Intelligent diagnosis guidance method, system and intelligent diagnosis guidance robot

Abstract

This application is applicable to the field of medical robot technology, and particularly relates to an intelligent triage method, system and intelligent triage robot. The method includes: integrating dynamic information by obtaining department attribute data and multimodal operation records of a target patient; performing data feature fusion based on the multimodal operation records and department attribute data to determine a triage adaptation feature group of the target patient; calling a historical triage database according to the triage adaptation feature group of the target patient to obtain the number of first triage instances with abnormalities in the historical triage database; screening in the historical triage database through the triage adaptation feature group of the target patient to obtain the number of second triage instances associated with the triage adaptation feature group; constructing a triage conflict coefficient according to the number of first triage instances and the number of second triage instances, and determining the triage information of the target patient based on the triage conflict coefficient, so as to achieve specific and accurate triage based on real-time features and historical data and improve the patient's medical experience.

Description

Technical Field

[0001] This application belongs to the technical field of medical robots, and particularly relates to an intelligent triage method, system and intelligent triage robot. Background Art

[0002] Intelligent triage is a key technology for clinical decision-making support, aiming to achieve accurate triage path recommendations by integrating multimodal medical data of patients. By analyzing dynamic indicators such as symptom descriptions, physical sign dimensions, and department resource allocations, it provides a decision-making basis for matching the visiting departments, and is widely used in scenarios such as triage path optimization, medical resource scheduling, and improvement of visiting efficiency.

[0003] Traditional triage methods have significant limitations, including insufficient data fusion capabilities, limited model processing capabilities, low department adaptability, and lack of effective triage decision-making means, resulting in low clinical triage efficiency, improper allocation of medical resources, and at the same time restricting the optimization space for personalized visits and affecting the patient's visiting experience. Summary of the Invention

[0004] Embodiments of this application provide an intelligent triage method, system and intelligent triage robot, which can solve the problems that triage lacks effective triage decision-making means, resulting in low clinical triage efficiency, improper allocation of medical resources, and at the same time restricting the optimization space for personalized visits and affecting the patient's visiting experience.

[0005] In a first aspect, embodiments of this application provide an intelligent triage method, including:

[0006] Obtain department attribute data and multimodal operation records of a target patient;

[0007] Perform data feature fusion based on the multimodal operation records and the department attribute data to determine a triage adaptation feature group of the target patient;

[0008] Obtain a historical triage database, retrieve the first triage instances with triage anomalies in the historical triage database, and obtain the number of the first triage instances; wherein the historical triage database includes a number of historical triage instances, and each historical triage instance records the physical sign data of the patient and the department attribute data of the department where the patient visits;

[0009] Screen the second triage instances associated with the triage adaptation feature group from all the first triage instances, and obtain the number of the second triage instances; wherein, the number of the second triage instances is the triage instances with triage anomalies where the matching degree between the department attribute data of the visiting department, the physical sign data of the patient and the triage adaptation feature group reaches a preset matching threshold;

[0010] Construct a triage conflict coefficient based on the first triage instance quantity and the second triage instance quantity, and determine the guiding diagnosis information of the target patient according to the triage conflict coefficient.

[0011] In the technical solution described above in the embodiments of the present application, at least the following technical effects are achieved:

[0012] The intelligent guiding diagnosis method provided by the embodiments of the present application comprehensively integrates the dynamic information of the patient's physical signs, behaviors, and department resource allocation by obtaining department attribute data and the multimodal operation records of the target patient. Data feature fusion is performed based on the multimodal operation records and department attribute data to determine the triage adaptation feature group of the target patient, providing data support for subsequent analysis. The historical triage database is called based on the triage adaptation feature group of the target patient to obtain the quantity of the first triage instances with abnormalities in the historical triage database; the quantity of the second triage instances associated with the triage adaptation feature group is obtained by screening in the historical triage database through the triage adaptation feature group of the target patient, realizing guiding diagnosis analysis based on historical data. A triage conflict coefficient is constructed according to the quantity of the first triage instances and the quantity of the second triage instances, and the guiding diagnosis information of the target patient is determined according to the triage conflict coefficient, achieving specialized and accurate guiding diagnosis based on real-time features and historical data and improving the patient's medical experience.

[0013] In a second aspect, the embodiments of the present application provide an intelligent guiding diagnosis system, including:

[0014] An obtaining unit, configured to obtain department attribute data and the multimodal operation records of the target patient;

[0015] A fusion unit, configured to perform data feature fusion according to the multimodal operation records and the department attribute data to determine the triage adaptation feature group of the target patient;

[0016] A retrieval unit, configured to obtain a historical triage database, retrieve the first triage instances with triage abnormalities in the historical triage database to obtain the quantity of the first triage instances; where the historical triage database includes a number of historical triage instances, and each historical triage instance records the physical sign data of the patient seeking medical treatment and the department attribute data of the department where the patient seeking medical treatment goes;

[0017] A screening unit, configured to screen the second triage instances associated with the triage adaptation feature group from all the first triage instances to obtain the quantity of the second triage instances; where the quantity of the second triage instances is the triage instances with triage abnormalities whose matching degree between the department attribute data of the department where the patient goes and the physical sign data of the patient seeking medical treatment and the triage adaptation feature group reaches a preset matching threshold;

[0018] A triage guidance unit, configured to construct a triage conflict coefficient based on the first triage instance quantity and the second triage instance quantity, and determine triage guidance information for the target patient according to the triage conflict coefficient.

[0019] In a third aspect, an embodiment of the present application provides an intelligent triage robot, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the method described in any one of the above aspects is implemented.

[0020] In a fourth aspect, an embodiment of the present application provides a computer program product. When the computer program product runs on an intelligent triage robot, the intelligent triage robot is enabled to execute the method described in any one of the above aspects.

[0021] It can be understood that the beneficial effects of the above second to fourth aspects can be referred to the relevant descriptions in the above aspects, and will not be elaborated here. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present application. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0023] Figure 1 is a flowchart of an intelligent triage method provided by an embodiment of the present application;

[0024] Figure 2 is an operation diagram of an intelligent triage method provided by an embodiment of the present application;

[0025] Figure 3 is a structural diagram of an intelligent triage system provided by an embodiment of the present application;

[0026] Figure 4 is a structural diagram of an intelligent triage robot provided by an embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0027] In the following description, specific details such as specific system structures and technologies are presented for the purpose of illustration rather than limitation, so as to thoroughly understand the embodiments of the present application. However, those skilled in the art should clearly understand that the present application can also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to avoid unnecessary details from interfering with the description of the present application.

[0028] It should be understood that, as used in the specification of this application and the appended claims, the term "comprising" indicates the presence of the described features, integers, steps, operations, elements and / or components, but does not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or their combinations.

[0029] It should also be understood that the term "and / or" as used in the specification of this application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0030] As used in the specification of this application and the appended claims, the term "if" may be construed, depending on the context, as "when", "once", "in response to determining", or "in response to detecting". Similarly, the phrases "if determined" or "if the described condition or event is detected" may be construed, depending on the context, to mean "once determined", "in response to determining", "once the described condition or event is detected", or "in response to detecting the described condition or event".

[0031] In addition, in the description of the specification of this application and the appended claims, the terms "first", "second", "third", etc. are used only for descriptive distinction and should not be construed as indicating or implying relative importance.

[0032] Reference to "one embodiment" or "some embodiments" or the like described in the specification of this application means that a particular feature, structure, or characteristic described in connection with that embodiment is included in one or more embodiments of this application. Thus, statements such as "in one embodiment", "in some embodiments", "in other some embodiments", "in still other embodiments", etc. that appear in different places in this specification are not necessarily all referring to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "comprising", "including", "having", and their variants all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0033] Traditional triage methods have significant limitations, including insufficient data fusion capabilities, limited model processing capabilities, low department matching fitness, and a lack of effective triage decision-making means, resulting in low clinical triage efficiency, improper allocation of medical resources, and at the same time restricting the optimization space for personalized medical treatment and affecting the patient's medical experience.

[0034] To solve the above problems, the embodiments of the present application provide an intelligent triage method, system and intelligent triage robot. In this method, by obtaining department attribute data and multimodal operation records of a target patient, the dynamic information of the patient's physical signs, behaviors and department resource allocation is comprehensively integrated. Data feature fusion is performed based on the multimodal operation records and department attribute data to determine the triage adaptation feature group of the target patient, providing data support for subsequent analysis. The historical triage database is called according to the triage adaptation feature group of the target patient to obtain the number of abnormal first triage instances in the historical triage database; the second triage instances associated with the triage adaptation feature group are obtained by screening in the historical triage database through the triage adaptation feature group of the target patient, realizing triage analysis based on historical data. According to the number of the first triage instances and the number of the second triage instances, a triage conflict coefficient is constructed, and the triage information of the target patient is determined based on the triage conflict coefficient, realizing specific and accurate triage based on real-time features and historical data and improving the patient's medical experience.

[0035] The intelligent triage method provided by the embodiments of the present application can be applied to an intelligent triage robot. At this time, the intelligent triage robot is the execution subject of the intelligent triage method provided by the embodiments of the present application, and the embodiments of the present application do not impose any restrictions on the specific type of the intelligent triage robot.

[0036] For example, the intelligent triage robot can be various types of robots. For example, the intelligent triage robot can be a humanoid robot, a counter robot, a wearable robot, a rail robot, a vehicle-mounted robot. The intelligent triage robot can include but is not limited to mobile devices, communication devices, display devices, sensor devices, etc.

[0037] To better understand the intelligent triage method provided by the embodiments of the present application, the following provides an exemplary introduction to the specific implementation process of the intelligent triage method provided by the embodiments of the present application.

[0038] Figure 1 Fig. shows a schematic flowchart of the intelligent triage method provided by the embodiments of the present application. Figure 2 Fig. shows a running flowchart of the intelligent triage method provided by the embodiments of the present application. The intelligent triage method includes:

[0039] S100, obtain department attribute data and multimodal operation records of a target patient.

[0040] It can be understood that the department attribute data refers to the description of the characteristics and capabilities of each department in a hospital, which may include the professional fields of the departments, the configuration of medical equipment, the professional level of the doctor team, the success rate of handling historical cases, the types of diseases that the departments are good at treating, etc. The department attribute data can be extracted from the hospital's management system or constructed in the form of a knowledge graph to describe the medical capabilities and resource distribution of the departments. The multi-modal operation records of the target patient refer to various types of data generated during the patient's medical treatment process, which may include the physiological data, behavioral data, medical history data of the target patient, and the operation records generated during the interaction process. The multi-modal operation records can be sourced from multiple data sources such as the hospital's electronic medical record system, sensor devices, and mobile applications.

[0041] Exemplarily, when obtaining the department attribute data, it can be directly extracted through the hospital's information system or queried from a pre-constructed knowledge base. When obtaining the multi-modal operation records of the target patient, the data of the patient at different time periods and in different scenarios can be collected and integrated. For example, the physiological data and medical history of the patient can be extracted from the electronic medical record system, and the real-time monitoring data, interaction records of the patient, etc. can be obtained from the sensor devices. It can provide basic data support for subsequent data feature fusion and analysis, and provide an important basis for the decision-making of the intelligent triage system.

[0042] S200, perform data feature fusion based on the multi-modal operation records and the department attribute data to determine the triage adaptation feature group of the target patient.

[0043] It can be understood that multimodal operation records can include the patient's physiological data, behavioral data, medical history, etc. at different times and occasions. The multimodal operation records can be temporally aligned and semantically parsed to extract operation behavior feature vectors and mark abnormal operations. Temporal alignment ensures that data from different time periods are consistent on the same time scale, while semantic parsing converts this data into a computable and analyzable format. A department professional ability map can be constructed based on department attribute data. The process of constructing a department professional ability map involves encoding information such as the historical case processing success rate of the department into three-dimensional features to form a network graph, where the nodes represent the key ability features of the department. After feature extraction and department ability map construction are completed, the behavior feature vectors of the patient can be cross-modally associated and mapped with the department ability map. By expanding the dimensions of the behavior feature vectors and projecting them into the same feature space as the department ability map, the dynamic semantic similarity between each behavior feature vector and the department nodes is calculated. After the dynamic semantic similarity is subjected to piecewise normalization processing, a standardized association weight is formed, which is used to construct a behavior-department association matrix. The dynamic attention mechanism can be used to assign weights to the abnormal operation marks and the three-dimensional feature encoding of the department ability map, calculate the feature coupling degree, and comprehensively perform non-linear fusion on the behavior-department association matrix and the feature coupling degree to generate a triage adaptation feature group reflecting the patient-department adaptation degree, which is conducive to providing core data support for subsequent triage decisions and helping to determine the most suitable department for the target patient.

[0044] In a possible implementation manner, S200, perform data feature fusion based on the multimodal operation records and department attribute data to determine the triage adaptation feature group of the target patient, including:

[0045] S210, perform temporal alignment and semantic parsing on the multimodal operation records to extract operation behavior feature vectors and abnormal operation marks.

[0046] It can be understood that temporal alignment refers to unifying data from different time points and different devices on the time axis to achieve consistency of data on the same time scale. For example, the blood pressure data and heart rate data of the target patient may be collected at different time points, and they can be aligned to the same time axis through interpolation or other methods for subsequent analysis. Temporal alignment can improve the availability of data and avoid analysis errors caused by inconsistent time.

[0047] Semantic parsing refers to the structured processing of unstructured or semi-structured data to extract key feature information. For example, the consultation content of the target patient may be recorded in natural language, and key disease descriptions, treatment suggestions, etc. need to be extracted through natural language processing techniques; the patient's examination items may be recorded in text or code form and need to be parsed into specific examination types and result data. Semantic parsing can convert complex data into computable and analyzable feature vectors, facilitating subsequent feature fusion.

[0048] Based on time series alignment and semantic parsing, operation behavior feature vectors can be extracted. The operation behavior feature vector is a numerical representation of the patient's operation record, which can reflect the behavior pattern of the target patient during the medical treatment process. In addition, there may be abnormal operation marks in the operation record, and the abnormal operation marks can be detected through predefined rules. The abnormal operation mark is an indication of the abnormal situation of the patient's health status or behavior pattern and can be used for subsequent weight assignment and feature fusion.

[0049] S220, constructing a department professional ability graph based on department attribute data; among them, the department professional ability graph includes three-dimensional feature encoding of the historical case processing success rate.

[0050] It can be understood that the process of constructing the department professional ability graph includes feature extraction and encoding of the department attribute data. Among them, the historical case processing success rate is one of the core features of the department attribute data, reflecting the actual performance of the department in a specific disease or treatment field. The historical case processing success rate can be encoded in multiple dimensions, such as based on dimensions such as disease type, treatment method, and patient age, to form three-dimensional feature encoding, which can more comprehensively describe the professional ability of the department and provide rich information for subsequent feature fusion.

[0051] Exemplarily, when constructing the department professional ability graph, the department attribute data can be cleaned and preprocessed to remove redundant information and unify the data format. Key feature dimensions can be extracted through statistical analysis of the historical case processing success rate. For example, the success rate of different departments can be statistically analyzed by disease type, the success rate can be statistically analyzed by treatment method (such as surgery, drug treatment), or the success rate can be statistically analyzed by patient age group. These statistical results are encoded into three-dimensional feature vectors, with each dimension corresponding to a specific feature, such as the disease type dimension, the treatment method dimension, the patient age dimension, etc. The encoded feature vectors can be integrated with other department attribute data (such as doctor professional level, medical equipment configuration, etc.) to form the department professional ability graph. The department professional ability graph reflects the distribution of capabilities and resources among departments in a graph structure, providing basic data support for subsequent cross-modal association mapping.

[0052] S230. Perform cross-modal association mapping between the operation behavior feature vector and the department professional ability graph to generate a behavior-department association matrix.

[0053] It can be understood that cross-modal association mapping refers to associating data from different modalities (patient behavior and department ability) and calculating the similarity or correlation between them.

[0054] When performing cross-modal association mapping, the operation behavior feature vector and the department professional ability graph can be encoded into the same feature space. The dynamic semantic similarity between each operation behavior feature vector and each node in the department professional ability graph can be calculated. The dynamic semantic similarity means dynamically adjusting the weights according to different feature dimensions to reflect the importance of different features for the association. The calculated dynamic semantic similarity can be subjected to piecewise normalization processing to convert it into a standardized association weight. Piecewise normalization can eliminate the dimensional differences between different feature dimensions, making the similarity values comparable within the same range. By arranging the association weights according to the node relationship between the patient behavior feature vector and the department professional ability graph, a behavior-department association matrix is generated. The behavior-department association matrix is a two-dimensional matrix, where the rows represent the operation behavior feature vectors of the patients, the columns represent the nodes in the department professional ability graph, and each element in the matrix represents the association strength between the patient behavior and the department ability. By generating the behavior-department association matrix, the behavior characteristics of the patients can be quantitatively associated with the professional abilities of the departments, providing an important basis for determining the subsequent triage adaptation feature group. The cross-modal association mapping method can comprehensively consider multiple dimensions of patient behavior and department ability, improving the accuracy and scientificity of triage decisions.

[0055] Optionally, S230. By performing cross-modal association mapping between the operation behavior feature vector and the department professional ability graph to generate a behavior-department association matrix, it includes:

[0056] S231. Expand the dimension of the operation behavior feature vector and project the operation behavior feature vector into the same feature space as the department professional ability graph.

[0057] It can be understood that dimension expansion refers to increasing the dimension of the operation behavior feature vector to the same number of dimensions as the department professional ability map through mathematical methods (such as zero padding, interpolation, or feature mapping), so that the two can be compared and analyzed in the same feature space. For example, if the feature space of the department professional ability map includes three dimensions: disease type, treatment method, and patient age, while the operation behavior feature vector only contains two dimensions: registration frequency and types of examination items, then dimension expansion can be completed by adding new feature dimensions (such as patient age) or expanding existing features (such as mapping the registration frequency to the disease type dimension). Dimension expansion enables the operation behavior feature vector to be comparable with the department professional ability map in the same feature space, providing a basis for subsequent calculation of dynamic semantic similarity.

[0058] S232, Based on the mapping relationship in the preset symptom-department association rule base, calculate the dynamic semantic similarity between each operation behavior feature vector and the node of the department professional ability map.

[0059] It can be understood that the symptom-department association rule base is a pre-constructed knowledge base used to describe the association relationship between different symptoms and departments. For example, the rule base may associate the symptom of "fever" with departments such as "Internal Medicine" and "Respiratory Medicine", or associate the symptom of "fracture" with departments such as "Orthopedics" and "Emergency Department". The symptom-department association rule base can be constructed through expert experience and historical case analysis. When calculating the dynamic semantic similarity, the corresponding mapping relationship can be found in the symptom-department association rule base according to the features in the operation behavior feature vector (such as the patient's chief complaint symptoms, examination results, etc.). Based on the mapping relationship and the node features in the department professional ability map, calculate the dynamic semantic similarity between the operation behavior feature vector and the node of the department professional ability map. Dynamic semantic similarity refers to dynamically adjusting the weights according to different feature dimensions to reflect the importance of different features for the association in a specific context. For example, for the symptom of fever, the weight of the historical case treatment success rate of Internal Medicine may be higher, while for the symptom of fracture, the weight of the historical case treatment success rate of Orthopedics may be higher. Through the calculation of dynamic semantic similarity, the association strength between the patient behavior features and the department capabilities can be quantified, providing a basis for subsequent construction of the association matrix.

[0060] S233, Perform piecewise normalization on the dynamic semantic similarity to generate a standardized association weight.

[0061] It can be understood that due to the possible different feature dimensions or calculation methods, the numerical range of dynamic semantic similarity may vary greatly. For example, the similarity values of some features may be concentrated between 0 and 1, while the similarity values of other features may be distributed between 0 and 100. To eliminate this dimensional difference, the dynamic semantic similarity can be processed by piecewise normalization. Piecewise normalization means standardizing the similarity values according to different intervals (segments). For example, the similarity values can be divided into three intervals: low, medium, and high, and normalization is performed separately within each interval so that all similarity values are mapped to a unified range (such as between 0 and 1). Piecewise normalization can preserve the distribution characteristics of the similarity values while eliminating the dimensional differences between different dimensions, making the similarity values comparable and interpretable. The normalized value is the standardized correlation weight, which is used to represent the correlation strength between the target patient's operation behavior feature vector and the department's capabilities.

[0062] S234. Construct an N×M sparse correlation matrix based on the standardized correlation weight, and perform temporal noise reduction on the sparse correlation matrix to generate a behavior-department correlation matrix; where N is the total number of dimensions of the operation behavior feature vector, and M is the total number of nodes in the department professional ability graph.

[0063] It can be understood that the standardized correlation weight is a numerical representation of the correlation strength between the operation behavior feature vector and the nodes in the department professional ability graph. Based on the standardized correlation weight, an N×M correlation matrix can be constructed, where N represents the total number of dimensions of the operation behavior feature vector (such as registration frequency, types of examination items, chief complaint symptoms, etc.), and M represents the total number of nodes in the department professional ability graph (such as internal medicine, surgery, orthopedics, etc.). Since there is not a strong correlation between all operation behavior features and all department nodes, this correlation matrix is usually sparse, that is, most of the elements in the matrix are zero or close to zero. To reduce the influence of noise in the sparse matrix, it is necessary to perform temporal noise reduction on the correlation matrix. Temporal noise reduction means using time series analysis methods or filtering techniques (such as moving average filtering, low-pass filtering, etc.) to smooth the weight values in the matrix and eliminate the influence of random noise or outliers. For example, if a weight value fluctuates violently in a short period of time, it may be regarded as noise and can be smoothed into a more stable value through filtering techniques. The correlation matrix after noise reduction processing is the final behavior-department correlation matrix. The behavior-department correlation matrix can clearly quantify the correlation relationship between the patient's operation behavior characteristics and the department's capabilities, providing an important basis for subsequent triage decisions.

[0064] S240. Assign weights to the abnormal operation markers and the three-dimensional feature encoding through the dynamic attention mechanism, and calculate the feature coupling degree.

[0065] It can be understood that the abnormal operation mark is an identification of the abnormal behavior of the patient during the medical treatment process, reflecting the possible health risks or behavioral abnormalities of the patient. The three-dimensional feature encoding is a multi-dimensional representation of the success rate of historical case processing in the department's professional ability map, such as encoding based on dimensions such as disease type, treatment method, and patient age. The feature coupling degree refers to the association strength between the abnormal operation mark and the three-dimensional feature encoding, and is used to quantify the matching degree between the patient's abnormal behavior and the department's ability.

[0066] Exemplarily, the dynamic attention mechanism is a model that can dynamically adjust the weight allocation according to the context of the input data. Its core idea is to adaptively weight the contribution degrees of different features. When calculating the feature coupling degree, the abnormal operation mark and the three-dimensional feature encoding can be input into the dynamic attention mechanism. The dynamic attention mechanism determines the association degree between each abnormal operation mark and each three-dimensional feature encoding by calculating the attention scores between the two. Based on the attention scores, dynamic weight allocation is performed on the abnormal operation mark and the three-dimensional feature encoding, that is, different weight values are assigned to different features to reflect their importance in the calculation of the feature coupling degree. According to the weight allocation result, the feature coupling degree between the abnormal operation mark and the three-dimensional feature encoding is calculated. The feature coupling degree is a numerical index used to quantify the matching degree between the patient's abnormal behavior and the department's ability, providing an important basis for subsequent feature fusion.

[0067] S250, according to the behavior-department association matrix and the feature coupling degree, non-linearly fuse the multi-modal operation record features and the department attribute features to generate the triage adaptation feature group of the target patient.

[0068] It can be understood that non-linear fusion refers to using non-linear transformation methods to integrate features from different modalities to generate a feature group that can comprehensively reflect the adaptability between the patient and the department.

[0069] Exemplarily, the behavior-department association matrix can be integrated with the feature coupling degree to generate a comprehensive association feature matrix. For example, the feature coupling degree can be used as a weight to weightedly adjust the behavior-department association matrix, enhancing the influence of features with strong relevance to abnormal behaviors on the final feature group. Input the multi-modal operation record features and department attribute features into a non-linear fusion model, such as models like neural network, decision tree, or support vector machine. The non-linear fusion model can capture the complex non-linear relationships between features, such as the interaction between the patient's behavior features and department capacity features. Through the calculation of the non-linear fusion model, the multi-modal operation record features and department attribute features are integrated to generate a triage adaptation feature group for the target patient. The triage adaptation feature group is a comprehensive feature vector that contains information on the patient's behavior features, abnormal behavior markers, department capacity features, and their association relationships, and can provide comprehensive and accurate data support for subsequent triage decisions. By generating the triage adaptation feature group, the scientificity and accuracy of triage decisions can be improved, ensuring that patients can be assigned to the most suitable department for treatment.

[0070] S300. Obtain a historical triage database, retrieve the first triage instances with triage anomalies in the historical triage database, and obtain the number of the first triage instances; wherein the historical triage database includes a number of historical triage instances, and each historical triage instance records the physical sign data of the patient seeking medical treatment and the department attribute data of the department where the patient seeking medical treatment goes.

[0071] It can be understood that the historical triage database is a data set storing the triage records of past patients, and each record is a historical triage instance. The historical triage instance records the physical sign data of the patient and the department attribute data of the assigned department. The physical sign data may include the patient's physiological indicators (such as blood pressure, heart rate, body temperature), medical history information (such as past diseases, allergy history), and behavior data (such as the number of registration times, the number of examination items), etc. The department attribute data may include the professional field of the department (such as internal medicine, surgery, pediatrics), equipment configuration (such as CT machine, ultrasound instrument), and the level of the doctor team (such as doctor qualifications, surgical success rate), etc. In the historical triage database, the first triage instance refers to the situation where the department assigned to the patient does not match the physical sign data, that is, the historical triage instance with triage anomalies. For example, a patient with heart disease is assigned to a non-cardiovascular department, or a patient who needs complex surgery is assigned to a department with insufficient equipment. Retrieving the first triage instance is to identify possible problems in historical triage, so as to optimize the triage process, improve triage accuracy and patient safety.

[0072] In a possible implementation manner, S300, obtaining a historical triage database, retrieving the first triage instances with triage anomalies in the historical triage database, and obtaining the number of the first triage instances, includes:

[0073] S310, obtain the historical triage database.

[0074] It can be understood that obtaining the historical triage database means reading a data set containing historical triage instances from a storage system. The historical triage database can be stored in a relational database (such as MySQL, PostgreSQL), a non-relational database (such as MongoDB), or a distributed file system (such as HDFS). When obtaining the historical triage database, the historical triage instances can be loaded into memory or local storage through a database query interface (such as an SQL query) or a data import tool (such as an ETL tool) for subsequent retrieval and analysis.

[0075] S320, retrieve the historical triage database according to a preset abnormal determination criterion, obtain a first triage instance with triage abnormalities in the historical triage database, and count the number of first triage instances to obtain the number of first triage instances.

[0076] It can be understood that the preset abnormal determination criterion is a rule or condition for judging whether there are triage abnormalities in historical triage instances. The abnormal determination criterion can be formulated based on professional medical guidelines or historical triage experience. For example, the abnormal determination criterion can include the following conditions:

[0077] The patient's vital sign data indicates that they need to be prioritized (such as high-risk pregnancy, acute myocardial infarction), but are assigned to a non-priority treatment department.

[0078] The patient requires specific equipment or technology (such as cardiac stent surgery, tumor radiotherapy), but is assigned to a department with insufficient equipment or technology.

[0079] The patient's medical history data indicates a higher need for a specific department (such as diabetes requires an endocrinology department), but is assigned to another department.

[0080] The patient's medical history data indicates that they have not received any treatment after registration and are reassigned to another department.

[0081] During the retrieval process, historical triage instances can be traversed one by one, and each instance can be evaluated according to the abnormal determination criteria. If an instance meets at least one of the abnormal determination criteria, it is marked as the first triage instance with triage abnormalities. For example, for a record, it can be checked whether the patient's physical sign data matches the department attribute data. If it matches, it is marked as normal; otherwise, it is marked as abnormal. After the retrieval is completed, the number of all the first triage instances marked as abnormal is counted to obtain the number of the first triage instances. The number of the first triage instances reflects the number of cases with triage abnormalities in the historical triage, providing data support for the subsequent optimization of the triage process. For example, if the number of the first triage instances is high, it indicates that there may be major problems in the existing triage process, and the reasons need to be further analyzed and the triage rules need to be optimized; if the number of the first triage instances is low, it indicates that the existing triage process is relatively reasonable. By counting the number of the first triage instances, a scientific basis can be provided for the triage dynamics.

[0082] S400. Screen the second triage instances associated with the triage adaptation feature group from all the first triage instances to obtain the number of the second triage instances. Among them, the number of the second triage instances is the triage instances with triage abnormalities where the matching degree between the department attribute data of the visiting department, the patient's physical sign data and the triage adaptation feature group reaches a preset matching threshold.

[0083] It can be understood that the first triage instance is a triage abnormal instance in the historical triage database that meets the abnormal determination criteria; the second triage instance is a further screened triage abnormal instance associated with the triage adaptation feature group from the first triage instances. The triage adaptation feature group is a feature vector used to reflect the adaptability between the patient and the department, including the patient's physical sign data, behavior data, medical history data, department attribute data, etc. The preset matching threshold is a numerical standard used to measure the matching degree between the department attribute data and the patient's physical sign data and the triage adaptation feature group. For example, an instance with a matching degree greater than 0.8 is considered a successful match. The purpose of screening the second triage instances is to identify the situations highly similar to the triage adaptation feature group of the current target patient among the historical triage abnormal instances, so as to discover potential triage rules or problems, and then optimize the triage process. The number of the second triage instances reflects the association degree between the historical abnormal instances and the triage adaptation feature group of the current target patient, providing a reference basis for the triage decision.

[0084] In a possible implementation manner, S400. Screen the second triage instances associated with the triage adaptation feature group from all the first triage instances to obtain the number of the second triage instances, including:

[0085] S410. According to the triage adaptation feature group of the target patient, construct the department attribute data matching condition and the physical sign data correlation degree constraint condition.

[0086] It can be understood that constructing the matching conditions for department attribute data means defining the department matching rules for screening historical triage instances based on the department attribute dimensions in the triage adaptation feature group (such as department specialty field, equipment configuration, doctor team level). For example, if the department attribute of "cardiovascular disease" is included in the triage adaptation feature group of the target patient, the matching condition can be set to screen the records in historical instances where the department specialty field includes the diagnosis and treatment ability of cardiovascular disease. The constraint condition for the correlation degree of sign data means defining the calculation method and threshold range of sign similarity based on the patient sign dimensions in the triage adaptation feature group (such as physiological indicators like blood pressure, heart rate, body temperature, etc.). For example, the cosine similarity algorithm is used to calculate the similarity between the sign vector of the target patient and the sign vector of the historical instance, and the lower limit of the similarity is set to 0.7.

[0087] The matching conditions for department attribute data can be constructed through a vector space model. After encoding the department attribute features into multi-dimensional vectors, the Euclidean distance is used to measure the matching degree. The constraint condition for the correlation degree of sign data can be combined with the standardized sign data, and the correlation degree is calculated through the Pearson correlation coefficient or the dynamic time warping algorithm (for time-series sign data).

[0088] S420, based on the matching conditions for department attribute data and the constraint conditions for the correlation degree of sign data, screen all the first triage instances to obtain the second triage instances, and count the number of the second triage instances to obtain the number of the second triage instances.

[0089] It can be understood that the matching conditions for department attribute data and the constraint conditions for the correlation degree of sign data can be applied to the data set of all the first triage instances through a database query engine or a data filtering algorithm. For each first triage instance, double verification can be performed: verify whether its department attribute meets the matching conditions, for example, judge whether the department specialty field includes the target disease type through boolean logic; secondly, calculate the correlation degree score of the sign data, for example, after aligning the sign data of the historical instance with the sign data of the target patient, use a predefined similarity algorithm to calculate the numerical result. When and only when both conditions are met (such as department matching degree ≥ 0.8 and sign correlation degree ≥ 0.7), then this first triage instance is marked as a second triage instance.

[0090] Exemplarily, a distributed computing framework (such as Spark) can be used to perform parallel screening on large-scale historical data. The data set of all the first triage instances can be sharded and loaded into memory, and the filtering conditions are applied to each data shard. The condition matching and correlation degree calculation are completed through MapReduce operations, and finally the total number of the second triage instances that meet the conditions is counted through an aggregation operation. The obtained number of the second triage instances will be used as one of the core parameters for calculating the subsequent triage abnormal risk probability.

[0091] S500. Construct a triage conflict coefficient based on the number of first triage instances and the number of second triage instances, and determine the triage information of the target patient based on the triage conflict coefficient.

[0092] It can be understood that the triage conflict coefficient is a quantitative indicator used to characterize the decision-making difference degree between the triage adaptation feature group of the current target patient and the historical triage abnormal instances. The number of first triage instances reflects the overall scale of historical triage abnormal instances, and the number of second triage instances characterizes the number of abnormal instances highly correlated with the characteristics of the current target patient. When constructing the triage conflict coefficient, the distribution difference between the first triage instance and the second triage instance in the multi-dimensional feature space can be modeled through a probability density distribution function, and the conflict degree can be quantified using information theory methods. The determination process of the triage information can select the department with the lowest triage conflict coefficient as the recommended target based on the correlation between the triage conflict coefficient and the misdiagnosis risk of the department, so as to reduce the risk of triage errors.

[0093] In a possible implementation manner, S500. Construct a triage conflict coefficient based on the number of first triage instances and the number of second triage instances, and determine the triage information of the target patient, including:

[0094] S510. Obtain the number of triage instances in the historical triage database.

[0095] It can be understood that the number of triage instances refers to the total amount of all triage instance records in the historical triage database, including normal triage instances and abnormal triage instances. The number of triage instances can be directly obtained through database statistical functions (such as the COUNT operation in SQL) or quickly read through pre-generated metadata indexes. The number of triage instances provides a reference value for subsequent probability density distribution modeling, which is used to standardize the statistical ratios of the first triage instance and the second triage instance. For example, when the historical triage database contains 100,000 records, the number of triage instances is 100,000, and the number of triage instances will be used as the denominator to calculate the occurrence probability of abnormal instances.

[0096] S520. Map the number of first triage instances and the number of triage instances to a multi-dimensional feature space to generate a first probability density distribution function corresponding to the number of first triage instances.

[0097] It can be understood that the multi-dimensional feature space is a mathematical modeling space composed of multiple feature dimensions of historical triage instances, such as department specialty matching degree, physical sign data deviation degree, equipment requirement satisfaction degree, etc. The first probability density distribution function P(x) reflects the aggregation pattern of historical triage abnormal instances in the feature space. The distribution pattern of historical triage abnormal instances (the first triage instance) in the multi-dimensional feature space can be quantified through probability density modeling.

[0098] Exemplarily, the first probability density distribution function P(x) can be calculated by the following formula: , where: is the number of the first triage instances, i.e., the total number of abnormal instances; represents the -th abnormal instance's coordinates in the -dimensional feature space; is the kernel bandwidth parameter, which controls the smoothness of density estimation; is the kernel function (such as Gaussian kernel ; is the probability density value at any point in the feature space. The features of each first triage instance can be encoded: numerical features (such as blood pressure values) directly retain the original values; categorical features (such as department types) are converted into binary vectors through one-hot encoding; text features (such as medical history descriptions) are mapped into low-dimensional dense vectors through word embedding (such as Word2Vec); then, standardization processing is performed to eliminate the dimensional difference. For example, for the j-th dimensional feature x j , its mean value is calculated as μ j and the standard deviation is σ j ; the standardized value is . The Gaussian kernel function is selected to ensure the continuous differentiability of the probability density function. The bandwidth is optimized through cross-validation to minimize the integrated mean square error (MISE). For the -dimensional space, the bandwidth matrix is often simplified to a diagonal matrix , where is initialized by the Silverman criterion: , where is the feature standard deviation. For any point in the feature space, its probability density is obtained by the weighted sum of the kernel functions of all abnormal instances. For example, in a three-dimensional space (department ability score , body temperature , heart rate ), the density of a certain point is calculated as: ; where, is the feature value of the -th abnormal instance.

[0099] S530, map the number of the second triage instances and the number of triage instances to the multi-dimensional feature space to generate the second probability density distribution function corresponding to the number of the second triage instances.

[0100] It can be understood that the second probability density distribution function is used to quantify the distribution pattern of abnormal instances (second triage instances) highly associated with the characteristics of the target patient in the multi-dimensional feature space. The second triage instances are a subset of the first triage instances and need to meet the department attribute matching condition and the physical sign data correlation constraint condition. The second probability density distribution function Q(x) focuses on historical abnormal instances highly related to the characteristics of the current target patient, and its density peak reflects the "high-risk feature combination" that may lead to triage conflicts. By comparing with the first probability density distribution function P(x) (such as calculating the KL divergence), the deviation degree of the current patient's characteristics from the historical abnormal pattern can be accurately identified, providing a fine-grained risk assessment for triage decisions. For example, if Q(x) is significantly higher than P(x) in the region of "high medical history complexity + low department response time", it indicates that such feature combinations need to be avoided preferentially.

[0101] Exemplarily, the mathematical form of the second probability density distribution function is similar to that of the first probability density distribution function. However, there are differences in data sources and parameter settings. The second probability density distribution function can be obtained through the following formula: ; where is the number of second triage instances; represents the feature vector of the th second triage instance; is the kernel bandwidth of the second distribution, usually smaller than the bandwidth of the first distribution (due to less data volume and more concentrated distribution).

[0102] S540. Calculate the relative entropy ratio of the first probability density distribution function and the second probability density distribution function to generate a feature sensitivity discriminant function.

[0103] It can be understood that the first probability density distribution function and the second probability density distribution function respectively represent the probability distributions of feature variables under different conditions. For example, the first distribution may correspond to the distribution of physiological parameters (such as blood pressure, heart rate) of a patient group in a certain department, and the second distribution may correspond to the distribution of patient groups in another department. The relative entropy (Kullback-Leibler divergence) is an asymmetric index to measure the difference between two probability distributions, and its definition is the difference between the logarithmic expectations of the two distributions. The feature sensitivity discriminant function is a mathematical expression constructed based on this ratio, used to evaluate the discrimination ability of specific features in triage among different departments. For example, if the relative entropy ratio of a certain physiological feature significantly deviates from 1, it indicates that there is a significant distribution difference of this feature among patients in the two types of departments and can be marked as a highly sensitive feature by the discriminant function.

[0104] Optionally, S540. Calculate the relative entropy ratio of the first probability density distribution function and the second probability density distribution function to generate a feature sensitivity discriminant function, including:

[0105] S541. Determine the mixed distribution of the first probability density distribution and the second probability density distribution, and calculate the symmetric divergence and the asymmetric divergence between the first probability density distribution and the second probability density distribution according to the mixed distribution.

[0106] It can be understood that the mixed distribution is an intermediate distribution generated by weighted averaging of two original distributions. For example, it is defined as M(x) = 1 / 2 (P(x) + Q(x)), where P(x) and Q(x) are the first and second probability density distribution functions respectively. The symmetric divergence usually refers to the Jensen-Shannon divergence (JS divergence), and its calculation method is JS(P||Q) = 0.5KL(P||M) + 0.5KL(Q||M). By introducing the mixed distribution M, the divergence becomes symmetric. The asymmetric divergence directly uses the original relative entropy KL(P||Q) or KL(Q||P).

[0107] The parameters of the mixed distribution M can be estimated based on the sample data of P and Q. For example, for a Gaussian mixture model, the expectation-maximization algorithm can be used to optimize the weight coefficients. Subsequently, calculate the relative entropy between P and M, and between Q and M respectively, and obtain the JS divergence value through weighted summation. At the same time, independently calculate the original relative entropy of P with respect to Q as the asymmetric divergence. These two divergence values quantify the distribution differences from symmetric and asymmetric perspectives respectively, providing a multi-dimensional measurement basis for constructing a discriminant function in the subsequent steps.

[0108] S542. Calculate the forward and backward relative entropy differences between the first probability density distribution function and the second probability density distribution function, and dynamically generate a weight factor according to the relative entropy differences.

[0109] It can be understood that the forward relative entropy difference refers to the difference between KL(P||Q) and KL(Q||P), and the backward relative entropy difference is the difference between KL(Q||P) and KL(P||Q). For example, when KL(P||Q) = 1.2 and KL(Q||P) = 0.8, the forward difference is 0.4 and the backward difference is -0.4. The generation of the dynamic weight factor can be achieved by setting threshold conditions: if the forward difference exceeds a preset threshold (such as 0.5), then assign a higher weight to the symmetric divergence; if the backward difference is significant, then increase the weight coefficient of the asymmetric divergence.

[0110] Exemplarily, a sigmoid function can be used to map the difference value to the interval (0, 1) as the weight. For example, the weight α = 1 / (1 + exp(-k*(D f -D t ))), where k is the slope parameter, D f is the forward relative entropy difference, D tis the difference threshold. This dynamic adjustment mechanism enables the discriminant function to adapt to the data distribution characteristics of different features. For example, for features with obvious directional differences (such as significantly older patients in a certain department), it strengthens the contribution of the asymmetric divergence, while for features with a high degree of distribution overlap, it relies on the symmetric divergence.

[0111] S543, weight the weight factor with the symmetric divergence and the asymmetric divergence for weighted fusion to generate a feature sensitivity discriminant function.

[0112] It can be understood that the symmetric divergence (JS) and the asymmetric divergence (KL) can be weighted and fused in the weighted fusion process by a linear combination method. For example, set the discriminant function as F = α * JS(P||Q)+(1 - α) * KL(P||Q), where α is the generated dynamic weight factor. When α approaches 1, the feature sensitivity discriminant function mainly reflects the symmetric characteristics of the distribution difference; when α approaches 0, it focuses on the asymmetric difference. The JS and KL can be normalized to avoid dimensional differences, such as dividing both by their historical maximum values or using z-score standardization. The finally generated discriminant function value can be directly used as the feature sensitivity score for sorting or threshold screening.

[0113] S550, integrate the feature sensitivity discriminant function in the multi-dimensional feature space to obtain a triage conflict coefficient, and determine the triage information of the target patient based on the triage conflict coefficient.

[0114] It can be understood that the triage conflict coefficient is an index obtained by quantifying the overall performance of the feature sensitivity discriminant function in the multi-dimensional feature space, which is used to measure the conflict degree of the feature combination of the target patient in the triage decisions of different departments. Each dimension in the multi-dimensional feature space corresponds to a patient feature (such as blood pressure, age, medical history index), and the value of the feature sensitivity discriminant function represents the discrimination ability of this feature in triage. By integrating the function values in the feature space, the triage uncertainty under the combined action of multiple features can be comprehensively evaluated. For example, if the discriminant function values of multiple features (such as high blood pressure, high blood sugar) of a certain patient in different departments vary greatly, the integration result (triage conflict coefficient) is higher, indicating that there is a greater conflict in the triage decision. The generation of triage information needs to combine the conflict coefficient with the department priority, and finally select the department with the least conflict as the recommended target.

[0115] Optionally, S550, integrate the feature sensitivity discriminant function in the multi-dimensional feature space to obtain a triage conflict coefficient, and determine the triage information of the target patient based on the triage conflict coefficient, including:

[0116] S551, define the department feature distribution and the integration domain range of each department in the multi-dimensional feature space; among them, the integration domain range is determined by the feature value distribution of historical triage instances.

[0117] It can be understood that the departmental feature distribution refers to the statistical distribution model of patient features in historical triage instances for each department. For example, the blood pressure and age of patients in the cardiology department may follow a Gaussian distribution with specific mean and variance, while the vital capacity and cough frequency of patients in the respiratory department may conform to a Poisson distribution. The integration domain range is the eigenvalue integration interval defined separately for each department, covering the eigenvalue distribution range of the historical cases of that department. For each department, based on its historical case data, probability distribution fitting (such as Gaussian mixture model, kernel density estimation) or non-parametric statistical methods (such as histogram binning) can be used to construct the probability density function of multi-dimensional features. For example, the blood pressure feature of the cardiology department may be fitted to a normal distribution with a mean of 130 mmHg and a standard deviation of 15, while the cough frequency of the respiratory department may be modeled as a Poisson distribution with λ = 3. For each feature dimension, the upper and lower limits of integration are determined according to the statistical characteristics of the department's historical data (such as 5%-95% quantiles, 3σ principle). For example, if the 95% quantile of the age of cardiology patients is 65 years old, the integration domain can be set as [18, 65], excluding extreme outliers (such as centenarians). For discrete features (such as disease codes), the integration domain range is the enumeration set of all possible values. The definition of the independence of the departmental feature distribution avoids the dilution of department-specific data by the global integration domain. For example, the height range of pediatric patients is significantly different from that of the orthopedics department. Defining the integration domain separately can improve the integration accuracy. The integration domain range driven by historical data ensures that the calculation focuses on the actual clinical scenario and avoids invalid calculations for impossible feature combinations (such as neonatal prostate-specific antigen detection).

[0118] S552, perform multiple integral operations on the feature sensitivity discriminant function within the department integration domain range according to the departmental feature distribution of each department to generate the triage conflict coefficient between the target patient and each department.

[0119] It can be understood that the triage conflict coefficient is a scalar value calculated independently for each department, indicating the degree of matching conflict between the target patient features and the historical case distribution of that department. For each department, within its exclusive integration domain, perform multiple integrals on the feature sensitivity discriminant function where represents the department number,[[]]END]] is the patient feature. The triage conflict coefficient can be calculated through the following integral formula: where is the integration domain of the th dimensional feature of the department. Randomly sample a large number of feature points within the department integration domain, calculate the mean value of the discriminant function values and multiply by the integration domain volume. For continuous features, polynomial approximation is used, and for discrete features, direct summation is performed. For the multi-department scenario, a distributed computing framework (such as Spark) can be used to synchronously calculate the conflict coefficients of each department.

[0120] S553. Select the corresponding department with the lowest triage conflict coefficient as the triage information for the target patient according to the triage conflict coefficient between the target patient and each department.

[0121] It can be understood that the triage conflict coefficient is a quantitative index of the matching conflict degree between the characteristic combination of the target patient and the historical case distribution of each department. The lower its value indicates the higher the matching degree between the patient characteristics and the department, and the lower the risk of abnormal triage. The triage conflict coefficients between the target patient and each department can be sorted in ascending order, and the department corresponding to the minimum value can be directly selected. If the conflict coefficients of multiple departments are the same (such as both the cardiology department and the neurology department are 0.3), dynamic decision-making can be carried out by reading and comparing the real-time load (number of patients waiting in line) and resource availability (equipment idle rate) of the departments from the department attribute data. For example, when the conflict coefficients are the same, the department with the shorter current waiting queue is preferentially recommended to optimize resource allocation.

[0122] Exemplarily, assuming that the conflict coefficient mapping of the target patient is {Cardiology: 0.2, Respiratory Medicine: 0.5, Orthopedics: 0.7}, then the cardiology department can be directly output as the triage information for the target patient; if the mapping is {Endocrinology: 0.3, Nephrology: 0.3}, the real-time load of the departments can be read and compared from the department attribute data. For example, the current number of patients waiting in line in the endocrinology department is 10 and that in the nephrology department is 5, then the nephrology department is output as the triage information for the target patient to balance the load. By combining the quantification of conflict intensity with dynamic rules, it is beneficial to make the triage result take into account accuracy, efficiency, clinical safety, achieve specific and precise triage based on real-time characteristics and historical data, and improve the patient's medical experience.

[0123] Corresponding to the intelligent triage method in the above embodiment, the embodiment of the present application also provides an intelligent triage system, and each unit of this system can implement each step of the intelligent triage method. Figure 3 The structural block diagram of the intelligent triage system provided by the embodiment of the present application is shown. For the sake of convenience of description, only the parts related to the embodiment of the present application are shown.

[0124] Refer to Figure 3 This intelligent triage system includes:

[0125] An acquisition unit, configured to acquire department attribute data and multimodal operation records of the target patient;

[0126] A fusion unit, configured to perform data feature fusion according to the multimodal operation records and the department attribute data to determine the triage adaptation feature group of the target patient;

[0127] A retrieval unit, configured to obtain a historical triage database, retrieve first triage instances with triage anomalies in the historical triage database, and obtain the number of first triage instances; wherein the historical triage database includes a number of historical triage instances, and each historical triage instance records the physical sign data of the patient and the department attribute data of the department where the patient seeks medical treatment.

[0128] A screening unit, configured to screen second triage instances associated with the triage adaptation feature group from all the first triage instances, and obtain the number of second triage instances; wherein the number of second triage instances is the triage instances with triage anomalies where the matching degree between the department attribute data of the department where the patient seeks medical treatment, the physical sign data of the patient and the triage adaptation feature group reaches a preset matching threshold.

[0129] A triage guidance unit, configured to construct a triage conflict coefficient according to the number of first triage instances and the number of second triage instances, and determine the triage guidance information of the target patient based on the triage conflict coefficient.

[0130] It should be noted that the information interaction, execution process, etc. between the above systems / units are based on the same concept as the method embodiments of the present application. For their specific functions and the technical effects brought, please refer to the method embodiment part specifically, and will not be elaborated here.

[0131] Those skilled in the art can clearly understand that for the convenience and conciseness of description, only the above division of each functional unit and module is used as an example. In actual applications, the above functions can be allocated to different functional units and modules according to needs, that is, the internal structure of the system is divided into different functional units or modules to complete all or part of the above-described functions. Each functional unit and module in the embodiment can be integrated in a processing unit, or each unit module exists physically alone, or two or more unit modules are integrated in one unit. The above integrated unit can be implemented in the form of hardware or in the form of a software functional unit. In addition, the specific names of each functional unit and module are only for the convenience of mutual distinction and do not limit the protection scope of the present application. The specific working process of the units and modules in the above system can refer to the corresponding process in the foregoing method embodiment, and will not be elaborated here.

[0132] The embodiment of the present application also provides an intelligent triage robot. Figure 4 It is a schematic structural diagram of the intelligent triage robot provided by an embodiment of the present application. As Figure 4 shown, the intelligent triage robot 6 of this embodiment includes: at least one processor 60 ( Figure 4 only one is shown here), at least one memory 61 ( Figure 4(only one is shown in the figure) and a computer program 62 stored in the at least one memory 61 and executable on the at least one processor 60. When the processor 60 executes the computer program 62, the intelligent diagnosis guiding robot 6 implements the steps in any of the above-mentioned intelligent diagnosis guiding method embodiments, or the intelligent diagnosis guiding robot 6 implements the functions of each unit in the above-mentioned system embodiments.

[0133] Exemplarily, the computer program 62 can be divided into one or more units. The one or more units are stored in the memory 61 and executed by the processor 60 to complete this application. The one or more units can be a series of computer program instruction segments capable of completing specific functions, and these instruction segments are used to describe the execution process of the computer program 62 in the intelligent diagnosis guiding robot 6.

[0134] For example, the intelligent diagnosis guiding robot can be various types of robots. The intelligent diagnosis guiding robot may include, but is not limited to, a processor 60 and a memory 61. Those skilled in the art can understand that Figure 4 merely an example of the intelligent diagnosis guiding robot 6, which does not constitute a limitation on the intelligent diagnosis guiding robot 6. It may include more or fewer components than shown in the figure, or combine certain components, or different components. For example, it may also include input / output devices, network access devices, buses, etc.

[0135] The processor 60 can be a central processing unit (CPU), and the processor 60 can also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0136] The memory 61 may be an internal storage unit of the intelligent medical guide robot 6 in some embodiments, such as the hard disk or memory of the intelligent medical guide robot 6. The memory 61 may also be an external storage device of the intelligent medical guide robot 6 in some other embodiments, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, etc. equipped on the intelligent medical guide robot 6. Further, the memory 61 may also include both the internal storage unit and the external storage device of the intelligent medical guide robot 6. The memory 61 is used to store an operating system, application programs, a boot loader, data, and other programs, such as the program code of the computer program. The memory 61 may also be used to temporarily store data that has been output or will be output.

[0137] An embodiment of the present application further provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps in any of the above method embodiments are implemented.

[0138] An embodiment of the present application provides a computer program product, and when the computer program product runs on the intelligent medical guide robot, the intelligent medical guide robot implements the steps in any of the above method embodiments.

[0139] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above method embodiments of the present application, a computer program can be used to instruct relevant hardware to complete. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps in the above method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, an executable file, or some intermediate form, etc. The computer-readable medium may at least include: any entity or device capable of carrying the computer program code to the intelligent medical guide robot, a recording medium, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunication signal, and a software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk, or an optical disc, etc. In some jurisdictions, according to legislation and patent practice, the computer-readable medium may not be an electrical carrier signal and a telecommunication signal.

[0140] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For parts not described or recorded in a certain embodiment, reference may be made to the relevant descriptions of other embodiments.

[0141] Those of ordinary skill in the art can realize that the units and algorithm steps of each example described in combination with the embodiments disclosed herein can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. A professional technician can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0142] In the embodiments provided in this application, it should be understood that the disclosed intelligent diagnosis guidance system / intelligent diagnosis guidance robot and method can be implemented in other ways. For example, the intelligent diagnosis guidance system / intelligent diagnosis guidance robot embodiments described above are merely illustrative. For example, the division of the units is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed couplings or direct couplings or communication connections to each other can be through some interfaces, and the indirect couplings or communication connections of devices or units can be in electrical, mechanical or other forms.

[0143] The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0144] The above-described embodiments are only used to illustrate the technical solutions of this application, rather than to limit them; although this application has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that they can still modify the technical solutions recorded in the foregoing embodiments, or perform equivalent replacements for some of the technical features; and these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of each embodiment of this application, and should all be included in the protection scope of this application.

Claims

1. An intelligent diagnosis guidance method, characterized in that, Including: Obtain department attribute data and multimodal operation records of the target patient; Perform data feature fusion based on the multimodal operation records and the department attribute data to determine a triage adaptation feature group of the target patient; Obtain a historical triage database, retrieve first triage instances with triage anomalies in the historical triage database, and obtain the number of first triage instances; wherein the historical triage database includes a number of historical triage instances, and each historical triage instance records the physical sign data of the patient and the department attribute data of the department where the patient seeks medical treatment; Screen second triage instances associated with the triage adaptation feature group from all the first triage instances to obtain the number of second triage instances; wherein the number of second triage instances is the triage instances with triage anomalies where the matching degree between the department attribute data of the department where the patient seeks medical treatment, the physical sign data of the patient and the triage adaptation feature group reaches a preset matching threshold; Construct a triage conflict coefficient based on the number of first triage instances and the number of second triage instances, and determine the guiding information of the target patient according to the triage conflict coefficient; wherein the triage conflict coefficient is a scalar value calculated independently for each department, indicating the matching conflict degree between the characteristics of the target patient and the historical case distribution of the department; Among them, the screening of second triage instances associated with the triage adaptation feature group from all the first triage instances to obtain the number of second triage instances includes: Construct department attribute data matching conditions and physical sign data correlation degree constraint conditions according to the triage adaptation feature group of the target patient; Screen among all the first triage instances based on the department attribute data matching conditions and the physical sign data correlation degree constraint conditions to obtain second triage instances, and count the number of the second triage instances to obtain the number of second triage instances.

2. The intelligent diagnosis guiding method according to claim 1, wherein The performing data feature fusion based on the multimodal operation records and the department attribute data to determine the triage adaptation feature group of the target patient includes: Perform temporal alignment and semantic parsing on the multimodal operation records, and extract operation behavior feature vectors and abnormal operation marks; Construct a department professional ability graph based on the department attribute data; wherein the department professional ability graph includes three-dimensional feature encodings of the historical case processing success rate; Perform cross-modal association mapping between the operation behavior feature vectors and the department professional ability graph to generate a behavior-department association matrix; Perform weight assignment on the abnormal operation marks and the three-dimensional feature encodings through a dynamic attention mechanism, and calculate the feature coupling degree; According to the behavior-department association matrix and the feature coupling degree, perform non-linear fusion on the multimodal operation record features and the department attribute features to generate the triage adaptation feature group of the target patient.

3. The intelligent diagnosis guidance method according to claim 2, wherein The performing cross-modal association mapping between the operation behavior feature vectors and the department professional ability graph to generate a behavior-department association matrix includes: Perform dimension expansion on the operation behavior feature vectors, and project the operation behavior feature vectors into the same feature space as the department professional ability graph; Based on the mapping relationships in the preset symptom - department association rule base, calculate the dynamic semantic similarity between each of the operation behavior feature vectors and the nodes of the department professional ability graph; Perform piece - wise normalization on the dynamic semantic similarity to generate a standardized association weight; Construct an N×M sparse association matrix according to the standardized association weight, and perform temporal noise reduction on the sparse association matrix to generate a behavior - department association matrix; where N is the total number of dimensions of the operation behavior feature vectors, and M is the total number of nodes of the department professional ability graph.

4. The intelligent medical guidance method according to claim 1, characterized in that, The obtaining the historical triage database, retrieving the first triage instances with triage anomalies in the historical triage database, and obtaining the number of the first triage instances includes: Obtain the historical triage database; Retrieve the historical triage database according to the preset anomaly determination criteria, obtain the first triage instances with triage anomalies in the historical triage database, and count the number of the first triage instances to obtain the number of the first triage instances.

5. The intelligent diagnosis guiding method according to claim 1, wherein The constructing a triage conflict coefficient according to the number of the first triage instances and the number of the second triage instances, and determining the triage information of the target patient based on the triage conflict coefficient includes: Obtain the number of triage instances in the historical triage database; Map the number of the first triage instances and the number of triage instances to a multi - dimensional feature space to generate a first probability density distribution function corresponding to the number of the first triage instances; Map the number of the second triage instances and the number of triage instances to a multi - dimensional feature space to generate a second probability density distribution function corresponding to the number of the second triage instances; Calculate the relative entropy ratio of the first probability density distribution function and the second probability density distribution function to generate a feature sensitivity discriminant function; Integrate the feature sensitivity discriminant function over the multi - dimensional feature space to obtain a triage conflict coefficient, and determine the triage information of the target patient based on the triage conflict coefficient.

6. The intelligent diagnosis guidance method according to claim 5, wherein The calculating the relative entropy ratio of the first probability density distribution function and the second probability density distribution function to generate a feature sensitivity discriminant function includes: Determine the mixed distribution of the first probability density distribution and the second probability density distribution, and calculate the symmetric divergence and the asymmetric divergence between the first probability density distribution and the second probability density distribution according to the mixed distribution; Calculate the forward and reverse relative entropy differences between the first probability density distribution function and the second probability density distribution function, and dynamically generate a weight factor according to the relative entropy difference; Perform weighted fusion of the weight factor with the symmetric divergence and the asymmetric divergence to generate a feature sensitivity discriminant function.

7. The intelligent medical guidance method according to claim 5, wherein The integrating the feature sensitivity discriminant function over the multi - dimensional feature space to obtain a triage conflict coefficient, and determining the triage information of the target patient based on the triage conflict coefficient includes: Define the department feature distribution and the integration domain range of each department in the multi - dimensional feature space; where the integration domain range is determined by the feature value distribution of historical triage instances; Performing multiple integral operations on the feature sensitivity discriminant function within the department integral domain according to the department feature distribution of each department to generate a triage conflict coefficient between the target patient and each department; According to the triage conflict coefficient between the target patient and each department, selecting the corresponding department with the lowest triage conflict coefficient as the triage guidance information for the target patient.

8. An intelligent medical guidance system, characterized in that, For implementing the method according to any one of claims 1 to 7, the intelligent triage system includes: An acquisition unit, configured to acquire department attribute data and multi-modal operation records of a target patient; A fusion unit, configured to perform data feature fusion according to the multi-modal operation records and the department attribute data to determine a triage adaptation feature group of the target patient; A retrieval unit, configured to acquire a historical triage database, retrieve first triage instances with triage anomalies in the historical triage database, and obtain the number of first triage instances; wherein the historical triage database includes a plurality of historical triage instances, and each historical triage instance records the physical sign data of the patient and the department attribute data of the department where the patient seeks medical treatment; A screening unit, configured to screen second triage instances associated with the triage adaptation feature group from all the first triage instances, and obtain the number of second triage instances; wherein the number of second triage instances is the number of triage instances with triage anomalies where the matching degree between the department attribute data of the department where the patient seeks medical treatment, the physical sign data of the patient and the triage adaptation feature group reaches a preset matching threshold; A triage guidance unit, configured to construct a triage conflict coefficient according to the number of first triage instances and the number of second triage instances, and determine the triage guidance information of the target patient based on the triage conflict coefficient.

9. An intelligent medical guidance robot, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

Images