Medical record quality assessment method, system, terminal and storage medium

Abstract

The present invention provides a medical record quality assessment method, system, terminal and storage medium, the method comprising: obtaining sample medical records, and digitally encoding the sample medical records to obtain a digital code set; generating a training data set based on the digital code set, and training an entity recognition model based on the training data set until the entity recognition model converges; inputting the medical record information of the hospital to be assessed into the converged entity recognition model to perform entity recognition to obtain an entity recognition result, and determining the medical record timeliness information of the hospital to be assessed based on the entity recognition result; and performing a medical record quality assessment on the hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result. The present invention can effectively determine the medical record timeliness information of the hospital to be assessed based on the entity recognition result, and can effectively assess the medical record quality of the entire hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result.

Description

Technical Field

[0001] The present invention relates to the field of medical technology, and in particular to a medical record quality assessment method, system, terminal and storage medium. Background Art

[0002] With the rapid development of science and technology, the quality management level of hospitals is getting higher and higher. Medical records, as an important carrier of medical behavior records, have always been the main source of medical quality management data and information. They are also the basis for various clinical specialties to carry out scientific research. Their importance is self-evident. With the release of the "Medical Record Management Quality Control Indicators (2021 Edition)" by the National Health Commission, the issue of medical record quality assessment has received more and more attention.

[0003] In the existing medical record quality assessment process, manual methods are generally used to assess the quality of a single medical record, which cannot effectively assess the overall medical record quality of the hospital. Summary of the Invention

[0004] The purpose of the embodiments of the present invention is to provide a medical record quality assessment method, system, terminal and storage medium, aiming to solve the problem that the existing medical record quality assessment method cannot effectively assess the overall medical record quality of the hospital.

[0005] The embodiment of the present invention is implemented as follows: a medical record quality assessment method, the method comprising:

[0006] Obtaining sample medical records, and digitally encoding the sample medical records to obtain a digital code set;

[0007] Generating a training data set according to the digital code set, and performing model training on an entity recognition model according to the training data set until the entity recognition model converges;

[0008] Inputting the medical record information of the hospital to be evaluated into the converged entity recognition model to perform entity recognition, obtaining an entity recognition result, and determining the medical record timeliness information of the hospital to be evaluated based on the entity recognition result;

[0009] A medical record quality assessment is performed on the hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result.

[0010] Preferably, the medical record timeliness information includes the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours;

[0011] The formula used to determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result includes:

[0012] The rate of admission records completed within 24 hours = the number of inpatient medical records completed within 24 hours of admission / the total number of inpatient medical records during the same period × 100%;

[0013] The number of inpatient medical records completed within 24 hours of the patient's admission was calculated as the number of records with the patient's admission date less the number of records with the patient's admission date completed within 24 hours of the admission date.

[0014] The completion rate of surgical records within 24 hours = the number of inpatient medical records with surgical records completed within 24 hours after surgery / the total number of inpatient medical records of surgical patients in the same period × 100%;

[0015] The number of inpatient medical records with surgical records completed within 24 hours after surgery = the recording time of the surgical record form - the number of medical records with surgical time less than 24 hours;

[0016] Completion rate of discharge records within 24 hours = number of medical records completed within 24 hours after discharge / total number of medical records of patients discharged during the same period × 100%;

[0017] The number of medical records completed within 24 hours after the patient's discharge = the recording time of the discharge record form - the number of medical records with the patient's discharge date < 24 hours;

[0018] Completion rate of the first page of medical records within 24 hours = the number of medical records with the first page completed within 24 hours after the patient is discharged / the total number of medical records of patients discharged during the same period × 100%;

[0019] The number of medical records with the first page completed within 24 hours after the patient is discharged = the recording time of the medical record first page document record form minus the number of medical records with the patient's discharge date completed within 24 hours;

[0020] The total number of medical records of inpatients during the same period = the number of records in the corresponding medical record homepage record table within the query start time and end time.

[0021] Preferably, the formula used to evaluate the medical record quality of the hospital to be evaluated based on the medical record timeliness information includes:

[0022] Medical record quality assessment results = a1 admission record completion rate within 24 hours + a2 surgical record completion rate within 24 hours + a3 discharge record completion rate within 24 hours + a4 medical record homepage completion rate within 24 hours;

[0023] Among them, a1, a2, a3, and a4 are the preset weights for the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours, respectively;

[0024] If the medical record quality assessment result is greater than the quality assessment threshold, it is determined that the medical record quality assessment of the hospital to be assessed is qualified.

[0025] Preferably, the digital encoding of the sample medical records to obtain a digital code set includes:

[0026] Obtaining a medical record statement in the sample medical record;

[0027] If the medical record statement contains characters and the sentence length of the medical record statement is less than the preset character length, the characters in the medical record statement are respectively coded and mapped;

[0028] The medical record statement after code mapping is data encapsulated to obtain the digital code set.

[0029] Preferably, the encoding and mapping of the characters in the medical record statements respectively includes:

[0030] Match each character with the pre-stored encoding mapping table;

[0031] If the character successfully matches the encoding mapping table, the character is mapped with the matched encoding;

[0032] If the character does not match the code mapping table, mapping the character with a preset code;

[0033] Matching the label of each character with the encoding mapping table respectively;

[0034] If the tag matches the code mapping table successfully, the matched code is mapped to the tag.

[0035] Preferably, generating a training data set according to the digital code set includes:

[0036] Extracting the sentence features and labels of each medical record sentence in the digital code set to obtain feature codes and label codes;

[0037] The feature code and the label code are tensor-encapsulated to obtain a feature tensor and a label tensor, and the feature tensor and the label tensor are dataset-encapsulated to obtain the training dataset.

[0038] Preferably, the performing model training on the entity recognition model according to the training data set until the entity recognition model converges includes:

[0039] Inputting the training data set into the entity recognition model to perform entity recognition, and determining the total number of correct entities, the number of recognized entities, and the number of real entities based on the entity recognition results;

[0040] Calculating the loss value, precision, recall rate, and F1 value of the entity recognition model based on the total number of correct entities, the number of recognized entities, and the number of real entities, and updating the parameters of the entity recognition model based on the loss value, wherein the F1 value is used to represent the harmonic mean of the precision of the entity recognition model and the recall rate;

[0041] When the number of iterations of the entity recognition model is greater than a threshold number, the entity recognition model is determined to be converged, and the loss value, accuracy, recall rate and F1 value of the entity recognition model of each iteration are obtained respectively;

[0042] An image is drawn based on the obtained loss value, accuracy, recall rate and F1 value of the entity recognition model to obtain a loss line graph, an accuracy line graph, a recall line graph and an F1 line graph.

[0043] Another object of an embodiment of the present invention is to provide a medical record quality assessment system, the system comprising:

[0044] A digital coding module is used to obtain sample medical records and digitally encode the sample medical records to obtain a digital code set;

[0045] A model training module, configured to generate a training data set according to the digital code set, and perform model training on an entity recognition model according to the training data set until the entity recognition model converges;

[0046] An entity recognition module is used to input the medical record information of the hospital to be evaluated into the converged entity recognition model to perform entity recognition, obtain an entity recognition result, and determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result;

[0047] The quality assessment module is used to perform a medical record quality assessment on the hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result.

[0048] Another object of an embodiment of the present invention is to provide a terminal device, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the above method when executing the computer program.

[0049] Another object of an embodiment of the present invention is to provide 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 of the above method are implemented.

[0050] In an embodiment of the present invention, a digital coding set is obtained by digitally encoding sample medical records. A training data set can be effectively generated based on the digital coding set. By inputting the medical record information of the hospital to be evaluated into a converged entity recognition model for entity recognition, entities in the medical record information can be effectively extracted. Based on the entity recognition results, the medical record timeliness information of the hospital to be evaluated can be effectively determined. Based on the medical record timeliness information, the medical record quality of the hospital to be evaluated as a whole can be effectively evaluated to obtain a medical record quality evaluation result. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 is a flow chart of a medical record quality assessment method provided by a first embodiment of the present invention;

[0052] Figure 2 is a flow chart of a medical record quality assessment method provided by a second embodiment of the present invention;

[0053] Figure 3 is a schematic structural diagram of a medical record quality assessment system provided by a third embodiment of the present invention;

[0054] Figure 4 It is a schematic structural diagram of a terminal device provided in the fourth embodiment of the present invention. DETAILED DESCRIPTION

[0055] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0056] In order to illustrate the technical solution of the present invention, specific embodiments are provided below.

[0057] Example 1

[0058] See also Figure 1 , is a flowchart of a medical record quality assessment method provided by the first embodiment of the present invention. The medical record quality assessment method can be applied to any terminal device or system. The medical record quality assessment method includes the following steps:

[0059] Step S10, obtaining sample medical records, and digitally encoding the sample medical records to obtain a digital code set;

[0060] Among them, medical record data is obtained from the medical information management system of each hospital, and the document ID of each sample medical record in the medical record data is used as the unique ID to be stored in a preset database. The data table in the preset database includes data such as document ID, document category, document content and discharge time.

[0061] In this step, the preset database can be set as required, for example, the preset database can be set as an hbase database. Preferably, by digitally encoding the sample medical records, each sample medical record can be effectively converted into digital text, which facilitates subsequent data processing operations on each sample medical record.

[0062] Step S20, generating a training data set according to the digital code set, and performing model training on the entity recognition model according to the training data set until the entity recognition model converges;

[0063] Among them, the entity recognition model is trained through the training data set, so that the converged entity recognition model can effectively perform entity recognition on the input medical records;

[0064] Optionally, in this step, the performing model training on the entity recognition model according to the training data set until the entity recognition model converges includes:

[0065] Inputting the training data set into the entity recognition model for entity recognition, and determining the total number of correct entities, the number of recognized entities, and the number of real entities based on the entity recognition results; wherein, by determining the total number of correct entities, the number of recognized entities, and the number of real entities in the current iteration round, the calculation of the loss value, precision, recall rate, and F1 value in the subsequent current iteration round is guaranteed;

[0066] Calculating the loss value, precision, recall rate, and F1 value of the entity recognition model based on the total number of correct entities, the number of recognized entities, and the number of real entities, and updating the parameters of the entity recognition model based on the loss value, wherein the F1 value is used to represent the harmonic mean of the precision and recall rate of the entity recognition model;

[0067] When the number of iterations of the entity recognition model is greater than a number threshold, the entity recognition model is determined to be converged, and the loss value, accuracy, recall rate, and F1 value of the entity recognition model of each iteration are respectively obtained; wherein the number threshold can be set according to needs, for example, the number threshold can be set to 10 times, 20 times, or 30 times;

[0068] An image is drawn based on the obtained loss value, accuracy, recall rate and F1 value of the entity recognition model to obtain a loss line graph, an accuracy line graph, a recall line graph and an F1 line graph; wherein, by drawing the loss line graph, the accuracy line graph, the recall line graph and the F1 line graph, it is convenient for the user to understand the changing status of the loss value, accuracy, recall rate and F1 value during the training process of the entity recognition model.

[0069] Step S30: Input the medical record information of the hospital to be evaluated into the converged entity recognition model to perform entity recognition, obtain an entity recognition result, and determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result;

[0070] During the entity recognition model training process, a medical knowledge graph encompassing over 5 million medical terms and their relationships was incorporated into the model training, enabling the model to fully learn the medical entity knowledge in the task data. This significantly improved performance across all tasks compared to solutions based on convolutional neural networks and long-short-term memory networks. The entity recognition model training was based on a standard medical named entity recognition task, encompassing nine entity categories: disease (dis), clinical manifestation (sym), drug (dru), medical device (equ), medical procedure (pro), body (bod), medical test item (ite), microbiology (mic), and department (dep).

[0071] In this step, the disease name and code, operation name and operation code of each electronic medical record text in the medical record information of the hospital to be evaluated are extracted based on the disease recognition model, and the patient name, hospitalization number, admission date, discharge date, admission department, operation department, discharge department and other information of each electronic medical record text in the medical record information of the hospital to be evaluated are extracted based on the entity recognition model to obtain the entity recognition results, among which the admission record form includes information such as patient name, patient hospitalization number, patient admission date, patient department and recording time.

[0072] Optionally, in this embodiment, the medical record timeliness information includes the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours;

[0073] The formula used to determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result includes:

[0074] The rate of admission records completed within 24 hours = the number of inpatient medical records completed within 24 hours of admission / the total number of inpatient medical records during the same period × 100%;

[0075] The number of inpatient medical records completed within 24 hours of the patient's admission was calculated as the number of records with the patient's admission date less the number of records with the patient's admission date completed within 24 hours of the admission date.

[0076] The completion rate of surgical records within 24 hours = the number of inpatient medical records with surgical records completed within 24 hours after surgery / the total number of inpatient medical records of surgical patients in the same period × 100%;

[0077] The number of inpatient medical records with surgical records completed within 24 hours after surgery = the recording time of the surgical record form - the number of medical records with surgical time less than 24 hours;

[0078] Completion rate of discharge records within 24 hours = number of medical records completed within 24 hours after discharge / total number of medical records of patients discharged during the same period × 100%;

[0079] The number of medical records completed within 24 hours after the patient's discharge = the recording time of the discharge record form - the number of medical records with the patient's discharge date < 24 hours;

[0080] Completion rate of the first page of medical records within 24 hours = the number of medical records with the first page completed within 24 hours after the patient is discharged / the total number of medical records of patients discharged during the same period × 100%;

[0081] The number of medical records with the first page completed within 24 hours after the patient is discharged = the recording time of the document record sheet on the first page of the medical record minus the number of medical records with the patient's discharge date completed within 24 hours;

[0082] The total number of medical records of inpatients during the same period = the number of records in the corresponding medical record homepage record table within the query start time and end time.

[0083] Step S40, performing a medical record quality assessment on the hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result;

[0084] The formula used to evaluate the medical record quality of the hospital to be evaluated based on the medical record timeliness information includes:

[0085] Medical record quality assessment results = a1 admission record completion rate within 24 hours + a2 surgical record completion rate within 24 hours + a3 discharge record completion rate within 24 hours + a4 medical record homepage completion rate within 24 hours;

[0086] Among them, a1, a2, a3, and a4 are the preset weights for the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours, respectively;

[0087] If the medical record quality assessment result is greater than the quality assessment threshold, the medical record quality assessment of the hospital to be assessed is determined to be qualified;

[0088] Among them, the quality assessment threshold can be set according to needs. If the medical record quality assessment result is greater than the quality assessment threshold, it is judged that the timeliness of the doctors in the hospital to be assessed in handling medical records is high. If the medical record quality assessment result is less than or equal to the quality assessment threshold, it is judged that the medical record quality assessment of the hospital to be assessed is unqualified, and the timeliness of the doctors in the hospital to be assessed in handling medical records is low.

[0089] In this embodiment, by digitally encoding the sample medical records, a digital code set is obtained. Based on the digital code set, a training data set can be effectively generated. By inputting the medical record information of the hospital to be evaluated into the converged entity recognition model for entity recognition, the entities in the medical record information can be effectively extracted. Based on the entity recognition results, the medical record timeliness information of the hospital to be evaluated can be effectively determined. Based on the medical record timeliness information, the medical record quality of the hospital to be evaluated as a whole can be effectively evaluated to obtain a medical record quality evaluation result. In this embodiment, information related to the medical record writing timeliness index can be extracted from the vast medical records of the hospital, and the indicator numerical results representing the medical record quality evaluation results can be automatically generated, thereby improving the quality of the hospital's medical record management.

[0090] Example 2

[0091] See also Figure 2 , is a flow chart of a medical record quality assessment method provided by a second embodiment of the present invention. This embodiment is used to further refine step S10 in the first embodiment, including the steps of:

[0092] Step S11, obtaining the medical record statement in the sample medical record;

[0093] Here, sentences in each sample medical record are obtained to obtain the medical record sentence. In this step, each line contains a word and its corresponding label. Words and labels are separated by "\t", and sentences are separated by blank lines.

[0094] In this step, after obtaining the medical record sentences, it is determined whether there are characters in each medical record sentence to determine whether the medical record sentence is a blank line. If it is determined that the medical record sentence is not a blank line, it is determined whether the sentence length of the medical record sentence is less than the preset character length.

[0095] Step S12: if the medical record statement contains characters and the length of the medical record statement is less than the preset character length, the characters in the medical record statement are coded and mapped respectively;

[0096] Among them, if there are characters in the medical record statement, it is determined that the medical record statement is not a blank line, and when the sentence length of the medical record statement is less than the preset character length, the characters in the medical record statement are encoded and mapped. The preset character length can be set according to needs;

[0097] In this step, if there are no characters in the medical record statement, the medical record statement is determined to be a blank line, that is, it means that the current medical record statement has ended, and the mapping result is saved.

[0098] Optionally, in this step, encoding and mapping the characters in the medical record statement respectively includes:

[0099] Matching each character with a pre-stored encoding mapping table; wherein the encoding mapping table stores the correspondence between different characters and corresponding encodings;

[0100] If the character successfully matches the encoding mapping table, the character is mapped with the matched encoding;

[0101] If the character does not match the code mapping table, the preset code is mapped to the character; wherein the preset code can be set as required. Optionally, in this step, if the character does not match the code mapping table, the current character can be replaced according to the preset character, and mapped according to the code corresponding to the replaced preset character. The preset character can be set as required, for example, the preset character can be set to "UNK";

[0102] Matching the label of each character with the encoding mapping table respectively; wherein the encoding mapping table also stores the correspondence between different labels and corresponding codes;

[0103] If the tag matches the encoding mapping table successfully, the matched encoding is mapped to the tag; optionally, if the tag does not match the encoding mapping table successfully, the tag is mapped according to a specified encoding, which can be set according to requirements.

[0104] Step S13, performing data encapsulation on the medical record statement after code mapping to obtain the digital code set;

[0105] By performing data encapsulation on the medical record statements after code mapping, the medical record statements can be effectively converted into a digital code set of a specified format; optionally, in this embodiment, with respect to step S20, generating a training data set based on the digital code set includes:

[0106] Extracting the sentence features and labels of each medical record sentence in the digital code set to obtain feature codes and label codes;

[0107] Performing tensor encapsulation on the feature code and the label code to obtain a feature tensor and a label tensor, and performing dataset encapsulation on the feature tensor and the label tensor to obtain the training dataset;

[0108] Among them, by performing feature extraction on the encoding corresponding to the sentence features and labels of each medical record sentence, the feature encoding and label encoding corresponding to each medical record sentence can be effectively obtained. By performing tensor encapsulation on the feature encoding and label encoding, the feature encoding and label encoding can be effectively converted into tensors. And by performing data set encapsulation on the feature tensor and label tensor, the tensor can be effectively converted into a training data set.

[0109] In this embodiment, whether each medical record statement is a blank line is determined by judging whether there are characters in each medical record statement. If there are characters in the medical record statement, it is determined that the medical record statement is not a blank line, and when the statement length of the medical record statement is less than the preset character length, the characters in the medical record statement are coded and mapped. By data encapsulating the medical record statement after coding mapping, the medical record statement can be effectively converted into a digital code set of a specified format.

[0110] Example 3

[0111] See also Figure 3 , is a schematic diagram of the structure of a medical record quality assessment system 100 provided in a third embodiment of the present invention, comprising: a digital encoding module 10, a model training module 11, an entity recognition module 12, and a quality assessment module 13, wherein:

[0112] The digital coding module 10 is used to obtain sample medical records and digitally encode the sample medical records to obtain a digital code set.

[0113] Optionally, the digital encoding module 10 is further used to: obtain a medical record statement in the sample medical record;

[0114] If the medical record statement contains characters and the sentence length of the medical record statement is less than the preset character length, the characters in the medical record statement are respectively coded and mapped;

[0115] The medical record statement after code mapping is data encapsulated to obtain the digital code set.

[0116] Furthermore, the digital encoding module 10 is further configured to: match each character with a pre-stored encoding mapping table;

[0117] If the character successfully matches the encoding mapping table, the character is mapped with the matched encoding;

[0118] If the character does not match the code mapping table, mapping the character with a preset code;

[0119] Matching the label of each character with the encoding mapping table respectively;

[0120] If the tag matches the code mapping table successfully, the matched code is mapped to the tag.

[0121] The model training module 11 is used to generate a training data set according to the digital code set, and perform model training on the entity recognition model according to the training data set until the entity recognition model converges.

[0122] Optionally, the model training module 11 is further configured to: extract the sentence features and the codes corresponding to the labels of each medical record sentence in the digital code set, respectively, to obtain feature codes and label codes;

[0123] The feature code and the label code are tensor-encapsulated to obtain a feature tensor and a label tensor, and the feature tensor and the label tensor are dataset-encapsulated to obtain the training dataset.

[0124] Furthermore, the model training module 11 is further configured to: input the training data set into the entity recognition model for entity recognition, and determine the total number of correct entities, the number of recognized entities, and the number of real entities based on the entity recognition results;

[0125] Calculating the loss value, precision, recall rate, and F1 value of the entity recognition model based on the total number of correct entities, the number of recognized entities, and the number of real entities, and updating the parameters of the entity recognition model based on the loss value, wherein the F1 value is used to represent the harmonic mean of the precision of the entity recognition model and the recall rate;

[0126] When the number of iterations of the entity recognition model is greater than a threshold number, the entity recognition model is determined to be converged, and the loss value, accuracy, recall rate and F1 value of the entity recognition model of each iteration are obtained respectively;

[0127] An image is drawn based on the obtained loss value, accuracy, recall rate and F1 value of the entity recognition model to obtain a loss line graph, an accuracy line graph, a recall line graph and an F1 line graph.

[0128] The entity recognition module 12 is used to input the medical record information of the hospital to be evaluated into the converged entity recognition model to perform entity recognition, obtain entity recognition results, and determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition results.

[0129] Optionally, in this embodiment, the medical record timeliness information includes the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours;

[0130] The formula used to determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result includes:

[0131] The rate of admission records completed within 24 hours = the number of inpatient medical records completed within 24 hours of admission / the total number of inpatient medical records during the same period × 100%;

[0132] The number of inpatient medical records completed within 24 hours of the patient's admission was calculated as the number of records with the patient's admission date less the number of records with the patient's admission date completed within 24 hours of the admission date.

[0133] The completion rate of surgical records within 24 hours = the number of inpatient medical records with surgical records completed within 24 hours after surgery / the total number of inpatient medical records of surgical patients in the same period × 100%;

[0134] The number of inpatient medical records with surgical records completed within 24 hours after surgery = the recording time of the surgical record form - the number of medical records with surgical time less than 24 hours;

[0135] Completion rate of discharge records within 24 hours = number of medical records completed within 24 hours after discharge / total number of medical records of patients discharged during the same period × 100%;

[0136] The number of medical records completed within 24 hours after the patient's discharge = the recording time of the discharge record form - the number of medical records with the patient's discharge date < 24 hours;

[0137] Completion rate of the first page of medical records within 24 hours = the number of medical records with the first page completed within 24 hours after the patient is discharged / the total number of medical records of patients discharged during the same period × 100%;

[0138] The number of medical records with the first page completed within 24 hours after the patient is discharged = the recording time of the document record sheet on the first page of the medical record minus the number of medical records with the patient's discharge date completed within 24 hours;

[0139] The total number of medical records of inpatients during the same period = the number of records in the corresponding medical record homepage record table within the query start time and end time.

[0140] The quality assessment module 13 is configured to perform a medical record quality assessment on the hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result.

[0141] Optionally, in this embodiment, the formula used to evaluate the medical record quality of the hospital to be evaluated based on the medical record timeliness information includes:

[0142] Medical record quality assessment results = a1 admission record completion rate within 24 hours + a2 surgical record completion rate within 24 hours + a3 discharge record completion rate within 24 hours + a4 medical record homepage completion rate within 24 hours;

[0143] Among them, a1, a2, a3, and a4 are the preset weights for the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours, respectively;

[0144] If the medical record quality assessment result is greater than the quality assessment threshold, it is determined that the medical record quality assessment of the hospital to be assessed is qualified.

[0145] In this embodiment, by digitally encoding the sample medical records, a digital code set is obtained. Based on the digital code set, a training data set can be effectively generated. By inputting the medical record information of the hospital to be evaluated into the converged entity recognition model for entity recognition, entities in the medical record information can be effectively extracted. Based on the entity recognition results, the medical record timeliness information of the hospital to be evaluated can be effectively determined. Based on the medical record timeliness information, the medical record quality of the hospital to be evaluated as a whole can be effectively evaluated to obtain a medical record quality evaluation result.

[0146] Example 4

[0147] Figure 4 This is a structural block diagram of a terminal device 2 provided in the fourth embodiment of the present application. Figure 4 As shown, the terminal device 2 of this embodiment includes: a processor 20, a memory 21, and a computer program 22 stored in the memory 21 and executable on the processor 20, such as a program for the medical record quality assessment method. When the processor 20 executes the computer program 22, the steps of each embodiment of the medical record quality assessment method described above are implemented.

[0148] Exemplarily, the computer program 22 may be divided into one or more modules, which are stored in the memory 21 and executed by the processor 20 to implement the present application. The one or more modules may be a series of computer program instruction segments capable of performing specific functions, which are used to describe the execution process of the computer program 22 in the terminal device 2. The terminal device may include, but is not limited to, a processor 20 and a memory 21.

[0149] The processor 20 may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA), other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. A general-purpose processor may be a microprocessor or any conventional processor.

[0150] The memory 21 may be an internal storage unit of the terminal device 2, such as a hard disk or memory of the terminal device 2. The memory 21 may also be an external storage device of the terminal device 2, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the terminal device 2. Furthermore, the memory 21 may include both an internal storage unit of the terminal device 2 and an external storage device. The memory 21 is used to store the computer program and other programs and data required by the terminal device. The memory 21 may also be used to temporarily store data that has been output or is about to be output.

[0151] In addition, the functional modules in the various embodiments of the present application may be integrated into a single processing unit, or each unit may exist physically separately, or two or more units may be integrated into a single unit. The aforementioned integrated units may be implemented in the form of hardware or software functional units.

[0152] If the integrated module 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. Among them, the computer-readable storage medium can be non-volatile or volatile. Based on this understanding, the present application implements all or part of the process in the above-mentioned embodiment method, and can also be completed by instructing the relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium, and the computer program can implement the steps of the above-mentioned various method embodiments when executed by the processor. 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, etc. The computer-readable storage medium may include: any entity or device that can carry computer program code, recording medium, U disk, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), electric carrier signal, telecommunication signal and software distribution medium, etc. It should be noted that the content contained in computer-readable storage media can be appropriately increased or decreased according to the requirements of legislation and patent practices in the jurisdiction. For example, in some jurisdictions, according to legislation and patent practices, computer-readable storage media do not include electrical carrier signals and telecommunications signals.

[0153] The above-described embodiments 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, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. These modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present application, and should all be included in the scope of protection of the present application.

Claims

1. A medical record quality assessment method, characterized in that: The method comprises: Obtaining sample medical records, and digitally encoding the sample medical records to obtain a digital code set; Generating a training data set according to the digital code set, and performing model training on an entity recognition model according to the training data set until the entity recognition model converges; Inputting the medical record information of the hospital to be evaluated into the converged entity recognition model to perform entity recognition, obtaining an entity recognition result, and determining the medical record timeliness information of the hospital to be evaluated based on the entity recognition result; Performing a medical record quality assessment on the hospital to be assessed based on the medical record timeliness information to obtain a medical record quality assessment result; The performing model training on the entity recognition model according to the training data set until the entity recognition model converges includes: Inputting the training data set into the entity recognition model to perform entity recognition, and determining the total number of correct entities, the number of recognized entities, and the number of real entities based on the entity recognition results; Calculating the loss value, precision, recall rate, and F1 value of the entity recognition model based on the total number of correct entities, the number of recognized entities, and the number of real entities, and updating the parameters of the entity recognition model based on the loss value, wherein the F1 value is used to represent the harmonic mean of the precision of the entity recognition model and the recall rate; When the number of iterations of the entity recognition model is greater than a threshold number, the entity recognition model is determined to be converged, and the loss value, accuracy, recall rate and F1 value of the entity recognition model of each iteration are obtained respectively; An image is drawn based on the obtained loss value, accuracy, recall rate and F1 value of the entity recognition model to obtain a loss line graph, an accuracy line graph, a recall line graph and an F1 line graph.

2. The medical record quality assessment method according to claim 1, wherein: The medical record timeliness information includes the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours; The formula used to determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result includes: The rate of admission records completed within 24 hours = the number of inpatient medical records completed within 24 hours of admission / the total number of inpatient medical records during the same period × 100%; The number of inpatient medical records completed within 24 hours of the patient's admission was calculated as the number of records with the patient's admission date less the number of records with the patient's admission date completed within 24 hours of the admission date. The completion rate of surgical records within 24 hours = the number of inpatient medical records with surgical records completed within 24 hours after surgery / the total number of inpatient medical records of surgical patients in the same period × 100%; The number of inpatient medical records with surgical records completed within 24 hours after surgery = the recording time of the surgical record form - the number of medical records with surgical time less than 24 hours; Completion rate of discharge records within 24 hours = number of medical records completed within 24 hours after discharge / total number of medical records of patients discharged during the same period × 100%; The number of medical records completed within 24 hours after the patient's discharge = the recording time of the discharge record form - the number of medical records with the patient's discharge date < 24 hours; Completion rate of the first page of medical records within 24 hours = the number of medical records with the first page completed within 24 hours after the patient is discharged / the total number of medical records of patients discharged during the same period × 100%; The number of medical records with the first page completed within 24 hours after the patient is discharged = the recording time of the medical record first page document record form minus the number of medical records with the patient's discharge date completed within 24 hours; The total number of medical records of inpatients during the same period = the number of records in the corresponding medical record homepage record table within the query start time and end time.

3. The medical record quality assessment method according to claim 2, wherein: The formula used to evaluate the medical record quality of the hospital to be evaluated based on the medical record timeliness information includes: Medical record quality assessment results = a1 admission record completion rate within 24 hours + a2 surgical record completion rate within 24 hours + a3 discharge record completion rate within 24 hours + a4 medical record homepage completion rate within 24 hours; Among them, a1, a2, a3, and a4 are the preset weights for the completion rate of admission records within 24 hours, the completion rate of surgical records within 24 hours, the completion rate of discharge records within 24 hours, and the completion rate of the medical record homepage within 24 hours, respectively; If the medical record quality assessment result is greater than the quality assessment threshold, it is determined that the medical record quality assessment of the hospital to be assessed is qualified.

4. The medical record quality assessment method according to claim 1, wherein: The digital coding of the sample medical records to obtain a digital coding set includes: Obtaining a medical record statement in the sample medical record; If the medical record statement contains characters and the sentence length of the medical record statement is less than the preset character length, the characters in the medical record statement are respectively coded and mapped; The medical record statement after code mapping is data encapsulated to obtain the digital code set.

5. The medical record quality assessment method according to claim 4, wherein: The encoding and mapping of the characters in the medical record statements respectively includes: Match each character with the pre-stored encoding mapping table; If the character successfully matches the encoding mapping table, the character is mapped with the matched encoding; If the character does not match the code mapping table, mapping the character with a preset code; Matching the label of each character with the encoding mapping table respectively; If the tag matches the code mapping table successfully, the matched code is mapped to the tag.

6. The medical record quality assessment method according to claim 5, wherein: Generating a training data set according to the digital code set includes: Extracting the sentence features and labels of each medical record sentence in the digital code set to obtain feature codes and label codes; The feature code and the label code are tensor-encapsulated to obtain a feature tensor and a label tensor, and the feature tensor and the label tensor are dataset-encapsulated to obtain the training dataset.

7. A medical record quality assessment system, characterized in that: The system comprises: A digital coding module is used to obtain sample medical records and digitally encode the sample medical records to obtain a digital code set; A model training module, configured to generate a training data set according to the digital code set, and perform model training on an entity recognition model according to the training data set until the entity recognition model converges; An entity recognition module is used to input the medical record information of the hospital to be evaluated into the converged entity recognition model to perform entity recognition, obtain an entity recognition result, and determine the medical record timeliness information of the hospital to be evaluated based on the entity recognition result; A quality assessment module, configured to perform a medical record quality assessment on the hospital to be assessed based on the medical record timeliness information, and obtain a medical record quality assessment result; The model training module is further configured to: input the training data set into the entity recognition model for entity recognition, and determine the total number of correct entities, the number of recognized entities, and the number of real entities based on the entity recognition results; Calculating the loss value, precision, recall rate, and F1 value of the entity recognition model based on the total number of correct entities, the number of recognized entities, and the number of real entities, and updating the parameters of the entity recognition model based on the loss value, wherein the F1 value is used to represent the harmonic mean of the precision of the entity recognition model and the recall rate; When the number of iterations of the entity recognition model is greater than a threshold number, the entity recognition model is determined to be converged, and the loss value, accuracy, recall rate and F1 value of the entity recognition model of each iteration are obtained respectively; An image is drawn based on the obtained loss value, accuracy, recall rate and F1 value of the entity recognition model to obtain a loss line graph, an accuracy line graph, a recall line graph and an F1 line graph.

8. A terminal 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 steps of the method according to any one of claims 1 to 6 are implemented.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.

Images