CN-121983329-A - Data screening and matching system and method based on lithiasis specific disease data model

CN121983329ACN 121983329 ACN121983329 ACN 121983329ACN-121983329-A

Abstract

The invention relates to a data screening and matching system and a method based on a lithiasis specific disease data model, which belong to the technical field of data screening and comprise a data acquisition module, a data preprocessing module, a data screening module, a report display module and a report output module; the system comprises a data acquisition module for realizing standardized multi-source data acquisition, a data preprocessing module for realizing standardized multi-source data normalization and structured extraction, a data screening and matching module for screening and matching patient data of patient data passing through the data preprocessing module, a report display module comprising a front end interface and a rear end interface, and a report output module for supporting one-key export report and being used as data input of a scientific research platform. The invention has the beneficial effects of 1, accurate matching, 2, high-efficiency response, 3, flexible expansion, 4 and clinical value.

Inventors

JI BING
ZHANG QIBO
ZHU WUHUI
LIU ZONGTAO
WANG FENGYAO

Assignees

青岛市第三人民医院

Dates

Publication Date: 20260505
Application Date: 20251211

Claims (10)

1. The data screening and matching system based on the lithiasis specific disease data model is characterized by comprising a data acquisition module, a data preprocessing module, a data screening module, a report display module and a report output module; the data acquisition module is used for realizing standardized multi-source data acquisition; Collecting patient data from the HIS system, pushing the patient data to a message queue, wherein the patient data comprises structured data and unstructured data; The data preprocessing module is used for realizing normalization and structured extraction of standardized multi-source data; Patient data is obtained from a message queue, and is stored into a lithiasis specific disease data model library after data cleaning, normalization and structural extraction and field mapping are carried out, wherein the data cleaning comprises removing duplicates, deletions or abnormal values, the normalization and structural extraction is carried out by unifying the patient data into a specific format, unstructured text in the patient data is analyzed by utilizing a Natural Language Processing (NLP) technology, image characteristics are associated, and the field mapping is carried out by unifying heterogeneous data into lithiasis specific disease data model fields; The data screening and matching module calls the intelligent screening and matching engine to screen and match the patient data passing through the data preprocessing module according to the diagnosis requirement; the report display module comprises a front-end interface and a back-end interface, wherein the front-end interface is used for realizing an interactive chart based on a front-end visual library, and the back-end interface is used for dynamically acquiring associated patient data; And the report output module automatically generates a patient queue according to the screening matching result, supports the duplication elimination and priority ordering of the patient queue, supports one-key export of the report, and outputs the report in a standardized format which can be used as data input of a scientific research platform.
2. The data screening matching system based on the lithiasis specific disease data model according to claim 1, wherein the structured data comprises pathology records, medical history information, examination records, diagnosis records, demographic information, surgical treatment, medical advice records, diagnosis and treatment overview, and the unstructured data comprises CT image original files, surgical record text and doctor diagnosis notes.
3. The data screening and matching system based on the lithiasis specific disease data model according to claim 2, wherein the data acquisition module adopts a distributed message queue to realize high concurrence data reception.
4. The data screening matching system based on a lithiasis specific disease data model according to claim 3, wherein said lithiasis specific disease data model library is a postgresQL-based extended spatiotemporal database, said outliers include a lithiasis size exceeding a physiological range, said unstructured text includes a calcium oxalate type of a lithiasis examination result in a lithiasis examination, said image features include a location, a density of an examination lithiasis in a urinary system Dan Yingxiang, said lithiasis specific disease data model fields include an examination view, an examination conclusion, whether there is a lithiasis, a lithiasis site, a lithiasis size, a lithiasis number, an echo feature, an examination date, an examination specimen name, an examination result raw value, an examination result.
5. The data screening and matching system based on the lithiasis specific disease data model of claim 4, wherein the data preprocessing module realizes automatic field extraction based on a rule engine and a pre-training NLP model and supports custom cleaning rule configuration.
6. The data screening and matching system based on the lithiasis specific disease data model according to claim 5, wherein said intelligent screening and matching engine adopts a condition tree algorithm which defines query logic through parent-child node levels.
7. The data screening and matching system based on the lithiasis specific disease data model of claim 6, wherein the data screening and matching module adopts a memory database to cache high-frequency query rules and combines a distributed computing framework to realize parallel screening and matching.
8. The data screening and matching system based on the lithiasis specific disease data model of claim 7, wherein the front-end interface provides a time-axis-based presentation data function for integrating patient visit records, treatment progress and recurrence trend curves in a time dimension and a diagrammatical presentation data function for supporting thermodynamic diagram presentation of lithiasis component distribution, wherein the diagrammatical presentation data function analyzes metabolic index changes, and wherein the back-end dynamically acquires associated patient data through the GraphQL interface.
9. The data screening matching system based on a lithiasis specific disease data model of claim 8, wherein said prioritization includes ranking by recurrence risk score, and said standardized format includes CSV, FHIR Bundle.
10. A data screening and matching method based on a lithiasis specific disease data model is characterized by comprising the following steps: S1, a data acquisition module acquires original patient data from an HIS system through an API and pushes the original patient data to a message queue; s2, the data preprocessing module acquires original patient data from the message queue, cleans the data and stores the cleaned data into a lithiasis specific disease data model library (a space-time database based on PostgreSQL expansion); S3, the intelligent screening and matching engine calls data and matching rules in a lithiasis specific disease data model library, executes multi-condition query, and writes the result into a cache; s4, the report display module reads the screening matching result from the cache, renders the chart in real time and feeds back the chart to the front-end interface; And S5, generating a queue file by the report output module according to the final matching result, and transmitting the queue file to a specified scientific research platform through an SFTP protocol.

Description

Data screening and matching system and method based on lithiasis specific disease data model Technical Field The invention relates to the technical field of data screening, in particular to a data screening and matching system and method based on a lithiasis specific disease data model. Background Urinary system stones, also known as urolithiasis, are one of the most common diseases in urology, and are the leading ones among hospitalized patients. The prevalence rate of urinary tract stones in different areas in China is different from 1.5% to 18%, the overall southern area is higher than the northern area, the annual new incidence rate is (150-200)/10 ten thousand, and 25% of patients need hospitalization. In recent years, the incidence rate of urinary tract stones worldwide has an increasing trend, and the recurrence rate of 5-10 years can reach more than 50%. The current hospital data depends on a general database system, data required by clinical research are not subjected to specialized splitting and integration, and a special disease data model aiming at lithiasis is lacked, so that the data screening efficiency is low, the matching precision is insufficient, multiple dimension data of a patient cannot be dynamically associated (such as the patient is diagnosed and inquired according to the time dimension, the lithiasis component, the treatment scheme, the recurrence factor and the like cannot be accurately found), and the data cannot be counted and analyzed. In the prior art, keyword matching or a simple classification algorithm is mostly adopted, and accurate research queue construction and data export analysis are difficult to support. The following problems are presented in detail: 1. The existing general system has the defects that a large amount of unstructured data need to be integrated manually aiming at the lithiasis data, the time consumption is long, and the lithiasis data set is not standardized, mapped, processed and calculated or structured, so that the data with aliases or the unstructured data cannot be hit accurately during retrieval or statistics. 2. The data query result is not matched, the existing data query only has single patient stock data, the patient data can be displayed only in a data table form, the patient data is required to be screened and analyzed in time sequence and medical record dimension under the clinical research scene, and if the patient data cannot be stored in a multi-dimension mode, the query data result is not matched with the expectation. And relational databases employ tabular forms to store data and thus become slower when processing large amounts of data. 3. The query expansibility is insufficient, namely the current data query mode only supports full-text or advanced query, and the flexibility and expansibility of supporting multi-condition combined query (such as 'calculus component + recurrence times + recurrence later than a certain date + not pregnant woman') cannot be supported according to medical record storage. 4. The information integration degree is low, the current database lacks a patient integrated treatment view and a patient panoramic view (such as a time axis for showing treatment progress), and doctors need to manually correlate data from original records in a cross-platform manner, so that the operation is inconvenient and time-consuming. Disclosure of Invention The invention aims to provide a data screening and matching system and method based on a lithiasis specific disease data model, so as to solve the problems in the background technology. On one hand, the invention provides a data screening and matching system based on a lithiasis specific disease data model, which comprises a data acquisition module, a data preprocessing module, a data screening module, a report display module and a report output module; the data acquisition module is used for realizing standardized multi-source data acquisition; Collecting patient data from the HIS system, pushing the patient data to a message queue, wherein the patient data comprises structured data and unstructured data; The data preprocessing module is used for realizing normalization and structured extraction of standardized multi-source data; Patient data is obtained from a message queue, and is stored into a lithiasis specific disease data model library after data cleaning, normalization and structural extraction and field mapping are carried out, wherein the data cleaning comprises removing duplicates, deletions or abnormal values, the normalization and structural extraction is carried out by unifying the patient data into a specific format, unstructured text in the patient data is analyzed by utilizing a Natural Language Processing (NLP) technology, image characteristics are associated, and the field mapping is carried out by unifying heterogeneous data into lithiasis specific disease data model fields; The data screening and matching module calls the intelligent screening and matching