Doctor-patient communication discourse feature analysis method and system based on corpus

Abstract

The invention provides a corpus-based doctor-patient communication discourse feature analysis method and system, and belongs to the technical field of artificial intelligence and doctor-patient communication. The method comprises the following steps: S100, capturing voice conversation records of a doctor and a patient in real time; s200, when a session unit is determined from the voice session conversation record, judging the validity of the session unit; when the session unit is valid, returning to the step S100; when the session unit is invalid, entering a step S300; and S300, based on the context of the session unit, searching an auxiliary session linguistic from a preset doctor-patient communication corpus, and providing the auxiliary session linguistic to the doctor to continue the session unit. The system comprises a voice capture module, a voice-to-text module, a session unit identification module, a session validity judgment module, an auxiliary corpus search module and a corpus display module. The real-time voice interaction process between the doctor and the patient can be assisted, and the doctor-patient communication efficiency is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of artificial intelligence and doctor-patient communication, and in particular relates to a corpus-based doctor-patient communication discourse feature analysis method and system, a computer-readable storage medium for implementing the method, a computer program product and an electronic device. Background Art

[0002] The corpus is a database dedicated to research that integrates a large amount of language information used in real situations. The doctor-patient conversation corpus not only has unique institutional conversation characteristics, but also has specific social and cultural attributes and psychological cognitive attributes, which provides different paths for analyzing important interpersonal pragmatic phenomena such as conflicts in doctor-patient conversations, the expression of doctor-patient communication modalities, and the construction of harmonious doctor-patient relationships.

[0003] Doctor-patient communication is a process of information transmission of medical understanding. It is carried out for the health needs of patients, so that both doctors and patients can fully and effectively express their understanding, willingness and requirements of medical activities. Doctor-patient communication is two-way. Interaction, complementarity and mutual understanding in doctor-patient communication are the prerequisites for harmonious doctor-patient communication.

[0004] Sufficient communication between doctors and patients can enable doctors to obtain sufficient pre-diagnosis information and make more accurate diagnosis and treatment recommendations. However, in actual clinical settings, patients may have various concerns and be unwilling to provide detailed information to doctors’ inquiries. Or, due to problems with the doctor’s own questioning methods and emotional expressions, patients may not know how to answer specifically or may resist answering.

[0005] In addition, in a clinical environment, doctors need to deal with a large number of patients and conduct a large number of inquiries. Due to their limited energy, they are likely to become fatigued over a long period of time, resulting in unprofessional subsequent inquiries and inappropriate questions, which may make doctor-patient communication ineffective or even intensify conflicts. These may affect the efficiency of doctor-patient communication and make it impossible to fully achieve the goals of interaction, complementarity and mutual understanding in doctor-patient communication, and doctors may not even realize that these problems have arisen. Summary of the invention

[0006] The present invention proposes a corpus-based doctor-patient communication discourse feature analysis method and system, a computer-readable storage medium, a computer program product and an electronic device for implementing the method, which are proposed to solve the above-mentioned technical problems.

[0007] In a first aspect of the present invention, a corpus-based method for analyzing doctor-patient communication discourse features is proposed, wherein the method is used to assist the real-time voice interaction process between doctors and patients.

[0008] Specifically, the method comprises the following steps:

[0009] S100: Real-time capture of voice conversations between doctors and patients;

[0010] S200: When a conversation unit is determined from the voice conversation record, determining the validity of the conversation unit;

[0011] When the session unit is valid, return to step S100;

[0012] When the session unit is invalid, proceed to step S300;

[0013] S300: Based on the context of the conversation unit, searching for auxiliary conversation data from a preset doctor-patient communication corpus, and providing the auxiliary conversation data to the doctor to continue the conversation unit;

[0014] Wherein, the one conversation unit includes the doctor's inquiry sentence and the patient's answer sentence;

[0015] The determination of the validity of the session unit in step S200 specifically includes:

[0016] When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

[0017] The step S300 specifically includes:

[0018] Based on the doctor's inquiry sentence pattern included in the conversation unit, an auxiliary conversation corpus is searched from a preset doctor-patient communication corpus, and the auxiliary conversation corpus expands the doctor's inquiry sentence pattern.

[0019] Expanding the inquiry sentence into a selection inquiry sentence or a judgment inquiry sentence containing multiple options;

[0020] The selection inquiry sentence or the judgment inquiry sentence is provided to the doctor to continue the conversation unit.

[0021] After providing the auxiliary conversation material to the doctor, the doctor adjusts the auxiliary conversation material and then asks the patient a new inquiry sentence pattern to continue the conversation unit.

[0022] After providing the auxiliary conversation material to the doctor to continue the conversation unit, the method further includes:

[0023] S400: receiving a new answer sentence pattern to the new inquiry sentence pattern from the patient;

[0024] S500: Determine whether the matching degree between the new inquiry sentence and the new answer sentence is higher than a preset threshold;

[0025] If yes, then updating the adjusted auxiliary conversation corpus into the preset doctor-patient communication corpus;

[0026] If not, the doctor skips the current conversation unit, enters the next voice conversation, and returns to step S100.

[0027] In the specific implementation of the present invention, the method is implemented based on an electronic device including a human-computer interaction interface, wherein the electronic device includes a speech capture unit, a speech-to-text unit, and a corpus display unit;

[0028] The voice capture unit captures the voice conversation record between the doctor and the patient in real time;

[0029] The speech-to-text unit converts the speech conversation record into a text conversation record;

[0030] The corpus display unit is used to display auxiliary conversation corpus, and the auxiliary conversation corpus is text conversation corpus.

[0031] In a second aspect of the present invention, the method can be implemented by a system, a server, a cluster, etc., including a cluster or combination of physical machines, virtual machines, and electronic equipment. The system also includes a speech capture module, a speech-to-text module, a conversation unit recognition module, a conversation validity judgment module, an auxiliary corpus search module, and a corpus display module;

[0032] The voice capture module is used to capture the voice conversation record between the doctor and the patient in real time;

[0033] The speech-to-text unit converts the speech conversation record into a text conversation record;

[0034] The conversation unit identification module is used to identify the conversation units of the text conversation record, each conversation unit includes a doctor's inquiry sentence and a patient's answer sentence to the inquiry sentence;

[0035] The session validity judgment module is used to judge the validity of the session unit; the session unit obtained at this time can be called the original session unit;

[0036] When the original conversation unit is invalid, the auxiliary corpus search module is started, and the auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation corpus to the doctor to continue the conversation unit;

[0037] When the matching degree between the inquiry sentence and the answer sentence of the original conversation unit is higher than the preset threshold, the doctor is prompted to skip the current conversation unit and enter the next voice conversation.

[0038] The corpus display unit is used to display auxiliary conversation corpus, and the auxiliary conversation corpus is text conversation corpus.

[0039] The session validity judgment module is used to judge the validity of the session unit, specifically including:

[0040] When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

[0041] The auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation corpus to the doctor to continue the conversation unit, specifically including:

[0042] The auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the doctor's inquiry sentence pattern included in the conversation unit, and the auxiliary conversation corpus expands the doctor's inquiry sentence pattern to a selection inquiry sentence pattern or a judgment inquiry sentence pattern containing multiple options;

[0043] The selection inquiry sentence or the judgment inquiry sentence is provided to the doctor to continue the conversation unit.

[0044] After providing the auxiliary conversation material to the doctor to continue the conversation unit,

[0045] The speech capture module continues to capture the patient's new answer sentence pattern to the new inquiry sentence pattern;

[0046] The conversation validity judgment module continues to judge whether the new conversation unit composed of the new inquiry sentence pattern and the new answer sentence pattern is valid;

[0047] When the new conversation unit is invalid, the doctor is prompted to skip the current conversation unit and enter the next voice conversation.

[0048] The above-mentioned corpus-based method for analyzing the characteristics of doctor-patient communication discourse can be automatically implemented through various forms of electronic devices through computer program instructions; the computer program instructions can be stored in storage media of different forms and loaded into computer electronic devices for execution.

[0049] Therefore, in the third aspect of the present invention, a computer-readable storage medium is also provided for storing computer instructions. When the computer instructions are executed on an electronic device, the electronic device executes a corpus-based method for analyzing doctor-patient communication discourse features of the first aspect.

[0050] In a fourth aspect of the present invention, a computer device is further proposed, comprising a processor and a memory, wherein the memory is used to store instructions, and the processor is used to call the instructions in the memory, so that the computer device executes a corpus-based doctor-patient communication discourse feature analysis method of the first aspect mentioned above.

[0051] In a fifth aspect of the present invention, a computer program product is also proposed, the product comprising a computer program, and when the computer program is executed, a corpus-based method for analyzing doctor-patient communication discourse features in the first aspect is implemented.

[0052] The technical solution of the present invention can be used to assist the real-time voice interaction process between doctors and patients. Specifically, the technical solution of the present invention can capture the voice conversation record between doctors and patients in real time through electronic devices. When the conversation unit determined from the voice conversation record is invalid, the electronic device can keenly identify that there may be a problem (invalid) in the current conversation unit before the doctor is aware of the problem, and then promptly provide artificial intelligence auxiliary corpus to remind the doctor and assist the doctor to expand and improve the questioning method and then continue the current conversation, thereby obtaining more effective doctor-patient communication information, so that doctor-patient communication can achieve the goals of full interaction, complementarity and mutual understanding.

[0053] Further advantages of the present invention will be further reflected in detail in the specific embodiments section in conjunction with the drawings of the specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0055] Figure 1 This is a schematic diagram of the main process of a corpus-based doctor-patient communication discourse feature analysis method according to an embodiment of the present invention.

[0056] Figure 2 yes Figure 1 A more specific example of the implementation flow chart of the method embodiment

[0057] Figure 3 This is a schematic diagram of a doctor-patient communication scene before the technical solution of the present invention is implemented

[0058] Figure 4 This is a flow chart of a doctor-patient communication scenario after implementing the technical solution of the present invention.

[0059] Figure 5 This is a schematic diagram of the functional modules of a corpus-based doctor-patient communication discourse feature analysis system according to an embodiment of the present invention. DETAILED DESCRIPTION

[0060] Here, exemplary embodiments will be described in detail, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments that are the same as the present application. Instead, they are only examples of devices and methods that are the same as some aspects of the present application as detailed in the attached claims.

[0061] See also Figure 1 , Figure 1 The main flow of a method for analyzing doctor-patient communication discourse characteristics based on a corpus according to an embodiment of the present invention is shown, which mainly includes steps S100-S300 (for ease of reading, the attached Figure 1 The steps are as follows:

[0062] S100: Real-time capture of voice conversations between doctors and patients;

[0063] S200: When a conversation unit is determined from the voice conversation record, determining the validity of the conversation unit;

[0064] When the session unit is valid, return to step S100;

[0065] When the session unit is invalid, proceed to step S300;

[0066] S300: Based on the context of the conversation unit, searching for auxiliary conversation data from a preset doctor-patient communication corpus, and providing the auxiliary conversation data to the doctor to continue the conversation unit;

[0067] Wherein, the one conversation unit includes the doctor's inquiry sentence and the patient's answer sentence;

[0068] The determination of the validity of the session unit in step S200 specifically includes:

[0069] When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

[0070] Next, combine Figure 2 ,right Figure 1 Each of the steps is described in detail.

[0071] In actual implementation, each step of the method embodiment corresponds to a corresponding function execution unit or module, which is implemented through computer electronic equipment.

[0072] For example, the method is implemented based on an electronic device including a human-computer interaction interface, wherein the electronic device includes a speech capture unit, a speech-to-text unit, and a corpus display unit;

[0073] The voice capture unit captures the voice conversation record between the doctor and the patient in real time;

[0074] The speech-to-text unit converts the speech conversation record into a text conversation record;

[0075] The corpus display unit is used to display auxiliary conversation corpus, and the auxiliary conversation corpus is text conversation corpus.

[0076] In the embodiments of the present application, the term "module" or "unit" refers to a computer program or a part of a computer program with a predetermined function, and works together with other related parts to achieve a predetermined goal, and can be implemented in whole or in part by using software, hardware (such as processing circuits or memories) or a combination thereof. Similarly, a processor (or multiple processors or memories) can be used to implement one or more modules or units. In addition, each module or unit can be part of an overall module or unit that includes the function of the module or unit.

[0077] At the same time, in the specific implementation of this application, if user-related data is involved, when the embodiment of this application is applied to a specific product or technology, the user's permission or consent must be obtained, and the collection, use and processing of relevant data must comply with relevant laws, regulations and standards of relevant countries and regions.

[0078] Specifically in each embodiment of the present invention, although it involves the collection of personal user information, most of the collected personal user information does not involve personal identity identification, and is only used for general communication corpus analysis samples, and will not disclose personal privacy. Some information involving personal privacy (such as name) will be anonymized in the actual corpus analysis.

[0079] Combination Figure 2 , in a clinical environment, after the doctor and the patient start a conversation, step S100 is to capture the voice conversation record between the doctor and the patient in real time through the voice capture unit;

[0080] This embodiment is introduced by taking face-to-face conversation as an example, but it can be understood that the technical solution of the present invention is also suitable for remote consultation scenarios.

[0081] Thus, the voice conversation 1 opened is as follows:

[0082] Doctor: Question 1 Patient: Answer 1 (voice form)

[0083] Next, step S200 converts the voice conversation record into a text conversation record through the voice-to-text unit, that is, text conversation 1 is as follows:

[0084] Doctor: Question 1 Patient: Answer 1 (Text format)

[0085] Of course, in an actual clinical environment, the conversation between a doctor and a patient is usually continuous and includes multiple conversation units. Therefore, the electronic device needs to continuously monitor the conversation record between the two and identify the conversation units of the text conversation record, each of which includes the doctor's questioning sentence and the patient's answer sentence to the questioning sentence.

[0086] The above process takes the identified doctor's inquiry sentence (doctor: question 1) and the patient's answer sentence (patient: answer 1) to the inquiry sentence as an example.

[0087] Next, the validity of the session unit is judged by a session validity judgment unit;

[0088] Specifically, judging the validity of the session unit in step S200 specifically includes:

[0089] When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

[0090] Examples of valid or invalid session units will be introduced in the following examples. Figure 2 The continuity of the embodiments will not be expanded for the time being.

[0091] exist Figure 2 When the conversation unit is valid, the process returns to step S100, that is, when the conversation unit (doctor: question 1 - patient: answer 1) is valid, the process continues to capture the voice conversation record between the doctor and the patient in real time, determines a conversation unit from the voice conversation record, and continues to determine whether the conversation unit is valid or invalid;

[0092] Figure 2 As shown in FIG. 1 , when a conversation unit (doctor: question 1 - patient: answer 1) is valid, the next conversation unit (doctor: question 2 - patient: answer 2) is identified and the process of steps S100 - S300 is repeated.

[0093] When the session unit is invalid, the process goes to step S300;

[0094] S300: Based on the context of the conversation unit, searching for auxiliary conversation data from a preset doctor-patient communication corpus, and providing the auxiliary conversation data to the doctor to continue the conversation unit.

[0095] In short, in the implementation scenario of the present invention, the invalid conversation unit means that the patient's answer cannot meet the doctor's expectations of the current question. For example, the patient may have various concerns and be unwilling to give detailed information to the doctor's questions, or the doctor may have problems with the way he asks questions or his emotional expression, so that the patient does not know how to answer specifically or resists answering;

[0096] More importantly, in a clinical environment, doctors need to deal with a large number of patients and conduct a large number of inquiries. Due to their limited energy, they are likely to become fatigued over a long period of time, resulting in unprofessional subsequent inquiries and inappropriate questions, which may make doctor-patient communication ineffective or even intensify conflicts. These may affect the efficiency of doctor-patient communication and make it impossible to fully achieve the goals of interaction, complementarity and mutual understanding in doctor-patient communication, and doctors may not even realize that these problems have arisen.

[0097] Figure 2 As shown in the figure, when the conversation unit (doctor: question 1 - patient: answer 1) is invalid, an auxiliary conversation corpus is searched from a preset doctor-patient communication corpus based on the doctor's inquiry sentence pattern included in the conversation unit, and the auxiliary conversation corpus expands the doctor's inquiry sentence pattern to a selection inquiry sentence pattern or a judgment inquiry sentence pattern containing multiple options;

[0098] The selection inquiry sentence or the judgment inquiry sentence is provided to the doctor to continue the conversation unit.

[0099] Figure 2 An example of the selection inquiry sentence or judgment inquiry sentence is (doctor asks 11), and the patient's answer is (patient answers 11).

[0100] In most cases, after extended questioning based on the auxiliary conversation material, patients are mostly able to give more detailed answers, thereby making the conversation unit composed of ((doctor asks 11)-(patient answers 11)) effective.

[0101] In order to better explain the practical application effect of the above method embodiment, Figure 3 and Figure 3 , further introducing the schematic diagram of the application scenario of the technical solution of the present invention in actual clinical doctor-patient communication.

[0102] Figure 3 The following is a routine communication text between doctors and patients in a specific context (obstetrics and gynecology clinic). For ease of description, Figure 3This is called the "raw session record", Figure 3 This is a schematic diagram of a doctor-patient communication scene before the technical solution of the present invention is implemented. An excerpt of the communication content is as follows:

[0103] Q1 Doctor: What is your name?

[0104] A1 Patient: Li ××, I have a question for you, doctor. My husband and I have been married for 5 years, but we have no children. We have always wanted one, and my husband has some health problems. Now we are finally getting pregnant, but I am very nervous. I am afraid that the child will not be healthy or have problems. Can you tell me what to do?

[0105] Q2 Doctor: What’s the problem with your spouse?

[0106] Patient A2: It’s hard to say, so don’t ask.

[0107] Q3 Doctor: If it is a genetic disease or something, please tell us. It involves the health of your child. If it is other privacy issues, we will not ask.

[0108] Patient A3: I understand. Don’t ask me anymore. I’m just asking about the child.

[0109] exist Figure 3 In the embodiment in , there are actually two problems, but three conversation units can be identified.

[0110] Session Unit 1:

[0111] Doctor: Q1

[0112] Patient: A1 (Answer 1)

[0113] It can be seen that in conversation unit 1, the patient answered the doctor's questions accurately, so conversation unit 1 is effective.

[0114] Session Unit 2:

[0115] Doctor: Q2

[0116] Patient: A2 (Answer 2)

[0117] It can be seen that in conversation unit 2, the patient tactfully avoided the doctor's questions, so conversation unit 2 is invalid.

[0118] Conversation Unit 3:

[0119] Doctor: Q3

[0120] Patient: A3 (Answer 3)

[0121] It can be seen that in conversation unit 3, the patient directly rejected / avoided the doctor's questions, and therefore, conversation unit 3 was invalid.

[0122] It is understandable that in the above scenario, the doctor cannot continue to ask questions about question 2, otherwise it will intensify the conflict. Therefore, there are certain problems with the above questioning method, which makes it impossible for the doctor to obtain accurate and reliable information, and the patient's answer cannot meet the doctor's expectations for the current question.

[0123] Moreover, judging from the fact that the doctor asked the same question twice in a row and failed to obtain an effective answer, the doctor may not have realized that there was a problem with his questioning method at this time.

[0124] In the above embodiment, the validity of the conversation unit is determined by the method automatically, specifically by the conversation validity determination module (unit) of the electronic device, which is executed imperceptibly in the background and will not affect the actual conversation process of the current doctor or patient.

[0125] Specifically, the session validity judgment module is used to judge the validity of the session unit, specifically including:

[0126] When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

[0127] The matching degree between the inquiry sentence and the answer sentence of the conversation unit is a numerical value used to measure the matching degree of "question-answer". For example, 100 points represent a perfect match and 0 points represent a complete mismatch. When the matching degree is lower than 60 points, the conversation unit is judged to be invalid. When the matching degree is higher than 60 points, the conversation unit is judged to be valid.

[0128] The matching degree between the inquiry sentence and the answer sentence of the conversation unit can be achieved through natural language processing technology (NLP), word segmentation technology, corpus self-learning technology, etc., and specifically can be achieved through a machine model trained with large-scale corpus, such as a neural network model, especially a self-feedback neural network model. This belongs to the prior art in this field and the present invention will not elaborate on this.

[0129] As a comparison, Figure 4 A schematic diagram showing a flow chart of a doctor-patient communication scenario after the technical solution of the present invention is implemented.

[0130] exist Figure 4 As a comparison, when the method identifies conversation unit 2 and determines that conversation unit 2 is invalid, the method proceeds to step S300 for intervention.

[0131] Specifically,

[0132] Session Unit 2:

[0133] Doctor: Q2 (Question 2) What’s the problem with your spouse?

[0134] Patient: A2 (Answer 2): It’s hard to say. Don’t ask.

[0135] It can be seen that in conversation unit 2, the patient euphemistically avoided the doctor's question, that is, "answered irrelevant questions". At this time, the matching degree between the inquiry sentence and the answer sentence of the conversation unit is lower than the preset threshold, and the conversation unit is invalid.

[0136] In specific implementation, at this time, the human-computer interaction interface of the electronic device will give the doctor some kind of prompt to display auxiliary corpus to assist the doctor in continuing to ask questions.

[0137] Specifically, the electronic device searches for auxiliary conversation data from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation data to the doctor to continue the conversation unit.

[0138] For example, the electronic device may search for auxiliary conversation corpus Q22 from a preset doctor-patient communication corpus based on the context of the conversation unit (including: pregnancy, pre-pregnancy check-up, inquiry, inquiry of both spouses, inquiry of husband's medical history, etc.);

[0139] Specifically, the auxiliary conversation data Q22 expands the doctor's inquiry sentence pattern, and expands the inquiry sentence pattern into a selection inquiry sentence pattern or a judgment inquiry sentence pattern containing multiple options;

[0140] The original doctor's question structure is simply "What's wrong with your loved one?" This question is too broad and may result in not knowing how to answer it specifically.

[0141] After step S300, Figure 4 The further expanded question sentence structure for question 1 of the above conversation unit is as follows:

[0142] Q22: “How about you introduce your spouse’s basic situation? How old is he? Is there any family history of genetic diseases, such as XXX, UUU, ZZZ, etc.?”

[0143] Or, do you have any bad habits, such as smoking...drinking...and ****?

[0144] Or, are there any other obvious characteristics in the family? For example...

[0145] Because in our experience, it is usually the parent who has ………… that is likely to result in the child being ………….

[0146] Under question 22, the original simple interrogative sentence is expanded into a selection interrogative sentence or a judgment interrogative sentence containing multiple options, making the question clearer. The patient can clearly know what to answer, what the consequences will be, etc., and is likely to be willing to provide more accurate information.

[0147] In a specific application, patients were mostly willing to provide further answers, e.g. Figure 4 Shown:

[0148] Patient A22: It’s like this… My wife is xx years old and often… There’s nothing wrong with other things, but the main thing is… Just like you said, there are some…

[0149] It can be seen that after the above reminders, doctors can obtain more doctor-patient communication discourses for question 2, achieving the goal of full interaction, complementarity and mutual understanding in doctor-patient communication.

[0150] At this time, it can be understood that after providing the selection inquiry sentence or judgment inquiry sentence (A22) to the doctor to continue the conversation unit, the patient's new answer sentence (A22) to the new inquiry sentence can be received;

[0151] Of course, as a preference, after the electronic device provides and displays the above-mentioned auxiliary language materials, the doctor can also make adjustments according to the actual context and finally obtain an expanded interrogation sentence pattern, that is, after providing the auxiliary conversation language materials to the doctor, the doctor adjusts the auxiliary conversation language materials and uses them as a new interrogation sentence pattern to ask the patient to continue the conversation unit.

[0152] In the above embodiment, the matching degree between the patient's new answer sentence (A22) to the new inquiry sentence and the new answer sentence (Q22) to the new inquiry sentence exceeds the preset threshold, that is, the new conversation unit (Q22-A22) is valid.

[0153] Of course, in some scenarios, it is possible that even if the electronic device prompts to provide auxiliary corpus and the doctor adjusts the auxiliary corpus to continue the current conversation, the patient is still unwilling to respond to the question, resulting in the patient's new answer sentence (A22) of the new inquiry sentence and the new answer sentence (Q22') of the new inquiry sentence still having a low degree of match, that is, the new conversation unit (Q22-A22') is still invalid.

[0154] At this point, considering the particularity of the clinical scenario and the emotions of both parties, it is not appropriate to continue asking the current question. Instead, you can choose to skip the current question and continue the current doctor-patient communication.

[0155] Specifically, after providing the auxiliary conversation material to the doctor to continue the conversation unit, the method further includes:

[0156] S400: receiving a new answer sentence pattern to the new inquiry sentence pattern from the patient;

[0157] S500: Determine whether the matching degree between the new inquiry sentence and the new answer sentence is higher than a preset threshold;

[0158] If yes, then updating the adjusted auxiliary conversation corpus into the preset doctor-patient communication corpus;

[0159] If not, the doctor skips the current conversation unit, enters the next voice conversation, and returns to step S100.

[0160] It can be seen that in the above method embodiment, the technical solution of the present invention can capture the voice conversation record between the doctor and the patient in real time through an electronic device. When the conversation unit determined from the voice conversation record is invalid, the electronic device can keenly identify that there may be a problem (invalid) in the current conversation unit before the doctor is aware of the problem, and then promptly provide artificial intelligence auxiliary corpus to remind the doctor and assist the doctor to expand and improve the questioning method and then continue the current conversation, thereby obtaining more effective doctor-patient communication information, so that the doctor-patient communication can achieve the goals of full interaction, complementarity and mutual understanding.

[0161] Based on the method embodiment, Figure 5 A schematic diagram showing the functional module composition of a corpus-based doctor-patient communication discourse feature analysis system according to an embodiment of the present invention.

[0162] exist Figure 5 In the figure, a corpus-based doctor-patient communication discourse feature analysis system is shown, the system also includes a speech capture module, a speech-to-text module, a conversation unit recognition module, a conversation validity judgment module, an auxiliary corpus search module and a corpus display module;

[0163] The voice capture module is used to capture the voice conversation record between the doctor and the patient in real time;

[0164] The speech-to-text unit converts the speech conversation record into a text conversation record;

[0165] The conversation unit identification module is used to identify the conversation units of the text conversation record, each conversation unit includes a doctor's inquiry sentence and a patient's answer sentence to the inquiry sentence;

[0166] The session validity determination module is used to determine the validity of the session unit;

[0167] When the conversation unit is invalid, the auxiliary corpus search module is started, and the auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation corpus to the doctor to continue the conversation unit;

[0168] The corpus display unit is used to display auxiliary conversation corpus, and the auxiliary conversation corpus is text conversation corpus.

[0169] The session validity judgment module is used to judge the validity of the session unit, specifically including:

[0170] When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

[0171] The auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation corpus to the doctor to continue the conversation unit, specifically including:

[0172] The auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the doctor's inquiry sentence pattern included in the conversation unit, and the auxiliary conversation corpus expands the doctor's inquiry sentence pattern to a selection inquiry sentence pattern or a judgment inquiry sentence pattern containing multiple options;

[0173] The selection inquiry sentence or the judgment inquiry sentence is provided to the doctor to continue the conversation unit.

[0174] After providing the auxiliary conversation material to the doctor to continue the conversation unit,

[0175] The speech capture module continues to capture the patient's new answer sentence pattern to the new inquiry sentence pattern;

[0176] The conversation validity judgment module continues to judge whether the new conversation unit composed of the new inquiry sentence pattern and the new answer sentence pattern is valid;

[0177] When the new conversation unit is invalid, the doctor is prompted to skip the current conversation unit and enter the next voice conversation.

[0178] Of course, when the matching degree between the inquiry sentence and the answer sentence of the original conversation unit is higher than the preset threshold, the doctor can be prompted to skip the current conversation unit and enter the next voice conversation to avoid asking the same question repeatedly.

[0179] The principles and corresponding effects of the system embodiment can be found in the corresponding steps of the aforementioned method embodiment, and the present invention will not be repeated here.

[0180] In the above introduction, as those skilled in the art will appreciate, the block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities may be implemented in software form, or implemented in one or more hardware modules or integrated circuits, or implemented in different networks and / or processor devices and / or microcontroller devices.

[0181] The steps of the method or algorithm described in conjunction with the embodiments disclosed herein can be implemented directly using hardware, a software module executed by a processor, or a combination of the two. The software module can be placed in a random access memory (RAM), a memory, a read-only memory (ROM), an electrically programmable ROM, an electrically erasable programmable ROM, a register, a hard disk, a removable disk, a CD-ROM, or any other form of storage medium known in the art. Therefore, more embodiments also include a computer readable storage medium provided having a computer program stored thereon, and when the computer program is executed by an electronic device including a processor, the steps of a method for analyzing the characteristics of a doctor-patient communication discourse based on a corpus as described in the above embodiments are implemented.

[0182] For other technologies, principles, algorithms or models not elaborated in detail in this application, please refer to the prior art.

[0183] The foregoing has shown and described the method embodiments and system of the present invention, but it is understandable to those skilled in the art that various changes, modifications, substitutions and variations may be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A corpus-based method for analyzing the characteristics of doctor-patient communication discourse, which is used to assist the real-time voice interaction process between doctors and patients, characterized in that: The method comprises the following steps: S100: Real-time capture of voice conversations between doctors and patients; S200: When a conversation unit is determined from the voice conversation record, determining the validity of the conversation unit; When the session unit is valid, return to step S100; When the session unit is invalid, proceed to step S300; S300: Based on the context of the conversation unit, searching for auxiliary conversation data from a preset doctor-patient communication corpus, and providing the auxiliary conversation data to the doctor to continue the conversation unit; Wherein, the one conversation unit includes the doctor's inquiry sentence and the patient's answer sentence; The determination of the validity of the session unit in step S200 specifically includes: When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

2. A corpus-based doctor-patient communication discourse feature analysis method as claimed in claim 1, characterized in that: The step S300 specifically includes: Based on the doctor's inquiry sentence pattern included in the conversation unit, an auxiliary conversation corpus is searched from a preset doctor-patient communication corpus, and the auxiliary conversation corpus expands the doctor's inquiry sentence pattern to expand the inquiry sentence pattern into a selection inquiry sentence pattern or a judgment inquiry sentence pattern containing multiple options; The selection inquiry sentence or the judgment inquiry sentence is provided to the doctor to continue the conversation unit.

3. The method for analyzing doctor-patient communication discourse characteristics based on a corpus as claimed in claim 1, characterized in that: The step S300 further includes: After providing the auxiliary conversation material to the doctor, the doctor adjusts the auxiliary conversation material and then asks the patient a new inquiry sentence pattern to continue the conversation unit.

4. A corpus-based doctor-patient communication discourse feature analysis method as claimed in claim 3, characterized in that: After providing the auxiliary conversation material to the doctor to continue the conversation unit, the method further includes: S400: receiving a new answer sentence pattern to the new inquiry sentence pattern from the patient; S500: Determine whether the matching degree between the new inquiry sentence and the new answer sentence is higher than a preset threshold; If yes, then updating the adjusted auxiliary conversation corpus into the preset doctor-patient communication corpus; If not, the doctor skips the current conversation unit, enters the next voice conversation, and returns to step S100.

5. The method for analyzing doctor-patient communication discourse characteristics based on a corpus as claimed in claim 1, characterized in that: The method is implemented based on an electronic device including a human-computer interaction interface, wherein the electronic device includes a speech capture unit, a speech-to-text unit, and a corpus display unit; The voice capture unit captures the voice conversation record between the doctor and the patient in real time; The speech-to-text unit converts the speech conversation record into a text conversation record; The corpus display unit is used to display auxiliary conversation corpus, and the auxiliary conversation corpus is text conversation corpus.

6. A corpus-based doctor-patient communication discourse feature analysis system, characterized in that: The system also includes a speech capture module, a speech-to-text module, a conversation unit recognition module, a conversation validity judgment module, an auxiliary corpus search module, and a corpus display module; The voice capture module is used to capture the voice conversation record between the doctor and the patient in real time; The speech-to-text unit converts the speech conversation record into a text conversation record; The conversation unit identification module is used to identify the conversation units of the text conversation record, each conversation unit includes a doctor's inquiry sentence and a patient's answer sentence to the inquiry sentence; The session validity determination module is used to determine the validity of the session unit; When the conversation unit is invalid, the auxiliary corpus search module is started, and the auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation corpus to the doctor to continue the conversation unit; The corpus display unit is used to display auxiliary conversation corpus, and the auxiliary conversation corpus is text conversation corpus.

7. A corpus-based doctor-patient communication discourse feature analysis system as claimed in claim 6, characterized in that: The session validity judgment module is used to judge the validity of the session unit, specifically including: When the matching degree between the inquiry sentence pattern and the answer sentence pattern of the conversation unit is higher than a preset threshold, the conversation unit is valid; otherwise, the conversation unit is invalid.

8. The corpus-based doctor-patient communication discourse feature analysis system according to claim 6, characterized in that: The auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the context of the conversation unit, and provides the auxiliary conversation corpus to the doctor to continue the conversation unit, specifically including: The auxiliary corpus search module searches for auxiliary conversation corpus from a preset doctor-patient communication corpus based on the doctor's inquiry sentence pattern included in the conversation unit, and the auxiliary conversation corpus expands the doctor's inquiry sentence pattern to a selection inquiry sentence pattern or a judgment inquiry sentence pattern containing multiple options; The selection inquiry sentence or the judgment inquiry sentence is provided to the doctor to continue the conversation unit.

9. The corpus-based doctor-patient communication discourse feature analysis system according to claim 6, characterized in that: After providing the auxiliary conversation data to the doctor to continue the conversation unit, the speech capture module continues to capture the patient's new answer sentence pattern to the new inquiry sentence pattern; The conversation validity judgment module continues to judge whether the new conversation unit composed of the new inquiry sentence pattern and the new answer sentence pattern is valid; When the new conversation unit is invalid, the doctor is prompted to skip the current conversation unit and enter the next voice conversation.

10. The corpus-based doctor-patient communication discourse feature analysis system according to claim 6, characterized in that: When the matching degree between the inquiry sentence and the answer sentence of the original conversation unit is higher than the preset threshold, the doctor is prompted to skip the current conversation unit and enter the next voice conversation.

Images