Clinical test data automatic conversion method and device, electronic equipment and storage medium

Abstract

The invention discloses a clinical test data automatic conversion method and device, electronic equipment and a storage medium, and the method comprises the steps: obtaining a research flow chart from a clinical test scheme, and generating periodic visit information data; obtaining CRF from the clinical test scheme, and generating visit task unit data; generating scheme full-text analysis data based on the scheme full-text; obtaining scheme knowledge graph data based on the scheme knowledge graph; based on the modular component tool, the EMR structured component and the EDC structured component, generating a structural relationship between an EMR data structure and an EDC data structure; based on the modular component tool, the EMR data structure and the EDC data structure, generating a mapping relation between the EMR project data structure and the EDC project data structure; and based on the modular component tool, the EMR project data and the EDC project data, generating an association relationship between the EMR project data and the EDC project data. According to the method and the device, the form design defects of EMR-to-EDC automatic mapping and format automatic conversion caused by human factors are reduced, and the form generation efficiency is improved.

Description

Technical Field

[0001] The present application relates to the technical field of computers, and in particular to a method, device, electronic device and storage medium for automatic conversion of clinical trial data. Background Art

[0002] With the continuous development of the Internet and computer technology, electronic data capture technology (EDC) is increasingly being used in clinical trials. The collection of clinical trial data is the core content of the drug clinical research center. Real, accurate, timely and standardized data collection can significantly improve the quality of clinical trials and shorten the research cycle.

[0003] The clinical trial electronic data capture (EDC) system is applicable to the core information system of drug clinical trials, medical randomized controlled trials and medical cohort studies. Its core purpose is to record the information of subjects and form electronic follow-up forms.

[0004] In the EDC system, eCRF is usually used to replace paper CRF to collect and manage clinical trial data. In an expected usage scenario, it is possible to use machine learning technology to build the EMR and EDC data structure relationship (Structure relationship) that meets the project needs based on the input clinical trial plan, plan knowledge graph, modular component tools, EMR structured components, and EDC structured components, and to build the data transmission relationship (Transmission relationship) between EMR and EDC based on the mapping relationship (Mappingrelationship) between EMR and EDC. Through EMR, the eCRF form database that meets the EDC standard and the data association relationship (Associationrelationship) between EMR and EDC are automatically output, thereby achieving the process of automatic mapping and automatic format conversion between EMR and EDC.

[0005] The goal of this method is to improve the accuracy and efficiency of clinical trial data and reduce the time and error rate of manual data entry. By automatically mapping data in the EMR to the EDC, the opportunity for manual intervention can be reduced, thereby improving the reliability and consistency of the data.

[0006] The information disclosed in this background technology section is only intended to increase the understanding of the overall background of the application and should not be regarded as an admission or any form of suggestion that the information constitutes the prior art already known to a person skilled in the art. Summary of the invention

[0007] In order to solve the problems existing in the prior art, the present application provides a method, device, electronic device and storage medium for automatic conversion of clinical trial data.

[0008] In order to achieve the above objectives, this application adopts the following technical solutions:

[0009] In a first aspect, the present invention provides a method for automatically converting clinical trial data, comprising:

[0010] Obtaining a research flowchart from the clinical trial protocol to generate periodic visit data information, wherein the research flowchart includes a periodic visit information block, a visit task information block, and an association relationship information block between the periodic visit and the visit task;

[0011] Generate CRF periodic visit matrix data information based on the CRF;

[0012] Extracting the full text of the clinical trial protocol from the protocol to generate full text data information of the protocol, wherein the full text data of the protocol includes the trial field, trial phase, and indication;

[0013] Extracting data information corresponding to the knowledge graph from the solution knowledge graph;

[0014] Perform text parsing on the research flowchart, CRF, full text of the solution, and solution knowledge graph to obtain data information corresponding to the modular component tool;

[0015] Based on modular component tools, EMR structured components, and EDC structured components, construct the structural relationship between EMR data structures and EDC data structures;

[0016] Based on modular component tools, EMR data structure, and EDC data structure, build the mapping relationship between EMR project data structure and EDC project data structure;

[0017] Based on modular component tools, EMR project data structure, and EDC project data structure, build the association relationship between EMR project data and EDC project data;

[0018] The research flowchart, CRF, full text of the plan, and plan knowledge graph respectively include the correspondence, association, and indirect relationship information of periodic visits-visit tasks-execution time-execution items-task units and task sub-units.

[0019] Furthermore, the study flow chart is obtained from the clinical trial protocol to generate periodic visit data information, including:

[0020] Conduct structural analysis of the clinical trial protocol to obtain a study flow chart;

[0021] Perform text analysis on the research flowchart and separate the periodic visit information blocks;

[0022] Extract periodic visit information from the periodic visit information block.

[0023] Furthermore, the periodic visit information includes at least one of the visit cycle, visit date, baseline, periodic visit, and window period; and / or, based on a text sequence annotation model, the periodic visit information block is identified and normalized in multiple languages ​​to extract the periodic visit information.

[0024] Furthermore, the study flow chart is obtained from the clinical trial protocol to generate periodic visit data information, including:

[0025] Perform text analysis on the research flowchart and split the visit task information blocks;

[0026] Identify a visit task from the visit task information block, and match it with a standard visit task in the knowledge graph to obtain a first candidate visit task set;

[0027] Parse the execution time and execution items corresponding to each visiting task in the first candidate visiting task set to generate visiting task-execution time correspondence information and visiting task-execution item correspondence information.

[0028] Furthermore, the study flow chart is obtained from the clinical trial protocol to generate periodic visit data information, including:

[0029] Perform text analysis on the research flowchart and separate the association relationship information blocks between periodic visits and visit tasks;

[0030] The association relationship information block between the periodic visits and the visit tasks is subjected to text processing to generate periodic visit-visit task correspondence information.

[0031] Furthermore, the study flow chart is obtained from the clinical trial protocol to generate periodic visit data, including:

[0032] The periodic visit information, the visit task-execution time correspondence information, the visit task-execution item correspondence information, and the periodic visit-visit task correspondence information are integrated to generate the periodic visit data.

[0033] Furthermore, the CRF periodic visit data is generated based on the CRF of the clinical trial plan, specifically including:

[0034] The CRF of the clinical trial plan is disassembled to obtain at least two of the visit task-filling rules, information block-task unit, visit task-task subunit, and periodic visit information; based on the visit task, visit requirements, and at least two of the periodic visit information, the CRF periodic visit data is generated.

[0035] Further, extracting the full text data of the protocol from the clinical trial protocol to generate the full text data of the protocol specifically includes:

[0036] Obtaining full text data from the clinical trial protocol;

[0037] Querying the solution knowledge graph based on the research process data to obtain a second candidate visit task set applicable to the solution data;

[0038] Based on the second candidate visit task set, the full-text data information of the plan is generated.

[0039] Furthermore, the periodic visits in the periodic visit data information cannot be missing, and at most one piece of data is missing in the visit tasks and execution items; at most two pieces of data are missing in the CRF visit matrix data, and the task unit and the task sub-unit cannot be missing at the same time.

[0040] In a second aspect, the present invention provides an automatic generation device for clinical trial data, comprising:

[0041] A first data generation module is used to obtain a research flowchart from a clinical trial protocol to generate a first data generation module, wherein the research flowchart includes a periodic visit information block, a periodic visit-visit task information block, and a visit task-information block association relationship information block;

[0042] A second data generation module is generated based on a CRF data form of a clinical trial protocol, wherein the CRF data form includes an association relationship information block of a visit task-filling rule, an information block-task unit, and a visit task-task subunit;

[0043] A third data generation module is used to obtain the overall design, overall steps, and overall efficacy from the clinical trial protocol to generate the third data generation module, wherein the full text of the clinical trial protocol includes the association relationship information blocks of the overall design, overall steps, and overall efficacy;

[0044] A fourth data generation module, used to obtain association relationships, corresponding relationships, and indirect relationships from the solution knowledge graph to generate a fourth data generation module, wherein the solution knowledge graph includes association relationship information blocks of association relationships, corresponding relationships, and indirect relationships;

[0045] The fifth data generation module is based on the association between the research flowchart, CRF, full text of the clinical trial plan, and the knowledge graph of the plan; combined with the modular component tool, compared with the EMR structured component and the EDC structured component, and analyzed according to the project template requirements to generate the fifth data generation module, and the constructed structural relationship is the structural relationship between the EMR data structure and the EDC data structure;

[0046] The sixth data generation module is based on the association between the research flowchart, CRF, full text of the clinical trial plan, and the knowledge graph of the plan; combined with the modular component tool, the EMR data structure is constructed and the EDC data structure is constructed, and the sixth data generation module is generated according to the project template requirement analysis, and the constructed mapping relationship is the mapping relationship between the EMR project data structure and the EDC project data structure;

[0047] Based on the full text of the solution, research flowchart, CRF content domain, and solution knowledge graph, the requirements for modular component tools and project templates are analyzed to generate the relationship between EMR and EDC.

[0048] In a third aspect, the present invention provides an electronic device, comprising:

[0049] at least one processor;

[0050] and a memory storing instructions, which, when executed by the at least one processor, causes the at least one processor to execute the method for automatically generating clinical trial data as described in any one of the above.

[0051] In a fourth aspect, the present invention provides a machine-readable storage medium storing executable instructions, which, when executed, enable the machine to perform the method for automatically generating clinical trial data as described in any one of the above.

[0052] Compared with the prior art, this application has the following beneficial effects:

[0053] The automatic generation method of automatic mapping and automatic format conversion from EMR to EDC in this application extracts and fuses corresponding data from four dimensions: research flowchart, CRF, full text of the plan, and plan knowledge graph. The obtained fused data covers the corresponding relationship information of periodic visits-visit tasks-execution time-execution items-task units and sub-task units, and then, under the guidance of the plan knowledge algorithm, a personalized CRF visit matrix form can be intelligently generated; and the automatic generation method of automatic mapping and automatic format conversion from EMR to EDC based on machine learning can reduce the design deficiencies and defects of automatic mapping and automatic format conversion from EMR to EMR caused by human factors, experience factors, etc., and improve the efficiency of form generation for automatic mapping and automatic format conversion from EMR to EDC. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a schematic diagram of an application scenario of an automatic generation method and device for automatic mapping and automatic format conversion from EMR to EDC according to an embodiment of the present application;

[0055] Figure 2 It is a logical framework diagram of an automatic generation method for automatic mapping and automatic format conversion from EMR to EDC according to an embodiment of the present application;

[0056] Figure 3 is a flow chart of an automatic generation method for automatic mapping and automatic format conversion from EMR to EDC according to an embodiment of the present application;

[0057] Figure 4 It is a module diagram of an automatic generation device for automatic mapping and format conversion from EMR to EDC according to an embodiment of the present application;

[0058] Figure 5 It is a hardware structure diagram of an electronic device according to an embodiment of the present application. DETAILED DESCRIPTION

[0059] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments.

[0060] The terms "first", "second", "third", "fourth", "fifth", "sixth", etc. (if any) in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein, for example. In addition, the terms "including" and "corresponding to" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0061] A clinical trial protocol is a document that describes how a clinical trial will be conducted, including the objectives, design, methods, statistical considerations, and organization of the trial. It also provides the background and reasons for conducting the study, the research questions to be addressed, and considerations of ethical issues to ensure the safety of participants and the integrity of the data collected.

[0062] Reference Figure 2 and Figure 3, introduces an embodiment of the method for automatic conversion of clinical trial data of the present application, namely, an automatic generation method for automatic mapping from EMR to EDC and automatic format conversion.

[0063] In this embodiment, the method includes:

[0064] S101a. Obtain the study flowchart from the clinical trial protocol to generate periodic visit information data.

[0065] Specifically, the clinical trial protocol can be structured and parsed to obtain the research flowchart therein. The research flowchart includes various types of information blocks, and each information block may contain corresponding notes. It should be noted that in this embodiment, the acquisition of the research flowchart mentioned can be to obtain two parts: ① the information block itself; ② the notes corresponding to the information block.

[0066] In this embodiment, the research flow chart includes a periodic visit (Visit) information block, a visit task (Form) information block, and a relationship between periodic visits and visit tasks (Visit-Form) information block. Table I demonstrates a common research flow chart.

[0067] Table I

[0068]

[0069] Table I includes registration information, trial phase, informed consent, inclusion criteria, issuance of subject identity cards, demographics and medical history, previous treatment and concomitant treatment, adverse events, quality of life, echocardiogram; baseline, visit date (cycle 1 (D1, D2, D4, D8, D15), cycle 2 (D1, D8, D15), cycle 3 (D1, D8, D15)) and final steps and other cycle visit information blocks; the symbol "O" represents the visit task that needs to be performed during the corresponding visit cycle. For example, the existence of "O" in the binary positioning information block of "baseline-registration information" means that during the baseline visit, the subject needs to be asked about the registration information. Therefore, the information block where the symbol "O" is located can be considered as the associated information block of the corresponding cycle visit and the visit task.

[0070] In the process of generating periodic visit information data, visit task unit data, full-text analysis data of the program, and program knowledge graph data, the text of the research flowchart obtained in advance can be parsed first, and the periodic visit information blocks, visit task information blocks, and the relationship information blocks between periodic visits and visit tasks, task units, sub-task units, overall design, overall steps, and overall efficacy can be split out.

[0071] Taking the research flowchart shown in Table I above as an example, the process of text parsing is the parsing and analysis of the table structure, distinguishing the header row (multiple rows) corresponding to the visit task information block Form, the header column (multiple columns) corresponding to the periodic visit information block Visit, and the table content corresponding to the periodic visit and visit task association relationship information block Visit-Form (across rows and columns). In addition, the note information that may appear in the header row / column / table content can also be matched.

[0072] Further:

[0073] ① For periodic visit information blocks

[0074] The periodic visit information block may include the period to which the visit belongs, the visit date, the baseline, the periodic visit, and at least one visit information in the window period. In one embodiment, the periodic visit information block may be identified and normalized for multilingual texts based on a text sequence annotation model to extract the periodic visit information from the periodic visit information block.

[0075] Exemplarily, the multilingual texts here can be periodic visit texts in at least two languages ​​of a country, such as Chinese and Tibetan periodic visit texts; or they can be periodic visit texts in the native languages ​​of at least two countries, such as Chinese and English periodic visit texts.

[0076] ②For the visit task information block

[0077] First, the visit tasks can be identified from the visit task information block and matched with the standard visit tasks in the protocol knowledge algorithm to obtain the first candidate visit task set.

[0078] The scheme knowledge map is a structured semantic knowledge base used to describe concepts and their interrelationships in the physical world. In each embodiment of the present application, the data construction of the scheme knowledge map used can use natural language processing (NLP, Natural Language Processing) technology to perform statistical analysis on historical clinical trial schemes and historical eCRFs, identify the trial field, trial stage, indication, key visit tasks, and collect eCRF form information, and then complete the preliminary knowledge map construction through entity disambiguation, normalization, relationship mining and other technologies; further, the scheme knowledge map can also be modified and improved according to CDASH / SDTM standards, domain expert knowledge, etc.

[0079] In one embodiment, when identifying visit tasks, the visit tasks can be extracted based on the above-mentioned solution knowledge graph through NLP technologies such as word segmentation, named entity recognition, multi-pattern matching algorithm, text vectorization and similarity calculation, and time information extraction.

[0080] Secondly, for each visit task in the first visit task candidate set, the NLP technology mentioned in the above visit task identification can also be used to parse the task requirement information for each visit task in the visit task information block. The task requirement information here can include execution time and execution item, so that the visit task-execution time (Execution-Timepoint) correspondence information and the visit task-execution item (Execution-Index) correspondence information can be generated.

[0081] ③ Information blocks on the relationship between periodic visits and visit tasks

[0082] Specifically, the association relationship information block between periodic visits and visit tasks can be processed textually to generate periodic visit-visit task (Visit-Form) correspondence information.

[0083] After completing the above processing of the periodic visit information block, the visit task information block, and the association relationship information block between the periodic visit and the visit task, four sets of information are obtained: periodic visit information, visit task-execution time correspondence information, visit task-execution item correspondence information, and periodic visit-visit task correspondence information.

[0084] Subsequently, these four sets of information are integrated to generate periodic visit information, visit task unit information, program design, program steps, and overall efficacy, including the correspondence information of periodic visit-visit task-execution time-execution item-task unit-subtask unit (Visit-Form-Timepoint-Index).

[0085] In the visit cycle information data, the periodic visit cannot be missing, and at most one piece of data is missing in the visit task and execution item. That is, for the periodic visit information data, its possible data structure is:

[0086] ① Periodic visit-visit task-execution time-execution item;

[0087] ② Periodic visit-visit task (blank)-execution time-execution item;

[0088] ③ Periodic visit-visit task-execution time (blank)-execution item;

[0089] ④ Periodic visit-visit task-execution time-execution item (blank);

[0090] ⑤ Periodic visit-visit task (blank)-execution time (blank)-execution item;

[0091] ⑥ Periodic visit-visit task-execution time (blank)-execution item (blank), where "blank" means that the data of the corresponding element is empty.

[0092] S102a, generating visit task unit data based on the CRF disassembly of the clinical trial plan.

[0093] In step S101a, the focus is on the research flowchart in the clinical trial plan to extract the visit tasks and their related information. For the clinical trial plan, there may be potential visit task information that is not covered by the research flowchart.

[0094] Specifically, the full text of the clinical trial protocol can be scanned to obtain at least two of the visit tasks, visit requirements, and periodic visit information, and based on at least two of the visit tasks, visit subtasks, visit requirements, and periodic visit information, periodic visit data information can be generated.

[0095] The full text group data of the protocol can also include the corresponding relationship information of periodic visit-visit task-visit subtask-execution time-execution item. However, for the full text of the clinical trial protocol, at most two pieces of data are missing in the full text data of the protocol, and the visit task and the execution item cannot be missing at the same time. That is, for the full text data of the protocol, its possible data structure is:

[0096] ① Periodic visit-visit task-execution time-execution item;

[0097] ② Periodic visit-visit task (blank)-execution time-execution item;

[0098] ③ Periodic visit-visit task-execution time (blank)-execution item;

[0099] ④ Periodic visit-visit task-execution time-execution item (blank);

[0100] ⑤ Periodic visit (blank) - visit task (blank) - execution time - execution item;

[0101] ⑥ Periodic visit (blank) - visit task - execution time (blank) - execution item;

[0102] ⑦ Periodic visit (blank) - visit task - execution time - execution item (blank);

[0103] ⑧ Periodic visit-visit task (blank)-execution time (blank)-execution item, where "blank" means that the data of the corresponding element is empty.

[0104] In the specific process of generating full-text data of the plan, similarly, we can first extract the visit tasks here based on the above-mentioned plan knowledge graph through NLP technologies such as word segmentation, named entity recognition, multi-mode matching algorithm, text vectorization and similarity calculation, and time information extraction, and match the extracted visit tasks with the standard visit tasks in the plan knowledge graph to obtain the corresponding candidate visit task set (visit tasks, visit subtasks).

[0105] Secondly, for each visit task in the visit task set here, the NLP technology mentioned in the above visit task extraction can also be used to parse the corresponding visit requirements (also including information such as execution time and execution items). In one embodiment, the visit requirements obtained from the full text of the clinical trial protocol may include at least one of inclusion criteria, exclusion criteria, execution time points, laboratory execution items, medication record requirements, and evaluation scale information.

[0106] Finally, the periodic visit information can be extracted from the full text of the clinical trial protocol based on the text sequence annotation model, and the correspondence information of periodic visit-visit task-execution time-execution item-CRF sub-template-CRF visit matrix can be generated from the full text of the clinical trial protocol based on text processing.

[0107] S103a, generating overall design, overall steps, and overall efficacy information data based on the clinical trial plan.

[0108] S104a, integrating the visit cycle information data, the visit task information data, and the solution text analysis data to obtain association relationship data of the solution knowledge graph.

[0109] In the specific data fusion process, we can first obtain the data relationship corresponding to the modular component tools based on the solution knowledge graph, the full text of the solution, CRF, and the research flowchart.

[0110] In different embodiments, the order of merging the research flowchart, CRF, and the full text of the solution may also be different. For example, the three sets of data may be fused simultaneously; or, the research flowchart data may be fused with the CRF data first, and then the fused data may be fused with the full text of the solution.

[0111] S104a, obtaining the association relationship of the solution knowledge graph based on the full text of the solution, the solution knowledge graph, CRF and research flowchart data.

[0112] S105a. Constructing an EMR data structure and an EDC data structure based on the modular component tool, the EMR structured component and the EDC structured component.

[0113] Under the guidance of the solution knowledge graph and modular component tools, based on the corresponding relationship information of periodic visits-visit tasks-execution time-execution items-task units-subtask units in the fused data, a personalized EMR data structure and the corresponding structural relationship of EDC data can be generated.

[0114] In some application scenarios, under the guidance of the solution knowledge graph, the generated automatic mapping and format conversion table from EMR to EDC can be represented in data exchange format files such as JSON and XML; further, under the guidance of the solution knowledge graph, the automatic mapping and format conversion table structure from EMR to EDC can be translated into a final output file that can be recognized and imported by the EDC system.

[0115] Reference Figure 4 , introduces an embodiment of the automatic generation device of clinical trial data of the present application, that is, an automatic generation device for automatic mapping from EMR to EDC and automatic format conversion.

[0116] In this embodiment, the device includes a first data generation module U21, a second data generation module U22, a third data generation module U23, a fourth data generation module U24, a fifth data generation module U25, a sixth data generation module U26, and an EMR to EDC automatic mapping and format automatic conversion generation module U27.

[0117] The first data generation module U21 is used to obtain a research flowchart from the clinical trial plan to generate periodic visit data information, wherein the research flowchart includes a periodic visit information block, a visit task-information block, and a periodic visit-visit task information block;

[0118] The second data generation module U22 is used to obtain CRF data matrix information from the clinical trial plan to generate periodic visit task unit data information, wherein CRF includes visit task-filling rules, information block-task unit, visit task-task subunit data information;

[0119] The third data generation module U23 is used to disassemble the full text of the clinical trial protocol to generate data information after the full text of the protocol is parsed, wherein the full text of the protocol includes overall design information, overall step information, and overall efficacy information data;

[0120] The fourth data generation module U24 is used to disassemble the association relationship, corresponding relationship, and indirect relationship corresponding to the full text of the solution, CRF, and research flowchart from the solution knowledge graph to generate data information after the solution knowledge graph is parsed, wherein the parsed solution knowledge graph includes association relationship, corresponding relationship, and indirect relationship data information;

[0121] The fifth data generation module U25 is used to generate a structure relationship (Structure relationship) between the EMR data structure and the EDC data structure from the modular component tool, the EMR structured component, and the EDC structured component;

[0122] The sixth data generation module U26 is used to generate a mapping relationship between the EMR data structure and the EDC data structure from the modular component tool, the EMR data structure, and the EDC data structure; and also a transmission relationship between the EMR project data structure and the EDC project data structure;

[0123] Based on the mapping relationship (Mapping relationship) and transmission relationship (Transmission relationship) of the EMR project data structure and the EDC project data structure, the association relationship (Association relationship) between the EMR project data and the EDC project data is constructed.

[0124] In one embodiment, the first data generation module U21 is specifically used to perform structured analysis on the clinical trial plan to obtain a research flowchart; and to perform text analysis on the research flowchart and split it into periodic visit information blocks, periodic visit-visit task information blocks, and visit task-information blocks; and to extract periodic visit information from the periodic visit information blocks.

[0125] In one embodiment, the periodic visit information includes at least one of the period to which the visit belongs, the visit date, the baseline, the periodic visit, and the window period.

[0126] In one embodiment, the first data generation module U21 is specifically used to recognize and normalize multilingual texts of periodic visit information blocks based on a text sequence annotation model to extract periodic visit information.

[0127] In one embodiment, the first data generation module U21 is specifically used to perform text parsing on the research flowchart and split out the visit task information block; identify the visit task from the visit task information block and match it with the standard visit task in the scheme knowledge graph to obtain a candidate visit task set; and parse the execution time and execution items corresponding to each visit task in the candidate visit task set to generate visit task-execution time correspondence information and visit task-execution item correspondence information.

[0128] In one embodiment, the first data generation module U21 is specifically used to perform text analysis on the research flowchart and split out the association relationship information blocks between periodic visits and visit tasks; and perform text processing on the association relationship information blocks between periodic visits and visit tasks to generate periodic visit-visit task correspondence information.

[0129] In one embodiment, the first data generation module U21 is specifically used to integrate periodic visit information, visit task-execution time correspondence information, visit task-execution item correspondence information, and periodic visit-visit task correspondence information to generate a periodic visit information data set.

[0130] In one embodiment, the second data generation module U22 is specifically used to scan the full text of the clinical trial protocol to obtain at least two of the visit tasks, visit requirements, and periodic visit information in the CRF matrix; and generate a CRF periodic visit data set based on the visit tasks, visit requirements, and at least two of the periodic visit information.

[0131] In one embodiment, the visit requirements include at least one of inclusion criteria, exclusion criteria, execution time point, laboratory test items, medication record requirements, and assessment scale information.

[0132] In one embodiment, the third data generation module U23 is specifically used to extract full-text data of the clinical trial protocol from the protocol; query the protocol knowledge graph based on the full-text protocol to obtain a candidate visit task set applicable to the protocol metadata; and generate a full-text protocol data set based on the candidate visit task set.

[0133] In one embodiment, the fourth data generation module U24 is specifically used to integrate the research flowchart, CRF data matrix, and full text of the solution based on the solution knowledge graph; as well as the solution knowledge graph data to obtain a solution knowledge graph data set.

[0134] In one embodiment, the visit cycle in the solution knowledge graph data set cannot be missing, and at most one piece of metadata is missing in the visit task and execution item; at most two pieces of metadata are missing in the CRF data set, and the visit task and execution item cannot be missing at the same time.

[0135] In one embodiment, the modular component tool, EMR structured component, and EDC structured component generation module U25 are specifically used for EDC content domain, solution knowledge graph, CRF matrix data, and research flowchart to obtain the association relationship between the structural relationship of EMR and EDC data structures.

[0136] In one embodiment, the modular component tool, EMR data structure, and EDC data structure generation module U26 are specifically used for EDC content domain, solution knowledge graph, CRF matrix data, and research flowchart to obtain the mapping relationship and transmission relationship between EMR and EDC data structures.

[0137] In one embodiment, the modular component tool, EMR project data, and EDC project data generation module U27 are specifically used for EDC content domains, solution knowledge graphs, CRF matrix data, and research flowcharts to obtain the association relationship between EMR and EDC data structures, and translate the CRF in the EMR data exchange format into the EDC format CRF.

[0138] As above Figures 1 to 3 , the method for automatic mapping and automatic format conversion from EMR to EDC according to the embodiment of this specification is described. The details mentioned in the above description of the method embodiment are also applicable to the automatic mapping and automatic format conversion automatic generation device from EMR to EDC in the embodiment of this specification. The above automatic mapping and automatic format conversion automatic generation device from EMR to EDC can be implemented by hardware, or by software, or by a combination of hardware and software.

[0139] Reference Figure 5 , shows a hardware structure diagram of an electronic device according to an embodiment of the present specification, the electronic device 30 may include at least one processor 31, a memory 32 (such as a non-volatile memory), a memory 33 and a communication interface 34, and the at least one processor 31, the memory 32, the memory 33 and the communication interface 34 are connected together via a bus 35. At least one processor 31 executes at least one computer-readable instruction stored or encoded in the memory 32.

[0140] It should be understood that the computer executable instructions stored in the memory 32, when executed, cause at least one processor 31 to perform the above combined operations in various embodiments of the present specification. Figures 1 to 3 Describes the various operations and functions.

[0141] In the embodiments of the present specification, the electronic device 30 may include, but is not limited to, personal computers, server computers, workstations, desktop computers, laptop computers, notebook computers, mobile electronic devices, smart phones, tablet computers, cellular phones, personal digital assistants (PDAs), handheld devices, messaging devices, wearable electronic devices, consumer electronic devices, and the like.

[0142] According to one embodiment, a program product such as a machine-readable medium is provided. The machine-readable medium may have instructions (i.e., the above-mentioned elements implemented in the form of software), which, when executed by a machine, causes the machine to perform the above-mentioned combination of various embodiments of this specification. Figures 1 to 5 Specifically, a system or device equipped with a readable storage medium may be provided, on which a software program code implementing the functions of any of the above-mentioned embodiments is stored, and a computer or processor of the system or device reads and executes the instructions stored in the readable storage medium.

[0143] In this case, the program code itself read from the machine-readable medium can realize the function of any one of the above embodiments, and thus the machine-readable code and the machine-readable storage medium storing the machine-readable code constitute part of this specification.

[0144] Examples of readable storage media include floppy disks, hard disks, magneto-optical disks, optical disks (such as CD-ROM, CD-R, CD-RW, DVD-ROM, DVD-RAM, DVD-RW, DVD-RW), magnetic tapes, non-volatile memory cards, and ROMs. Alternatively, the program code may be downloaded from a server computer or a cloud via a communication network.

[0145] Those skilled in the art should understand that the various embodiments disclosed above can be modified and altered in various ways without departing from the essence of the invention. Therefore, the protection scope of this specification should be defined by the appended claims.

[0146] It should be noted that not all steps and units in the above-mentioned processes and system structure diagrams are necessary, and some steps or units can be ignored according to actual needs. The execution order of each step is not fixed and can be determined as needed. The device structure described in the above-mentioned embodiments can be a physical structure or a logical structure, that is, some units may be implemented by the same physical client, or some units may be implemented by multiple physical clients, or some components in multiple independent devices may be implemented together.

[0147] In the above embodiments, the hardware unit or module can be implemented mechanically or electrically. The hardware unit or processor can also include programmable logic or circuits (such as general-purpose processors or other programmable processors), which can be temporarily set by software to complete the corresponding operations. The specific implementation (mechanical method, or dedicated permanent circuit, or temporarily set circuit) can be determined based on cost and time considerations.

[0148] The specific embodiments described above in conjunction with the accompanying drawings describe exemplary embodiments, but do not represent all embodiments that can be implemented or fall within the scope of protection of the claims. The term "exemplary" used throughout this specification means "used as an example, instance or illustration" and does not mean "preferred" or "having advantages" over other embodiments. For the purpose of providing an understanding of the described technology, the specific embodiments include specific details. However, these technologies can be implemented without these specific details. In some instances, in order to avoid making the concepts of the described embodiments difficult to understand, well-known structures and devices are shown in block diagram form.

[0149] The above description of the present disclosure is provided to enable any person of ordinary skill in the art to implement or use the present disclosure. Various modifications to the present disclosure will be apparent to those of ordinary skill in the art, and the general principles corresponding to the present disclosure may be applied to other variations without departing from the scope of protection of the present disclosure. Therefore, the present disclosure is not limited to the examples and designs described herein, but is consistent with the widest range of principles and novel features disclosed herein.

Claims

1. A method for automatically converting clinical trial data, characterized in that: include: Obtaining a research flowchart from the clinical trial protocol to generate periodic visit data information, wherein the research flowchart includes a periodic visit information block, a visit task information block, and an association relationship information block between the periodic visit and the visit task; Generate CRF periodic visit matrix data information based on the CRF; Extracting the full text of the clinical trial protocol from the protocol to generate full text data information of the protocol, wherein the full text data of the protocol includes the trial field, trial phase, and indication; Extracting data information corresponding to the knowledge graph from the solution knowledge graph; Perform text parsing on the research flowchart, CRF, full text of the solution, and solution knowledge graph to obtain data information corresponding to the modular component tool; Based on modular component tools, EMR structured components, and EDC structured components, construct the structural relationship between EMR data structures and EDC data structures; Based on modular component tools, EMR data structure, and EDC data structure, build the mapping relationship between EMR project data structure and EDC project data structure; Based on modular component tools, EMR project data structure, and EDC project data structure, build the association relationship between EMR project data and EDC project data; The research flowchart, CRF, full text of the plan, and plan knowledge graph respectively include the correspondence, association, and indirect relationship information of periodic visits-visit tasks-execution time-execution items-task units and task sub-units.

2. The method for automatic conversion of clinical trial data according to claim 1, characterized in that: Obtain study flow chart from clinical trial protocol to generate periodic visit data information, including: Conduct structural analysis of the clinical trial protocol to obtain a study flow chart; Perform text analysis on the research flowchart and separate the periodic visit information blocks; Extract periodic visit information from the periodic visit information block.

3. The method for automatic conversion of clinical trial data according to claim 2, characterized in that: The periodic visit information includes at least one of the visit cycle, visit date, baseline, periodic visit, and window period; and / or, based on a text sequence annotation model, multilingual text recognition and normalization are performed on the periodic visit information block to extract the periodic visit information.

4. The method for automatic conversion of clinical trial data according to claim 2 or 3, characterized in that: Obtain study flow chart from clinical trial protocol to generate periodic visit data information, including: Perform text analysis on the research flowchart and split the visit task information blocks; Identify a visit task from the visit task information block, and match it with a standard visit task in the knowledge graph to obtain a first candidate visit task set; Parse the execution time and execution items corresponding to each visiting task in the first candidate visiting task set to generate visiting task-execution time correspondence information and visiting task-execution item correspondence information.

5. The method for automatic conversion of clinical trial data according to claim 4, characterized in that: Obtain study flow chart from clinical trial protocol to generate periodic visit data information, including: Perform text analysis on the research flowchart and separate the association relationship information blocks between periodic visits and visit tasks; The association relationship information block between the periodic visits and the visit tasks is subjected to text processing to generate periodic visit-visit task correspondence information.

6. The method for automatic conversion of clinical trial data according to claim 5, characterized in that: Obtain study flow chart from clinical trial protocol to generate period visit data, including: The periodic visit information, the visit task-execution time correspondence information, the visit task-execution item correspondence information, and the periodic visit-visit task correspondence information are integrated to generate the periodic visit data.

7. The method for automatic conversion of clinical trial data according to claim 1, characterized in that: Generate CRF periodic visit data based on the CRF of the clinical trial plan, including: The CRF of the clinical trial plan is disassembled to obtain at least two of the visit task-filling rules, information block-task unit, visit task-task subunit, and periodic visit information; based on the visit task, visit requirements, and at least two of the periodic visit information, the CRF periodic visit data is generated.

8. The method for automatic conversion of clinical trial data according to claim 1, characterized in that: Extracting the full text data of the protocol from the clinical trial protocol to generate the full text data of the protocol, specifically including: Obtaining full text data from the clinical trial protocol; Querying the solution knowledge graph based on the research process data to obtain a second candidate visit task set applicable to the solution data; Based on the second candidate visit task set, the full-text data information of the plan is generated.

9. The method for automatic conversion of clinical trial data according to claim 1, characterized in that: The periodic visits in the periodic visit data information cannot be missing, and at most one piece of data is missing in the visit tasks and execution items; at most two pieces of data are missing in the CRF visit matrix data, and the task unit and the task sub-unit cannot be missing at the same time.

10. An automatic generation device for clinical trial data, characterized in that: include: A first data generation module is used to obtain a research flowchart from a clinical trial protocol to generate a first data generation module, wherein the research flowchart includes a periodic visit information block, a periodic visit-visit task information block, and a visit task-information block association relationship information block; A second data generation module is generated based on a CRF data form of a clinical trial protocol, wherein the CRF data form includes an association relationship information block of a visit task-filling rule, an information block-task unit, and a visit task-task subunit; A third data generation module is used to obtain the overall design, overall steps, and overall efficacy from the clinical trial protocol to generate the third data generation module, wherein the full text of the clinical trial protocol includes the association relationship information blocks of the overall design, overall steps, and overall efficacy; A fourth data generation module, used to obtain association relationships, corresponding relationships, and indirect relationships from the solution knowledge graph to generate a fourth data generation module, wherein the solution knowledge graph includes association relationship information blocks of association relationships, corresponding relationships, and indirect relationships; The fifth data generation module is based on the association between the research flowchart, CRF, full text of the clinical trial plan, and the knowledge graph of the plan; combined with the modular component tool, compared with the EMR structured component and the EDC structured component, and analyzed according to the project template requirements to generate the fifth data generation module, and the constructed structural relationship is the structural relationship between the EMR data structure and the EDC data structure; The sixth data generation module is based on the association between the research flowchart, CRF, full text of the clinical trial plan, and the knowledge graph of the plan; combined with the modular component tool, the EMR data structure is constructed and the EDC data structure is constructed, and the sixth data generation module is generated according to the project template requirement analysis, and the constructed mapping relationship is the mapping relationship between the EMR project data structure and the EDC project data structure; Based on the full text of the solution, research flowchart, CRF content domain, and solution knowledge graph, the requirements for modular component tools and project templates are analyzed to generate the relationship between EMR and EDC.

11. An electronic device, characterized in that include: at least one processor; and a memory storing instructions, which, when executed by the at least one processor, enable the at least one processor to execute the method for automatically generating clinical trial data according to any one of claims 1 to 9.

12. A machine-readable storage medium, characterized in that: It stores executable instructions, which, when executed, enable the machine to perform the method for automatically generating clinical trial data as described in any one of claims 1 to 9.

Images