Clinical test medical record table generation method and electronic equipment

Through natural language processing model analysis and template matching technology, clinical trial medical records are automatically generated, solving the problem of low efficiency and accuracy of medical records creation in the existing technology, and achieving more efficient and accurate medical records generation.

CN119069058APending Publication Date: 2024-12-03长沙通诺信息科技有限责任公司
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411194489.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-28
Publication Date
2024-12-03

AI Technical Summary

Technical Problem

Medical record creation in existing clinical trials is less efficient and accurate, relying on manual operations and difficult to adapt to the complexity and diversity of different trial protocols.

Method used

The experimental plan is analyzed using a trained natural language processing model, matches the templates in the preset template library, and fills the templates based on the target text information to generate an electronic medical record to improve creation efficiency and accuracy.

Benefits of technology

Through automated medical record generation, the efficiency and accuracy of clinical trial medical record creation is significantly improved, adapting to the complexity and diversity of different trial plans, and reducing human errors.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119069058A_ABST
    Figure CN119069058A_ABST
Patent Text Reader

Abstract

The invention is suitable for the technical field of medical data processing, and provides a clinical test medical record table generation method and electronic equipment. Analyzing the test scheme by using a trained natural language processing model to obtain target text information; matching the target text information with a preset template library, and judging whether the matching is successful or not; a plurality of templates are arranged in the preset template library; if the matching is successful, determining the successfully matched template as a target template; filling a target template based on the target text information to obtain an electronic medical record table; the method is beneficial to improving the efficiency and accuracy of medical record table creation.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application belongs to the field of medical data processing technology, and in particular, relates to a method for generating a clinical trial medical record form and an electronic device. Background Art

[0002] In the existing clinical trial process, the creation of medical records mainly relies on manual operations, which is not only time-consuming but also prone to errors, affecting the quality and efficiency of the trial data. Manual creation of medical records requires professionals to have high professional qualities and undergo special training. In addition, the complexity and diversity of different trial plans also increase the difficulty of medical record creation. Therefore, how to improve the efficiency and accuracy of medical record creation is a problem that needs to be solved at present. Summary of the invention

[0003] The embodiments of the present application provide a method and electronic device for generating a clinical trial medical record, which can solve the problem of low efficiency and accuracy in creating medical records in the current clinical trial process.

[0004] In a first aspect, an embodiment of the present application provides a method for generating a clinical trial medical record form, comprising:

[0005] Obtain trial protocols;

[0006] Parsing the test plan using a trained natural language processing model to obtain target text information;

[0007] Matching the target text information with a preset template library to determine whether the match is successful; the preset template library has multiple templates;

[0008] If the match is successful, the successfully matched template is determined as the target template;

[0009] The target template is filled based on the target text information to obtain an electronic medical record.

[0010] In a possible implementation manner of the first aspect, obtaining the test plan includes:

[0011] Get the voice data input by the user;

[0012] Converting the voice data into text data;

[0013] The test plan is generated according to the text data.

[0014] In a possible implementation manner of the first aspect, before parsing the test scheme using the trained natural language processing model to obtain target text information, the method further includes:

[0015] adding an output layer to the natural language processing model to form an adjusted model;

[0016] Constructing a fine-tuning dataset; the data in the fine-tuning dataset are all labeled;

[0017] Pre-training the adjusted model using a text corpus to obtain an initial model;

[0018] The initial model is retrained using the fine-tuning data set, and during the training process, the model parameters are optimized using a back-propagation algorithm to obtain the trained natural language processing model.

[0019] In a possible implementation manner of the first aspect, matching the target text information with a preset template library and determining whether the match is successful includes:

[0020] According to a similarity algorithm or a rule matching algorithm, searching in the preset template library whether there is a template matching the target text information;

[0021] If it exists, the match succeeds; otherwise, the match fails;

[0022] The method further comprises:

[0023] When the match fails, the target template is generated according to the target text information and predefined rules.

[0024] In a possible implementation manner of the first aspect, before filling the target template based on the target text information to obtain the electronic medical record, the method further includes:

[0025] The target template is subjected to data verification according to a preset rule base to obtain a verified target template.

[0026] In a possible implementation manner of the first aspect, filling a target template based on the target text information to obtain an electronic medical record form includes:

[0027] Filling a target template based on the target text information to obtain first medical record information;

[0028] Acquire image data and test data for recording the behavior of the subject participating in the test process corresponding to the clinical trial;

[0029] According to the trial purpose and / or evaluation index of the clinical trial, identifying abnormal data in the image data;

[0030] Determine a time period corresponding to the abnormal data;

[0031] Deleting the test data corresponding to the time period to obtain the retained test data;

[0032] Extracting and obtaining second medical record information according to the retained test data;

[0033] The electronic medical record form is generated according to the first medical record information and the second medical record information.

[0034] In a possible implementation manner of the first aspect, filling a target template based on the target text information to obtain an electronic medical record form includes:

[0035] Filling a target template based on the target text information to obtain first medical record information;

[0036] Acquire image data and test data for recording the behavior of the subject participating in the test process corresponding to the clinical trial;

[0037] According to the trial purpose and / or evaluation index of the clinical trial, identifying abnormal data in the image data, and determining the subjects with the abnormal data as candidate subjects, and determining the remaining subjects as first target subjects;

[0038] When the number of the first target subjects is less than the sample size threshold, screening out second target subjects from the candidate subjects;

[0039] Extracting third medical record information based on the test data of the first target subject and the test data of the second target subject;

[0040] The electronic medical record form is generated according to the first medical record information and the third medical record information.

[0041] In a possible implementation of the first aspect, when the number of the first target subjects is less than a sample number threshold, screening out second target subjects from the candidate subjects includes:

[0042] When the number of the first target subjects is less than a sample size threshold, adjusting the sample size threshold according to the test data of the candidate subjects;

[0043] Determine the number of supplementary elections based on the adjusted sample size threshold and the number of the first target subjects;

[0044] Based on the supplementary selection number and the test data of the alternative subjects, a second target subject is screened out from the alternative subjects.

[0045] In a possible implementation of the first aspect, adjusting the sample quantity threshold according to the test data of the candidate subject includes:

[0046] Calculating the effect size of the clinical trial based on the test data of the candidate subjects;

[0047] An adjusted sample size threshold is determined based on the effect size and a preset effect size threshold.

[0048] In a second aspect, an embodiment of the present application provides an electronic device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the method for generating a clinical trial medical record described in any one of the first aspects above is implemented.

[0049] In a third aspect, an embodiment of the present application provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method for generating a clinical trial medical record form described in any one of the first aspects above is implemented.

[0050] Compared with the prior art, the embodiments of the present invention have the following beneficial effects:

[0051] The embodiment of the present application utilizes a trained natural language processing model to parse the test protocol, obtains target text information, matches the target text information with a pre-built preset template library, uses the successfully matched template as the target template, fills the target template based on the target text information, and obtains an electronic medical record, thereby improving the efficiency of creating clinical trial medical records; on the other hand, through natural language processing technology analysis, it can better adapt to the complexity and diversity of different test protocols, which is conducive to improving the accuracy of creating clinical trial medical records. BRIEF DESCRIPTION OF THE DRAWINGS

[0052] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0053] Figure 1 It is a flowchart of a method for generating a clinical trial medical record provided in an embodiment of the present application;

[0054] Figure 2 is a flowchart of a method for generating a clinical trial medical record form provided in another embodiment of the present application;

[0055] Figure 3 is a flowchart of a method for generating a clinical trial medical record form provided in another embodiment of the present application;

[0056] Figure 4is a flowchart of a method for generating a clinical trial medical record form provided in another embodiment of the present application;

[0057] Figure 5 is a flowchart of a method for generating a clinical trial medical record form provided in another embodiment of the present application;

[0058] Figure 6 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0059] In the following description, specific details such as specific system structures, technologies, etc. are provided for the purpose of illustration rather than limitation, so as to provide a thorough understanding of the embodiments of the present application. However, it should be clear to those skilled in the art that the present application may also be implemented in other embodiments without these specific details. In other cases, detailed descriptions of well-known systems, devices, circuits, and methods are omitted to prevent unnecessary details from obstructing the description of the present application.

[0060] It should be understood that when used in the present specification and the appended claims, the term "comprising" indicates the presence of described features, wholes, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or combinations thereof.

[0061] It should also be understood that the term “and / or” used in the specification and appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0062] As used in the specification and appended claims of this application, the term "if" can be interpreted as "when" or "uponce" or "in response to determining" or "in response to detecting", depending on the context. Similarly, the phrase "if it is determined" or "if [described condition or event] is detected" can be interpreted as meaning "uponce it is determined" or "in response to determining" or "uponce [described condition or event] is detected" or "in response to detecting [described condition or event]", depending on the context.

[0063] In addition, in the description of the present application specification and the appended claims, the terms "first", "second", "third", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance.

[0064] References to "one embodiment" or "some embodiments" etc. described in the specification of this application mean that one or more embodiments of the present application include specific features, structures or characteristics described in conjunction with the embodiment. Therefore, the statements "in one embodiment", "in some embodiments", "in some other embodiments", "in some other embodiments", etc. that appear in different places in this specification do not necessarily refer to the same embodiment, but mean "one or more but not all embodiments", unless otherwise specifically emphasized in other ways. The terms "including", "comprising", "having" and their variations all mean "including but not limited to", unless otherwise specifically emphasized in other ways.

[0065] An embodiment of the present application discloses a method for generating a clinical trial medical record. Figure 1 As shown, the method comprises the steps of:

[0066] S110, obtain the test plan. Exemplarily, the above test plan can be input into the system in any of the following ways: manual input, file upload and voice input. Manual input means that the user directly types the test plan in the text input box of the system. File upload means that the user uploads the document of the test plan, and then the system converts the electronic document into a text-based test plan based on OCR (Optical Character Recognition) technology. Voice input means that the user inputs voice through a microphone, and the system uses voice recognition technology to convert the voice into text. Specifically: obtain the voice data input by the user, convert the voice data into text data, and generate the test plan based on the text data. In this embodiment, the above clinical trial plan is a detailed document that describes the plan and design of the clinical trial, and may include information such as the purpose of the trial, design type, and evaluation indicators.

[0067] S120, using a trained natural language processing (NLP) model to parse the test plan and obtain the target text information. The above-mentioned natural language processing model (NLP model) has been trained in advance using a relevant corpus. The NLP model can be BERT (Bidirectional Encoder Representations from Transformers) or spaCy, etc. BERT is a language representation model. spaCy is an open source natural language processing library for processing and analyzing text data. It provides many functions, including word segmentation, part-of-speech tagging, named entity recognition, dependency syntax analysis, etc. The target text information includes information such as the purpose of the experiment, design type, and evaluation indicators.

[0068] Before using the NLP model for parsing, the test plan needs to be preprocessed, segmented, sentence-by-sentence, and word-by-word. Specifically, preprocessing includes text normalization (converting the text to a uniform format, such as converting all letters to lowercase), removing special characters (removing special characters and extra blank characters in the text), and other operations. You can use line breaks or paragraph marks (such as two consecutive line breaks) to split the text into paragraphs. You can use the sentence segmentation function of the spaCy library to split the text into sentences. Then, perform word segmentation on each sentence to split the sentence into individual words or tokens.

[0069] The analysis process of the NLP model on the test plan mainly includes named entity recognition, relationship extraction, merging and optimizing the key information extraction results. Named entity recognition is to use the NLP model to identify named entities and identify key information in the text. Relationship extraction is to further extract the relationship between entities after extracting named entities. For example, in the test design, the purpose of the test and the corresponding evaluation indicators are identified. Then the extracted named entities and relationships are merged and optimized to generate structured key information, which is the target text information.

[0070] S130, match the target text information with the preset template library to determine whether the match is successful. The preset template library has multiple templates. If the match is successful, execute step S140. If the match fails, execute step S150. In the above step S130, the preset template library can be searched for a template matching the target text information according to a similarity algorithm or a rule matching algorithm.

[0071] The preset template library contains different types of medical record templates, which are classified and stored according to different test types and requirements. The preset template library can be pre-built, and each template in the preset template library corresponds to a specific test type and requirement. Each template contains a set of features and structured fields for matching parsing results.

[0072] This embodiment uses an example to illustrate the calculation process of the similarity algorithm:

[0073] First, define the representation of the template and the parsing results: vectorize or structure the template and the target text information parsed from the test plan. For example, you can use the bag-of-words model, TF-IDF vectorization, or convert the template and parsing results into feature vectors. Then select a similarity measurement method, such as cosine similarity, Jaccard similarity, edit distance, etc. The specific measurement method can be determined according to the data representation type and matching requirements. Then calculate the similarity between each template and the target text information, such as calculating cosine similarity. The template with a similarity greater than the preset similarity threshold is used as the target template. If there are multiple templates with a similarity greater than the preset similarity threshold, the template with the greatest similarity is used as the final target template.

[0074] Regarding the setting of the preset similarity threshold, it can be adjusted and optimized according to the specific application scenario and data characteristics, such as determining the minimum similarity value through the evaluation of the experiment and the validation set. According to the complexity of the actual data set and template, this threshold can be adjusted and optimized, and the specific value can be determined according to the size of the template library, the complexity of the template, and the diversity of the experimental scheme information. In this embodiment, the value interval of the preset similarity threshold is [0.6, 0.9].

[0075] The rule matching algorithm may be a keyword-based matching. For example, when the maximum number of keywords matching the template is greater than a preset threshold, it indicates a successful match, otherwise it indicates a failed match.

[0076] S140: Determine the successfully matched template as the target template.

[0077] S150, generate a target template according to the target text information and predefined rules. Specifically, in the case of a matching failure, the unmatched target text information can be filled into the basic template according to the predefined rules. If there is no suitable field in the basic template, a new field is created and filled with the corresponding information. The above predefined rules are used to dynamically generate new templates. These rules define how to map key information to template fields, such as: the purpose of the experiment is mapped to the "Purpose" field of the template, the experimental design information should be located in the "Design" part of the template, and the evaluation indicator information should be in a table format, listing various indicators and their descriptions.

[0078] S160, filling the target template based on the target text information to obtain an electronic medical record, that is, the information such as the test purpose, design type, and evaluation index in the target text information will be automatically filled into the corresponding position in the template.

[0079] The above-mentioned electronic medical record is a standardized document used to record the subject information, trial process and results in clinical trials. It usually includes personal information, medical history, observations and data during the trial, etc. The information and structure in the medical record are designed according to the clinical trial protocol. The above-mentioned electronic medical record can be exported in document format or table format, or stored in a database to support subsequent query, entry, statistics and analysis.

[0080] This embodiment uses a trained natural language processing model to parse the test plan, obtains target text information, matches the target text information with a pre-built preset template library, uses the successfully matched template as the target template, fills the target template based on the target text information, and obtains an electronic medical record, thereby improving the efficiency of creating clinical trial medical records; on the other hand, through natural language processing technology analysis, it can better adapt to the complexity and diversity of different test plans, which is conducive to improving the accuracy of creating clinical trial medical records.

[0081] In some optional embodiments, in the above Figure 1 Based on the corresponding embodiment, the training process of the natural language processing model includes the following steps:

[0082] Train natural language processing models using large, general-purpose text corpora.

[0083] If the natural language processing model is a BERT model, BERT uses the Transformer architecture to learn the deep features of the text. BERT considers the contextual information on both the left and right sides at the same time during the pre-training stage. Specifically, the BERT model better understands the context of the language based on the MLM (Masked Language Model) task, and the BERT model understands the relationship between sentences based on the NSP (Next Sentence Prediction) task. In the NSP task, the model receives two sentences, one of which is the previous context and the other is the next context. The model needs to predict whether the second sentence is the next context of the first sentence. This task helps the model learn the logical relationship between sentences and enhances its understanding of the text structure. MLM and NSP are both specific tasks used in the BERT pre-training stage, and together they help BERT learn the complex features of the language. MLM enables BERT to deeply understand the context of each word, while NSP allows BERT to understand the order and logical relationship between sentences.

[0084] In some optional embodiments, in the above Figure 1 Based on the corresponding embodiment, Figure 2 As shown, step S161 is formed on the basis of step S160:

[0085] The verified target template is filled based on the target text information to obtain an electronic medical record.

[0086] Before step S161, the method further includes:

[0087] S170, performing data verification on the target template according to a preset rule base to obtain a verified target template.

[0088] The above-mentioned data verification process of this embodiment may include: checking whether all required items have been filled in, whether the data format is correct, whether the data range is reasonable, etc. For example, check whether the age field is within a reasonable range (such as 0 to 120 years old). Ensure that the value of the weight field is within a reasonable medical range (such as 30kg to 200kg), etc. If there are empty fields in the required fields in the medical record form, prompt the user to fill in these empty fields. Verify whether all ID fields in the medical record form conform to the expected format: such as ensuring that the date field follows a specific format (such as YYYY-MM-DD), checking whether the numeric field is a valid number (such as the age field should be an integer), and checking whether the text field conforms to the expected text format (such as the ID card number).

[0089] The above data verification process also includes verifying whether the data between multiple records of the same patient are consistent, such as checking whether the gender field of the same patient in different records is consistent, checking whether the date of birth field of the same patient in different records is consistent, etc. It is also necessary to verify whether the logical relationship between the data is correct, such as ensuring that the follow-up date is later than the initial visit date, etc. It is also necessary to analyze the filling rate of certain fields. If the filling rate of a field is low, prompt the user to reconfirm or improve the filling method.

[0090] This embodiment can ensure the accuracy, completeness and consistency of the data in the filled template by performing data verification on the target template.

[0091] In some optional embodiments, based on the above embodiments, after step S160, the following steps are further included:

[0092] Optimize data entry for the preset template library based on historical data, including adjusting entry fields and indicating common incorrect filling methods.

[0093] In one possible implementation, if the filling rate of an input field is low, the expression of the input field is adjusted, such as adjusting it to an expression with the same meaning but a higher frequency of words obtained in network statistics; or taking measures such as prompting users to reconfirm or improve the filling method. If the number of filling errors in an input field exceeds the threshold, a warning prompt is given for the input position of the field, and a correct example is provided. The optimization suggestions can be prompted to the user through the system interface, and the user can modify them according to the prompts.

[0094] Another embodiment of the present application discloses another method for generating a clinical trial medical record. Figure 3 As shown, this embodiment is Figure 1 On the basis of the corresponding embodiment, before step S110, the following steps are further included:

[0095] S101, adding an output layer to the natural language processing model to form an adjusted model.

[0096] S102, construct a fine-tuning dataset. The data in the fine-tuning dataset are all labeled.

[0097] S103, pre-training the adjusted model using the text corpus to obtain an initial model.

[0098] S104, retraining the initial model using the fine-tuning data set, optimizing the model parameters through the back propagation algorithm during the training process, and obtaining a trained natural language processing model.

[0099] In this embodiment, the above-mentioned natural language processing model is a BERT model. This embodiment adjusts the structure of the BERT model to adapt to key information extraction tasks, mainly referring to making some customized modifications based on the original BERT model to better complete specific text processing tasks, such as named entity recognition (NER) or text classification.

[0100] For example, since the raw output of the BERT model is a sequence of hidden state representations, for specific tasks such as NER, it is necessary to predict the entity category of each word from these hidden states. Therefore, a fully connected layer (usually a linear layer) is added on top of the BERT model to map the output of BERT to the output dimension required for the task. For example, in the NER task, each word may correspond to multiple predefined entity categories, and the output layer outputs the probability of each word belonging to each category. Alternatively, for the NER task, in addition to adding a classification layer, it is also necessary to ensure that the model can recognize specific patterns of word combinations in the text. This usually involves using specific algorithms to identify and classify entity boundaries, such as biomedical entities, geographic locations, etc. Therefore, the above output layer in this case includes a named entity recognition layer.

[0101] For text classification tasks, the design of the output layer is determined by the number of categories. For example, if there is a binary classification problem, the output layer may contain only one neuron to output the probability of belonging to a certain category; for multi-classification problems, the number of neurons in the output layer will match the number of categories. Depending on the task requirements, it may also be necessary to adjust the internal architecture of the BERT model, such as changing the number of Transformer layers, adjusting the attention mechanism, etc. In addition to adjusting the network structure, it is also necessary to define a loss function suitable for a specific task, such as the cross entropy loss commonly used in NER, which is used to train the model to correctly classify entities.

[0102] After adjusting the model structure, the model is fine-tuned using task-specific labeled data, and the model parameters are optimized through the back-propagation algorithm to minimize the model's prediction error on task-specific data. The data annotation process can be: collecting a large amount of clinical trial protocol text, cleaning the data and unifying the format; defining the annotation standards for key information based on the specific content of the clinical trial protocol text, and using professional annotation tools and teams to manually annotate the collected data.

[0103] Another embodiment of the present application discloses another method for generating a clinical trial medical record. Figure 4 As shown, in this embodiment, based on any of the above embodiments, step S160 includes:

[0104] S1601, filling in a target template based on target text information to obtain first medical record information.

[0105] S1602, acquiring image data and test data for recording the behavior of the subjects participating in the test process corresponding to the clinical trial.

[0106] S1603, identifying abnormal data in the image data according to the trial purpose and / or evaluation indicators of the clinical trial.

[0107] S1604, determine the time period corresponding to the abnormal data.

[0108] S1605, deleting the test data corresponding to the above time period to obtain the retained test data.

[0109] S1606, extracting the second medical record information based on the retained test data.

[0110] S1607, generating an electronic medical record form according to the first medical record information and the second medical record information.

[0111] Among them, the above test data may be data collected by sensors. For example, the clinical trial may require the subject to go through three test stages before the test is completed. During the test, some subjects may find someone else to complete some test stages, that is, cheating. In this case, for example, it can be determined by the facial data recorded in the image data. For example, it is determined that some subjects asked others to complete the first two test stages. In this case, the test data of the corresponding time period needs to be discarded to ensure the accuracy of the test analysis and the generated electronic medical record.

[0112] Another embodiment of the present application discloses another method for generating a clinical trial medical record. Figure 5 As shown, this embodiment is Figures 1 to 3 Based on any corresponding embodiment, step S160 includes:

[0113] S1601, filling in a target template based on target text information to obtain first medical record information.

[0114] S1602, acquiring image data and test data for recording the behavior of the subjects participating in the test process corresponding to the clinical trial.

[0115] S1608, according to the clinical trial purpose and / or evaluation index, identify abnormal data in the image data, and determine the subjects with abnormal data as candidate subjects, and determine the remaining subjects as first target subjects. The remaining subjects are subjects without abnormal data in the image data.

[0116] S1609: When the number of first target subjects is less than the sample number threshold, second target subjects are selected from the candidate subjects.

[0117] S16010: Extract third medical record information based on the test data of the first target subject and the test data of the second target subject.

[0118] S16011, generate an electronic medical record form based on the first medical record information and the third medical record information.

[0119] When the number of the first target subjects is greater than or equal to the sample quantity threshold, the third medical record information is directly extracted based on the test data of the first target subjects.

[0120] In this embodiment, when the number of first target subjects is less than the sample size threshold, second target subjects are selected from the candidate subjects to supplement the number of subjects as clinical trial objects, which can ensure the smooth implementation of the clinical trial and the availability of the test results, and also ensure the availability of the electronic medical record. Among them, the screening method for selecting the second target subject from the candidate subjects can be: the first N candidate subjects whose test data corresponding to certain specific test items have the smallest difference with the test data corresponding to the first target subject are selected as the second target subjects. The first N candidate subjects with the smallest deviation of the above abnormal data can also be selected as the second target subjects.

[0121] Among them, the above-mentioned difference can be the absolute value of the difference, the above-mentioned abnormal data deviation degree can be the duration corresponding to the abnormal data, and N is the difference between the sample quantity threshold and the number of the first target subjects.

[0122] In some optional embodiments, based on the above embodiments, step S1609 includes:

[0123] S16091: When the number of the first target subjects is less than the sample size threshold, the sample size threshold is adjusted according to the test data of the alternative subjects.

[0124] S16092, determining the number of supplementary selections according to the adjusted sample number threshold and the number of the first target subjects, wherein the number of supplementary selections may be the difference between the adjusted sample number threshold and the number of the first target subjects.

[0125] S16093, based on the number of supplementary selections and the test data of the alternative subjects, select the second target subject from the alternative subjects.

[0126] Specifically, when the number of the first target subjects is less than the sample size threshold, if a relatively obvious difference can be observed between the experimental group and the control group (or the two experimental groups) under the current number, then the current number only needs to be added with a smaller sample size, and the significance test effect can be observed without reaching the sample size threshold. In this way, the sample size threshold can be adjusted to reduce the sample size, thereby reducing the amount of data that the system needs to process, which is conducive to improving the efficiency of generating electronic medical records.

[0127] Optionally, the above step S16091 includes:

[0128] When the number of first target subjects is less than the sample size threshold, the effect size of the clinical trial is calculated based on the test data of alternative subjects.

[0129] Determine the adjusted sample size threshold based on the effect size and the preset effect size threshold.

[0130] Specifically, effect size is a statistical measure of the extent to which an experimental intervention or treatment affects an outcome variable. It provides a quantitative assessment of the size of the intervention effect and is a key indicator for evaluating the importance of research results. Effect size is usually defined as the difference between an experimental group and a control group (or two experimental groups), or the difference before and after an intervention. Effect size is an important component of statistical inference, providing information about the practical significance of hypothesis test results. Effect size does not depend on sample size, and even in small sample studies, significant effects can be observed if the effect size is large.

[0131] For example, the effect size indicator may be Cohen's d (a standardized measure of the difference in means between two independent sample groups), a standardized mean difference, a response ratio, or the like.

[0132] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0133] It should be noted that the above-mentioned method embodiments disclosed in the present application can be freely combined, and the technical solutions obtained after the free combination are also within the protection scope of the present application.

[0134] Corresponding to the method for generating a clinical trial medical record table described in the above embodiment, another embodiment of the present application discloses a device for generating a clinical trial medical record table. The device for generating a clinical trial medical record table includes:

[0135] The test plan acquisition module obtains the test plan.

[0136] The target text information acquisition module uses the trained natural language processing model to parse the test plan and obtain the target text information.

[0137] The template matching module matches the target text information with the preset template library to determine whether the match is successful. The preset template library contains multiple templates.

[0138] The target template determination module determines the successfully matched template as the target template if the match is successful.

[0139] The electronic medical record acquisition module fills the target template based on the target text information to obtain the electronic medical record.

[0140] For the deformable embodiments related to the corresponding embodiments of the device, reference may be made to the description of the above-mentioned method embodiments, and this embodiment will not be described in detail here.

[0141] It should be noted that the information interaction, execution process, etc. between the above-mentioned devices / units are based on the same concept as the method embodiment of the present application. Their specific functions and technical effects can be found in the method embodiment part and will not be repeated here.

[0142] The technicians in the relevant field can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example for illustration. In practical applications, the above-mentioned function allocation can be completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above. The functional units and modules in the embodiment can be integrated in a processing unit, or each unit can exist physically separately, or two or more units can be integrated in one unit. The above-mentioned integrated unit can be implemented in the form of hardware or in the form of software functional units. In addition, the specific names of the functional units and modules are only for the convenience of distinguishing each other, and are not used to limit the scope of protection of this application. The specific working process of the units and modules in the above-mentioned system can refer to the corresponding process in the aforementioned method embodiment, which will not be repeated here.

[0143] The present application also provides an electronic device, such as Figure 6 As shown, the electronic device 60 includes: at least one processor 601, a memory 602, and a computer program 603 stored in the memory and executable on the at least one processor, and the processor implements the steps in any of the above method embodiments when executing the computer program.

[0144] An embodiment of the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps in the above-mentioned method embodiments can be implemented.

[0145] An embodiment of the present application provides a computer program product. When the computer program product runs on a mobile terminal, the mobile terminal can implement the steps in the above-mentioned method embodiments when executing the computer program product.

[0146] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the present application implements all or part of the processes in the above-mentioned embodiment method, which can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a computer-readable storage medium. When the computer program is executed by the processor, the steps of the above-mentioned various method embodiments can be implemented. Among them, the computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may at least include: any entity or device that can carry the computer program code to the camera / electronic device, recording medium, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, RandomAccess Memory), electric carrier signal, telecommunication signal and software distribution medium. For example, a USB flash drive, a mobile hard disk, a magnetic disk or an optical disk. In some jurisdictions, according to legislation and patent practice, computer-readable media cannot be electric carrier signals and telecommunication signals.

[0147] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described or recorded in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0148] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0149] In the embodiments provided in the present application, it should be understood that the disclosed devices / network equipment and methods can be implemented in other ways. For example, the device / network equipment embodiments described above are merely schematic. For example, the division of the modules or units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0150] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0151] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for generating a clinical trial medical record form, It is characterized in that include: Obtain trial protocols; Parsing the test plan using a trained natural language processing model to obtain target text information; Matching the target text information with a preset template library to determine whether the match is successful; The preset template library has multiple templates; If the match is successful, the successfully matched template is determined as the target template; The target template is filled based on the target text information to obtain an electronic medical record.

2. The method for generating a clinical trial medical record form according to claim 1, It is characterized in that The acquisition test plan includes: Get the voice data input by the user; Converting the voice data into text data; The test plan is generated according to the text data.

3. The method for generating a clinical trial medical record form according to claim 1, It is characterized in that Before parsing the test plan using the trained natural language processing model to obtain target text information, the method further includes: adding an output layer to the natural language processing model to form an adjusted model; Constructing a fine-tuning dataset; the data in the fine-tuning dataset are all labeled; Pre-training the adjusted model using a text corpus to obtain an initial model; The initial model is retrained using the fine-tuning data set, and during the training process, the model parameters are optimized using a back-propagation algorithm to obtain the trained natural language processing model.

4. The method for generating a clinical trial medical record form according to claim 1, It is characterized in that The step of matching the target text information with a preset template library and determining whether the match is successful includes: According to a similarity algorithm or a rule matching algorithm, searching in the preset template library whether there is a template matching the target text information; If it exists, the match succeeds; otherwise, the match fails; The method further comprises: When the match fails, the target template is generated according to the target text information and predefined rules.

5. The method for generating a clinical trial medical record form according to claim 1, It is characterized in that Before filling the target template based on the target text information to obtain the electronic medical record, the method further includes: The target template is subjected to data verification according to a preset rule base to obtain a verified target template.

6. The method for generating a clinical trial medical record according to any one of claims 1 to 5, It is characterized in that The step of filling the target template based on the target text information to obtain an electronic medical record form includes: Filling a target template based on the target text information to obtain first medical record information; Acquire image data and test data for recording the behavior of the subject participating in the test process corresponding to the clinical trial; According to the trial purpose and / or evaluation index of the clinical trial, identifying abnormal data in the image data; Determine a time period corresponding to the abnormal data; Deleting the test data corresponding to the time period to obtain the retained test data; Extracting and obtaining second medical record information according to the retained test data; The electronic medical record form is generated according to the first medical record information and the second medical record information.

7. The method for generating a clinical trial medical record according to any one of claims 1 to 5, It is characterized in that The step of filling the target template based on the target text information to obtain an electronic medical record form includes: Filling a target template based on the target text information to obtain first medical record information; Acquire image data and test data for recording the behavior of the subject participating in the test process corresponding to the clinical trial; According to the trial purpose and / or evaluation index of the clinical trial, identifying abnormal data in the image data, and determining the subjects with the abnormal data as candidate subjects, and determining the remaining subjects as first target subjects; When the number of the first target subjects is less than the sample size threshold, screening out second target subjects from the candidate subjects; Extracting third medical record information based on the test data of the first target subject and the test data of the second target subject; The electronic medical record form is generated according to the first medical record information and the third medical record information.

8. The method for generating a clinical trial medical record form according to claim 7, It is characterized in that When the number of the first target subjects is less than the sample number threshold, screening out second target subjects from the candidate subjects comprises: When the number of the first target subjects is less than the sample size threshold, adjusting the sample size threshold according to the test data of the candidate subjects; Determining the number of supplementary elections based on the adjusted sample size threshold and the number of the first target subjects; Based on the supplementary selection number and the test data of the alternative subjects, a second target subject is screened out from the alternative subjects.

9. The method for generating a clinical trial medical record form according to claim 8, It is characterized in that The step of adjusting the sample quantity threshold according to the test data of the candidate subject comprises: Calculating the effect size of the clinical trial based on the test data of the candidate subjects; An adjusted sample size threshold is determined based on the effect size and a preset effect size threshold.

10. An electronic device, It is characterized in that The method comprises a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the method according to any one of claims 1 to 9 when executing the computer program.

Citation Information

Cited By

  • Photovoltaic installation data processing method and system, electronic equipment and storage medium

    CN119720967A

  • Data processing method, system and device, electronic equipment, medium and program product

    CN121235750A