An automatic coding method and system for disease diagnosis

Abstract

The application discloses a disease diagnosis automatic coding method and system, and belongs to the technical field of medical event recognition. The method comprises the following steps: acquiring and structuring preprocessed disease diagnosis documents, displaying the document directory and content; using a diagnosis coding model to extract diagnosis names, evidences and surgical operations from the documents, and displaying the diagnosis names, evidences and surgical operations in a multi-window in a highlighting mode; supporting auditing, commenting, editing, merging and traceability positioning of the diagnosis and evidence information; automatically generating a diagnosis coding result, sorting the diagnosis coding result, modifying the main diagnosis and adjusting the sequence of the associated evidences; the system supports DRG / DIP simulation grouping display, and assists in coding sequence decision-making; the coding result can be saved, collected and uploaded to a related system, and a complete coding closed-loop process is constructed. The application can greatly reduce the work burden of coders, improve medical insurance settlement efficiency and data quality, and has good clinical application value and popularization prospect.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of medical event recognition, and particularly relates to a disease diagnosis automatic coding method and system. BACKGROUND

[0002] Under the background of rapid development of medical and health informatization, standardization of disease diagnosis terminology has become a key link to improve medical quality and management efficiency. However, in actual medical documents, due to the habit of doctors' expression, various types of documents, and free recording forms, a large amount of diagnosis information has problems such as ambiguous description, non-standard language, and non-uniform structure, which seriously restricts the subsequent structured processing, intelligent analysis, and auxiliary decision-making application.

[0003] To cope with the above challenges, automatic disease diagnosis coding technology is proposed and gradually becomes one of the key solutions to improve the efficiency of medical information processing. The technology aims to accurately identify diagnosis information from complex and unstructured medical records or medical documents, and match it to the national standard diagnosis coding system (such as ICD-10, ICD-11, etc.), to realize semantic standardization and structuring. However, the existing automatic disease diagnosis coding technology still faces the following multiple difficulties:

[0004] Firstly, disease diagnosis in documents is often expressed in a non-standardized and context-dependent manner, relying on a large amount of semantics, context, and evidence support, which is difficult to extract through rules or simple models. Secondly, when relying solely on natural language processing models for diagnosis extraction and coding matching, accuracy and explainability are difficult to balance, and "black box" errors are likely to occur, affecting clinical trust. In addition, the coding result also needs to support manual review, evidence tracing, diagnosis merging / sorting, and other advanced operations to meet the complex needs of actual clinical and medical insurance supervision.

[0005] To solve the above problems, an automatic disease diagnosis coding method and system integrating document structure display, diagnosis automatic extraction, evidence tracing, efficient interaction, coding recommendation, and manual review are needed. SUMMARY

[0006] In view of the above defects of the prior art, the purpose of the present application is to provide an automatic disease diagnosis coding method and system, which can reduce the burden of manual coding, improve the efficiency of medical insurance settlement, and solve the problem of weak interaction between coders and automatic coding systems in the prior art.

[0007] The technical problem solved by the present application adopts the following technical solutions:

[0008] In a first aspect, an automatic disease diagnosis coding method comprises the following steps:

[0009] Step S1: Obtain the complete content of the disease diagnosis document through a standard interface and preprocess it to generate the structure and directory of the disease diagnosis document; display the directory in the directory window and the disease diagnosis document in the document interaction window; click on the corresponding directory item in the directory window, and the document interaction window will quickly locate to the starting position of the disease diagnosis document corresponding to the directory item;

[0010] Step S2: Start from the starting position of the disease diagnosis document and browse the complete content of the disease diagnosis document in chronological order; the diagnosis name extracted from the disease diagnosis document by the diagnosis coding model, the diagnosis evidence and the important information of the surgical operation will be highlighted in different presentation modes in the document interaction window; review and associate the important information in the document interaction window;

[0011] Step S3: The diagnosis name extracted from the disease diagnosis document by the diagnosis coding model and its position in the document, the diagnosis code corresponding to the diagnosis name are also displayed in the diagnosis list interaction window and the diagnosis code management window respectively; the diagnosis codes are sorted according to the importance of the diagnosis, and the main diagnosis code is placed in the first place; edit the diagnosis name and diagnosis code, and merge the diagnosis in the diagnosis list interaction window and the diagnosis code management window;

[0012] Step S4: The diagnosis coding model extracts the diagnosis evidence from the disease diagnosis document, and associates the diagnosis name with the diagnosis evidence; the evidence list corresponding to each diagnosis and the specific position of each evidence in the document are displayed in the evidence management window; review and edit the evidence and its relationship with the diagnosis name in the evidence management window;

[0013] Step S5: Manually adjust the arrangement order of the diagnosis code and modify the main diagnosis item; while adjusting the arrangement order of the diagnosis code, the display order of the disease diagnosis and its evidence fragments associated with the code will also be automatically updated;

[0014] Step S6: After calling the DRG / DIP grouper to accurately group according to the diagnosis code order, the grouping situation of different diagnosis codes of the same patient as the main diagnosis can be displayed for user selection and decision-making;

[0015] Step S7: Save, collect, review and revise the disease diagnosis and its standard code, corresponding evidence to the local disease diagnosis coding system, and synchronize upload to the medical record or medical record system through the submission function, and complete the diagnosis coding task.

[0016] Further, in step S2, the medical record interaction window is used to display the disease diagnosis medical record content, which covers the medical record and test report content of the whole process of patient visit, hospitalization, treatment and discharge, and is presented in a time sequence and according to the importance of the medical record according to the requirements of the coder.

[0017] Further, in step S3, the diagnosis coding model automatically extracts all disease diagnoses in the medical record. When a disease diagnosis appears in multiple places in the medical record, the system will classify it into the corresponding medical record block group according to the latest time information, and the remaining information will be displayed in the form of a list. When clicking on any information, the medical record interaction window automatically locates the medical record content corresponding to the disease diagnosis information, and highlights the disease diagnosis in the medical record.

[0018] Further, in step S3, the diagnosis name is edited, including adding, deleting and modifying:

[0019] Add: Select any text segment in the medical record interaction window to add it as a disease diagnosis item. The newly added disease diagnosis item and its evidence segments will be highlighted in different colors in the medical record interaction window. The system will automatically identify the added content and record the detailed information of the addition operation.

[0020] Delete: Delete a disease diagnosis item in the diagnosis list interaction window. The system will automatically delete all evidence segments associated with it. The deleted disease diagnosis item and evidence segments will be removed from the highlight display in the medical record interaction window. The system will identify the deleted content and record the relevant information of the deletion.

[0021] Modify: When the disease diagnosis expression in the medical record is not standardized, modify the relevant disease diagnosis item in the diagnosis list interaction window. The modified disease diagnosis item will be removed from the original highlight display in the medical record interaction window. The system will identify the modified content and record the detailed information of the modification result.

[0022] Further, in step S3, the method for diagnosis merging operation of the diagnosis name is:

[0023] There are multiple diagnoses and corresponding evidence segments in the medical record, but according to the coding rules, these diagnoses need to be merged into one code. Click to select these diagnoses, and the system will automatically merge these diagnoses and corresponding evidence segments, and give the coding result of the merged diagnosis. Revise the diagnosis name and coding result after merging, and the system will identify the merged diagnosis and record the information of the merging operation.

[0024] Further, in step S3, the editing operation of the diagnosis coding includes the following types:

[0025] Add: Click the Add Coding function in the function bar module, select the Add Reason, add the corresponding diagnosis in the Diagnosis List Interaction Window, and add the corresponding coding result in the Diagnosis Coding Management Window. The system will automatically identify the added diagnosis coding and record the details of the addition operation.

[0026] Delete: When there is a diagnosis in the medical record document but no corresponding evidence to support it, select the Delete Coding function in the function bar module, delete the corresponding disease diagnosis item and coding result, and remove the display in the Diagnosis List Interaction Window and Diagnosis Coding Management Window. The system will automatically identify and record the deleted diagnosis coding.

[0027] Modify: When the diagnosis in the medical record document matches the evidence but the coding result is inaccurate, modify the coding result. The system will automatically identify the modified diagnosis coding and record the details of the modification.

[0028] Further, in step S4, the diagnosis coding model extracts all evidence fragments related to the disease diagnosis from the disease diagnosis document, including symptoms and signs, test results, examination results, treatment feedback, and dynamic monitoring that are important evidence for the disease diagnosis. When associating the disease diagnosis with the evidence fragments, the time constraint relationship between the two will be strictly checked, i.e., the document creation time of the evidence source must not be later than the document creation time of the corresponding disease diagnosis;

[0029] When a disease diagnosis corresponds to multiple evidence fragments, these evidence fragments will be displayed in a list format for clear viewing and management. When the user clicks on any diagnosis, the evidence list associated with the diagnosis will be fully presented. When clicking on any evidence fragment in the list, the document interaction window automatically locates to the document content corresponding to the evidence fragment and highlights the evidence fragment in the document.

[0030] Further, in step S4, the evidence fragments are edited, including adding, deleting, and modifying:

[0031] Add: In the document interaction window, select any text fragment and add it as one or more existing disease diagnosis evidence fragments, provided that the creation time of the selected fragment is not later than the creation time of the corresponding diagnosis. The added evidence fragments will be highlighted in different colors in the document interaction window. The system will automatically identify the added content and record the details of the addition operation.

[0032] Delete: In the evidence management window, select a certain evidence fragment for deletion. The deleted evidence fragment will be removed from the highlight display in the document interaction window. The system will identify the deleted content and record the relevant information of the deletion.

[0033] Modification: When the expression of evidence fragments in the document is not standardized, select a certain evidence fragment in the evidence management window for editing, the system will automatically update the content of the evidence fragment in all positions; the modified evidence fragment will be removed from the original highlight in the document interaction window, the system will identify the modified content and record the detailed information of the modification result.

[0034] In a second aspect, an automatic coding system for disease diagnosis is used to implement the automatic coding method for disease diagnosis described in any of the above aspects, comprising:

[0035] A directory module is used to display the directory items of the disease diagnosis documents read from the standard interface in the form of a directory tree in the directory window. Clicking on a superior directory item will expand the subdirectory items of the directory item. Clicking on a specific document directory item will directly jump to the corresponding document content in the document interaction module.

[0036] A medical record document structured module is used to process medical record documents, structure them according to their content and document type using regular expressions or natural language processing models, and output the results in json format for subsequent processing and use.

[0037] A document interaction module is used to display the disease diagnosis documents read from the standard interface using the document interaction window, synchronously highlight the disease diagnosis and evidence fragments in the diagnosis and evidence interaction module, and support selecting a document fragment and appending it to the disease diagnosis or evidence fragment in the diagnosis and evidence interaction module.

[0038] A diagnosis and evidence interaction module is used to display all disease diagnoses and evidence fragments and their sources in the disease diagnosis document using the diagnosis list interaction window and the evidence management window, and associate them according to their corresponding relationship. When a disease diagnosis corresponds to multiple evidence fragments, all evidence fragments are displayed in the list, each evidence fragment is attached with a source label, a reason label, and a delete button. After adding or deleting a disease diagnosis or evidence fragment, the changes will be synchronized to the document interaction module, and the error correction reason will be recorded.

[0039] An encoding module is used to display the disease diagnosis coding results in the disease diagnosis document using the diagnosis coding management window. The coding results need to be selected from the mapping dictionary published by the authoritative agency, and must be the most appropriate clinical diagnosis term.

[0040] The function bar module mainly provides various key functions for the system, including: automatic coding function, which automatically codes diseases based on a disease diagnosis coding model; delete coding function, which is used to delete all automatic coding results; add coding function, which facilitates manual coding by coders; merge coding function, which supports merging related codes; collection function, which supports marking typical cases for collection and recording the collection reason for subsequent focus and quick retrieval;

[0041] The database module is used to save and export the extraction and coding results given by the disease diagnosis coding model and the extraction and coding results after manual review. The saved data is used for model performance optimization.

[0042] In a third aspect, a terminal device includes a memory, a processor, and a disease diagnosis automatic coding method program stored in the memory and executable on the processor. When the processor executes the disease diagnosis automatic coding method program, the steps of any one of the disease diagnosis automatic coding methods described above are implemented.

[0043] The present application provides a disease diagnosis automatic coding method and system, which combines the intelligent extraction capability of large language models or deep learning models and multi-window interaction design, realizes disease diagnosis document structured analysis, diagnosis and evidence correlation, coding result management, DRG / DIP group simulation and other core functions. The system supports real-time positioning, highlight display and bidirectional linkage in multiple windows such as directory, document, diagnosis, evidence, coding, and annotation, greatly improving the efficiency and experience of manual review. At the same time, it has the operation functions of adding, deleting, modifying, merging and sorting diagnosis and coding, realizing fine management and control of the whole coding process. The system also has built-in coding mapping dictionary and group coefficient display functions to assist coders in making accurate decisions, supports local saving and synchronous uploading of coding results, and builds an efficient, intelligent and traceable coding closed-loop process. The method and system effectively alleviate the workload of coders, improve the efficiency of medical insurance settlement and the level of data standardization, provide a solid data foundation for medical quality control, clinical research, medical insurance payment and hospital performance evaluation, and promote the intelligent and fine development of medical coding. BRIEF DESCRIPTION OF DRAWINGS

[0044] Figure 1 is a flowchart of the automatic coding method of the present application.

[0045] Figure 2 is a relationship diagram of each system module of the present application. DETAILED DESCRIPTION

[0046] In order to make the purposes, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are some but not all of the embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative work fall within the protection scope of the present application.

[0047] The present application has strong structuring capability, intelligent extraction capability, rich interaction mechanism, and standard coding recommendation and revision mechanism, and can support clinical doctors and coders to conveniently participate in auditing and modifying operations while maintaining high accuracy, thereby significantly improving coding quality, compliance and efficiency.

[0048] Reference Figure 1 The present embodiment provides a disease diagnosis automatic coding method, comprising the following steps:

[0049] Step S1: obtaining the complete content of the disease diagnosis document through a standard interface and preprocessing to generate the structure and directory of the disease diagnosis document; displaying the directory in a directory window and the disease diagnosis document in a document interaction window, and clicking the corresponding directory item in the directory window to quickly position the document interaction window to the starting position of the disease diagnosis document corresponding to the directory item.

[0050] Specifically, the preprocessing process includes format normalization, paragraph segmentation, timestamp analysis and chapter title identification of the original document, etc., to ensure that the generated directory structure is clear and readable. The system supports automatic recognition of multiple document templates (such as admission records, discharge summaries, surgical records, etc.), and automatically classifies and organizes the directory according to the document type.

[0051] Step S2: browsing the complete disease diagnosis document content in the time sequence of the disease diagnosis document from the starting position of the disease diagnosis document; the diagnosis name, diagnosis evidence and important information of surgical operation extracted from the disease diagnosis document by the diagnosis coding model are highlighted in different presentation modes in the document interaction window, and the important information is audited, annotated and associated in the document interaction window.

[0052] Specifically, the diagnosis coding model can be an open source or closed source large language model or other natural language processing model that has been fine-tuned on the diagnosis coding task, and can be one model or multiple models. The model needs to have the related function of disease diagnosis coding.

[0053] Further optimize the technical scheme, the style of the highlighted display supports multiple colors and label formats, helping users quickly distinguish the relationship between diagnosis items and evidence, and the document interaction window is used to display disease diagnosis documents, including medical records, medical records and other types of medical documents. In the document interaction window, the specified document content can be annotated, the system will locate the annotation according to the source information of the document content, and display the annotation content and the source information in the annotation management window; When clicking any annotation, the document interaction window automatically jumps to the document content corresponding to the annotation, and the user can modify and delete the annotation. The version tracking mechanism is embedded in the document interaction window, which records the time and user information of each annotation or operation, and realizes traceable management of the entire coding process.

[0054] Step S3: The diagnosis name extracted from the disease diagnosis document by the diagnosis coding model and its position in the document, and the diagnosis code corresponding to the diagnosis name are also displayed in the diagnosis list interaction window and the diagnosis code management window respectively. The diagnosis codes are sorted according to the importance of the diagnosis, and the main diagnosis code is placed in the first place; In the diagnosis list interaction window and the diagnosis code management window, the diagnosis name and the diagnosis code are edited and merged.

[0055] Specifically: The diagnosis coding model automatically extracts all disease diagnoses in the medical record document. When a disease diagnosis appears in multiple places in the document, the system will classify it into the corresponding document block group according to the latest time source information, and the remaining source labels will be displayed in the form of a list; When clicking any source label, the document interaction window automatically locates the disease diagnosis source label corresponding to the document content, and highlights the disease diagnosis in the document.

[0056] Next, the diagnosis name is edited, including adding, deleting and modifying:

[0057] Add: Select any text segment in the document interaction window to add it as a disease diagnosis item; The newly added disease diagnosis item and its evidence segment will be highlighted in different colors in the document interaction window, and the system will automatically identify the added content and record the detailed information of the addition operation;

[0058] Delete: Delete a disease diagnosis item in the diagnosis list interaction window, the system will automatically delete all evidence segments associated with it; The deleted disease diagnosis item and evidence segment will be removed from the highlighted display in the document interaction window, and the system will identify the deleted content and record the relevant information of the deletion;

[0059] Modification: When the disease diagnosis expression in the document is not standardized, modify the relevant disease diagnosis item in the diagnosis list interaction window; the modified disease diagnosis item will be removed from the original highlight display in the document interaction window, the system will identify the modification content and record the detailed information of the modification result.

[0060] Further optimization of the technical solution, the method for diagnosing the name of the diagnosis is to have multiple diagnoses and corresponding evidence fragments in the medical record document, but according to the coding rules, these diagnoses need to be combined into one code, click to select these diagnoses, the system will automatically combine these diagnoses and corresponding evidence fragments, and give the coding result after combining the diagnosis, revise the diagnosis name and coding result after combining, the system identifies the combined diagnosis and records the information of the combination operation.

[0061] Further optimization of the technical solution, the editing operation of the diagnosis code includes the following types:

[0062] Add: Click the add code function in the function bar module, select the add reason, add the corresponding diagnosis in the diagnosis list interaction window, and add the corresponding coding result in the diagnosis coding management window, the system will automatically identify the added diagnosis code and record the detailed information of the addition operation;

[0063] Delete: When there is a diagnosis in the medical record document but no corresponding evidence support, select the delete code function in the function bar module, delete the corresponding disease diagnosis item and coding result, and remove the display in the diagnosis list interaction window and the diagnosis coding management window, the system will automatically identify and record the deleted diagnosis code;

[0064] Modification: When the diagnosis in the medical record document matches the evidence but the coding result is inaccurate, modify the coding result, the system will automatically identify the modified diagnosis code and record the detailed content of the modification.

[0065] The system provides a rollback mechanism, all editing, combining and deleting operations can be restored to the previous version with one key, ensuring operation safety.

[0066] Step S4: The diagnosis coding model extracts diagnosis evidence from the disease diagnosis document, and associates the diagnosis name and diagnosis evidence, the evidence list corresponding to each diagnosis and the specific position of each evidence in the document are displayed in the evidence management window; the evidence and its relationship with the diagnosis name are audited and edited in the evidence management window.

[0067] Specifically, the diagnostic coding model extracts all evidence fragments related to disease diagnosis from the disease diagnosis document, including symptoms and signs, test results, examination results, treatment feedback, and dynamic monitoring that are important evidence for disease diagnosis. When associating disease diagnosis with evidence fragments, the time constraint relationship between the two will be strictly checked, i.e. the creation time of the evidence source document cannot be later than the creation time of the corresponding disease diagnosis source document.

[0068] When a disease diagnosis corresponds to multiple evidence fragments, these evidence fragments will be displayed in the form of a list for clear viewing and management. When the user clicks on any diagnosis, the evidence list associated with the diagnosis will be presented in its entirety. When clicking on any evidence fragment in the list, the document interaction window automatically locates the content of the document corresponding to the evidence fragment and highlights the evidence fragment in the document.

[0069] Specifically, the system supports editing operations on evidence fragments, including adding, deleting, and modifying:

[0070] Add: In the document interaction window, select any text fragment and add it as an evidence fragment for one or more existing disease diagnoses, provided that the creation time of the selected fragment is not later than the creation time of the corresponding diagnosis. The added evidence fragment will be highlighted in different colors in the document interaction window. The system will automatically identify the added content and record the detailed information of the addition operation.

[0071] Delete: In the evidence management window, select a certain evidence fragment for deletion. The deleted evidence fragment will be removed from the highlighted display in the document interaction window. The system will identify the deleted content and record the relevant information of the deletion.

[0072] Modify: When the evidence fragment in the document is not expressed in a standard manner, select a certain evidence fragment in the evidence management window for editing. The system will automatically update the content of the evidence fragment in all locations. The modified evidence fragment will be removed from the original highlighted display in the document interaction window. The system will identify the modified content and record the detailed information of the modification result.

[0073] Step S5: Manually adjust the arrangement order of the diagnostic code and modify the principal diagnosis item. When adjusting the order of the diagnostic code, the display order of the disease diagnosis and its evidence fragments associated with the code will also be automatically updated.

[0074] Specifically, the system supports "drag-and-drop" adjustment operations to simplify the coding process. Each sorting operation is accompanied by an operation log for easy auditing.

[0075] Step S6: After the DRG / DIP grouper accurately groups according to the diagnostic coding sequence, the system can display the grouping situation of different diagnostic codes of the same patient as the principal diagnosis for users to choose and decide.

[0076] Specifically, when the user adjusts the coding sequence, the system can calculate and display the corresponding grouping coefficients based on different coding arrangements. Users can adjust the sequence in real time to view the grouping situation under different coding sequences.

[0077] Step S7: Save, collect, and revise the disease diagnosis and its standard code, corresponding evidence to the local disease diagnosis coding system, and synchronize upload to the medical record or medical record system through the submission function, completing the diagnosis coding task.

[0078] Further optimization of the technical solution, the disease diagnosis name given by the diagnosis coding model, the corresponding evidence segment and the coding result, the disease diagnosis name, the corresponding evidence segment and the coding result after the coding staff's review, the coding staff's annotation on the disease diagnosis document, the collected disease diagnosis document and the collection reason are saved in the database module for storage; using these mistakes and typical cases can optimize the performance of the diagnosis coding model and use the search enhancement generation method to correct the model results.

[0079] Specifically: the system will first accurately write the coding results processed by the coding staff into the hospital's medical record system database, the local disease diagnosis coding system database, and then synchronize to the database of the medical record or medical record system, ensuring that all updated coding information can be timely reflected in the medical record, providing reliable basis for clinical decision-making, medical insurance settlement, and patient management.

[0080] At the same time, the system will also record the diagnosis name, corresponding diagnosis evidence and diagnosis code initially generated by the diagnosis coding model, and record all error types and specific reasons marked by the coding staff in detail. In this way, not only can the performance of the diagnosis coding model be fully understood, but also common error patterns can be identified, laying a foundation for subsequent data analysis and model optimization. These detailed records support the performance evaluation of the diagnosis coding model from multiple angles, and are helpful for targeted adjustment and optimization, thereby continuously improving the accuracy and efficiency of the diagnosis coding model.

[0081] In addition, detailed records support internal quality control processes and external audit requirements. They provide necessary documentation for checking the compliance, accuracy and consistency of coding work, helping to improve the overall quality of medical services. By saving the original output of the diagnosis coding model and the correction records of the coder, the system provides the possibility for subsequent data analysis, making it possible to continuously improve the diagnosis coding model based on actual operation feedback. This method effectively combines the advantages of artificial intelligence and professional judgment, promoting the development of medical information management systems to a higher level, not only improving work efficiency, but also enhancing data reliability.

[0082] Reference Figure 2 The present application also provides an automatic coding system for disease diagnosis, which is used to implement any one of the above-mentioned automatic coding methods for disease diagnosis, comprising:

[0083] A directory module is used to display the directory items of the disease diagnosis documents read from the standard interface in the form of a directory tree in the directory window. Clicking on a superior directory item will expand the subdirectory items of the directory item. Clicking on a specific document directory item will directly jump to the corresponding document content in the document interaction module.

[0084] A medical record document structured module is used to process medical record documents. According to the content and type of the document, it uses regular expressions or natural language processing models to structure the document, and outputs the results in json format for subsequent processing.

[0085] A document interaction module is used to display the disease diagnosis documents read from the standard interface using the document interaction window, and synchronously highlight the disease diagnosis and evidence fragments in the diagnosis and evidence interaction module. It also supports selecting a document fragment and appending it to the disease diagnosis or evidence fragment in the diagnosis and evidence interaction module.

[0086] A diagnosis and evidence interaction module is used to display all disease diagnoses and evidence fragments and their sources in the disease diagnosis documents using the diagnosis list interaction window and the evidence management window, and associate them according to their corresponding relationship. When a disease diagnosis corresponds to multiple evidence fragments, all evidence fragments are displayed in the list, each with a source label, a reason label and a delete button. After adding or deleting a disease diagnosis or evidence fragment, the changes will be synchronized to the document interaction module, and the error correction reason will be recorded.

[0087] An encoding module is used to display the disease diagnosis coding results in the disease diagnosis documents using the diagnosis coding management window. The coding results must be selected from the mapping dictionary published by the authoritative agency, and must be the most appropriate clinical diagnosis term. The mapping dictionary is constructed based on the standard disease diagnosis terms and their corresponding codes published by the national functional department or professional association, ensuring the standardization and authority of the coding results.

[0088] A comment module is configured to display all comment segments and their sources in the disease diagnosis document using the comment management window, and to provide the changes to the document interaction module after adding, modifying or deleting the comment segments.

[0089] A function bar module is configured to provide various key functions for the system, including an automatic coding function configured to automatically code the disease in the case document based on the disease diagnosis coding model, a delete coding function configured to delete all automatic coding results, an add coding function configured to facilitate the coding staff to manually add coding, a merge coding function configured to support the merging of related coding, and a collection function configured to support the collection and marking of typical cases and record the collection reasons for subsequent focus and quick retrieval.

[0090] A database module is configured to save and export the extraction and coding results given by the disease diagnosis coding model and the extraction and coding results after manual review, and the saved data is used for the optimization of the model performance.

[0091] The embodiment provides a terminal device, which comprises a memory, a processor, and a program of a disease diagnosis automatic coding method stored in the memory and executable on the processor. When the processor executes the program of the disease diagnosis automatic coding method, the steps of any one of the disease diagnosis automatic coding methods are implemented.

[0092] The terminal device can comprise one or more processors, a memory, and a computer program stored in the memory and executable on the one or more processors, for example, a disease diagnosis automatic coding system. When the one or more processors execute the computer program, the steps of the disease diagnosis automatic coding method in the embodiment can be implemented. Alternatively, when the one or more processors execute the computer program, the functions of the modules in the disease diagnosis automatic coding system in the embodiment can be implemented, which is not limited here.

[0093] In an embodiment, the processor can be a central processing unit (CPU), and can also be other general-purpose processors, digital signal processors (DSP), application specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) 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.

[0094] In an embodiment, the memory can be an internal storage unit of the electronic device, such as a hard disk or a memory of the electronic device. The memory can also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device. Further, the memory can include both the internal storage unit and the external storage device of the electronic device. The memory is used to store a computer program and other programs and data required by the terminal device. The memory can also be used to temporarily store data that has been output or will be output.

[0095] Those skilled in the art can understand that the structure of the terminal device described above is only part of the structure related to the present application scheme, and does not constitute a limitation on the terminal device to which the present application scheme is applied. The specific terminal device can include more or fewer components than those described above, or combine certain components, or have a different arrangement of components.

[0096] In the description of the present specification, the description referring to the terms "one embodiment", "an example", "a specific example" and the like means that the specific features, structures, materials or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0097] Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present application, and not to limit them; although the present application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that: it can still modify the technical solutions recorded in the foregoing embodiments, or make equivalent replacement to part of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A disease diagnosis automatic coding method characterized by comprising: The method comprises the following steps: Step S1: obtaining the complete content of the disease diagnosis document through a standard interface and preprocessing to generate the structure and directory of the disease diagnosis document; displaying the directory in a directory window and the disease diagnosis document in a document interaction window; and clicking on a corresponding directory item in the directory window to quickly position the document interaction window to the starting position of the disease diagnosis document corresponding to the directory item; Step S2: browsing the complete disease diagnosis document content in the time sequence of the disease diagnosis document from the starting position of the disease diagnosis document; and highlighting the diagnosis name, diagnosis evidence and important information of surgical operation extracted from the disease diagnosis document by the diagnosis coding model in different presentation modes in the document interaction window, and auditing and associating the important information in the document interaction window; Step S3: the diagnosis name extracted from the disease diagnosis document by the diagnosis coding model and the position of the diagnosis name in the document, the diagnosis coding corresponding to the diagnosis name are also displayed in a diagnosis list interaction window and a diagnosis coding management window, the diagnosis coding is sorted according to the important order of the diagnosis, and the main diagnosis coding is arranged in the first place; and editing and diagnosis merging operations are performed on the diagnosis name and diagnosis coding in the diagnosis list interaction window and the diagnosis coding management window; Step S4: extracting the diagnosis evidence from the disease diagnosis document by the diagnosis coding model, associating the diagnosis name and the diagnosis evidence, displaying the evidence list corresponding to each diagnosis and the specific position of each evidence in the document in an evidence management window; and auditing and editing the evidence and the relationship between the evidence and the diagnosis name in the evidence management window; Step S5: manually adjusting the arrangement order of the diagnosis coding and modifying the main diagnosis item; and automatically updating the display order of the disease diagnosis and the evidence segment associated with the coding when the arrangement order of the diagnosis coding is adjusted; Step S6: after the DRG / DIP grouper is called according to the diagnosis coding order and accurate grouping is performed, the grouping conditions of different diagnosis codings of the same patient as the main diagnosis are displayed for the user to select and decide; Step S7: saving, collecting, auditing and revising the disease diagnosis and the standard coding and corresponding evidence to the local disease diagnosis coding system, and synchronously uploading the disease diagnosis and the standard coding and corresponding evidence to the medical record or medical record system through a submission function, and completing the diagnosis coding task.

2. The automatic coding method for disease diagnosis according to claim 1, characterized in that, In step S2, the document interaction window is used to display the disease diagnosis document content, which covers the medical document, test report content of the whole process of patient treatment, hospitalization, treatment and discharge, and is presented in a stream according to the time sequence of the document formation and the importance of the document according to the requirements of the coder.

3. The automatic coding method for disease diagnosis according to claim 1, characterized in that, In step S3, the diagnosis coding model automatically extracts all disease diagnoses in the medical record document. When a disease diagnosis appears in multiple places in the document, the system will classify it into the corresponding document block group according to the latest time of the source information, and the remaining source tags will be displayed in the form of a list. When any source tag is clicked, the document interaction window automatically positions to the document content corresponding to the disease diagnosis source tag, and highlights the disease diagnosis in the document.

4. The automatic coding method for disease diagnosis according to claim 3, wherein, In step S3, the diagnosis name is edited, including adding, deleting and modifying: Add: Select any text segment in the document interaction window to add it as a disease diagnosis item; the added disease diagnosis item and its evidence segments will be highlighted in different colors in the document interaction window, and the system will automatically identify the added content and record the detailed information of the addition operation; Delete: Delete a disease diagnosis item in the diagnosis list interaction window, and the system will automatically delete all evidence segments associated with it; the deleted disease diagnosis item and evidence segments will be removed from the highlight display in the document interaction window, and the system will identify the deleted content and record the relevant information of the deletion; Modify: When the disease diagnosis expression in the document is not standardized, modify the relevant disease diagnosis item in the diagnosis list interaction window; the modified disease diagnosis item will be removed from the original highlight display in the document interaction window, and the system will identify the modified content and record the detailed information of the modification result.

5. The automatic coding method for disease diagnosis according to claim 1, wherein, In step S3, the method for diagnosis merging operation of diagnosis name is: There are multiple diagnoses and corresponding evidence segments in the medical record document, but according to the coding rules, these diagnoses need to be merged into one code, click to select these diagnoses, the system will automatically merge these diagnoses and corresponding evidence segments, and give the coding result after merging, revise the diagnosis name and coding result after merging, the system identifies the merged diagnosis and records the information of the merging operation.

6. The automatic coding method for disease diagnosis according to claim 1, wherein, In step S3, the editing operation of diagnosis coding includes the following types: Add: Click the add coding function in the function bar module, select the added reason, add the corresponding diagnosis in the diagnosis list interaction window, and add the corresponding coding result in the diagnosis coding management window, the system will automatically identify the added diagnosis coding and record the detailed information of the addition operation; Delete: When there is a diagnosis in the medical record document but no corresponding evidence to support it, select the delete coding function in the function bar module, delete the corresponding disease diagnosis item and coding result, and remove the display in the diagnosis list interaction window and the diagnosis coding management window, the system will automatically identify and record the deleted diagnosis coding; Modify: When the diagnosis in the medical record document matches the evidence but the coding result is inaccurate, modify the coding result, the system will automatically identify the modified diagnosis coding and record the detailed content of the modification.

7. The automatic coding method for disease diagnosis according to claim 1, wherein, In step S4, the diagnosis coding model extracts all evidence segments related to disease diagnosis from the disease diagnosis document, including symptoms and signs, test results, examination results, treatment feedback, and dynamic monitoring that are important evidence for disease diagnosis; when associating disease diagnosis with evidence segments, the time constraint relationship between the two will be strictly checked, that is, the document creation time of the evidence source cannot be later than the document creation time of the corresponding disease diagnosis; When a disease diagnosis corresponds to multiple evidence fragments, these evidence fragments will be displayed in the form of a list for clear viewing and management; when the user clicks on any diagnosis, the evidence list associated with the diagnosis will be presented in full; when clicking on any evidence fragment in the list, the document interaction window automatically locates the evidence fragment corresponding to the document content, and highlights the evidence fragment in the document.

8. The automatic coding method for disease diagnosis according to claim 1, characterized in that, In step S4, the evidence fragments are edited, including adding, deleting and modifying: Adding: In the document interaction window, select any text fragment to add it as an evidence fragment of one or more existing disease diagnoses, provided that the creation time of the selected fragment is not later than the creation time of the corresponding diagnosis; the added evidence fragment will be highlighted in different colors in the document interaction window, and the system will automatically identify the added content and record the detailed information of the addition operation; Deleting: In the evidence management window, select a certain evidence fragment for deletion; the deleted evidence fragment will be removed from the highlight display in the document interaction window, and the system will identify the deleted content and record the relevant information of the deletion; Modifying: When the evidence fragment in the document is not expressed in a standard manner, select a certain evidence fragment in the evidence management window for editing; the system will automatically update the content of the evidence fragment in all locations; the modified evidence fragment will be removed from the original highlight display in the document interaction window, and the system will identify the modified content and record the detailed information of the modification result.

9. A disease diagnosis automatic coding system characterized by comprising: A method for implementing the disease diagnosis automatic coding method according to any one of claims 1-8, comprising: a directory module for displaying the directory items of the disease diagnosis documents read from the standard interface in the form of a directory tree in the directory window; clicking on a superior directory item will expand the subdirectory items of the directory item; clicking on a specific document directory item will directly jump to the corresponding document content in the document interaction module; a medical record document structuring module for processing medical record documents, structuring them according to their content and document type using regular expressions or natural language processing models, and outputting the results in json format for subsequent processing; a document interaction module for displaying the disease diagnosis documents read from the standard interface using the document interaction window, synchronously highlighting the disease diagnoses and evidence fragments in the diagnosis and evidence interaction module, and supporting the selection of a document fragment to be added to a disease diagnosis or evidence fragment in the diagnosis and evidence interaction module; a diagnosis and evidence interaction module for displaying all disease diagnoses and evidence fragments and their sources in the disease diagnosis documents using the diagnosis list interaction window and the evidence management window, and associating them according to their corresponding relationship; when a disease diagnosis corresponds to multiple evidence fragments with sources, all evidence fragments are displayed in the list, each evidence fragment is attached with a source label, a reason label and a delete button, and after adding or deleting a disease diagnosis or an evidence fragment, the changes will be provided to the document interaction module, and the error correction reasons will be recorded. The coding module is configured to display disease diagnosis coding results in a disease diagnosis document using the diagnosis coding management window, wherein the coding results are selected from a mapping dictionary issued by an authority and must be the most appropriate clinical diagnosis term; The function bar module provides a plurality of key functions for the system, including an automatic coding function for automatically coding a case document based on a disease diagnosis coding model, a delete coding function for deleting all automatic coding results, an add coding function for facilitating manual coding by a coder, a merge coding function for supporting the merging of related codes, and a collection function for supporting the collection and marking of typical cases and recording the collection reasons; The database module is configured to save and export extraction and coding results given by the disease diagnosis coding model and the extraction and coding results after manual review, and the saved data is used for the tuning of model performance.

10. A terminal device, comprising: The terminal device comprises a memory, a processor, and a program of a disease diagnosis automatic coding method stored in the memory and executable on the processor. When the processor executes the program of the disease diagnosis automatic coding method, the steps of the disease diagnosis automatic coding method according to any one of claims 1-8 are implemented.

Images