A method and device for constructing computerized clinical guidelines based on graphical representation

Abstract

The present invention discloses a method and device for constructing a computerized clinical guideline based on graphical representation, which is applied to a computerized clinical guideline model. The computerized clinical guideline model includes: a data layer, a model layer and an application layer. The method includes: obtaining text input in a clinical guideline database at the data layer, determining clinical data corresponding to the text input, and determining a target text input based on the text input and the clinical data; converting the target text input into a graphical diagnosis and treatment process at the model layer; and determining a decision recommendation result based on the graphical diagnosis and treatment process at the application layer. The decision recommendation result includes: at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction. In the above process, the diagnosis and treatment process is displayed in a graphical manner, and the determination process of the diagnosis and treatment process can be realized without the medical staff mastering the relevant language. The requirements for medical staff are relatively low, which is conducive to the dissemination and application of clinical guideline methods among medical staff.

Description

Technical Field

[0001] The present invention relates to the field of data processing technology, and in particular to a method and device for constructing computerized clinical guidelines based on graphical representation. Background Art

[0002] With the continuous advancement of medical informationization and diagnosis and treatment standardization, the use of computer technology to construct computer-understandable computerized guidelines from authoritative clinical guideline texts has become a core module in clinical decision support systems. Computerized guidelines can represent clinical knowledge and reasoning processes using computer logic, which can help clinical medical staff make diagnostic decisions and is an important method to improve the level of primary medical care.

[0003] The process from text-based clinical guidelines to a knowledge representation system for computerized guidelines is a difficult one. From the text side of clinical guidelines, standard medical terminology representation and reasoning rules are needed to facilitate information engineers to understand and build a reasonable and accurate computerized guideline representation model. From the guideline representation model presentation side, it is necessary to design concise and easy-to-understand logical statements and clinical guideline presentation methods to make the clinical decision support results based on the guideline representation model interpretable.

[0004] The current mainstream computerized clinical guideline representation models include GEM, GDL, etc. The platforms developed based on these clinical guideline models include LinkEHR, nedap, Template Designer, etc. Although they can accurately represent clinical knowledge in a logical rule language, medical staff need to master the language corresponding to the platform to realize the computerized representation of clinical guidelines, which is not conducive to the dissemination and use among medical staff. Summary of the invention

[0005] In view of this, the present invention provides a method and device for constructing computerized clinical guidelines based on graphical representation, which is used to solve the problem that the current mainstream computerized clinical guideline representation models include GEM, GDL, etc., and the platforms developed based on these clinical guideline models include LinkEHR, nedap, Template Designer, etc. Although they can accurately represent clinical knowledge in a logical rule language, medical staff need to master the language corresponding to the platform to implement clinical guidelines, which is not conducive to the dissemination and use among medical staff. The specific scheme is as follows:

[0006] A method for constructing a computerized clinical guideline based on graphical representation is applied to a computerized clinical guideline model, wherein the computerized clinical guideline model includes: a data layer, a model layer and an application layer, and the method includes:

[0007] Acquire text input in a clinical guideline database at the data layer, determine clinical data corresponding to the text input, and determine a target text input based on the text input and the clinical data;

[0008] Converting the target text input into a graphical diagnosis and treatment process at the model layer;

[0009] A decision recommendation result is determined at the application layer based on the graphical diagnosis and treatment process, wherein the decision recommendation result includes at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction.

[0010] The above method, optionally, determines the target text input based on the text input and the clinical data, including:

[0011] Integrating the text input with corresponding items in the clinical data to obtain a first text input;

[0012] Performing a terminology standardization operation on the first text input to obtain a second text input;

[0013] Perform text annotation on the second text input to obtain a target text input.

[0014] The above method, optionally, converts the target text input into a graphical diagnosis and treatment process at the model layer, including:

[0015] Converting the target text input into decision condition variables and logic statements;

[0016] Converting the decision condition variables into graphical variables;

[0017] The logical statements and the graphical variables are passed to the disease clinical guideline representation model to obtain a graphical diagnosis process.

[0018] The above method may optionally further include:

[0019] The clinical guideline representation model for the disease was validated.

[0020] The above method, optionally, determines the decision recommendation result based on the graphical diagnosis and treatment process at the application layer, including:

[0021] Perform clinical decision support path retrieval in the clinical guideline representation model based on the graphical diagnosis and treatment process;

[0022] Based on the search result, the decision recommendation result is determined.

[0023] A computerized clinical guideline construction device based on graphical representation is applied to a computerized clinical guideline model, wherein the computerized clinical guideline model includes: a data layer, a model layer and an application layer, and the device includes:

[0024] an acquisition and determination module, configured to acquire text input in a clinical guideline database at the data layer, determine clinical data corresponding to the text input, and determine a target text input based on the text input and the clinical data;

[0025] A conversion module, used for converting the target text input into a graphical diagnosis and treatment process at the model layer;

[0026] A determination module is used to determine a decision recommendation result based on the graphical diagnosis and treatment process at the application layer, wherein the decision recommendation result includes at least one of: diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction.

[0027] In the above device, optionally, the acquisition and determination module includes:

[0028] an integration unit, configured to integrate the text input with corresponding items in the clinical data to obtain a first text input;

[0029] A standardization unit, configured to perform a term standardization operation on the first text input to obtain a second text input;

[0030] The annotation unit is used to perform text annotation on the second text input to obtain a target text input.

[0031] In the above device, optionally, the conversion module comprises:

[0032] A first conversion unit, used for converting the target text input into decision condition variables and logic statements;

[0033] A second conversion unit, used for converting the decision condition variable into a graphical variable;

[0034] The first determining unit is used to pass the logical statement and the graphical variable to the disease clinical guideline representation model to obtain a graphical diagnosis process.

[0035] The above device may optionally further include:

[0036] A verification unit is used to verify the disease clinical guideline representation model.

[0037] In the above device, optionally, the determining module includes:

[0038] A retrieval unit, configured to perform clinical decision support path retrieval in the clinical guideline representation model based on the graphical diagnosis and treatment process;

[0039] The second determining unit is used to determine the decision recommendation result based on the search result.

[0040] Compared with the prior art, the present invention has the following advantages:

[0041] The present invention discloses a method and device for constructing a computerized clinical guideline based on graphical representation, which is applied to a computerized clinical guideline model, wherein the computerized clinical guideline model includes: a data layer, a model layer and an application layer, and the method includes: obtaining text input in a clinical guideline database at the data layer, determining clinical data corresponding to the text input, and determining a target text input based on the text input and the clinical data; converting the target text input into a graphical diagnosis and treatment process at the model layer; determining a decision recommendation result based on the graphical diagnosis and treatment process at the application layer, wherein the decision recommendation result includes: at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction. In the above process, the determination process of the diagnosis and treatment process can be realized without medical staff mastering the relevant language, and the requirements for medical staff are relatively low, which is conducive to the dissemination and application of clinical guideline methods among medical staff. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0043] Figure 1 A flowchart of a method for constructing a computerized clinical guideline based on graphical representation disclosed in an embodiment of the present application;

[0044] Figure 2 A schematic diagram of a computerized clinical guideline model disclosed in an embodiment of the present application;

[0045] Figure 3 A schematic diagram of a diagnosis and treatment decision process disclosed in an embodiment of the present application;

[0046] Figure 4 A schematic diagram of an automated disease typing process disclosed in an embodiment of the present application;

[0047] Figure 5 A schematic diagram of a treatment plan recommendation process disclosed in an embodiment of the present application;

[0048] Figure 6 A schematic diagram of a complication prediction process disclosed in an embodiment of the present application;

[0049] Figure 7 This is a structural block diagram of a computerized clinical guideline construction device based on graphical representation disclosed in an embodiment of the present application. DETAILED DESCRIPTION

[0050] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0051] The above description of the disclosed embodiments enables one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to the embodiments shown herein, but rather to the widest scope consistent with the principles and novel features disclosed herein.

[0052] The present invention discloses a method and device for constructing computerized clinical guidelines based on graphical representation, which is used to convert unstructured clinical practice guidelines into a computer-interpretable disease clinical guideline identification model that can be verified, and can generate corresponding code files, which can be embedded in a clinical decision support system to achieve at least one of diagnosis and treatment decision support, automated disease typing, treatment plan recommendation and complication prediction methods based on the disease clinical guideline representation model, wherein the clinical practice guidelines CPG (Clinical Practical Guidelines) refer to multiple groups of clinical guidelines developed by the system, which help doctors and patients make appropriate treatment, selection, and decision-making health care services for specific clinical problems, and are used to improve medical quality and control medical expenses. Computer-interpretable guidelines CIGs (Computer-interpretable guidelines) refer to the use of computer logic to represent the knowledge and reasoning relationships in clinical guidelines, and can be integrated and embedded in the clinical diagnosis and treatment information system, and can provide a reference knowledge source for decision-making. Clinical decision support CDS (Clinical Decision Support) refers to the use of computer field knowledge to assist medical personnel in completing clinical decision-making work. Clinical decision support mainly includes knowledge-based clinical decision support and data-based clinical decision support. The knowledge-based clinical decision support construction process is to establish a knowledge base for the diagnosis and treatment of related diseases and make clinical decisions based on the knowledge base; data-based clinical decision support uses artificial intelligence, probability statistics and other methods to enable computers to automatically make decisions.

[0053] The method is applied to a computerized clinical guideline model, wherein the computerized clinical guideline model includes: a data layer, a model layer and an application layer. The execution process of the method is as follows: Figure 1 The steps shown include:

[0054] S101, obtaining text input in a clinical guideline database at the data layer, determining clinical data corresponding to the text input, and determining a target text input based on the text input and the clinical data;

[0055] In an embodiment of the present invention, the computerized clinical guideline model is as follows: Figure 2As shown, the data layer includes: data acquisition module, data integration module, terminology standardization module and text annotation module, wherein the data acquisition module is completed using ETL (Extraction-Transformation-Loading) tools, wherein ETL is used for data extraction, transformation and loading. ETL is responsible for extracting data from distributed and heterogeneous data sources such as relational data and flat data files to a temporary intermediate layer for cleaning, transformation and integration, and finally loading it into a data warehouse or data mart, becoming the basis for online analytical processing and data mining. Based on the ETL, the clinical guideline text input in the clinical guideline database is obtained, and the clinical data corresponding to the text input is determined, wherein the determination process can be implemented based on an association relationship or an identifier. The specific determination process is not limited in the embodiment of the present invention. In the embodiment of the present invention, the text input submitted by the user or the clinical data input in the hospital information system HIS is passed to the middle layer for cleaning and integration, and then a result report PDF file containing sample information, decision support results, decision support evidence, decision support processes and rules, and related basis is returned according to the application requirements submitted by the user and stored in a temporary database. The ultimate goal of the above process is to be used for clinical decision support. Different application requirements are also different types of clinical decision support, and the corresponding files required for feedback for different clinical decision support scenarios are also different. For clinical diagnosis and treatment process support, what needs to be fed back is the diagnosis and treatment process roadmap after the integrated input data; what needs to be fed back for automated disease typing is the typing results, as well as the inference rules corresponding to the typing results.

[0056] When the user or HIS submits a download application, the ETL tool is used to export and complete the interaction; in the data integration module, the system defines eight types of clinical data, including demographic information, physical signs, examination results, test results, diagnosis, symptoms, medication orders, and other medical orders. According to the data type requirements of the clinical guideline representation model for a specific disease, the ETL tool is used to extract and integrate the above eight types of data from the HIS to obtain the first text input. In the terminology standardization module, the embodiment of the present invention uses the Chinese medical subject word list CMeSH, the OMAHA terminology set, and the "Common Clinical Medical Terms (2019 Edition)" to standardize the first text input to obtain the second text input. At the same time, the graphical guideline representation model from the model layer is automatically standardized and returned to the model editor for confirmation, so that the graphical guideline representation model constructed using the present invention has high authority and versatility. Among them, the Chinese medical subject word list CMeSH: translated and published by the Institute of Medical Information, Chinese Academy of Medical Sciences, includes: the National Library of Medicine's "Medical Subject Word List", "Chinese Traditional Medicine Subject Word List", and "Chinese Library Classification-Medical Professional Classification List".

[0057] Furthermore, the second text input is annotated to obtain the target text input, wherein the annotation principle can be set based on experience or specific circumstances, and is not specifically limited in the embodiment of the present invention.

[0058] S102, converting the target text input into a graphical diagnosis and treatment process at the model layer;

[0059] In the embodiment of the present invention, Figure 2 As shown, the model layer includes a rule organization module, a graphical representation module, a model building module and a model testing module. The rule organization module uses logical statements to process the annotated text. For example, "and" and "or" are mapped to "and" and "or". At the same time, each logical statement is connected with a decision condition (IF-THEN structure). The specific method is to assign a node ID to each logical statement, record the relationship with other node IDs, and ensure that each node can be connected to no less than two other nodes to represent the previous and next relationship. The connection between the two node IDs is the decision condition, and the output of the rule organization module is the node information containing the decision condition variables and the logical statement; in the graphical representation module, the present invention defines four atomic type nodes (Atomic Node), a composite type node (CompositeNode) and two connecting lines (Line) for describing clinical diagnosis and treatment pathways and representing clinical guideline knowledge. Atomic type nodes include:

[0060] 1. Explanation Node: Ex_xx indicates that it is used to describe the background knowledge that is not recorded in the clinical guidelines but belongs to clinical experience or clinical guidelines, so as to assist information engineers and medical personnel to better understand the graphical computerized guideline representation model;

[0061] 2. Data acquisition node (Enquiry Node): En_xx represents the node that connects to external data, including patient narration, electronic medical record system interface, doctor input data, and external knowledge base, etc.;

[0062] 3. Decision Node: De_nn represents the decision point in the clinical guidelines, which can correspond to the diagnostic reasoning process in the clinical guidelines by setting decision conditions;

[0063] 4. Action Node: Ac_xx is used to represent and model the guidance actions in clinical guidelines; composite nodes are represented by Co_xx, which are used to represent other clinical guideline knowledge that cannot be classified into the four atomic type nodes;

[0064] The connection cable includes:

[0065] 1. Sequence Line: It is used to connect sequential structures, indicating that the upper and lower nodes have a sequential relationship and can be executed in sequence;

[0066] 2. Decision Line: Li_xx indicates that starting from the decision node, it is connected to other different nodes according to the decision conditions to represent the clinical decision path in the clinical guidelines.

[0067] For each connection line, a functional logic statement is used to represent the reasoning relationship between the upper and lower nodes. The available logic judgment types include exist, find, or, etc., and support the AND / OR calculation of multiple logic statements. By defining a graphical identifier for each type of clinical knowledge and reasoning rule, the decision-making conditions can be intuitively and vividly converted into graphical variables, which has strong interpretability.

[0068] The model building module builds and saves the disease clinical guideline representation model for the model editor through a graphical user interface (GUI), and passes the logical statements and the graphical variables to the disease clinical guideline representation model to obtain a graphical diagnosis process. If subsequent users have objections to the current guideline representation model, they can contact the model editor according to the guideline note information and submit a model modification request to the graphical computerized clinical guideline representation model construction platform manager to obtain model modification permissions, so that the constructed graphical guideline representation model has flexibility and modifiability; in addition, in order to improve the traceability and authority of the computerized clinical guideline representation model, the present invention also defines the note information of the standardized clinical guideline representation model, including editor information (editor name, email address, unit, editing time), description (clinical guideline source and details, guideline representation model construction purpose, when it can be used, when it cannot be used, references), and text clinical guideline information (uploaded PDF file); in the model testing module, the disease clinical guideline representation model is input to test the terminology standardization, decision path connectivity, data interactivity and other issues using the test samples provided by the model editor or the virtual data automatically generated by the computer to ensure the usability of the graphical disease clinical guideline representation model.

[0069] S103. Determine, at the application layer, a decision recommendation result based on the graphical diagnosis and treatment process, wherein the decision recommendation result includes at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation, and complication prediction.

[0070] In the embodiment of the present invention, Figure 2 As shown in Figure 1, the application layer includes: diagnosis and treatment decision support, automated disease classification, treatment plan recommendation and complication prediction. Figure 3As shown, a data acquisition node is determined based on the disease clinical guideline representation model, a decision node is determined based on the data acquisition node, and whether the decision node meets the decision conditions, wherein the decision conditions can be set based on experience or specific circumstances, and are not specifically limited in the embodiments of the present invention. If the decision conditions are met, an action node 1 is generated, and a composite node is determined based on the action node 1, and the process ends; if the decision conditions are not met, an action node 2 is generated, and a composite node is determined based on the action node 1, and a complete inspection is performed, and the process ends; for the diagnosis and treatment decision support process, a clinical decision support path retrieval is performed in the graphical clinical guideline constructed by the present invention according to the input clinical data, and the final result is represented by outputting the medical examination items that should be completed at present and supplementing the relevant clinical data. Preferably,

[0071] GDL (Guideline Definition Language) is a guideline representation model based on openEHR. It is constructed by serially combining single rules represented as "when-then" and can be used to express complex clinical decisions.

[0072] Furthermore, for disease classification, Figure 4 As shown, a data acquisition node is determined based on a disease clinical guideline representation model, and a decision node 1 and a decision node 2 are determined based on the data acquisition node. For the decision node 1, it is determined whether the decision node 1 meets the decision condition, wherein the decision condition can be set based on experience or specific circumstances, and is not specifically limited in the embodiment of the present invention. If the decision node 1 meets the decision condition, an action node 1 is generated, and if the decision node 1 does not meet the decision condition, an action node 2 is generated. For decision node 1, if it meets the decision condition, a decision node 3 is generated, and if the decision node 3 meets the decision condition, an action node 3 is generated. The action node 1, the action node 2 and the action node 3 generate classification 1, classification 2, classification 3 and classification 4 based on the disease classification decision node 4, and finally end. For the disease classification process, according to the complete clinical data input, the graphical clinical guideline representation model constructed by the present invention is used to perform clinical decision support path retrieval, and the final result is represented by outputting the disease classification result obtained by judging the clinical guideline representation model, and returning the clinical decision support path and related decision reasoning process based on the graphical guideline representation.

[0073] Further, the treatment options are recommended as follows: Figure 5As shown, based on the disease clinical guideline identification model, a data acquisition node is determined, and decision node 1 and decision node 2 are determined based on the data acquisition node. For decision node 1, it is judged whether the decision node 1 meets the decision condition, wherein the decision condition can be vital signs, complications and allergic history, etc., which are not specifically limited in the embodiment of the present invention. If the decision node 1 meets the decision condition, treatment plan 1 is determined based on the decision node 1. If the decision node 1 does not meet the decision condition, treatment plan 1 is determined based on the decision node 1. For the decision node 2, it is judged whether the decision node 2 meets the decision condition. If the decision node 2 does not meet the decision condition, treatment plan 1 is determined based on the decision node 1. Determine treatment plan 2, if the decision node 2 meets the decision condition, determine decision node 3 based on the decision node 2, judge whether the decision node 3 meets the decision condition, if the decision node 3 meets the decision condition, determine treatment plan 3 based on the decision node 3, if the decision node 3 does not meet the decision condition, determine treatment plan 4 based on the decision node 3, and finally end. For the treatment plan recommendation process, according to the input complete clinical data and disease classification results, the graphical clinical guideline representation constructed by the present invention is used to perform clinical decision support path retrieval, and the final result is represented as the output of the treatment plan recommendation result based on the graphical guideline representation model, including drug, instrument and surgical treatment recommendations. This application method can be used multiple times along with the medical consultation link, and each time it is used, the patient's clinical data is modified and entered to obtain a dynamic treatment plan.

[0074] Furthermore, for the prediction of complications, Figure 6As shown, a data acquisition node is determined based on a disease clinical guideline representation model 1, and the data acquisition node determines a data acquisition node based on a disease clinical guideline representation model 2 that meets the decision conditions through a matching degree decision node. For the data acquisition node determined based on the disease squamous guideline representation model, decision nodes 1 and 2 are continued to be determined. For the decision node 1, it is determined whether the decision node 1 meets the decision conditions, wherein the decision conditions can be set based on experience or specific circumstances, and are not specifically limited in the embodiments of the present invention. If the decision conditions are met, type 1 is determined based on the decision node 1; if the decision conditions are not met, type 2 is determined based on the decision node 1; for the decision node 2, if the decision conditions are met, decision node 3 is determined based on the decision node 2; if the decision node 3 is loaded with decision conditions, type 3 is determined based on the decision node 3. , for the data acquisition node determined based on the disease clinical guideline representation model, continue to determine decision node 4 and decision node 5 based on the data acquisition node, for the decision node 4, if the decision node 4 meets the decision condition, determine the decision node 6 based on the decision node 4, if the decision node 6 meets the decision condition, determine the complication type 1 based on the decision node 3, if the decision node 4 does not meet the decision condition, determine the complication type 2 based on the decision node 4, for the decision node 5, if the decision node 5 meets the toilet cleaning condition, determine the complication type 3 based on the decision node 5, and finally end. For the complication prediction process, according to the input complete clinical data, automatically perform clinical decision support path retrieval in all graphical clinical guideline representation models constructed using the present invention, and the input is divided into two parts: 1. Main disease classification results (such as Figure 2 2. Complication prediction results. In the retrieval stage of other graphical guideline representation models, the input data is matched with the input data required by other guideline representation models, and only the guideline representation models with a high degree of matching are fuzzy searched, and other possible disease classifications are returned as complication prediction results.

[0075] In the embodiment of the present invention, for the above four recommendation results, at least one of them can be selected for recommendation based on specific circumstances, and no specific limitation is made in the embodiment of the present invention.

[0076] The present invention discloses a method for constructing a computerized clinical guideline based on graphical representation, which is applied to a computerized clinical guideline model, wherein the computerized clinical guideline model includes: a data layer, a model layer and an application layer, and the method includes: obtaining text input in a clinical guideline database at the data layer, determining clinical data corresponding to the text input, and determining a target text input based on the text input and the clinical data; converting the target text input into a graphical diagnosis and treatment process at the model layer; and determining a decision recommendation result based on the graphical diagnosis and treatment process at the application layer, wherein the decision recommendation result includes: at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction. In the above process, the determination process of the diagnosis and treatment process can be realized without medical staff mastering the relevant language, and the requirements for medical staff are relatively low, which is conducive to the dissemination and application of clinical guideline methods among medical staff.

[0077] Since the existing computerized guideline representation models only focus on the knowledge representation of clinical guidelines and ignore the standardized representation and detailed description of clinical diagnosis and treatment processes, it is not conducive to information engineers' understanding of clinical guideline knowledge, resulting in certain deviations in the constructed computerized guideline representation models, and is not conducive to the actual application of computerized clinical guideline representation models in clinical practice; in addition, most of the existing computerized guideline representation models can only be used within dedicated systems, and although some computerized guideline representation models can interact with HIS systems, they contain little clinical information and are not standardized, resulting in poor interactivity with medical information systems and low versatility; moreover, the existing computerized guideline representation models are mostly represented by complex computer logic languages, which is not conducive to clinical workers' correction and modification of computerized guideline representation models, and such computerized guideline representation models are difficult to display intuitively and have poor interpretability.

[0078] In view of the above problems, the present invention proposes a universal graphical evidence-based computerized clinical guideline representation model to solve the shortcomings of other computerized clinical guideline representation models, such as ignoring the diagnosis and treatment process, poor generality and interactivity, and difficulty in intuitive representation. Among them, the computerized clinical guideline representation model uses graphical identifiers to represent clinical knowledge, clinical diagnosis and treatment paths, and computerized clinical guideline representation models of reasoning rules. The graphical identifiers include nodes and connecting lines, corresponding to clinical knowledge, clinical recommended actions, clinical data acquisition, and reasoning decisions. The clinical guideline representation model is a graphical decision roadmap containing start and end points, which can be used to make decision reasoning based on given clinical data. The sample data types and standardized guideline representation model remark information set in the computerized guideline representation model, wherein the sample data types are standardized through medical terminology sets and assisted correction is performed during the model construction, testing, and use stages. The graphical computerized clinical guideline representation model is used to support clinical diagnosis and treatment process decisions. The guideline representation model constructed by extracting clinical diagnosis and treatment knowledge from text clinical guidelines can give the diagnosis and treatment steps that need to be improved at present according to given clinical data. The graphical computerized clinical guideline representation model is used for automated disease typing. A guideline representation model is constructed by extracting clinical disease classification knowledge from text clinical guidelines, and the disease classification results are obtained by inference through the guideline representation model based on given clinical data. A graphical computerized clinical guideline representation model is used to recommend treatment plans. A guideline representation model is constructed by extracting clinical disease classification knowledge from text clinical guidelines, and the recommended treatment plan for the current sample is obtained by inference through the guideline representation model based on given clinical data and disease classification results. It can be used in conjunction with automated disease classification using a graphical computerized clinical guideline representation model. Complication prediction is performed using a graphical computerized clinical guideline representation model. Clinical disease classification knowledge is extracted from text clinical guidelines to construct guideline representation models for multiple diseases. First, the target disease classification results are obtained based on the given data using technical point 4, and then the given data is matched with the input data type in other disease guideline representation models. In the guideline representation model with a higher degree of matching, technical point 4 is used to obtain the predicted complication classification results.

[0079] Furthermore, for the above-mentioned processing, similar type nodes and connecting lines can also be used to construct a graphical clinical guideline representation model. It is also possible to use logical expressions with similar functions rather than graphical identifiers to complete the construction of the clinical guideline representation model. Existing computerized guideline representation models such as GEM and GDL can also be used for automated disease typing and treatment plan recommendations. It is also possible to use probability statistics-based methods to learn from previously known sample data and use them for complication prediction. It is also possible to use machine learning, deep learning and other methods to construct a computerized guideline representation model from clinical guideline texts. It is also possible to use knowledge graph-based methods to construct a computerized clinical guideline representation model.

[0080] The embodiment of the present invention constructs a graphical clinical guideline representation method, using different types of atomic type nodes and connecting lines to represent clinical guideline knowledge and paths. For the connecting lines, logical statements are constructed to reflect association rules. This method can intuitively display the clinical decision support path based on the guidelines, and has strong interpretability; at the same time, the method can also be mapped to the actual diagnosis and treatment process, which is consistent with clinical practice. In addition, the present invention uses commonly used fields in electronic medical records as node data types, and the input and output of the guideline representation model conform to the data format of the electronic medical record system, supporting the construction of guideline knowledge representation models for multiple diseases. Therefore, this method can be embedded in a real diagnosis and treatment system and has strong versatility.

[0081] Based on the above-mentioned method for constructing a computerized clinical guideline based on graphical representation, an embodiment of the present invention further provides a computerized clinical guideline construction device based on graphical representation, and the device is applied to a computerized clinical guideline model, wherein the computerized clinical guideline model includes: a data layer, a model layer and an application layer, and the structural block diagram of the device is as follows Figure 7 As shown, including:

[0082] Acquisition and determination module 201 , conversion module 202 and determination module 203 .

[0083] in,

[0084] The acquisition and determination module 201 is used to acquire text input in the clinical guideline database at the data layer, determine clinical data corresponding to the text input, and determine a target text input based on the text input and the clinical data;

[0085] The conversion module 202 is used to convert the target text input into a graphical diagnosis and treatment process at the model layer;

[0086] The determination module 203 is used to determine a decision recommendation result based on the graphical diagnosis and treatment process at the application layer, wherein the decision recommendation result includes at least one of: diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction.

[0087] The present invention discloses a computerized clinical guideline construction device based on graphical representation, which is applied to a computerized clinical guideline model. The computerized clinical guideline model includes: a data layer, a model layer and an application layer. The method includes: obtaining text input in a clinical guideline database at the data layer, determining clinical data corresponding to the text input, and determining a target text input based on the text input and the clinical data; converting the target text input into a graphical diagnosis and treatment process at the model layer; determining a decision recommendation result based on the graphical diagnosis and treatment process at the application layer, and the decision recommendation result includes: diagnosis and treatment decision, automated disease classification, treatment plan recommendation and at least one of complication prediction. In the above process, the determination process of the diagnosis and treatment process can be realized without medical staff mastering the relevant language, and the requirements for medical staff are relatively low, which is conducive to the dissemination and application of clinical guideline methods among medical staff.

[0088] In the embodiment of the present invention, the acquisition and determination module 201 includes:

[0089] An integration unit 204 , a standardization unit 205 and a labeling unit 206 .

[0090] in,

[0091] The integration unit 204 is used to integrate the text input and the corresponding items in the clinical data to obtain a first text input;

[0092] The standardization unit 205 is used to perform a term standardization operation on the first text input to obtain a second text input;

[0093] The annotation unit 206 is used to perform text annotation on the second text input to obtain a target text input.

[0094] In the embodiment of the present invention, the conversion module 202 includes:

[0095] A first transforming unit 207 , a second transforming unit 208 and a first determining unit 209 .

[0096] in,

[0097] The first conversion unit 207 is used to convert the target text input into decision condition variables and logic statements;

[0098] The second conversion unit 208 is used to convert the decision condition variable into a graphical variable;

[0099] The first determining unit 209 is used to pass the logic statement and the graphical variable to the disease clinical guideline representation model to obtain a graphical diagnosis process.

[0100] In the embodiment of the present invention, the conversion module 202 further includes: a verification unit 210 .

[0101] in,

[0102] The verification unit 210 is used to verify the disease clinical guideline representation model.

[0103] In the embodiment of the present invention, the determining module 203 includes:

[0104] The retrieval unit 211 and the second determination unit 212 .

[0105] in,

[0106] The retrieval unit 211 is used to perform clinical decision support path retrieval in the clinical guideline representation model based on the graphical diagnosis and treatment process;

[0107] The second determining unit 212 is used to determine the decision recommendation result based on the search result.

[0108] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments, and the same or similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0109] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the presence of other identical elements in the process, method, article or device including the elements.

[0110] For the convenience of description, the above device is described as being divided into various units according to their functions. Of course, when implementing the present invention, the functions of each unit can be implemented in the same or multiple software and / or hardware.

[0111] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present invention can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art can be embodied in the form of a software product, which can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present invention or certain parts of the embodiments.

[0112] The above is a detailed introduction to a method and device for constructing a computerized clinical guideline based on graphical representation provided by the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation methods and application scopes. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for constructing a computerized clinical guideline based on graphical representation, characterized in that: Applied to a computerized clinical guideline model, wherein the computerized clinical guideline model comprises: a data layer, a model layer and an application layer, and the method comprises: Acquiring text input in a clinical guideline database at the data layer, and determining clinical data corresponding to the text input; Integrating the text input with corresponding items in the clinical data to obtain a first text input; Performing a terminology standardization operation on the first text input to obtain a second text input; Performing text annotation on the second text input to obtain a target text input; Converting the target text input into decision condition variables and logic statements; Converting the decision condition variables into graphical variables; Passing the logic statement and the graphical variable to the disease clinical guideline representation model to obtain a graphical diagnosis and treatment process; A decision recommendation result is determined at the application layer based on the graphical diagnosis and treatment process, wherein the decision recommendation result includes at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation and complication prediction.

2. The method according to claim 1, characterized in that Also includes: The clinical guideline representation model for the disease was validated.

3. The method according to claim 1, characterized in that Determining a decision recommendation result based on the graphical diagnosis and treatment process at the application layer includes: Perform clinical decision support path retrieval in the clinical guideline representation model based on the graphical diagnosis and treatment process; Based on the search result, the decision recommendation result is determined.

4. A computerized clinical guideline construction device based on graphical representation, characterized in that: Applied to a computerized clinical guideline model, wherein the computerized clinical guideline model comprises: a data layer, a model layer and an application layer, and the device comprises: an acquisition and determination module, configured to acquire text input in a clinical guideline database at the data layer, determine clinical data corresponding to the text input, and determine a target text input based on the text input and the clinical data; A first conversion unit, used for converting the target text input into decision condition variables and logic statements; A second conversion unit, used to convert the decision condition variable into a graphical variable; A first determining unit is used to pass the logic statement and the graphical variable to the disease clinical guideline representation model to obtain a graphical diagnosis and treatment process; A determination module, configured to determine a decision recommendation result based on the graphical diagnosis and treatment process at the application layer, wherein the decision recommendation result includes at least one of diagnosis and treatment decision, automated disease classification, treatment plan recommendation, and complication prediction; The acquisition and determination module comprises: an integration unit, configured to integrate the text input with corresponding items in the clinical data to obtain a first text input; A standardization unit, configured to perform a term standardization operation on the first text input to obtain a second text input; The annotation unit is used to perform text annotation on the second text input to obtain a target text input.

5. The device according to claim 4, characterized in that Also includes: A verification unit is used to verify the disease clinical guideline representation model.

6. The device according to claim 4, characterized in that The determination module comprises: A retrieval unit, configured to perform clinical decision support path retrieval in the clinical guideline representation model based on the graphical diagnosis and treatment process; The second determining unit is used to determine the decision recommendation result based on the search result.

Images