suitable for large-scale application and promotion; among them, for example, this method can be but not limited to running on the database construction end side, optionally, the database construction end can be but not limited to a server, it can be understood that the aforementioned execution subject does not constitute a limitation on the embodiments of the present application, and accordingly, the operation steps of this method can be but not limited to the following steps S1 to S6.

[0054] S1. Obtaining source database data of hemodynamics, wherein the source database data includes a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data; in specific implementation, the source database data may be but is not limited to being collected from the business systems of multiple hospitals (such as HIS, LIS and EMR systems) and different monitoring instruments, wherein the basic patient data includes information such as the patient's name, age, gender, etc., and the monitoring instrument data may include but is not limited to the patient's psychology, blood pressure measurement value, cardiac index, cardiac output, stroke volume, etc.; similarly, the imaging data may include but is not limited to data such as electrocardiogram and ultrasonic image; of course, the types of the aforementioned data may be specifically set according to actual use, and are not limited to the aforementioned examples; at the same time, the storage method of each data in the aforementioned source database data is taken as a database table, that is, in the business systems of different hospitals, the aforementioned basic patient data, monitoring instrument data, imaging data and electronic medical record data are stored in the form of database tables; such as a basic patient information table, a monitoring instrument data table, and the like.

[0055] Specifically, a database table is an important component of a database and a key structure for data storage. It is used to store, organize, and manage large amounts of data, and supports operations such as adding, deleting, modifying, and querying data. A database table is the basic unit in a relational database and is operated using the SQL language.

[0056] In this way, after collecting the aforementioned source database data, the data can be integrated so that a standardized hemodynamic database can be constructed based on it later; of course, the construction of the hemodynamic database is based on ETL technology, among which ETL (Extract, Transform, Load) is a key concept in data warehouse and big data analysis, representing the process of extraction (Extract), transformation (Transform) and loading (Load); in this way, the construction of the database is to obtain data from the business data of different hospitals, and then extract, integrate and transform the data, and finally load it into the multimodal database.

[0057] The data integration process may be, but is not limited to, steps S2 to S4 as shown below.

[0058] S2. Based on the sequence of medical processes, a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data are processed to obtain a number of initial medical data, and a time synchronization alignment process is performed on the number of initial medical data to obtain a number of medical data after the time synchronization alignment process; in specific applications, data combing is to comb the key medical processes, that is, to clarify the data flow according to the sequence of the medical processes. At the same time, the key data elements required for each link of the medical process, as well as the logical relationship and dependency relationship between the key data elements can also be identified. In this way, the data can be aligned with the medical process, thereby ensuring the accuracy and objectivity of the data to the greatest extent.

[0059] At the same time, as mentioned above, the source database data is collected from the business systems of different hospitals and different monitoring instruments, and the data sources are diverse. Therefore, in actual use, it is difficult to ensure the complete use of a unified time source and regular time calibration. In this way, the timestamps of data from different sources may have problems such as accuracy differences and time drift, which will lead to deviations or errors in the analysis results. Based on this, it is necessary to synchronize and align the data in time to eliminate the problem of difficulty in matching medical process data caused by time deviation as much as possible.

[0060] Optionally, one of the implementation methods of the time synchronization alignment and data combing of the following public data may be, but is not limited to, the following steps S21 to S24.

[0061] S21. According to the subject of the database table, data division processing is performed on some basic information of patients, some monitoring instrument data, some imaging data and some electronic medical record data in the source database data to obtain some divided medical data; in this embodiment, it has been explained above that each data in the source database data is stored in the form of a database table, then, data division can be performed according to the subject of the database table (i.e., the data subject); if the data subject is a case, then it is equivalent to dividing the basic information of patients, monitoring instrument data, imaging data and electronic medical record data of the same case into one piece of data, so as to realize data statistics with the case as the unit; for another example, if the data subject is a monitoring instrument, then the data division is performed with the monitoring instrument as the unit, that is, the data generated by one instrument is regarded as one piece of divided medical data; for another example, if the electronic medical record data contains an admission table, then an admission record in the table is regarded as one piece of data; of course, when the subject of the database table is different, the principle of data division is the same as the above example, and it will not be repeated here.

[0062] Furthermore, the present embodiment preferably uses cases as units for data division; of course, specific settings can be made according to actual use and are not limited to the aforementioned examples.

[0063] In this way, after completing the division of the aforementioned source database data, medical process alignment can be performed, and the process can be but is not limited to as shown in the following step S22.

[0064] S22. According to the sequence of medical procedures, the divided medical data are sorted and processed to obtain a number of initial medical data after data processing; in this embodiment, the sequence of medical procedures is the sequence in which data is generated during the medical process, such as first registering to generate registration data, then having a medical consultation to generate medical consultation data, then having an examination to generate examination data, and finally, taking medicine to generate medication data, etc.; based on this, the divided medical data can be sorted based on the medical procedure sequence, taking the patient as the unit, so as to clarify the data flow.

[0065] At the same time, each initial medical data can be sent to the sorting terminal so that the personnel corresponding to the sorting terminal can identify the key data of each initial medical data, as well as the logical relationship and dependency relationship between the key data, so as to upload the final sorted initial medical data to the sorting terminal; and the sorting terminal can send the final sorted initial medical data to the database construction end so that the database construction end can perform subsequent time synchronization alignment processing.

[0066] After the key medical processes are sorted out, time synchronization and alignment processing can be performed. In this embodiment, any initial medical data is used as an example to illustrate the process, and the process can be but is not limited to the following steps S22 and S23.

[0067] S23. For any initial medical data, determine the calibration time of the initial medical data.

[0068] S24. According to the calibration time, the generation time of each data in any of the initial medical data is subjected to time synchronization alignment processing, so as to obtain the aligned any of the initial medical data after the time synchronization alignment processing, wherein the alignment accuracy of different types of data in any of the initial medical data is different.

[0069] In this embodiment, the example can be but not limited to uniformly determining the calibration time according to data from different sources, where the data source refers to the business systems of different hospitals; in this way, the collected data can be time-synchronized and aligned (i.e., time correction) based on the determined calibration time; optionally, the alignment accuracy of different types of data in any initial medical data described in the example is different; based on this, time correction of different accuracies can be achieved for different types of data.

[0070] Furthermore, the following discloses a time synchronization alignment rule: (1) Patient information dimension (i.e., basic patient information): time alignment is performed with an accuracy of hours or minutes; (2) Patient time series data (such as monitoring instrument and imaging data, i.e., data that can reflect the outcome changes of the patient) is aligned with an accuracy of seconds; (3) Ultrasound images (i.e., data that reflect the organic changes of the patient) are aligned with an accuracy of milliseconds; at the same time, for data collected by monitoring instruments and other equipment, data correction can also be performed based on the instrument and measurement principle, such as using thermal dilution method, ultrasonic delay and other methods to correct the data.

[0071] In this way, through the aforementioned steps S23 and S24, the time synchronization alignment of each initial medical data can be completed, thereby minimizing the time deviation to the greatest extent.

[0072] Therefore, through the aforementioned steps S21 to S24, the data combing and time synchronization alignment of the source database data can be completed, thereby obtaining a number of medical data; then, the medical data can be stored, and the process can be but not limited to as shown in the following steps S3 to S5.

[0073] S3. Divide each medical data into structured data and unstructured data, and perform data standardization on each unstructured data to obtain each standardized unstructured data; in the present embodiment, this step is the process of combing the structured data and unstructured data in each medical data, wherein the structured data has a clear data structure and a standardized format, that is, each field has a predefined data type, such as an integer, a character string, a date, etc., therefore; structured data is easy to query and analyze; on the other hand, unstructured data, because it has no fixed format and clear structure of data, such as text, images, audio, video, etc.; therefore, unstructured data has various forms, and processing and analysis are relatively complex; based on this, the present embodiment needs to adopt different processing methods through the identification of structured and unstructured data.

[0074] Optionally, for example, each piece of medical data may be parsed to obtain the types and formats of different data in each piece of medical data, thereby achieving the division of structured data and unstructured data in each piece of medical data based on this.

[0075] After completing the division of structured data and unstructured data, the unstructured data can be standardized. That is, since the unstructured data in each piece of medical data is in various forms (such as data from different instruments, different systems, different data sources, different monitoring sites, and different sensing methods), it is necessary to standardize the heterogeneous data (i.e., unstructured data) in the medical data to achieve data unification. The standardization process of unstructured data is as follows: (1) determine the data type (i.e., first determine the data type, whether it is text, image, or video, etc.); (2) perform data cleaning (such as filling missing values, denoising, removing outliers, etc.); (3) data conversion, that is, converting unstructured data into structured data. Of course, the standardization of unstructured data is a common technology for data processing, and its principles will not be repeated one by one.

[0076] In this way, based on the aforementioned step S3, after completing the division of structured data and unstructured data in each piece of medical data and the standardization of unstructured data, data association can be performed to ensure the availability of the data; wherein, the data association process can be but is not limited to as shown in the following step S4.

[0077] S4. Perform data association processing on the structured data in each medical data and the corresponding standardized unstructured data to obtain associated structured data and associated unstructured data; in this embodiment, for any medical data, it is equivalent to associating its corresponding unstructured data with the key content (that is, structured data) in the medical data, so as to ensure that the complete data of any medical data can be queried during the query, thereby ensuring the availability of the data.

[0078] Furthermore, relevance is a key factor in improving database usability and query efficiency. Implementation methods include adding field redundancy to the data model (structured data) and adding key data fields to unstructured data. Of course, the specific association method can be set according to actual use and is not limited to the above examples.

[0079] In this way, after completing the data association, data storage can be performed, that is, data loading in the ETL process can be performed, which involves writing the processed data into the target database or data warehouse; wherein the data loading process can be but is not limited to as shown in the following step S5.

[0080] S5. Construct a multimodal database based on the integration of lake and warehouse, and store each associated structured data and each associated unstructured data in the multimodal database to obtain a hemodynamic database; in specific applications, this embodiment designs a data (warehouse) library technology architecture in combination with hemodynamic medical application requirements and system resources, and based on this, constructs a data (warehouse) library model, that is, constructs a multimodal database based on the integration of lake and warehouse (that is, including a data warehouse and a data lake); wherein, this embodiment can adopt different storage methods based on the data type, store different types of data, and perform storage optimization.

[0081] Optionally, the specific storage process may be but is not limited to the following steps S51 to S53.

[0082] S51. Determine the storage method of each structured data and each unstructured data; in this embodiment, the data storage types are structured data and unstructured data, and the distinction between the two is achieved by parsing medical data; at the same time, for example, the storage method of unstructured data is binary files, etc., while structured data is directly stored based on data warehouses; at the same time, this embodiment also constructs a data model to perform data processing of the aforementioned associated structured data and unstructured data, wherein the data processing process is shown in the following step S52.

[0083] S52. Construct a data model, and use the data model to process each associated structured data and each associated unstructured data to obtain each processed structured data and each processed unstructured data; in specific implementation, for example, but not limited to, the Kimba 11 model can be used to perform data processing to define data entities, attributes, relationships and data flows; wherein, the Kimba 11 model splits the data into different table requirements according to the target, and the data will be extracted into a fact-dimensional model, and the data source is converted into a fact table and a dimension table through ETL (Extract, Transform, Load) and imported into the data mart (i.e., the subsequent multimodal database); wherein, after the aforementioned data processing, the unstructured data will obtain the fact table and dimension table associated with it, and the structured data will be directly converted into the corresponding fact table and dimension table; based on this, after completing the processing of each associated structured data and unstructured data, the data can be stored, and the process can be but not limited to the following step S53.

[0084] S53. According to the storage method of each structured data and each unstructured data, each processed structured data and each processed unstructured data are stored in the multimodal database, so as to obtain a hemodynamic database after storage; in this embodiment, for example, each processed structured data is directly stored in the data warehouse in the multimodal database, and each processed unstructured data is converted into binary data to obtain a binary file corresponding to each processed unstructured data; then, each binary file is stored in the data lake in the multimodal database, so as to complete the storage of each processed structured data and each processed unstructured data after storage; in this way, unstructured data with excessive data volume (such as waveform data and image data) will be pre-integrated with the key data of the source database, and stored in a new binary storage method that adapts to big data, which can improve the convenience of associated query and the flexibility of data storage.

[0085] At the same time, when storing data, ETL technology can also be used for storage optimization, that is, ETL storage optimization, which can be done in the following ways, but is not limited to: (1) using data compression technology to reduce storage space requirements and reduce the time it takes for data to be transmitted over the network; (2) partitioning data according to access patterns (such as time, region, business type, etc.), which can speed up queries and simplify data management; (3) creating appropriate indexes to speed up queries, but avoid over-indexing, because indexes also take up additional storage space and may reduce performance when loading data; (4) for analytical queries, columnar storage can significantly improve query performance because they can skip irrelevant data and only read necessary columns; (5) performing data cleansing and normalization during the ETL process to eliminate duplicate and unnecessary data and reduce storage requirements; (6) extracting, transforming, and loading only data that has changed since the last ETL process, rather than processing the entire data set each time; (7) using in-memory databases or cache technology to handle the transformation process. process, which can greatly improve the processing speed; (8) Storage tiering is performed according to the access frequency and importance of data, that is, different databases are stored in different data media (such as cloud big data clusters or local data servers, etc.) according to data access frequency, data storage type and storage cost, so as to complete storage tiering, so as to improve access to data, balance performance and cost-effectiveness, and optimize the utilization of storage resources. In other words, infrequently used data is migrated to a slower but larger and cheaper storage layer, thereby improving storage capacity utilization and meeting different application requirements. For example, cache is suitable for storing data with high access frequency, while auxiliary storage is suitable for long-term storage of large amounts of data: (9) Develop a data retention strategy to migrate infrequently accessed data to lower-cost storage media, or regularly clean up outdated data: (10) Select an appropriate data structure to store data, such as using bitmap indexes, B-trees, etc.; (10) Upgrade the server's hardware resources according to actual needs, such as increasing CPU processing power, expanding memory capacity, using high-speed SSD hard drives, etc. At the same time, optimize network configuration, such as using high-speed network equipment and network acceleration technology to improve data transmission speed and reliability; (11) Use efficient storage formats in data warehouses, such as Parquet, ORC, etc. These formats support columnar storage and compression, which can significantly improve query speed and reduce storage costs; (12) Regularly monitor the performance indicators of the data warehouse (such as query response time, storage space utilization, etc.), and adjust the storage optimization strategy based on the monitoring results. At the same time, pay attention to the development of new technologies and new methods, and introduce and apply them to data warehouse storage optimization in a timely manner.

[0086] Furthermore, before loading the data into the multimodal database, data processing is required, such as data cleaning and data quality assurance; this includes removing duplicate data, repairing missing values, and processing outliers, so as to ensure the accuracy of the data; at the same time, after the data is loaded, deduplication processing is required; and a data backup and recovery mechanism can be established to regularly back up the lake warehouse and ensure that the data can be quickly restored in the event of a failure or data loss; in addition, after the multimodal database is put into use, it can be monitored and its performance continuously optimized; this includes monitoring the query performance, data quality and health of the multimodal database; based on this, monitoring can promptly discover and solve problems, thereby improving the performance and reliability of the multimodal database.

[0087] In addition, ETL optimization can be carried out throughout the entire database construction process, that is, data extraction optimization is performed during data extraction; specifically, it includes: incremental extraction: only extracting data that has changed since the last extraction, rather than extracting the entire amount; parallel extraction: if the source system supports it, data can be extracted from multiple sources in parallel; direct source connection: using database log mining or Change Data Capture (CDC) technology to capture changes directly from the source system; batch processing: merging small batches of data into large batches for processing to reduce the number of extractions.

[0088] Similarly, when converting data, optimization is performed; specifically, it includes: reducing conversion steps: simplifying the conversion logic in the ETL process and removing unnecessary steps; using faster algorithms: optimizing or selecting more efficient algorithms for data processing; data preprocessing: preprocessing the data before it enters the ETL process, such as data cleaning and formatting; in-memory computing: using in-memory databases or distributed computing frameworks (such as Spark) for conversion to increase processing speed; index optimization: creating indexes for commonly used fields during the conversion process to speed up queries.

[0089] Data loading optimization has been explained above and will not be elaborated on here. Furthermore, system and architecture optimization can be performed. Specifically, resource allocation: reasonably allocate resources of the database and ETL server to avoid resource competition; cache mechanism: use cache to store frequently accessed data to reduce the number of database accesses.

[0090] At the same time, ETL process management optimization can also be carried out; specifically including: monitoring and logging: implement real-time monitoring and detailed logging to quickly identify and solve problems; error handling: design a robust error handling mechanism to ensure that the ETL process can automatically recover when an error occurs; scheduling optimization: reasonably arrange the scheduling of ETL jobs to avoid resource contention during peak hours.

[0091] In this way, after ETL optimization, the efficiency of data loading and query can be improved.

[0092] Then, a query interface may be configured for the aforementioned database to implement data query through a unified query interface; wherein the query interface configuration process is shown in the following step S6.

[0093] S6. Configure a data query API interface for the hemodynamic database to complete the construction of the hemodynamic database after configuration; in this embodiment, by configuring the data query API interface, a rich data query interface can be provided to the outside, thereby facilitating the query, download and application of medical data; at the same time, in specific applications, taking the hemodynamic database as an example, incremental storage and full storage methods are used for data update processing; in this way, the space occupied by the database can be reduced.

[0094] Optionally, the delta is stored as:

[0095] Incremental loading only loads the newly added or changed data into the target system, rather than the entire data set. This approach can significantly improve the efficiency and performance of data processing; when processing incremental data, a variety of methods can be used to determine the increment, such as using timestamps, version numbers, triggers, and comparison tables; of course, different incremental determination methods can be selected according to actual use, and are not limited to the above examples.

[0096] Similarly, the full storage is:

[0097] Full loading means reloading or copying the entire data set to the target system each time it is processed. Although this method is simple and direct, it may increase processing time and resource consumption in a big data environment. It is generally used in scenarios where it is difficult to determine the increment (such as frequent data changes, no timestamps, etc.), or the amount of data is too small, and it is not worth or appropriate to use incremental storage.

[0098] Therefore, through the method for constructing a hemodynamic database described in detail in the aforementioned steps S1 to S6, the present invention realizes the integration and standardization of data of different structures and formats such as clinical system data and monitor data; at the same time, it also realizes the associated storage of unstructured data and structured data, and configures a query API interface for the database; in this way, it can be directly queried through the unified interface, thereby realizing the remote data query function, based on this, eliminating the need for the existing technology to install and download the database by itself, as well as the cumbersome data preprocessing process such as self-integration and standardization of data; thereby, the data analysis efficiency can be greatly improved while reducing the cost of data use; therefore, the present invention is very suitable for large-scale application and promotion.

[0099] like Figure 2As shown, the second aspect of this embodiment provides a hardware device for implementing the method for constructing a hemodynamic database described in the first aspect of the embodiment, including:

[0100] The data acquisition unit is used to acquire the source database data of hemodynamics, wherein the source database data includes some basic patient data, some monitoring instrument data, some imaging data and some electronic medical record data.

[0101] The data combing and correction unit is used to perform data combing processing on a number of patient basic data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data based on the medical process sequence to obtain a number of initial medical data, and perform time synchronization and alignment processing on the number of initial medical data to obtain a number of medical data after the time synchronization and alignment processing.

[0102] The data division unit is used to divide various medical data into structured data and unstructured data, and perform data standardization processing on various unstructured data to obtain various standardized unstructured data.

[0103] The data association unit is used to perform data association processing on the structured data in each medical data and the corresponding standardized unstructured data to obtain the associated structured data and the associated unstructured data.

[0104] The data storage unit is used to construct a multimodal database based on the lake and warehouse, and store each associated structured data and each associated unstructured data in the multimodal database to obtain a hemodynamic database.

[0105] The data storage unit is also used to configure a data query API interface for the hemodynamic database, so as to complete the construction of the hemodynamic database after configuration.

[0106] The working process, working details and technical effects of the device provided in this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0107] like Figure 3 As shown, the third aspect of this embodiment provides another device for constructing a hemodynamic database. Taking the device as an electronic device as an example, it includes: a memory, a processor and a transceiver that are communicatively connected in sequence, wherein the memory is used to store computer programs, the transceiver is used to send and receive messages, and the processor is used to read the computer program to execute the method for constructing a hemodynamic database as described in the first aspect of the embodiment.

[0108] For example, the memory may include, but is not limited to, random access memory (RAM), read-only memory (ROM), flash memory, first-in-first-out memory (FIFO) and / or first-in-last-out memory (FILO), etc. Specifically, the processor may include one or more processing cores, such as a 4-core processor, an 8-core processor, etc. The processor may be implemented in at least one hardware form of DSP (Digital Signal Processing), FPGA (Field-Programmable Gate Array), and PLA (Programmable Logic Array). At the same time, the processor may also include a main processor and a coprocessor. The main processor is a processor for processing data in the awake state, also known as a CPU (Central Processing Unit); the coprocessor is a low-power processor for processing data in the standby state.

[0109] In some embodiments, the processor may be integrated with a GPU (Graphics Processing Unit), which is responsible for rendering and drawing the content to be displayed on the display screen. For example, the processor may not be limited to a microprocessor of the STM32F105 series, a reduced instruction set computer (RISC) microprocessor, an X86 or other architecture processor, or a processor with an integrated embedded neural network processor (NPU); the transceiver may be, but is not limited to, a wireless fidelity (WIFI) wireless transceiver, a Bluetooth wireless transceiver, a general packet radio service technology (Genera l Packet Rad ioService, GPRS) wireless transceiver, a ZigBee protocol (a low-power LAN protocol based on the IEEE 802.15.4 standard, ZigBee) wireless transceiver, a 3G transceiver, a 4G transceiver and / or a 5G transceiver, etc. In addition, the device may also include, but is not limited to, a power module, a display screen and other necessary components.

[0110] The working process, working details and technical effects of the electronic device provided in this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0111] The fourth aspect of this embodiment provides a storage medium storing instructions for the method for constructing a hemodynamic database as described in the first aspect of the embodiment, that is, the storage medium stores instructions, and when the instructions are run on a computer, the method for constructing a hemodynamic database as described in the first aspect of the embodiment is executed.

[0112] The storage medium refers to a carrier for storing data, which may include but is not limited to a floppy disk, a CD, a hard disk, a flash memory, a USB flash drive and / or a memory stick, etc. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0113] The working process, working details and technical effects of the storage medium provided in this embodiment can be found in the first aspect of the embodiment and will not be described in detail here.

[0114] The fifth aspect of this embodiment provides a computer program product comprising instructions, which, when executed on a computer, causes the computer to execute the method for constructing a hemodynamic database as described in the first aspect of the embodiment, wherein the computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices.

[0115] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for constructing a hemodynamic database, characterized in that: include: Acquire source database data of hemodynamics, wherein the source database data includes some basic patient data, some monitoring instrument data, some imaging data and some electronic medical record data; Based on the medical process sequence, a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data are processed to obtain a number of initial medical data, and a number of initial medical data are processed for time synchronization and alignment to obtain a number of medical data after the time synchronization and alignment processing; Divide each medical data into structured data and unstructured data, and perform data standardization on each unstructured data to obtain each standardized unstructured data; Performing data association processing on the structured data in each medical data and the corresponding standardized unstructured data to obtain associated structured data and associated unstructured data; Constructing a multimodal database based on the integration of lake and warehouse, and storing each associated structured data and each associated unstructured data in the multimodal database to obtain a hemodynamic database; Configure the data query API interface for the hemodynamic database to complete the construction of the hemodynamic database after configuration.

2. The method according to claim 1, characterized in that The storage method of each data in the source database data is a database table, wherein, based on the medical process sequence, a number of patient basic data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data are sorted and processed to obtain a number of initial medical data, including: According to the subject of the database table, data segmentation processing is performed on a number of patient basic information, a number of monitoring instrument data, a number of imaging data, and a number of electronic medical record data in the source database data to obtain a number of segmented medical data; According to the sequence of medical procedures, each divided medical data is sorted and processed to obtain a number of initial medical data after data processing.

3. The method according to claim 1, characterized in that Performing time synchronization alignment processing on a plurality of initial medical data, so as to obtain a plurality of medical data after the time synchronization alignment processing, including: For any initial medical data, determining the calibration time of the initial medical data; According to the calibration time, the generation time of each data in any of the initial medical data is subjected to time synchronization alignment processing, so as to obtain the aligned any of the initial medical data after the time synchronization alignment processing, wherein the alignment accuracy of different types of data in any of the initial medical data is different.

4. The method according to claim 1, characterized in that: Storing each associated structured data and each associated unstructured data in the multimodal database to obtain an initial hemodynamic database, including: Determine the storage method for each structured data and each unstructured data; Constructing a data model, and using the data model to perform data processing on each associated structured data and each associated unstructured data, so as to obtain each processed structured data and each processed unstructured data; According to the storage method of each structured data and each unstructured data, each processed structured data and each processed unstructured data are stored in the multimodal database, so as to obtain a hemodynamic database after storage.

5. The method according to claim 4, characterized in that The storage methods of each unstructured data are: stored in the form of binary files; Wherein, according to the storage mode of each structured data and each unstructured data, each processed structured data and each processed unstructured data are stored in the multimodal database, including: Directly storing each processed structured data in a data warehouse in the multimodal database, and converting each processed unstructured data into binary data to obtain a binary file corresponding to each processed unstructured data; Each binary file is stored in the data lake in the multimodal database, so as to complete the storage of each processed structured data and each processed unstructured data after storage.

6. The method according to claim 5, characterized in that When storing each processed structured data and each processed unstructured data in a multimodal database, ETL technology is used to perform storage optimization.

7. The method according to claim 1, characterized in that The hemodynamic database adopts incremental storage and full storage methods to perform data update processing.

8. A device for constructing a hemodynamic database, characterized in that: include: A data acquisition unit, used for acquiring source database data of hemodynamics, wherein the source database data includes a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data; A data combing and correction unit is used to perform data combing processing on a number of basic patient data, a number of monitoring instrument data, a number of imaging data and a number of electronic medical record data in the source database data based on the medical process sequence to obtain a number of initial medical data, and perform time synchronization and alignment processing on the number of initial medical data to obtain a number of medical data after the time synchronization and alignment processing; A data division unit, used for dividing each medical data into structured data and unstructured data, and performing data standardization processing on each unstructured data to obtain each standardized unstructured data; A data association unit, used for performing data association processing on the structured data in each medical data and the corresponding standardized unstructured data to obtain associated structured data and associated unstructured data; A data storage unit is used to construct a multimodal database based on the lake-warehouse integration, and store each associated structured data and each associated unstructured data in the multimodal database to obtain a hemodynamic database; The data storage unit is also used to configure a data query API interface for the hemodynamic database, so as to complete the construction of the hemodynamic database after configuration.

9. An electronic device, characterized in that: include: A memory, a processor and a transceiver which are sequentially connected in communication, wherein the memory is used to store a computer program, the transceiver is used to send and receive messages, and the processor is used to read the computer program, and execute the method for constructing a hemodynamic database as described in any one of claims 1 to 7.

10. A computer program product comprising instructions, characterized in that When the instructions are executed on a computer, the computer is caused to execute the method for constructing a hemodynamic database according to any one of claims 1 to 7.

Images