An intelligent triage and cause early warning system and method for acute abdomen
The intelligent triage and etiology early warning system for acute abdominal pain utilizes multimodal fusion modeling and association rule mining algorithms to achieve accurate patient triage and department matching in public health emergencies, solving the problems of triage delays and resource shortages, and improving emergency response capabilities.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-03-27
- Publication Date
- 2026-06-30
AI Technical Summary
In existing technologies, when a large number of patients with abdominal pain seek medical treatment due to a sudden public health event, the triage efficiency is low, and delays and inaccurate classification are likely to occur. This results in critically ill patients not being able to receive priority treatment, while patients with mild symptoms occupy medical resources, leading to a shortage of medical resources.
An intelligent triage and etiology early warning system for acute abdominal pain is adopted, including a data acquisition module, an intelligent triage module, an etiology analysis and early warning module, and a terminal interaction module. Through multimodal fusion modeling and association rule mining algorithms, the system identifies the severity of patients and issues graded early warnings, achieving accurate triage of patient information and matching with the appropriate department.
It has improved triage efficiency, ensured that critically ill patients receive timely treatment, optimized the allocation of medical resources, prevented the spread of acute abdominal pain, and enhanced the timeliness and effectiveness of public health emergency response.
Smart Images
Figure CN122314296A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of medical technology, and more specifically, to an intelligent triage and etiology early warning system and method for acute abdominal pain. Background Technology
[0002] Acute abdomen is a general term for a group of clinical emergencies with acute abdominal pain as the core symptom. It encompasses a variety of conditions such as food poisoning, acute gastroenteritis, acute pancreatitis, and intestinal obstruction. It is characterized by rapid onset, rapid progression, complex causes, and large differences in the severity of the condition. Among these, scenarios such as food poisoning and mass abdominal pain can easily lead to a large number of patients flocking to medical institutions in a short period of time, putting enormous pressure on the diagnosis and treatment work.
[0003] Currently, in emergency response to public health emergencies, a large number of patients with abdominal pain seek medical attention within a short period after an outbreak. Existing triage models primarily rely on manual inquiry, physical examination, and preliminary laboratory tests, leading to a surge in workload for medical staff and increasing the risk of triage delays and inaccurate classification. This results in critically ill patients not receiving priority treatment, while patients with mild symptoms occupy high-quality medical resources, further exacerbating the strain on healthcare resources. Therefore, we propose an intelligent triage and etiology early warning system and method for acute abdominal pain. Summary of the Invention
[0004] The purpose of this invention is to provide an intelligent triage and etiology early warning system and method for acute abdominal pain, which aims to solve the problem of low triage efficiency in the emergency response to public health emergencies when a large number of patients flood in.
[0005] To solve the above-mentioned technical problems, the present invention provides the following technical solution: an intelligent triage and etiology early warning system for acute abdominal pain, comprising a data acquisition module, an intelligent triage module, an etiology analysis and early warning module, and a terminal interaction module; The data acquisition module is used to collect patient information, which includes basic information, symptom information, physical signs information, epidemiological history information, and auxiliary examination data. The intelligent triage module, based on patient information and a built-in acute abdominal triage model, enables the matching of patient severity levels with the appropriate department. The etiology analysis and early warning module is used to integrate batch patient information, extract key features and perform common analysis, identify concentrated etiologies, and issue graded early warnings based on the severity of the etiology. The terminal interaction module includes medical staff terminals, patient terminals, management terminals, and public health terminals, enabling patients to view, input, and interact with information.
[0006] Preferably, the data acquisition module includes a patient self-reporting terminal, a medical staff handheld terminal, and an examination equipment interface. The examination equipment interface is connected to the hospital's HIS, LIS, and PACS systems to achieve automatic synchronization of auxiliary examination data.
[0007] Preferably, the basic information includes the patient's age, gender, and name; the symptom information includes the location of abdominal pain, the nature of abdominal pain, the duration of abdominal pain, and accompanying symptoms; the physical signs information includes the patient's body temperature, blood pressure, and abdominal tenderness; the epidemiological history information includes shared food history, contact history, and the time of onset; and the auxiliary examination data includes complete blood count, stool routine examination, blood biochemistry, and abdominal imaging data.
[0008] Preferably, the acute abdominal triage model adopts a multimodal fusion modeling strategy, which integrates natural language processing, gradient boosting tree and graph neural network. It uses historical acute abdominal diagnosis and treatment data and emergency response data of public health emergencies as training samples. The input features include patient symptom information, physical signs information, epidemiological history information, auxiliary examination data and real-time hospital load indicators. The output is the patient's critical level and the corresponding department.
[0009] Preferably, the etiology analysis and early warning module includes a feature extraction unit, a commonality matching unit, and an early warning triggering unit; The feature extraction unit is used to extract common symptom information, physical sign information, epidemiological history information, and auxiliary examination data of patients. The commonality matching unit uses an association rule mining algorithm, combined with a medical knowledge graph, to identify common causes and types of causes. The warning triggering unit is used to issue graded warnings based on the number of patients and the proportion of patients with critical illness levels output by the acute abdominal triage model.
[0010] Preferably, the association rule mining algorithm uses extracted common patient symptom information, physical sign information, epidemiological history information, auxiliary examination data, and etiological types in the medical knowledge graph as the association analysis objects, and uses support and confidence as metrics to screen effective association rules.
[0011] Preferably, the specific formula for selecting effective association rules using support and confidence as metrics is as follows: Support formula: ,in, This is a collection of common symptom information, physical signs information, epidemiological history information, and auxiliary examination data extracted from patients. This represents the set of corresponding etiological types in the medical knowledge graph. To contain simultaneously The number of patient samples, The total number of patient information samples participating in the batch; confidence formula: ,in, For containing feature sets The number of patient samples. This invention also provides a method for intelligent triage and etiology prediction of acute abdominal pain, the method comprising the following steps: S1. Collect patients' basic information, symptom information, physical signs information, epidemiological history information and auxiliary examination data in batches through the data acquisition module; S2. Call the acute abdominal triage model to output the patient's critical severity level and the department to which they should go; S3. Identify concentrated causes through the etiology analysis and early warning module, and issue graded early warnings based on the severity of the causes.
[0012] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention solves the triage problem for patients with acute abdominal pain during public health emergencies, enabling accurate classification of the severity of patients' conditions and reasonable matching of departments for treatment. At the same time, it adapts to the real-time medical resource load of hospitals, optimizes the dynamic allocation efficiency of medical resources, avoids problems such as triage delays and inaccurate classification, and improves the overall efficiency of triage work.
[0013] 2. This invention can integrate various information on a large number of patients with acute abdominal pain, identify the concentrated causes of the disease, and provide graded early warnings, enabling medical institutions and public health departments to intervene in a timely manner, effectively preventing the spread of foodborne and infectious acute abdominal pain, and improving the timeliness and effectiveness of public health emergency response. Attached Figure Description
[0014] Figure 1 This is a schematic diagram of the system architecture of the present invention; Figure 2 This is a flowchart illustrating the method of this procedure. Detailed Implementation
[0015] Example 1: This example addresses the emergency response to public health emergencies. When a public health emergency occurs, a large number of patients with abdominal pain seek medical attention in a short period of time, which can easily lead to delays in triage and inaccurate classification, resulting in low triage efficiency. This example provides an intelligent triage and etiology early warning system for acute abdominal pain, including a data acquisition module, an intelligent triage module, an etiology analysis and early warning module, and a terminal interaction module. The data acquisition module is used to collect patient information, which includes basic information, symptom information, physical signs information, epidemiological history information, and auxiliary examination data. The intelligent triage module, based on patient information and a built-in acute abdominal triage model, enables the matching of patient severity levels with the appropriate department. The etiology analysis and early warning module is used to integrate batch patient information, extract key features and perform common analysis to identify concentrated etiologies and issue graded early warnings based on the severity of the etiology. The terminal interaction module includes medical staff terminals, patient terminals, management terminals, and public health terminals, enabling patients to view, input, and interact with information; In this embodiment, by integrating and analyzing batch patient data, the causes of acute abdominal pain that are concentrated or clustered can be identified in a timely manner. Based on the severity, a graded early warning system can be implemented, which facilitates early intervention by medical institutions and public health departments to prevent the spread of acute abdominal pain caused by foodborne or infectious diseases and improve public health emergency response capabilities.
[0016] Furthermore, the data acquisition module includes a patient self-service reporting terminal, a medical staff handheld terminal, and an interface for examination equipment. The examination equipment interface is connected to the hospital's HIS, LIS, and PACS systems to achieve automatic synchronization of auxiliary examination data. In this embodiment, a multi-terminal, multi-interface data collection method is adopted, which combines self-service filling, medical staff input, and system integration. This not only improves the convenience and efficiency of patient information collection, but also realizes data interoperability with the hospital's existing information systems, avoids manual re-entry of auxiliary examination data, reduces human error, and ensures the timeliness and accuracy of collected data.
[0017] Further, basic information includes the patient's age, gender, and name; symptom information includes the location, nature, duration, and accompanying symptoms of abdominal pain; physical signs information includes the patient's temperature, blood pressure, and abdominal tenderness; epidemiological history information includes shared food history, contact history, and onset time; and auxiliary examination data includes complete blood count, stool routine, blood biochemistry, and abdominal imaging data. In this embodiment, patient information in each dimension is clearly and meticulously defined, and the content and standards of data collection are standardized to ensure that the collected information is uniform and standardized, which facilitates subsequent processing of the data by the algorithm model and improves the accuracy of etiological analysis.
[0018] Furthermore, the acute abdominal pain triage model adopts a multimodal fusion modeling strategy, which integrates natural language processing, gradient boosting tree and graph neural network. It uses historical acute abdominal pain diagnosis and treatment data and emergency response data of public health emergencies as training samples. The input features include patient symptom information, physical signs information, epidemiological history information, auxiliary examination data and real-time hospital load indicators. The output is the patient's critical level and the corresponding department. In this embodiment, a multimodal fusion modeling strategy is adopted, which integrates natural language processing, gradient boosting trees, and graph neural networks. Based on historical acute abdominal pain diagnosis and treatment data and emergency response data of public health emergencies, the model has a more clinically relevant judgment ability. At the same time, the real-time load index of the hospital is included in the input features, so that the triage results not only match the patient's condition, but also adapt to the current status of hospital medical resources, achieving dual precision in criticality classification and department matching, and further optimizing the dynamic allocation of medical resources.
[0019] Furthermore, the etiology analysis and early warning module includes a feature extraction unit, a commonality matching unit, and an early warning triggering unit; The feature extraction unit is used to extract common symptom information, physical signs information, epidemiological history information, and auxiliary examination data of patients; The commonality matching unit uses an association rule mining algorithm, combined with a medical knowledge graph, to identify common causes and types of causes. The early warning triggering unit is used to issue graded early warnings based on the number of patients and the proportion of patients with critical illness levels output by the acute abdominal triage model. In this embodiment, common features are extracted from multi-dimensional information of a batch of patients. The association rule mining algorithm is combined with a medical knowledge graph. The algorithm is responsible for mining the relationship between common features and etiologies from the data level, while the medical knowledge graph provides professional medical theoretical support for association analysis. This ensures that the identified concentrated etiologies and types conform to clinical diagnosis and treatment guidelines, improving the scientificity and accuracy of etiology identification. At the same time, the number of patients and the proportion of critically ill patients are used as core indicators for triggering early warnings. This ensures that the early warning level matches the degree of harm and scope of impact of the etiology, avoiding early warning bias caused by a single indicator, and achieving precise and differentiated public health emergency intervention.
[0020] Furthermore, the association rule mining algorithm uses the extracted common symptoms, signs, epidemiological history, auxiliary examination data, and etiological types from the medical knowledge graph as the objects of association analysis, and uses support and confidence as metrics to screen effective association rules. In this embodiment, the mined association rules are quantitatively screened using support and confidence. Confidence reflects the degree of certainty of the association between common features and etiology. Together with support, association rules are screened from both universality and certainty dimensions, which further improves the accuracy of effective association rule screening and ensures the accuracy of etiology identification.
[0021] Furthermore, the specific formulas for using support and confidence as metrics to screen effective association rules are as follows: Support formula: ,in, This is a collection of common symptom information, physical signs information, epidemiological history information, and auxiliary examination data extracted from patients. This represents the set of corresponding etiological types in the medical knowledge graph. To contain simultaneously The number of patient samples, The total number of patient information samples participating in the batch; confidence formula: ,in, For containing feature sets The number of patient samples. The embodiments disclosed in this invention are preferred embodiments, but are not limited thereto. Those skilled in the art can easily understand the spirit of this invention based on the above embodiments and make different extensions and variations, but as long as they do not depart from the spirit of this invention, they are all within the protection scope of this invention.
Claims
1. An intelligent triage and etiology early warning system for acute abdominal pain, characterized in that, It includes a data acquisition module, an intelligent triage module, a pathogenesis analysis and early warning module, and a terminal interaction module; The data acquisition module is used to collect patient information, which includes basic information, symptom information, physical signs information, epidemiological history information, and auxiliary examination data. The intelligent triage module, based on patient information and a built-in acute abdominal triage model, enables the matching of patient severity levels with the appropriate department. The etiology analysis and early warning module is used to integrate batch patient information, extract key features and perform common analysis, identify concentrated etiologies, and issue graded early warnings based on the severity of the etiology. The terminal interaction module includes medical staff terminals, patient terminals, management terminals, and public health terminals, enabling patients to view, input, and interact with information.
2. The intelligent triage and etiology early warning system for acute abdominal pain according to claim 1, characterized in that, The data acquisition module includes a patient self-reporting terminal, a medical staff handheld terminal, and an examination equipment interface. The examination equipment interface is connected to the hospital's HIS, LIS, and PACS systems to achieve automatic synchronization of auxiliary examination data.
3. The intelligent triage and etiology early warning system for acute abdominal pain according to claim 1, characterized in that, The basic information includes the patient's age, gender, and name; the symptom information includes the location of abdominal pain, the nature of abdominal pain, the duration of abdominal pain, and accompanying symptoms; the physical signs information includes the patient's body temperature, blood pressure, and abdominal tenderness; the epidemiological history information includes shared food history, contact history, and the time of onset; and the auxiliary examination data includes complete blood count, stool routine examination, blood biochemistry, and abdominal imaging data.
4. The intelligent triage and etiology early warning system for acute abdominal pain according to claim 1, characterized in that, The acute abdominal pain triage model adopts a multimodal fusion modeling strategy, which integrates natural language processing, gradient boosting tree and graph neural network. It uses historical acute abdominal pain diagnosis and treatment data and emergency response data of public health emergencies as training samples. The input features include patient symptom information, physical signs information, epidemiological history information, auxiliary examination data and real-time hospital load indicators. The output is the patient's critical level and the corresponding department.
5. The intelligent triage and etiology early warning system for acute abdominal pain according to claim 1, characterized in that, The etiology analysis and early warning module includes a feature extraction unit, a common matching unit, and an early warning triggering unit; The feature extraction unit is used to extract common symptom information, physical sign information, epidemiological history information, and auxiliary examination data of patients. The commonality matching unit uses an association rule mining algorithm, combined with a medical knowledge graph, to identify common causes and types of causes. The warning triggering unit is used to issue graded warnings based on the number of patients and the proportion of patients with critical illness levels output by the acute abdominal triage model.
6. The intelligent triage and etiology early warning system for acute abdominal pain according to claim 5, characterized in that, The association rule mining algorithm uses extracted common patient symptom information, physical sign information, epidemiological history information, auxiliary examination data, and etiology types in medical knowledge graphs as the objects of association analysis, and uses support and confidence as metrics to screen effective association rules.
7. The intelligent triage and etiology early warning system for acute abdominal pain according to claim 6, characterized in that, The specific formula for selecting effective association rules using support and confidence as metrics is as follows: Support formula: ,in, This is a collection of common symptom information, physical signs information, epidemiological history information, and auxiliary examination data extracted from patients. This represents the set of corresponding etiological types in the medical knowledge graph. To contain simultaneously The number of patient samples, The total number of patient information samples participating in the batch; confidence formula: ,in, For containing feature sets The number of patient samples.
8. A method for intelligent triage and etiology early warning of acute abdominal pain, wherein the method employs an intelligent triage and etiology early warning system for acute abdominal pain as described in any one of claims 1-7 for triage and etiology early warning, characterized in that, The method includes the following steps: S1. Collect patients' basic information, symptom information, physical signs information, epidemiological history information and auxiliary examination data in batches through the data acquisition module; S2. Call the acute abdominal triage model to output the patient's critical severity level and the department to which they should go; S3. Identify concentrated causes through the etiology analysis and early warning module, and issue graded early warnings based on the severity of the causes.