Structured medical record generation method and device

Abstract

The invention discloses a structured medical record generation method and device, and the method comprises the steps: obtaining a pre-constructed text hot word bank which comprises a plurality of text hot words; obtaining target dialogue voice information; determining first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information; determining second dialogue text information based on the text hot word bank and the first dialogue text information; and generating the target structured medical record text based on the second dialogue text information. According to the invention, the accuracy and normalization of medical record generation can be improved.

Description

Technical Field

[0001] The present application relates to the field of artificial intelligence processing technology, and specifically to a method and device for generating structured medical records. Background Art

[0002] The heavy paperwork in the traditional medical record recording process also brings a heavy burden to the medical treatment process, including extended recording time, extended working hours, poor communication between doctors and patients, etc., which are important factors leading to doctor burnout and reduced quality of diagnosis and treatment. In addition, incomplete, non-standard, low-quality medical records, diagnostic differences, patient dissatisfaction, and even affect patient safety may occur. Therefore, it is urgent to provide a new technology that can simultaneously reduce the pressure of medical record writing for doctors and improve the efficiency and quality of recording. In recent years, in the field of automatic speech recognition (ASR), ASR systems have been rapidly deployed in many industries such as finance, healthcare, and transportation, showing their great application potential. However, the language models (LM) in the ASR system still have obvious limitations when recognizing domain-specific nouns, and often mistakenly transcribe them as common words in the vocabulary, which will seriously affect the accuracy of ASR results, limit its promotion and application in the medical field, and lead to low accuracy of medical record generation. Summary of the invention

[0003] The embodiments of the present application provide a structured medical record generation method and device, which can improve the accuracy of medical record generation.

[0004] In a first aspect, the present application provides a method for generating a structured medical record, comprising:

[0005] Acquire a pre-built text hot word library, wherein the text hot word library contains a plurality of text hot words;

[0006] Obtain target conversation voice information;

[0007] Determining first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information;

[0008] Determine second dialogue text information based on the text hot word library and the first dialogue text information;

[0009] Generate a target structured medical record text based on the second conversation text information.

[0010] Optionally, the determining the second dialogue text information based on the text hot word library and the first dialogue text information includes:

[0011] Acquire multiple text segmentations obtained by segmenting the first conversation text information;

[0012] At least part of the text segmentation in the first dialogue text information is replaced based on a plurality of the text hot words in the text hot word library to obtain the second dialogue text information.

[0013] Optionally, the replacing at least part of the text segmentation in the first dialogue text information based on a plurality of the text hot words in the text hot word library to obtain the second dialogue text information includes:

[0014] Calculating the similarity index of the text segmentation and each text hot word in the text hot word library respectively;

[0015] When the segmentation similarity index is not lower than a preset index value, the text segmentation corresponding to the segmentation similarity index in the first dialogue text information is replaced with the corresponding text hot word to obtain the second dialogue text information.

[0016] Optionally, respectively calculating the text segmentation and the segmentation similarity index of each text hot word in the text hot word library includes:

[0017] Obtain the total number of pinyin letters of the pinyin of the text segmentation and the number of pinyin letters of the pinyin of the text hot words;

[0018] Obtaining the longest identical letter sequence between the pinyin of the text segmentation and the pinyin of the text hot word, wherein the longest identical letter sequence exists in both the text segmentation and the text hot word;

[0019] The ratio of the number of letters in the longest identical letter sequence to the total number of pinyin letters is determined as the segmentation similarity index of the text segmentation and the text hot word.

[0020] Optionally, the determining, based on the target conversation voice information, first conversation text information corresponding to the target conversation voice information includes:

[0021] Inputting the target dialogue voice information into a speech recognition model to obtain third dialogue text information corresponding to the target dialogue voice information, wherein the speech recognition model is a whisper model or a FunASR model;

[0022] The first dialogue text information corresponding to the target dialogue voice information is determined based on the third dialogue text information.

[0023] Optionally, the determining, based on the third dialogue text information, the first dialogue text information corresponding to the target dialogue voice information includes:

[0024] Classifying each text segmentation in the third dialogue text information to obtain a word category of each text segmentation;

[0025] The text segmentation of the word categories in the third dialogue text information that belong to the preset category set is removed to obtain the first dialogue text information corresponding to the target dialogue voice information.

[0026] Optionally, acquiring the target conversation voice information includes:

[0027] Obtaining initial conversation voice information;

[0028] Splitting the initial conversation voice information into a plurality of conversation voice information segments;

[0029] Detecting background noise parameters of each of the dialogue voice information segments;

[0030] When the background noise parameter of the conversation voice information segment is lower than a preset parameter value, the conversation voice information segment is determined as the target conversation voice information.

[0031] In a second aspect, the structured medical record generating device provided by the present application includes:

[0032] A first acquisition module is used to acquire a pre-built text hot word library, wherein the text hot word library contains a plurality of text hot words;

[0033] The second acquisition module is used to acquire the target conversation voice information;

[0034] A first determining module, configured to determine first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information;

[0035] A second determination module, configured to determine second dialogue text information based on the text hot word library and the first dialogue text information;

[0036] A generation module is used to generate a target structured medical record text based on the second conversation text information.

[0037] In a third aspect, the electronic device provided in the present application includes a memory and a processor, the memory stores a computer program, and the processor is used to run the computer program in the memory to implement the steps in the structured medical record generation method provided in the present application.

[0038] In a fourth aspect, the computer-readable storage medium provided in the present application stores a plurality of instructions, which are suitable for loading by a processor to implement the steps in the structured medical record generation method provided in the present application.

[0039] In a fifth aspect, the computer program product provided in the present application includes a computer program or instructions, which, when executed by a processor, implements the steps in the structured medical record generation method provided in the present application.

[0040] In this application, compared with the related art, a pre-built text hot word library is obtained, wherein the text hot word library contains multiple text hot words; target conversation voice information is obtained; first conversation text information corresponding to the target conversation voice information is determined based on the target conversation voice information; second conversation text information is determined based on the text hot word library and the first conversation text information; and target structured medical record text is generated based on the second conversation text information. This application can improve the accuracy of medical record generation. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 is a scenario diagram of a structured medical record generation system provided in an embodiment of the present application;

[0043] Figure 2 It is a flowchart of an embodiment of a method for generating a structured medical record provided in an embodiment of the present application;

[0044] Figure 3 is a schematic diagram of a text hot word library in an embodiment of the structured medical record generation method provided in an embodiment of the present application;

[0045] Figure 4 It is a schematic diagram of calculating the word segmentation similarity index in one embodiment of the structured medical record generation method provided in the embodiment of the present application;

[0046] Figure 5 is a schematic diagram of converting first conversation text information into second conversation text information in an embodiment of the structured medical record generation method provided by an embodiment of the present application;

[0047] Figure 6 is a schematic diagram of the structure of a structured medical record generating device provided in an embodiment of the present application;

[0048] Figure 7 It is a schematic diagram of the structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0049] It should be noted that the principles of the present application are illustrated by implementing them in an appropriate computing environment. The following description is based on the illustrated specific embodiments of the present application and should not be considered as limiting other specific embodiments of the present application that are not described in detail herein.

[0050] In the following description of the present application, reference is made to “some embodiments”, which describe a subset of all possible embodiments, but it can be understood that “some embodiments” may be the same subset or different subsets of all possible embodiments and may be combined with each other without conflict.

[0051] In the following description of the present application, the terms "first\second\third" involved are merely used to distinguish similar objects and do not represent a specific ordering of the objects. It can be understood that "first\second\third" can be interchanged with a specific order or sequence where permitted, so that the embodiments of the present application described here can be implemented in an order other than that illustrated or described here.

[0052] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those commonly understood by those skilled in the art to which this application belongs. The terms used herein are only for the purpose of describing the embodiments of this application and are not intended to limit this application.

[0053] In order to improve the effect of structured medical record generation, the embodiments of the present application provide a structured medical record generation method, a structured medical record generation device, an electronic device, a computer-readable storage medium, and a computer program product. The structured medical record generation method can be executed by the structured medical record generation device, or by an electronic device integrated with the structured medical record generation device.

[0054] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of this application.

[0055] Please refer to Figure 1 , the present application also provides a structured medical record generation system, such as Figure 1 As shown, the structured medical record generation system electronic device 100, the electronic device 100 is integrated with the structured medical record generation device provided by the present application.

[0056] Among them, the electronic device 100 can be any device equipped with a processor and has processing capabilities, such as mobile electronic devices with processors such as smart phones, tablet computers, PDAs, laptops, smart speakers, or fixed electronic devices with processors such as desktop computers, televisions, servers, industrial equipment, etc.

[0057] In addition, if Figure 1 As shown, the structured medical record generation system may further include a memory 200 for storing original data, intermediate data, and result data.

[0058] In the embodiment of the present application, the memory 200 can be a cloud storage. Cloud storage is a new concept extended and developed from the concept of cloud computing. A distributed cloud storage system (hereinafter referred to as the storage system) refers to a storage system that uses cluster applications, grid technology, and distributed storage file systems to bring together a large number of different types of storage devices (storage devices are also called storage nodes) in the network through application software or application interfaces to work together and provide external data storage and business access functions.

[0059] At present, the storage method of the storage system is: create a logical volume, and when creating a logical volume, allocate physical storage space for each logical volume. The physical storage space may be composed of disks of a storage device or several storage devices. The client stores data on a logical volume, that is, stores the data on the file system. The file system divides the data into many parts, each of which is an object. The object contains not only data but also additional information such as data identification (ID, ID entity). The file system writes each object into the physical storage space of the logical volume, and the file system records the storage location information of each object, so that when the client requests to access the data, the file system can allow the client to access the data according to the storage location information of each object.

[0060] The process of the storage system allocating physical storage space to a logical volume is as follows: based on the estimated capacity of the objects stored in the logical volume (this estimate often has a large margin relative to the actual capacity of the objects to be stored) and the grouping of independent redundant disk arrays (RAID, Redundant Array of Independent Disks), the physical storage space is pre-divided into stripes. A logical volume can be understood as a stripe, thereby allocating physical storage space to the logical volume.

[0061] It should be noted that Figure 1The scenario diagram of the structured medical record generation system shown is merely an example. The structured medical record generation system and scenario described in the embodiment of the present application are intended to more clearly illustrate the technical solution of the embodiment of the present application, and do not constitute a limitation on the technical solution provided in the embodiment of the present application. A person of ordinary skill in the art can appreciate that with the evolution of the structured medical record generation system and the emergence of new business scenarios, the technical solution provided in the embodiment of the present application is equally applicable to similar technical problems.

[0062] It should be noted that the serial numbers of the following embodiments are not intended to limit the preferred order of the embodiments.

[0063] Please refer to Figure 2 , Figure 2 FIG. 1 is a flow chart of an embodiment of a method for generating a structured medical record provided in an embodiment of the present application. Figure 2 As shown, the process of the structured medical record generation method provided by this application is as follows:

[0064] 201. Obtain a pre-built text hot word library.

[0065] Among them, the text hot word library contains multiple text hot words.

[0066] In the embodiment of the present application, the audio of the medical consultation dialogue is obtained. Through standardized recording equipment, the audio of the medical consultation dialogue with sensitive information such as identity identification and home address hidden is obtained. According to international guidelines and expert consensus, a text hot word library is constructed. A clinical medicine professional text hot word library is formulated by clinical expert physicians based on international guidelines and expert consensus. The text hot word library consists of medical professional terms and their corresponding pinyin without tone markings, wherein the medical professional terms include but are not limited to common symptoms, auxiliary examination related terms (including but not limited to laboratory tests, imaging examinations, pathological histological examinations, endoscopy, electrocardiograms, etc.), related terms of commonly used clinical diagnostic techniques, disease names of various systems, anatomy-related terms, chemical names of commonly used drugs in various departments, trade names and their abbreviations, etc. Since the strategy of replacing hot words according to pronunciation is adopted, in order to reduce the situation where common homophones are replaced incorrectly, it is necessary to exempt hot words with common pronunciation and single-word professional vocabulary that overlaps with the pinyin of common characters, and remove these words.

[0067] like Figure 3 As shown, the text hot word library includes text hot words and their pinyins of multiple hot word categories. The multiple hot word categories may include common symptom categories, auxiliary examination categories, clinical common diagnostic technology categories, disease name categories of each system, anatomical categories, and drug categories. For example, the text hot words of the common symptom category include: dysphagia, chest pain, and jaundice.

[0068] In an embodiment of the present application, obtaining a pre-constructed text hot word library includes: obtaining a plurality of initial hot words uploaded manually, obtaining a plurality of consultation dialogue audios, performing speech recognition on the plurality of consultation dialogue audios to obtain a plurality of consultation text information, segmenting the consultation text information to obtain a plurality of consultation word segments, obtaining the occurrence frequency of each consultation word segment, determining the consultation word segments with occurrence frequencies higher than a preset frequency as popular word segments to obtain a plurality of popular word segments, determining whether there are popular word segments among the plurality of popular word segments that have the same pinyin as the initial hot words but are different from the initial hot words, and if there are popular word segments among the plurality of popular word segments that have the same pinyin as the initial hot words but are different from the initial hot words, deleting the initial hot words from the plurality of initial hot words to obtain a plurality of text hot words as the text hot word library.

[0069] This refers to deleting some hot words or professional terms with very common pronunciations. Since the model replaces hot words based on pronunciation, it is necessary to delete some hot words with many similar pronunciations from the corpus. For example, it is necessary to delete a medicine called "Shiyi" because "eleven", "suitable", etc. are also very common in conversations and may cause incorrect replacements. There are also drugs like "Nexium", "pancreatic neck", "pancreatic tail", and some single characters like liver, lung, etc.

[0070] 202. Obtain the target dialogue voice information.

[0071] In a specific embodiment, the target dialogue voice information may be the dialogue information between a doctor and a patient during a consultation.

[0072] In another specific embodiment, obtaining the target dialogue voice information includes:

[0073] (1) Obtain the initial dialogue voice information.

[0074] In an embodiment of the present application, the initial dialogue voice information may be the dialogue information between a doctor and a patient during a consultation.

[0075] (2) Split the initial dialogue voice information into multiple dialogue voice information segments.

[0076] In an embodiment of the present application, the initial dialogue voice information is split into multiple dialogue voice information segments of a preset length according to the preset length.

[0077] (3) Detect the background noise parameters of each dialogue voice information segment.

[0078] The background noise parameter is the background noise, also known as ambient noise, generally referring to the total noise in an electroacoustic system except the useful signal: including two parts, the noise of audio equipment and the noise of the playback environment. For example, the "rustling" sound in the TV sound except the program sound. Excessive background noise not only makes people irritable but also drowns out the weaker details in the sound.

[0079] (4) When the background noise parameter of the conversation voice information segment is lower than a preset parameter value, the conversation voice information segment is determined as the target conversation voice information.

[0080] Among them, the preset parameter value can be set according to the specific situation.

[0081] Furthermore, when the background noise parameter of the conversation voice information segment is lower than a preset parameter value, the conversation voice information segment is input into a speaker recognition model to obtain the number of speakers of the conversation voice information segment. If the number of speakers is 1, the conversation voice information segment is determined as the target conversation voice information.

[0082] 203. Determine first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information.

[0083] In the embodiment of the present application, determining the first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information includes:

[0084] (1) Inputting the target dialogue voice information into the speech recognition model to obtain third dialogue text information corresponding to the target dialogue voice information.

[0085] The speech recognition model is a whisper model or a FunASR model.

[0086] (2) Determine the first dialogue text information corresponding to the target dialogue voice information based on the third dialogue text information.

[0087] In a specific embodiment, the third dialogue text information is determined as the first dialogue text information corresponding to the target dialogue voice information.

[0088] In another specific embodiment, determining the first dialogue text information corresponding to the target dialogue voice information based on the third dialogue text information includes: classifying each text segmentation in the third dialogue text information to obtain the word category of each text segmentation; and removing the text segmentations whose word categories in the third dialogue text information belong to a preset category set to obtain the first dialogue text information corresponding to the target dialogue voice information.

[0089] In the embodiment of the present application, a segmentation model such as the Jieba segmentation model can be used to perform segmentation processing on the third dialogue text information to obtain multiple text segmentations.

[0090] In the embodiment of the present application, the preset category set includes multiple word categories. The preset category set may include: name category, address category, contact information category. The word category in the preset category set corresponds to the user's name, address, contact information and other sensitive information, which needs to be removed.

[0091] In an embodiment of the present application, a convolutional neural network is pre-trained, and each text segmentation in the third dialogue text information is input into the convolutional neural network for classification to obtain a word category of each text segmentation.

[0092] 204. Determine second dialogue text information based on the text hot word library and the first dialogue text information.

[0093] In the embodiment of the present application, determining the second dialogue text information based on the text hot word library and the first dialogue text information includes:

[0094] (1) Obtain multiple text segmentations obtained by segmenting the first conversation text information.

[0095] In the embodiment of the present application, a segmentation model such as the Jieba segmentation model can be used to perform segmentation processing on the first conversation text information to obtain multiple text segmentations.

[0096] (2) At least part of the text segmentation in the first dialogue text information is replaced based on a plurality of text hot words in the text hot word library to obtain second dialogue text information.

[0097] In the embodiment of the present application, at least part of the text segmentation in the first dialogue text information is replaced based on multiple text hot words in the text hot word library to obtain the second dialogue text information, including:

[0098] (1) Calculate the similarity index of the text segmentation and each hot word in the text hot word library respectively.

[0099] In a specific embodiment, the text segmentation and the segmentation similarity index of each text hot word in the text hot word library are calculated respectively, including:

[0100] First, the total number of pinyin letters of the pinyin of the text segmentation and the number of pinyin letters of the pinyin of the text hot words is obtained.

[0101] Records A and B represent text segmentation A and text hot word B respectively, and py(A) and py(B) represent the number of pinyin letters of the pinyin of text segmentation A and text hot word B respectively.

[0102] Secondly, the longest identical letter sequence between the pinyin of the text segmentation and the pinyin of the text hot word is obtained, wherein the longest identical letter sequence exists in both the text segmentation and the text hot word.

[0103] Specifically, multiple consecutive letters in the pinyin of text segmentation are concatenated as a first letter sequence to obtain multiple first letter sequences, multiple consecutive letters in the pinyin of text hot words are concatenated as a second letter sequence to obtain multiple second letter sequences, the same group of first letter sequences and second letter sequences are concatenated as an identical letter combination to obtain multiple groups of identical letter combinations, and multiple first letter sequences corresponding to the multiple groups of identical letter combinations are obtained, and the first letter sequence with the largest number of letters in the multiple first letter sequences corresponding to the multiple groups of identical letter combinations is determined as the longest identical letter sequence.

[0104] Finally, the ratio of the number of letters in the longest identical letter sequence to the total number of pinyin letters is determined as the segmentation similarity index of text segmentation and text hot words.

[0105] Specifically, record LCS(py(A), py(B)) to represent the number of letters in the longest identical letter sequence of the pinyin of text segment A and text hot word B.

[0106] Among them, the word segmentation similarity index similarity(A,B) satisfies the following formula:

[0107]

[0108] Among them, A and B represent text segmentation A and text hot word B respectively, py(A) and py(B) represent the number of pinyin letters of text segmentation A and text hot word B respectively, and LCS(py(A), py(B)) represents the number of letters of the longest identical letter sequence of the pinyin of text segmentation A and text hot word B.

[0109] like Figure 4 As shown, for example, text segmentation A is yihuo, text hot word B is yihuo, the number of pinyin letters of the pinyin of text segmentation A is 5, the number of pinyin letters of the pinyin of text hot word B is 5, the longest identical letter sequence of the pinyin of text segmentation A and text hot word B is yihuo, the number of letters in the longest identical letter sequence LCS(py(A), py(B)) is 5, and the segmentation similarity index similarity(A,B) is 1.

[0110] For example, text segmentation A is dimethyl suangua, and text hot word B is metformin. The number of pinyin letters of text segmentation A is 12, and the number of pinyin letters of text hot word B is 14. The longest identical letter sequence of the pinyin of text segmentation A and text hot word B is erjiasuangua, the number of letters in the longest identical letter sequence LCS(py(A), py(B)) is 12, and the segmentation similarity index similarity(A,B) is 0.92.

[0111] (2) When the word segmentation similarity index is not lower than a preset index value, the text segmentation corresponding to the word segmentation similarity index in the first dialogue text information is replaced with the corresponding text hot word to obtain the second dialogue text information.

[0112] In the embodiment of the present application, the preset indicator value is 1. In other embodiments, the preset indicator value may be other values.

[0113] like Figure 5 As shown, the first dialogue text information is: "That was caused by my brother", "brother" is a text segmentation word, and the text hot word corresponding to "brother" is "hiccups". "Brother" is replaced with "hiccups", and the second dialogue text information is: "That was caused by hiccups". Based on the same principle, the first dialogue text information is: "It's a problem with the tube" and the second dialogue text information is: "Problems with the esophagus" through hot word replacement.

[0114] The text segmentation corresponding to the segmentation similarity index in the first dialogue text information is replaced with the corresponding text hot word to obtain the second dialogue text information.

[0115] 205. Generate a target structured medical record text based on the second conversation text information.

[0116] In an embodiment of the present application, a pre-fine-tuned large language model in the medical field is accessed to build a structured medical record transcription model, and the target structured medical record text is automatically generated based on the second conversation text information.

[0117] Specifically, multiple second conversation text information is input into the structured medical record transcription model to automatically generate a target structured medical record text.

[0118] In the embodiment of the present application, multiple historical structured medical record texts are obtained, and the regional texts of each text region in the historical structured medical record texts are obtained to obtain multiple regional texts in the historical structured medical record texts. The multiple regional texts are used as a sample, and the historical structured medical record texts corresponding to the multiple regional texts are used as labels to construct a training set to train a structured medical record transcription model.

[0119] In order to facilitate better implementation of the structured medical record generation method provided in the embodiment of the present application, the embodiment of the present application also provides a structured medical record generation device based on the structured medical record generation method. The meanings of the terms are the same as those in the structured medical record generation method. For specific implementation details, please refer to the description in the above method embodiment.

[0120] Please refer to Figure 6 , Figure 6 A schematic diagram of the structure of a structured medical record generating device provided in an embodiment of the present application, the structured medical record generating device may include:

[0121] The first acquisition module 701 is used to acquire a pre-built text hot word library, wherein the text hot word library contains a plurality of text hot words;

[0122] The second acquisition module 702 is used to acquire the target conversation voice information;

[0123] A first determining module 703, configured to determine first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information;

[0124] A second determination module 704, configured to determine second dialogue text information based on the text hot word library and the first dialogue text information;

[0125] The generating module 705 is used to generate a target structured medical record text based on the second conversation text information.

[0126] The specific implementation of each of the above modules can be found in the previous embodiments and will not be described in detail here.

[0127] An embodiment of the present application also provides an electronic device, including a memory and a processor, wherein the processor is used to execute the steps in the structured medical record generation method provided in this embodiment by calling a computer program stored in the memory.

[0128] Please refer to Figure 7 , Figure 7 A schematic diagram of the structure of an electronic device provided in an embodiment of the present application.

[0129] The electronic device may include components such as a processor 101 with one or more processing cores, a memory 102 with one or more computer-readable storage media, a power supply 103, and an input unit 104. Those skilled in the art will appreciate that the electronic device structure shown in the figure does not constitute a limitation on the electronic device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently. Among them:

[0130] The processor 101 is the control center of the electronic device, which uses various interfaces and lines to connect various parts of the entire electronic device, and executes various functions of the electronic device and processes data by running or executing software programs and / or modules stored in the memory 102, and calling data stored in the memory 102. Optionally, the processor 101 may include one or more processing cores; optionally, the processor 101 may integrate an application processor and a modem processor, wherein the application processor mainly processes the operating system, user interface, and application programs, and the modem processor mainly processes wireless communications. It is understandable that the above-mentioned modem processor may not be integrated into the processor 101.

[0131] The memory 102 can be used to store software programs and modules. The processor 101 executes various functional applications and data processing by running the software programs and modules stored in the memory 102. The memory 102 may mainly include a program storage area and a data storage area, wherein the program storage area may store an operating system, an application required for at least one function (such as a sound playback function, an image playback function, etc.), etc.; the data storage area may store data created according to the use of the electronic device, etc. In addition, the memory 102 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device, a flash memory device, or other volatile solid-state storage devices. Accordingly, the memory 102 may also include a memory controller to provide the processor 101 with access to the memory 102.

[0132] The electronic device also includes a power supply 103 for supplying power to each component. Optionally, the power supply 103 can be logically connected to the processor 101 through a power management system, so as to manage charging, discharging, and power consumption through the power management system. The power supply 103 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0133] The electronic device may further include an input unit 104, which may be used to receive input digital or character information and generate keyboard, mouse, joystick, optical or trackball signal input related to user settings and function control.

[0134] Although not shown, the electronic device may also include a display unit, an image acquisition component, etc., which will not be described in detail here. Specifically in this embodiment, the processor 101 in the electronic device will load the executable code corresponding to one or more computer programs into the memory 102 according to the following instructions, and the processor 101 will execute the steps in the structured medical record generation method provided by the present application, such as:

[0135] Obtain a pre-built text hot word library, wherein the text hot word library includes a plurality of text hot words; obtain target conversation voice information; determine first conversation text information corresponding to the target conversation voice information based on the target conversation voice information; determine second conversation text information based on the text hot word library and the first conversation text information; and generate a target structured medical record text based on the second conversation text information.

[0136] It should be noted that the electronic device provided in the embodiment of the present application and the structured medical record generation method in the above embodiment belong to the same concept, and its specific implementation process is detailed in the above related embodiments and will not be repeated here.

[0137] The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program stored therein is executed on a processor of an electronic device provided in an embodiment of the present application, the processor of the electronic device executes the steps in the structured medical record generation method provided in the present application. The storage medium may be a magnetic disk, an optical disk, a read-only memory (ROM) or a random access memory (RAM), etc.

[0138] The present application also provides a computer program product or a computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium, and the processor executes the computer instructions, so that the computer device executes various optional implementations of the above-mentioned structured medical record generation method.

[0139] The above is a detailed introduction to a structured medical record generation method and device provided by the present application. Specific examples are used in this article to illustrate the principles and implementation methods of the present application. The description of the above embodiments is only used to help understand the method of the present application and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present application, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present application.

[0140] It should be noted that when the above embodiments of the present application are applied to specific products or technologies, the relevant data of the user is involved, and the user's permission or consent is required, and the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions.

Claims

1. A method for generating structured medical records, characterized in that: The structured medical record generation method comprises: Acquire a pre-built text hot word library, wherein the text hot word library contains a plurality of text hot words; Obtain target conversation voice information; Determining first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information; Determine second dialogue text information based on the text hot word library and the first dialogue text information; Generate a target structured medical record text based on the second conversation text information.

2. The structured medical record generation method according to claim 1, characterized in that: The determining the second dialogue text information based on the text hot word library and the first dialogue text information includes: Acquire multiple text segmentations obtained by segmenting the first conversation text information; At least part of the text segmentation in the first dialogue text information is replaced based on a plurality of the text hot words in the text hot word library to obtain the second dialogue text information.

3. The structured medical record generation method according to claim 2, characterized in that: The step of replacing at least part of the text segmentation in the first dialogue text information based on a plurality of the text hot words in the text hot word library to obtain the second dialogue text information includes: Calculating the similarity index of the text segmentation and each text hot word in the text hot word library respectively; When the segmentation similarity index is not lower than a preset index value, the text segmentation corresponding to the segmentation similarity index in the first dialogue text information is replaced with the corresponding text hot word to obtain the second dialogue text information.

4. The structured medical record generation method according to claim 3, characterized in that: The separately calculating the segmentation similarity index of the text segmentation and each text hot word in the text hot word library includes: Obtain the total number of pinyin letters of the pinyin of the text segmentation and the number of pinyin letters of the pinyin of the text hot words; Obtaining the longest identical letter sequence between the pinyin of the text segmentation and the pinyin of the text hot word, wherein the longest identical letter sequence exists in both the text segmentation and the text hot word; The ratio of the number of letters in the longest identical letter sequence to the total number of pinyin letters is determined as the segmentation similarity index of the text segmentation and the text hot word.

5. The structured medical record generation method according to claim 4, characterized in that: The determining, based on the target conversation voice information, first conversation text information corresponding to the target conversation voice information includes: Inputting the target dialogue voice information into a speech recognition model to obtain third dialogue text information corresponding to the target dialogue voice information, wherein the speech recognition model is a whisper model or a FunASR model; The first dialogue text information corresponding to the target dialogue voice information is determined based on the third dialogue text information.

6. The structured medical record generation method according to claim 5, characterized in that: The determining, based on the third dialogue text information, the first dialogue text information corresponding to the target dialogue voice information comprises: Classifying each text segmentation in the third dialogue text information to obtain a word category of each text segmentation; The text segmentation of the word categories in the third dialogue text information that belong to the preset category set is removed to obtain the first dialogue text information corresponding to the target dialogue voice information.

7. The method for generating structured medical records according to claim 6, characterized in that: The step of obtaining the target conversation voice information includes: Obtaining initial conversation voice information; Splitting the initial conversation voice information into a plurality of conversation voice information segments; Detecting background noise parameters of each of the dialogue voice information segments; When the background noise parameter of the conversation voice information segment is lower than a preset parameter value, the conversation voice information segment is determined as the target conversation voice information.

8. A structured medical record generating device, characterized in that: include: A first acquisition module is used to acquire a pre-built text hot word library, wherein the text hot word library contains a plurality of text hot words; The second acquisition module is used to acquire the target conversation voice information; A first determining module, configured to determine first dialogue text information corresponding to the target dialogue voice information based on the target dialogue voice information; A second determination module, configured to determine second dialogue text information based on the text hot word library and the first dialogue text information; A generation module is used to generate a target structured medical record text based on the second conversation text information.

9. An electronic device, characterized in that: It comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to run the computer program in the memory to execute the steps in the structured medical record generation method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in the structured medical record generating method according to any one of claims 1 to 7.

Images