Prediction system for early risk of sepsis

Abstract

The invention provides a sepsis early-stage risk prediction system, and relates to the technical field of sepsis risk prediction, the sepsis early-stage risk prediction system comprises a data acquisition module, a data preprocessing module, a risk assessment module, a result display module and a data management module, and the data acquisition module is used for acquiring vital signs, laboratory data and medical record data through vital sign acquisition, laboratory data acquisition and medical record data acquisition; the data is collected from three aspects; the data preprocessing module can process the data and find features related to the early stage of sepsis from the collected data; the risk assessment module firstly scores, and determines whether machine prediction or expert prediction is performed according to the score level; and the result display module displays the evaluation result through visual display, and generates a report according to patient demands. The system has the advantages that multi-dimensional data such as vital signs of a patient, laboratory examination and electronic medical records are comprehensively collected through the data acquisition module, data islands are broken, and a rich information basis is provided for accurate prediction.

Description

Technical Field

[0001] The present invention relates to the technical field of sepsis risk prediction, and in particular to a system for predicting early sepsis risk. Background Art

[0002] Sepsis is a systemic inflammatory response syndrome caused by infection, characterized by rapid disease progression and high mortality. Early and accurate prediction of the risk of sepsis is crucial for timely and effective treatment measures and improving patient prognosis. Currently, commonly used sepsis prediction methods in clinical practice include judgments based on scoring systems and biomarker detection. However, these methods have certain limitations, such as the limited accuracy of a single scoring system and insufficient specificity of biomarker detection, which make it difficult to meet the clinical demand for early and accurate prediction of sepsis. At the same time, existing technologies lack a comprehensive prediction system that integrates multiple data and performs intelligent analysis, and cannot fully utilize the patient's multi-dimensional information to achieve accurate early risk assessment. Summary of the Invention

[0003] The purpose of the present invention is to address the shortcomings of the existing technology and propose a system for predicting the early risk of sepsis. The technical solution adopted by the present invention is:

[0004] A system for predicting the early risk of sepsis includes a data acquisition module, a data preprocessing module, a risk assessment module, a result display module, and a data management module. The data acquisition module collects data from three aspects: vital signs, laboratory data, and medical records.

[0005] The data preprocessing module can process the data and find features related to early sepsis from the collected data;

[0006] The risk assessment module first scores and then determines whether it is a machine prediction or an expert prediction based on the score level;

[0007] The result display module displays the evaluation results through visual display and generates a report based on the patient's needs;

[0008] The data management module can store historical data, detect and query historical data, and correct data errors when they are found.

[0009] As an improvement, the data collection module includes a vital signs data collection module, a laboratory examination data collection module and an electronic medical record data collection module;

[0010] The vital signs data acquisition module collects the patient's heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation and other basic vital signs information in real time by connecting to bedside monitoring equipment;

[0011] The laboratory examination data acquisition module is connected to the hospital's laboratory information system to automatically obtain data on the patient's blood routine, blood biochemistry, coagulation function indicators, and infection-related biomarkers such as procalcitonin and C-reactive protein;

[0012] The electronic medical record data acquisition module extracts text information such as the patient's basic information, medical history, medication history, and current symptom description from the hospital's electronic medical record system.

[0013] As an improvement, the data preprocessing module includes a data cleaning module, a data standardization module and a feature engineering module;

[0014] The data cleaning module cleans the collected data to remove records with excessive duplication, errors or missing values;

[0015] The data standardization module standardizes data of different types and ranges to make various indicators comparable;

[0016] The feature engineering module extracts and selects features related to the early risk of sepsis from the raw data.

[0017] As an improvement, the risk assessment module includes a scoring calculation module, an expert rule judgment module and a machine learning model prediction module;

[0018] The scoring calculation module integrates common sepsis-related scoring systems, such as the rapid sequential organ failure assessment score, the systemic inflammatory response syndrome score, the national early warning score, etc.

[0019] The expert rule judgment module uses a trained machine learning model to predict the patient's early risk of sepsis;

[0020] The machine learning model prediction module establishes a rule base based on clinical expert experience.

[0021] As an improvement, the result display module includes a visualization interface and a report generation module;

[0022] The visual interface displays the patient's early sepsis risk assessment results in an intuitive and easy-to-understand manner;

[0023] The report generation module system can automatically generate a detailed sepsis early risk assessment report.

[0024] As an improvement, the data management module includes a data correction module, a data storage module and a data retrieval and query module;

[0025] The data correction module can correct and modify the data when the doctor finds data errors to ensure the accuracy of the data;

[0026] The data storage module uses a database management system to securely store the collected patient data;

[0027] The data retrieval and query module provides flexible data retrieval and query functions.

[0028] The beneficial effects of the present invention are:

[0029] The present invention provides a prediction system for the early risk of sepsis. Through a data acquisition module, it comprehensively collects multi-dimensional data such as patients' vital signs, laboratory tests, and electronic medical records, breaking down data silos and providing a rich information basis for accurate prediction.

[0030] Combining traditional scoring systems, machine learning models and expert rules, the early risk of sepsis is assessed from multiple perspectives, making up for the shortcomings of a single method and improving the accuracy and reliability of prediction.

[0031] The result display module presents the prediction results in an intuitive and easy-to-understand manner, allowing doctors to quickly understand the patient's condition; the automatic report generation function reduces doctors' workload and improves work efficiency.

[0032] The data management module ensures the secure storage and effective management of patient data. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 It is the overall flow chart of the present invention.

[0034] Figure 2 This is a flow chart of the data acquisition module of the present invention.

[0035] Figure 3 This is a flow chart of the data preprocessing module of the present invention.

[0036] Figure 4 This is a flow chart of the risk assessment module of the present invention.

[0037] Figure 5 The module flow chart of the structure of the present invention is displayed.

[0038] Figure 6 This is a flow chart of the data management module of the present invention.

[0039] In the figure: 1. Data acquisition module; 101. Vital signs data acquisition module; 102. Laboratory examination data acquisition module; 103. Electronic medical record data acquisition module; 2. Data preprocessing module; 201. Data cleaning module; 202. Data standardization module; 203. Feature engineering module; 3. Risk assessment module; 301. Score calculation module; 302. Expert rule judgment module; 303. Machine learning model prediction module; 4. Result display module; 401. Visualization interface; 402. Report generation module; 5. Data management module; 501. Data correction module; 502. Data storage module; 503. Data retrieval and query module. DETAILED DESCRIPTION

[0040] In order to make the content of the present invention more clearly understood, the technical solutions in the embodiments of the present invention will be described clearly and completely below with reference to the accompanying drawings in which the same components are denoted by the same reference numerals.

[0041] like Figure 1 As shown, it includes data collection module 1, data preprocessing module 2, risk assessment module 3, result display module 4, and data management module 5. Data collection module 1 collects data from three aspects through vital signs collection, laboratory data collection, and medical record data collection;

[0042] Data preprocessing module 2 can process the data and find features related to the early stage of sepsis from the collected data;

[0043] Risk assessment module 3 first scores and then determines whether it is a machine prediction or an expert prediction based on the score level;

[0044] The result display module 4 displays the evaluation results through visual display and generates reports based on patient needs;

[0045] The data management module 5 can store historical data, detect and query historical data, and correct data errors when they are found.

[0046] like Figure 2 As shown, the data collection module 1 includes a vital sign data collection module 101, a laboratory test data collection module 102 and an electronic medical record data collection module 103;

[0047] The vital signs data collection module 101 collects the patient's heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation and other basic vital signs information in real time by connecting to the bedside monitoring equipment. This data will be collected and stored at fixed time intervals;

[0048] The laboratory test data collection module 102 is connected to the hospital's laboratory information system to automatically obtain the patient's blood routine, blood biochemistry, coagulation function indicators and infection-related biomarkers such as procalcitonin and C-reactive protein. The data will be imported into the system in a timely manner after the test results are available;

[0049] The electronic medical record data collection module 103 extracts text information such as the patient's basic information, medical history, medication history, and current symptom description from the hospital's electronic medical record system, and uses natural language processing technology to structure the text data for subsequent analysis;

[0050] Through data collection module 1, multi-dimensional data such as patient vital signs, laboratory tests and electronic medical records are comprehensively collected, breaking down data silos and providing a rich information foundation for accurate predictions.

[0051] like Figure 3 As shown, the data preprocessing module 2 includes a data cleaning module 201, a data standardization module 202 and a feature engineering module 203;

[0052] The data cleaning module 201 cleans the collected data, removes duplicate, erroneous or missing records, and fills in a small number of missing values ​​using methods such as mean interpolation and regression interpolation based on the data characteristics; outliers are identified and processed using statistical methods;

[0053] The data standardization module 202 standardizes data of different types and ranges to make various indicators comparable. For example, vital sign data and laboratory test data are converted into standard normal distribution to facilitate subsequent algorithm calculations.

[0054] The feature engineering module 203 extracts and selects features related to the early risk of sepsis from the raw data. For example, it calculates the rate of change of vital signs and the dynamic trend of laboratory indicators. At the same time, it encodes some categorical variables, such as converting underlying diseases and infection types into numerical form.

[0055] like Figure 4 As shown, the risk assessment module 3 includes a score calculation module 301, an expert rule judgment module 302 and a machine learning model prediction module 303;

[0056] The scoring calculation module 301 integrates common sepsis-related scoring systems, such as the rapid sequential organ failure assessment score, the systemic inflammatory response syndrome score, and the national early warning score, and automatically calculates the corresponding score based on the patient's vital signs and laboratory test data, and displays the results in real time;

[0057] The expert rule judgment module 302 uses a trained machine learning model to predict the patient's early risk of sepsis. The model comprehensively considers the various features extracted by the data preprocessing module and outputs a risk probability value, which indicates the probability of the patient developing sepsis in the future.

[0058] The machine learning model prediction module 303 establishes a set of rule bases based on clinical expert experience. For example, when certain indicators of a patient reach a certain threshold, the system will issue a corresponding alarm and prompt the doctor to conduct further evaluation;

[0059] Combining traditional scoring systems, machine learning models and expert rules, the early risk of sepsis is assessed from multiple perspectives, making up for the shortcomings of a single method and improving the accuracy and reliability of prediction.

[0060] like Figure 5 As shown, the result display module 4 includes a visualization interface 401 and a report generation module 402;

[0061] Visual interface 401 displays the patient's early sepsis risk assessment results in an intuitive and easy-to-understand manner. For example, a dashboard is used to display various scores and risk probability values, and color or icons are used to indicate risk levels. Trend graphs of the patient's key vital signs and laboratory indicators are also displayed, making it easier for doctors to observe the progression of the disease.

[0062] The report generation module 402 system can automatically generate a detailed early sepsis risk assessment report, including the patient's basic information, various scoring results, risk probability, main indicators, and recommended monitoring and intervention measures. The report can be exported in PDF format for easy printing and archiving by doctors;

[0063] The result display module 4 presents the prediction results in an intuitive and easy-to-understand manner, allowing doctors to quickly understand the patient's condition; the automatic report generation function reduces the doctor's workload and improves work efficiency.

[0064] like Figure 6 As shown, the data management module 5 includes a data correction module 501, a data storage module 502 and a data retrieval and query module 503;

[0065] The data correction module 501 can correct and modify the data when the doctor finds data errors to ensure the accuracy of the data;

[0066] The data storage module 502 uses a database management system to securely store the collected patient data. The data will be backed up at regular intervals to prevent data loss. At the same time, the data will be encrypted to ensure patient privacy.

[0067] The data retrieval and query module 503 provides flexible data retrieval and query functions. Doctors can quickly find relevant sepsis risk assessment records and historical data based on conditions such as the patient's name, hospitalization number, and time range. It supports multi-condition combination queries to facilitate clinical research and data analysis.

[0068] The data management module 5 ensures the safe storage and effective management of patient data.

[0069] During use, the vital signs data collection module 101 first collects the patient's heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation and other basic vital signs information. At the same time, the laboratory test data collection module 102 connects with the hospital's laboratory information system to automatically obtain the patient's blood routine, blood biochemistry, coagulation function indicators and infection-related biomarkers such as procalcitonin. The electronic medical record data collection module 103 also extracts the patient's basic information, medical history, medication history and current symptom description from the hospital's electronic medical record system, thereby completing the collection of the required data.

[0070] The collected data is then cleaned by the data cleaning module 201 to remove duplicate, erroneous, or excessively missing records. The cleaned data is then sent to the data standardization module 202, which standardizes data of different types and ranges to make the various indicators comparable. The standardized data is then sent to the feature engineering module 203, which extracts and selects features related to the early risk of sepsis from the raw data.

[0071] The data processed by the data preprocessing module 2 is sent to the score calculation module 301, which automatically calculates the corresponding score based on the patient's vital signs and laboratory test data, and displays the results in real time. Then, based on the score level, it is determined whether to assign it to the expert rule judgment module 302 or the machine learning model prediction module 303. If the score level is relatively low, the machine learning model prediction module 303 will use the trained machine learning model to predict the patient's early sepsis risk. If certain indicators in the score level exceed a certain value, the data will be sent to the expert rule judgment module 302 for further evaluation by medical experts.

[0072] The data after the risk assessment module 3 completes the assessment will be transmitted to the visualization interface 401 for display, and a risk assessment report can be generated through the report generation module 402;

[0073] The result display module 4 will transmit the data and results previously detected by the system to the data storage module 502 for storage. Doctors and other personnel can then use the data retrieval and query module 503 to detect, query and retrieve historical data. At the same time, if doctors find errors in the data during query, they can modify and correct the erroneous data through the data correction module 501.

[0074] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention shall be included in the scope of protection of the present invention.

Claims

1. A system for predicting the early risk of sepsis, characterized in that: The system comprises a data collection module (1), a data preprocessing module (2), a risk assessment module (3), a result display module (4), and a data management module (5). The data collection module (1) collects data from three aspects, namely, vital signs collection, laboratory data collection, and medical record data collection. The data preprocessing module (2) is capable of processing the data and finding features related to the early stage of sepsis from the collected data; The risk assessment module (3) first scores and determines whether it is a machine prediction or an expert prediction based on the score level; The result display module (4) displays the evaluation results through visual display and generates a report according to the patient's needs; The data management module (5) can store historical data, detect and query historical data, and can also correct data errors when they are found.

2. A system for predicting early risk of sepsis according to claim 1, characterized in that: The data acquisition module (1) includes a vital sign data acquisition module (101), a laboratory examination data acquisition module (102) and an electronic medical record data acquisition module (103); The vital signs data acquisition module (101) collects the patient's basic vital signs information such as heart rate, blood pressure, body temperature, respiratory rate, blood oxygen saturation, etc. in real time by connecting to bedside monitoring equipment; The laboratory examination data acquisition module (102) is connected to the hospital's laboratory information system to automatically obtain the patient's blood routine, blood biochemistry, coagulation function indicators and infection-related biomarkers such as procalcitonin and C-reactive protein; The electronic medical record data collection module (103) extracts text information such as the patient's basic information, medical history, medication history, and current symptom description from the hospital's electronic medical record system.

3. The system for predicting the early risk of sepsis according to claim 1, characterized in that: The data preprocessing module (2) includes a data cleaning module (201), a data standardization module (202) and a feature engineering module (203); The data cleaning module (201) cleans the collected data to remove records with excessive duplication, errors or missing values; The data standardization module (202) standardizes data of different types and ranges to make various indicators comparable; The feature engineering module (203) extracts and selects features related to the early risk of sepsis from the raw data.

4. The system for predicting the early risk of sepsis according to claim 1, characterized in that: The risk assessment module (3) includes a scoring calculation module (301), an expert rule judgment module (302) and a machine learning model prediction module (303); The scoring calculation module (301) integrates common sepsis-related scoring systems, such as the rapid sequential organ failure assessment score, the systemic inflammatory response syndrome score, the national early warning score, etc. The expert rule judgment module (302) uses a trained machine learning model to predict the patient's early sepsis risk; The machine learning model prediction module (303) establishes a rule base based on clinical expert experience.

5. The system for predicting the early risk of sepsis according to claim 1, characterized in that: The result display module (4) includes a visualization interface (401) and a report generation module (402); The visualization interface (401) displays the patient's early sepsis risk assessment results in an intuitive and easy-to-understand manner; The report generation module (402) system can automatically generate a detailed sepsis early risk assessment report.

6. The system for predicting the early risk of sepsis according to claim 1, characterized in that: The data management module (5) includes a data correction module (501), a data storage module (502) and a data retrieval and query module (503); The data correction module (501) can correct and modify the data when the doctor finds data errors, thereby ensuring the accuracy of the data; The data storage module (502) uses a database management system to securely store the collected patient data; The data retrieval and query module (503) provides flexible data retrieval and query functions.

Images