CN118014078A - Follow-up method and related device - Google Patents
Follow-up method and related device Download PDFInfo
- Publication number
- CN118014078A CN118014078A CN202311630883.XA CN202311630883A CN118014078A CN 118014078 A CN118014078 A CN 118014078A CN 202311630883 A CN202311630883 A CN 202311630883A CN 118014078 A CN118014078 A CN 118014078A
- Authority
- CN
- China
- Prior art keywords
- follow
- patient
- language model
- question
- content
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
- G06N5/041—Abduction
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/22—Procedures used during a speech recognition process, e.g. man-machine dialogue
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
- G16H10/20—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for electronic clinical trials or questionnaires
Landscapes
- Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Medical Informatics (AREA)
- Computing Systems (AREA)
- Primary Health Care (AREA)
- Public Health (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Epidemiology (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application discloses a follow-up method and a related device, wherein the method comprises the following steps: carrying out follow-up interaction with the patient by utilizing the large language model to obtain feedback content of the patient in the follow-up interaction, wherein the follow-up interaction comprises multiple rounds of answers; and based on the feedback content, obtaining a follow-up result. Through the mode, the intelligent follow-up visit method and device can realize intelligent follow-up visit, and improve the follow-up visit quality and efficiency.
Description
Technical Field
The application relates to the technical field of follow-up access, in particular to a follow-up access method and a related device.
Background
Follow-up refers to an observation method by which a hospital communicates or otherwise regularly knows the change of the patient's condition and guides the patient's recovery to the patient who was at the hospital's visit. Follow-up is an important part of clinical work, and doctors can timely know the change of the disease condition of patients and give treatment advice by carrying out medical tracking on discharged patients or patients suffering from chronic diseases, and can schedule readmission treatment on patients with recurrent and worsened disease condition; on the other hand, doctors can conveniently track and observe patients, know prognosis conditions, long-term curative effects and clinical application effects of new technologies, master first-hand data and accumulate experience, and are favorable for development of scientific research work and improvement of business level and better serve the patients.
At present, the main modes of follow-up are postoperative ward follow-up, outpatient interview, home follow-up and the like. The modes all need manual recording and classified storage information by medical staff, and when data are required to be subjected to statistical analysis, the data are also required to be manually input into a computer again, so that the efficiency is reduced, and the quality is low.
Disclosure of Invention
The application provides at least one follow-up method and a related device.
In order to solve the technical problems, the application adopts a technical scheme that: there is provided a follow-up method comprising: carrying out follow-up interaction with the patient by utilizing the large language model to obtain feedback content of the patient in the follow-up interaction, wherein the follow-up interaction comprises multiple rounds of answers; and based on the feedback content, obtaining a follow-up result.
Thus, multiple rounds of answers are made with the patient using the large language model, and follow-up results are obtained based on the patient's feedback content in the follow-up interactions. On one hand, the large language model is utilized to carry out follow-up interaction with the patient, so that intelligent follow-up is realized, and the follow-up efficiency is improved. On the other hand, the large language model has strong understanding capability, logic thinking capability, reasoning capability and the like, so that the feedback content of a patient in the follow-up interaction process can be accurately interpreted, and the understanding of the feedback content of the patient is increased, so that a reply with detailed content, smooth expression and clear logic can be generated, the follow-up interaction with the patient is more professional, and the patient experience is more humanized; namely, the follow-up interaction is carried out with the patient by utilizing the large language model, so that the follow-up quality is improved.
Wherein the multiple rounds of questions include at least one set of follow-up questions, the follow-up questions being generated by the large language model and the first answers to the follow-up questions being fed back by the patient.
Thus, the follow-up question is generated by the large language model, and the large language model receives a first answer to the follow-up question by the patient, completing a set of follow-up answers with the patient.
Wherein the multiple rounds of questions further include at least one set of patient questions, the patient questions being patient questions fed back by the patient and second answers being generated for the patient questions by the large language model.
Therefore, the large language model has strong understanding capability, logic thinking capability, reasoning capability and the like, so that the large language model can understand and accurately read patient problems fed back by patients and can generate answers with detailed contents, smooth expression and clear logic; that is, the large language model can provide real-time, professional, accurate and humanized medical advice for the patient feedback problem of the patient, and the follow-up quality and efficiency are improved.
The method for obtaining the feedback content of the patient in the follow-up interaction comprises the following steps of: generating a follow-up problem by using the large language model, and taking the follow-up problem as interactive content; providing the interactive content to the patient and receiving feedback content of the patient for the interactive content; generating new interactive contents based on the newly received feedback contents by using the large language model, and re-executing the steps of providing the interactive contents to the patient, receiving the feedback contents of the patient for the interactive contents and the following steps until the follow-up visit end condition is met.
Therefore, the related conditions of the patient after operation or discharge can be more comprehensively known.
Wherein generating new interactive content based on the newly received feedback content using the large language model includes: analyzing the latest received feedback content by using a large language model to obtain an analysis result, wherein the analysis result comprises at least one of the following components: a first analysis result characterizing whether a first answer related to a previous follow-up question is included, a second analysis result characterizing whether a patient question is included; generating a new follow-up question based on the first analysis result by using the large language model as new interactive content in response to the analysis result including the first analysis result; in response to the analysis result comprising a second analysis result and the second analysis result comprising the patient question, a second answer to the patient question is generated using the large language model as new interactive content.
Therefore, the analysis result corresponding to the feedback content comprises a first analysis result, which indicates that the latest received feedback content is fed back by the patient aiming at the previous follow-up problem, and indicates that the feedback content of the patient is normal; at this time, follow-up is performed on the patient aiming at other follow-up problems so as to more comprehensively understand the related conditions of the patient after operation or discharge; when the analysis result comprises a second analysis result and the second analysis result is a patient-containing problem, indicating that the patient currently has a medical question; the large language model has strong understanding capability, logic thinking capability and reasoning capability, so that the large language model can understand and accurately read patient questions fed back by patients and can generate answers with detailed contents, smooth expression and clear logic; that is, the large language model can provide real-time, professional, accurate and humanized medical advice for the patient feedback problem of the patient, and the follow-up quality and efficiency are improved.
Wherein generating a new follow-up question based on the first analysis result using the large language model as new interactive content includes: acquiring new follow-up questions from the question library by using a large language model; and/or, in response to the newly received feedback content not including the first answer related to the previous follow-up question, adapting the description of the previous follow-up question as a new follow-up question using the large language model.
Thus, the feedback content does not contain a first answer related to the previous follow-up question, indicating that the latest received feedback content was not fed back by the patient for the previous follow-up question, indicating that the feedback content of the patient is abnormal; at this time, the description of the previous follow-up problem is adjusted by using the large language model to serve as a new follow-up problem, so that the follow-up problem is provided for the patient again, the accuracy of feedback content of the patient aiming at each follow-up problem is ensured, and the follow-up quality is improved.
Wherein, utilize the big language model to obtain the new follow-up question from the question bank, include: carrying out semantic understanding on the first answer by using the large language model to obtain a semantic result; selecting a new follow-up question with a preset association relation with a semantic result from a question library; and/or adjust the description of the previous follow-up question using the large language model, including: the location of at least some of the content in the previous follow-up question is adjusted using the large language model.
Thus, new follow-up questions are flexibly selected from the question bank.
Wherein, when the analysis result includes a first analysis result and a second analysis result, and the second analysis result is a patient question, the step of generating a second answer to the patient question by using the large language model is preferentially executed as new interactive content, and when the interaction to the patient question is ended, the step of generating a new follow-up question based on the first analysis result by using the large language model is executed again as new interactive content; and/or after generating a second answer to the patient question using the large language model as new interactive content, further comprising: the large language model is updated with patient questions and feedback content for the patient questions.
Thus, the large language model is updated with patient problems and feedback content for the patient problems to continuously iterate through the large language model with patient preferences.
Wherein generating follow-up questions using a large language model includes: acquiring disease information aimed at the follow-up visit by using a large language model, and searching out a problem set related to the disease information from a problem library; a follow-up question is obtained from the question set.
Thus, the purpose of follow-up is to learn about the patient post-surgery or post-discharge, and the patient's condition of relevance to be learned for each disease is different; therefore, the problem of the problem concentration related to the disease information is used as the follow-up problem, the follow-up interaction with the patient can be performed more specifically, and the follow-up quality is improved.
Wherein providing the interactive content to the patient and receiving feedback content of the patient for the interactive content, comprises: the interactive content is provided to the patient in a voice mode, and feedback content of the patient aiming at the interactive content is received in a voice mode.
Therefore, the large language model carries out follow-up on the patient in a voice mode, and intelligent voice follow-up is realized.
Wherein providing interactive content to the patient in a voice manner comprises: generating a first voice of the interactive content by using the large language model, and transmitting the first voice to the outbound device so that the outbound device transmits the first voice to the patient; or the interactive content is sent to the voice processing equipment, so that the voice processing equipment generates first voice for the interactive content and transmits the first voice to a patient through the outbound equipment; receiving feedback content of the patient aiming at the interactive content in a voice mode, wherein the feedback content comprises the following steps: receiving second voice fed back by the patient received by the outbound equipment, and identifying the second voice by using a large language model to obtain feedback content of the patient aiming at the interactive content; or receiving feedback content of the interaction content, which is sent by the voice processing equipment and is fed back by the patient, wherein the feedback content is obtained by the voice processing equipment through identifying the second voice fed back by the patient and received by the outbound equipment.
Accordingly, the execution subject that generates the first voice of the interactive content and that recognizes the second voice generation feedback content can be flexibly selected.
In order to solve the technical problems, the application adopts another technical scheme that: the follow-up device comprises an interaction module and a determination module, wherein the interaction module is used for carrying out follow-up interaction with a patient by utilizing a large language model to obtain feedback content of the patient in the follow-up interaction, and the follow-up interaction comprises multiple rounds of answers; the determining module is used for obtaining follow-up results based on feedback content.
In order to solve the technical problems, the application adopts another technical scheme that: there is provided an electronic device comprising a processor and a memory, the memory storing program instructions, the processor being adapted to execute the program instructions to implement the above-described follow-up method.
In order to solve the technical problems, the application adopts another technical scheme that: there is provided a computer readable storage medium for storing program instructions executable to implement the above-described follow-up method.
According to the technical scheme, the large language model is utilized to answer the patient for multiple rounds, and the follow-up result is obtained based on the feedback content of the patient in the follow-up interaction. On one hand, the large language model is utilized to carry out follow-up interaction with the patient, so that intelligent follow-up is realized, and the follow-up efficiency is improved. On the other hand, the large language model has strong understanding capability, logic thinking capability, reasoning capability and the like, so that the feedback content of a patient in the follow-up interaction process can be accurately interpreted, and the understanding of the feedback content of the patient is increased, so that a reply with detailed content, smooth expression and clear logic can be generated, the follow-up interaction with the patient is more professional, and the patient experience is more humanized; namely, the follow-up interaction is carried out with the patient by utilizing the large language model, so that the follow-up quality is improved.
Drawings
FIG. 1 is a flow chart of an embodiment of a follow-up method provided by the present application;
FIG. 2 is a flowchart illustrating an embodiment of the step S11 shown in FIG. 1;
FIG. 3 is a flow diagram of one embodiment of generating new interactive content based on newly received feedback content using a large language model provided by the present application;
FIG. 4 is a schematic structural diagram of an embodiment of a follow-up device according to the present application;
FIG. 5 is a schematic diagram of an embodiment of an electronic device according to the present application;
Fig. 6 is a schematic structural diagram of an embodiment of a computer readable storage medium according to the present application.
Detailed Description
The following describes embodiments of the present application in detail with reference to the drawings.
In the following description, for purposes of explanation and not limitation, specific details are set forth such as the particular system architecture, interfaces, techniques, etc., in order to provide a thorough understanding of the present application.
The term "and/or" is herein merely an association relationship describing an associated object, meaning that there may be three relationships, e.g., a and/or B, may represent: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship. Further, "a plurality" herein means two or more than two. In addition, the term "at least one" herein means any one of a plurality or any combination of at least two of a plurality, for example, including at least one of A, B, C, may mean including any one or more elements selected from the group consisting of A, B and C.
Referring to fig. 1, fig. 1 is a flowchart of an embodiment of a follow-up method according to the present application. It should be noted that, if there are substantially the same results, the embodiment of the present application is not limited to the flow sequence shown in fig. 1. As shown in fig. 1, the present embodiment includes:
step S11: and carrying out follow-up interaction with the patient by using the large language model to obtain feedback content of the patient in the follow-up interaction.
The method of the embodiment is used for carrying out follow-up interaction with the patient by using the large language model, so that intelligent follow-up is realized, and the follow-up efficiency and quality can be improved.
In the embodiment, the large language model is utilized to carry out follow-up interaction with the patient, and feedback content of the patient in the follow-up interaction is obtained; wherein the follow-up interaction comprises a plurality of rounds of answers. Multiple rounds of answers are performed with the patient using the large language model to complete the follow-up of the patient. Because the large language model has strong understanding capability, logic thinking capability, reasoning capability and the like, the feedback content of the patient in the follow-up interaction process can be accurately read, the understanding of the feedback content of the patient is increased, and accordingly a reply with detailed content, smooth expression and clear logic can be generated, so that follow-up interaction with the patient is more professional and patient experience is more humanized. Therefore, the large language model is utilized to carry out follow-up interaction with the patient, intelligent follow-up is realized, and the follow-up quality and efficiency are improved.
In an embodiment, the multiple rounds of questions include at least one set of follow-up questions, the follow-up questions being generated by the large language model and the first answers to the follow-up questions being fed back by the patient. That is, the follow-up question is generated by the large language model, and the large language model receives a first answer to the follow-up question by the patient, completing a set of follow-up answers with the patient. For example, a follow-up question generated by a large language model, "please you are? "is the first answer to follow-up questions" i am yes "fed back by the patient," please you are a? "and first answer" I am "is a set of follow-up answers.
The number of groups of follow-up access answers included in the multiple rounds of access answers is not limited, and the follow-up access answers can be specifically set according to actual use requirements; it should be noted that the number of sets of follow-up answers included in the multiple rounds of questions may be affected by the type of disease.
In a specific embodiment, the multiple round of questions further includes at least one set of patient questions, the patient questions being patient-fed back by the patient and a second answer being generated by the large language model for the patient questions. That is, in the process of carrying out follow-up interaction with a patient by using the large language model, not only is the patient answer to the follow-up question, but also the patient has medical doubt, and the problem is presented to the large language model; that is, during follow-up interactions with patients using the large language model, there will be at least one set of patient questions and answers. Because the large language model has strong understanding capability, logic thinking capability, reasoning capability and the like, the large language model can understand and accurately read patient questions fed back by patients and can generate answers with detailed contents, smooth expression and clear logic; that is, the large language model can provide real-time, professional, accurate and humanized medical advice for the patient feedback problem of the patient, and the follow-up quality and efficiency are improved.
The number of groups of patient questions and answers included in the multiple questions and answers is not limited, and the group number can be specifically set according to actual use requirements.
For example, the patient question of patient feedback "× how the drug was taken", the large language model generated a second answer "3 times a day, 2 granules once" for the patient question, "patient question" × how the drug was taken ", and a second answer" 3 times a day, 2 granules once "for the patient question as a set of patient questions.
Step S12: and based on the feedback content, obtaining a follow-up result.
In this embodiment, the follow-up result is obtained based on the feedback content of the patient in the follow-up interaction. In an embodiment, the feedback content of the patient in the follow-up interaction can be directly used as the follow-up result. Of course, in other embodiments, the feedback content of the patient in the follow-up interaction may be reduced, and the reduced feedback content of the patient in the follow-up interaction may be used as the follow-up result, which is not specifically limited herein.
And carrying out multiple rounds of answers on the patient by using the large language model, and obtaining a follow-up result based on feedback content of the patient in follow-up interaction. On one hand, the large language model is utilized to carry out follow-up interaction with the patient, so that intelligent follow-up is realized, and the follow-up efficiency is improved. On the other hand, the large language model has strong understanding capability, logic thinking capability, reasoning capability and the like, so that the feedback content of a patient in the follow-up interaction process can be accurately interpreted, and the understanding of the feedback content of the patient is increased, so that a reply with detailed content, smooth expression and clear logic can be generated, the follow-up interaction with the patient is more professional, and the patient experience is more humanized; namely, the follow-up interaction is carried out with the patient by utilizing the large language model, so that the follow-up quality is improved.
Referring to fig. 2, fig. 2 is a flowchart illustrating an embodiment of step S11 shown in fig. 1. It should be noted that, if there are substantially the same results, the embodiment of the present application is not limited to the flow sequence shown in fig. 2. As shown in fig. 2, the present embodiment includes:
step S21: and generating a follow-up problem by using the large language model, and taking the follow-up problem as interactive content.
In this embodiment, a follow-up question is generated using a large language model, and the follow-up question is used as the interactive content. And generating a follow-up problem by the large language model, and then interacting with the patient based on the generated follow-up problem to realize follow-up on the patient.
In one embodiment, the follow-up content is generated by using a large language model, specifically: acquiring disease information aimed at the follow-up visit by using a large language model, and searching out a problem set related to the disease information from a problem library; a follow-up question is obtained from the question set. The purpose of the follow-up is to know the relevant condition of the patient after operation or discharge, and the relevant condition of the patient to be known is different for each disease; therefore, the problem of the problem concentration related to the disease information is used as the follow-up problem, the follow-up interaction with the patient can be performed more specifically, and the follow-up quality is improved.
Step S22: the interactive content is provided to the patient and feedback content is received from the patient for the interactive content.
In this embodiment, the interactive contents are provided to the patient, and feedback contents of the patient for the interactive contents are received. The large language model provides the generated follow-up questions to the patient, who may respond back to the received follow-up questions.
In one embodiment, the interactive content may be provided to the patient literally and feedback content of the patient for the interactive content received literally. That is, the large language model carries out the follow-up of the patient by the text mode, and the text mode can be short messages, weChat and the like, and is not limited herein.
In other embodiments, the interactive content may also be provided to the patient in a voice manner and feedback content of the patient for the interactive content may be received in a voice manner. That is, the large language model carries out follow-up on the patient through a voice mode, so as to realize intelligent voice follow-up, and the voice mode can be a WeChat voice call, an outbound device making a call, and the like, and is not limited herein.
In one embodiment, the interactive content is provided to the patient through an outbound device. In this case, the speech corresponding to the interactive content may be generated by the large language model itself or may be generated by the speech processing device. Specifically, under the condition that the large language model generates the interactive content by itself, the interactive content is provided to the patient in a voice mode, specifically: generating a first voice of the interactive content by using the large language model, and transmitting the first voice to the outbound device so that the outbound device transmits the first voice to the patient; that is, the large language model itself has the capability of converting the interaction content generated by itself into the first voice, and when the outbound device receives the first voice of the interaction content sent by the large language model, the first voice is sent to the outbound device, so that the patient communicating with the outbound device receives the follow-up problem, and the voice follow-up to the patient is realized.
Further, feedback content of the patient for the interactive content is received in a voice mode, specifically: receiving second voice fed back by the patient received by the outbound equipment, and identifying the second voice by using a large language model to obtain feedback content of the patient aiming at the interactive content; that is, the large language model itself has the capability of converting voice data into text content, and when the large language model receives the second voice from the patient feedback sent by the outbound device, the second voice is recognized to obtain the feedback content of the patient for the interactive content.
Or specifically, in the case that the voice processing device generates the first voice of the interactive content, the interactive content is provided to the patient in a voice manner, specifically: transmitting the interactive content to the voice processing equipment so that the voice processing equipment generates first voice for the interactive content and transmits the first voice to a patient through the outbound equipment; that is, the interactive content generated by the large language model is sent to the voice processing device, the voice processing device generates the first voice of the interactive content, and when the outbound device receives the first voice of the interactive content sent by the voice processing device, the patient communicating with the outbound device receives the follow-up problem, so that the follow-up of the patient is realized.
Further, feedback content of the patient for the interactive content is received in a voice mode, specifically: receiving feedback content of the interaction content, which is sent by the voice processing equipment and is fed back by the patient, wherein the feedback content is obtained by the voice processing equipment through identifying second voice fed back by the patient and received by the outbound equipment; that is, the large language model itself does not have the ability to convert speech data into text content, and the second speech received from the patient is speech-recognized by the speech processing device.
Step S23: generating new interactive contents based on the newly received feedback contents by using the large language model, and re-executing the steps of providing the interactive contents to the patient, receiving the feedback contents of the patient for the interactive contents and the following steps until the follow-up visit end condition is met.
In this embodiment, a new interactive content is generated based on the newly received feedback content using the large language model, and the providing of the interactive content to the patient is re-performed, and the feedback content of the patient for the interactive content and the subsequent steps are received until the follow-up condition is satisfied.
In an embodiment, as shown in fig. 3, fig. 3 is a flow chart of an embodiment of generating new interactive content based on newly received feedback content by using a large language model, and the method specifically includes the following sub-steps:
step S31: and analyzing the latest received feedback content by using the large language model to obtain an analysis result.
In the embodiment, the latest received feedback content is analyzed by using a large language model to obtain an analysis result; wherein the analysis result includes at least one of: a first analysis result characterizing a first answer related to a previous follow-up question, a second analysis result characterizing whether the patient question is included. In the case that the analysis result contains a first analysis result of a first answer related to a previous follow-up question, step S32 is performed; in case the analysis result comprises a second analysis result of the patient problem, step S33 is performed.
Step S32: in response to the analysis result including the first analysis result, a new follow-up question is generated as new interactive content based on the first analysis result using the large language model.
In this embodiment, in response to the analysis result including the first analysis result, a new follow-up question is generated as new interactive content based on the first analysis result using the large language model. The analysis result of the corresponding feedback content comprises a first analysis result, which shows that the latest received feedback content is fed back by the patient aiming at the previous follow-up problem, and shows that the feedback content of the patient is normal; at this time, the patient is followed for other follow-up questions to more fully understand the patient's post-operative or post-discharge conditions.
In one embodiment, a large language model is used to generate a new follow-up question based on the first analysis result, so as to be used as new interaction content, specifically: and acquiring new follow-up questions from the question library by using a large language model. Patients are followed for other follow-up questions to more fully understand the patient's post-operative or post-discharge conditions.
In one embodiment, a new question may be randomly selected from a question library as a follow-up question. In other embodiments, a large language model is used to obtain new follow-up questions from a question bank, specifically: carrying out semantic understanding on the first answer by using the large language model to obtain a semantic result; and selecting a follow-up question with a new preset association relation with the semantic result from the question library. For example: when the feedback content of the patient aiming at the follow-up problem A in the problem library is 'no', the follow-up problem B is selected from the problem library to serve as a new follow-up problem, and when the feedback content of the patient aiming at the follow-up problem A in the problem library is 'yes', the follow-up problem C is selected from the problem library to serve as a new follow-up problem.
In one embodiment, in response to the newly received feedback content not including a first answer related to a previous follow-up question, the description of the previous follow-up question is adjusted using the large language model as a new follow-up question. The feedback content does not contain a first answer related to the previous follow-up question, which indicates that the latest received feedback content is that the patient does not feed back for the previous follow-up question, and indicates that the feedback content of the patient is abnormal; at this time, the description of the previous follow-up problem is adjusted by using the large language model to serve as a new follow-up problem, so that the follow-up problem is provided for the patient again, the accuracy of feedback content of the patient aiming at each follow-up problem is ensured, and the follow-up quality is improved.
In one embodiment, the description of the previous follow-up question is adjusted by using a large language model, specifically: the location of at least some of the content in the previous follow-up question is adjusted using the large language model. That is, the description of the follow-up question is adjusted by adjusting the position of the content in the follow-up question. For example, the previous follow-up question is "do you eat" and the new follow-up question is "do you eat? ".
Step S33: in response to the analysis result comprising a second analysis result and the second analysis result comprising the patient question, a second answer to the patient question is generated using the large language model as new interactive content.
In this embodiment, in response to the analysis result including the second analysis result and the second analysis result including the patient question, a second answer to the patient question is generated as new interactive content using the large language model. When the analysis result comprises a second analysis result and the second analysis result is a patient-containing problem, indicating that the patient currently has a medical question; the large language model has strong understanding capability, logic thinking capability and reasoning capability, so that the large language model can understand and accurately read patient questions fed back by patients and can generate answers with detailed contents, smooth expression and clear logic; that is, the large language model can provide real-time, professional, accurate and humanized medical advice for the patient feedback problem of the patient, and the follow-up quality and efficiency are improved.
In one embodiment, after generating a second answer to the patient question using the large language model as new interactive content, the large language model is updated using the patient question and feedback content for the patient question to iteratively update the large language model using the patient preferences.
In an embodiment, in a case where the analysis results include the first analysis result and the second analysis result, and the second analysis result is a question of the patient, the step of generating the second answer to the question of the patient using the large language model as new interactive contents is preferentially performed, and in a case where the interaction to the question of the patient is ended, the step of generating the new follow-up question based on the first analysis result using the large language model as new interactive contents is performed again.
Referring to fig. 4, fig. 4 is a schematic structural diagram of an embodiment of a follow-up device according to the present application. The follow-up device 40 comprises an interaction module 41 and a determination module 42. The interaction module 41 is used for carrying out follow-up interaction with the patient by utilizing the large language model, so as to obtain feedback content of the patient in the follow-up interaction, wherein the follow-up interaction comprises multiple rounds of answers; the determining module 42 is configured to obtain a follow-up result based on the feedback content.
Wherein the multiple rounds of questions include at least one set of follow-up questions, the follow-up questions being generated by the large language model and the first answers to the follow-up questions being fed back by the patient.
Wherein the multiple round of questions further includes at least one set of patient questions, the patient questions being patient feedback by the patient and second answers being generated by the large language model for the patient questions.
The interaction module 41 is configured to perform a follow-up interaction with the patient by using the large language model, and obtain feedback content of the patient in the follow-up interaction, and specifically includes: generating a follow-up problem by using the large language model, and taking the follow-up problem as interactive content; providing the interactive content to the patient and receiving feedback content of the patient for the interactive content; generating new interactive contents based on the newly received feedback contents by using the large language model, and re-executing the steps of providing the interactive contents to the patient, receiving the feedback contents of the patient for the interactive contents and the following steps until the follow-up visit end condition is met.
Wherein, the interaction module 41 is configured to generate new interaction content based on the newly received feedback content by using the large language model, and specifically includes: analyzing the latest received feedback content by using a large language model to obtain an analysis result, wherein the analysis result comprises at least one of the following components: a first analysis result characterizing whether a first answer related to a previous follow-up question is included, a second analysis result characterizing whether a patient question is included; generating a new follow-up problem as new interactive content according to the first analysis result; in response to the analysis result comprising a second analysis result and the second analysis result comprising the patient question, a second answer to the patient question is generated using the large language model as new interactive content.
The interaction module 41 is configured to generate a new follow-up question based on the first analysis result by using the large language model, so as to serve as new interaction content, and specifically includes: acquiring new follow-up questions from the question library by using a large language model; and/or, in response to the newly received feedback content not including the first answer related to the previous follow-up question, adapting the description of the previous follow-up question as a new follow-up question using the large language model.
Wherein, the interaction module 41 is configured to obtain a new follow-up question from the question library by using a large language model, and includes: carrying out semantic understanding on the first answer by using the large language model to obtain a semantic result; selecting a new follow-up question with a preset association relation with a semantic result from a question library; and/or, the interaction module 41 is configured to adjust the description of the previous follow-up question by using the large language model, including: the location of at least some of the content in the previous follow-up question is adjusted using the large language model.
The method comprises the steps of generating a second answer to a patient problem by using a large language model to serve as new interactive content under the condition that the analysis results comprise a first analysis result and a second analysis result, and generating a new follow-up problem based on the first analysis result by using the large language model to serve as new interactive content under the condition that interaction to the patient problem is finished; and/or the follow-up device further comprises an updating module 43, wherein the updating module 43 is used for generating a second answer to the patient question by using the large language model to serve as new interaction content, and specifically comprises: the large language model is updated with patient questions and feedback content for the patient questions.
The interaction module 41 is configured to generate a follow-up question by using a large language model, and specifically includes: acquiring disease information aimed at the follow-up visit by using a large language model, and searching out a problem set related to the disease information from a problem library; a follow-up question is obtained from the question set.
The interaction module 41 is configured to provide the interaction content to the patient, and receive feedback content of the patient for the interaction content, and specifically includes: the interactive content is provided to the patient in a voice mode, and feedback content of the patient aiming at the interactive content is received in a voice mode.
The interaction module 41 is configured to provide the interaction content to the patient in a voice manner, and specifically includes: generating a first voice of the interactive content by using the large language model, and transmitting the first voice to the outbound device so that the outbound device transmits the first voice to the patient; or the interactive content is sent to the voice processing equipment, so that the voice processing equipment generates first voice for the interactive content and transmits the first voice to a patient through the outbound equipment; the interaction module 41 is configured to receive feedback content of the patient for the interaction content in a voice manner, and specifically includes: receiving second voice fed back by the patient received by the outbound equipment, and identifying the second voice by using a large language model to obtain feedback content of the patient aiming at the interactive content; or receiving feedback content of the interaction content, which is sent by the voice processing equipment and is fed back by the patient, wherein the feedback content is obtained by the voice processing equipment through identifying the second voice fed back by the patient and received by the outbound equipment.
Referring to fig. 5, fig. 5 is a schematic structural diagram of an embodiment of an electronic device according to the present application. The electronic device 50 comprises a memory 51 and a processor 52 coupled to each other, the processor 52 being adapted to execute program instructions stored in the memory 51 for carrying out the steps of any of the above-described embodiments of the follow-up method. In one particular implementation scenario, electronic device 50 may include, but is not limited to: the microcomputer and the server, and the electronic device 50 may also include a mobile device such as a notebook computer and a tablet computer, which is not limited herein.
In particular, the processor 52 is configured to control itself and the memory 51 to implement the steps of any of the above-described embodiments of the follow-up method. The processor 52 may also be referred to as a CPU (Central Processing Unit ). The processor 52 may be an integrated circuit chip having signal processing capabilities. The Processor 52 may also be a general purpose Processor, a digital signal Processor (DIGITAL SIGNAL Processor, DSP), an Application SPECIFIC INTEGRATED Circuit (ASIC), a Field-Programmable gate array (Field-Programmable GATE ARRAY, FPGA) or other Programmable logic device, a discrete gate or transistor logic device, a discrete hardware component. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. In addition, the processor 52 may be commonly implemented by an integrated circuit chip.
Referring to fig. 6, fig. 6 is a schematic structural diagram of an embodiment of a computer readable storage medium according to the present application. The computer readable storage medium 60 of an embodiment of the present application stores program instructions 61 that when executed implement the method provided by any embodiment of the present application of the follow-up method and any non-conflicting combination. Wherein the program instructions 61 may form a program file stored in the above-mentioned computer readable storage medium 60 in the form of a software product for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the application. And the aforementioned computer-readable storage medium 60 includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, an optical disk, or other various media capable of storing program codes, or a terminal device such as a computer, a server, a mobile phone, a tablet, or the like.
If the technical scheme of the application relates to personal information, the product applying the technical scheme of the application clearly informs the personal information processing rule before processing the personal information and obtains the autonomous agreement of the individual. If the technical scheme of the application relates to sensitive personal information, the product applying the technical scheme of the application obtains individual consent before processing the sensitive personal information, and simultaneously meets the requirement of 'explicit consent'. For example, a clear and remarkable mark is set at a personal information acquisition device such as a camera to inform that the personal information acquisition range is entered, personal information is acquired, and if the personal voluntarily enters the acquisition range, the personal information is considered as consent to be acquired; or on the device for processing the personal information, under the condition that obvious identification/information is utilized to inform the personal information processing rule, personal authorization is obtained by popup information or a person is requested to upload personal information and the like; the personal information processing rule may include information such as a personal information processor, a personal information processing purpose, a processing mode, and a type of personal information to be processed.
The foregoing description is only of embodiments of the present application, and is not intended to limit the scope of the application, and all equivalent structures or equivalent processes using the descriptions and the drawings of the present application or directly or indirectly applied to other related technical fields are included in the scope of the present application.
Claims (14)
1. A method of follow-up, the method comprising:
carrying out follow-up interaction with a patient by using a large language model to obtain feedback content of the patient in the follow-up interaction, wherein the follow-up interaction comprises multiple rounds of answers;
And obtaining a follow-up result based on the feedback content.
2. The method of claim 1, wherein the multiple rounds of questions comprise at least one set of follow-up questions, the follow-up questions being generated by the large language model and the first answers to the follow-up questions being fed back by the patient.
3. The method of claim 2, wherein the multiple rounds of questions and answers further comprise at least one set of patient questions and answers, the patient questions and answers being patient questions fed back by the patient and second answers being generated by the large language model for the patient questions.
4. A method according to any one of claims 1 to 3, wherein the follow-up interaction with the patient using the large language model results in feedback content of the patient in the follow-up interaction, comprising:
Generating a follow-up problem by using the large language model, and taking the follow-up problem as interactive content;
Providing the interactive content to the patient, and receiving feedback content of the patient for the interactive content;
And generating new interactive contents based on the newly received feedback contents by utilizing the large language model, and re-executing the steps of providing the interactive contents to the patient, and receiving the feedback contents of the patient aiming at the interactive contents and the subsequent steps until the follow-up ending condition is met.
5. The method of claim 4, wherein generating new interactive content based on the feedback content received last using the large language model comprises:
Analyzing the latest received feedback content by utilizing the large language model to obtain an analysis result, wherein the analysis result comprises at least one of the following components: a first analysis result characterizing whether a first answer related to a previous said follow-up question is included, a second analysis result characterizing whether a patient question is included;
Generating a new follow-up question based on the first analysis result by using the large language model as new interaction content in response to the analysis result including the first analysis result;
in response to the analysis result including the second analysis result and the second analysis result including the patient question, generating a second answer to the patient question using the large language model as the new interactive content.
6. The method of claim 5, wherein generating new follow-up questions as new interactive contents based on the first analysis result using the large language model comprises:
Acquiring new follow-up questions from a question library by using the large language model; and/or the number of the groups of groups,
In response to the most recently received feedback content not including a first answer related to a previous said follow-up question, adapting a description of the previous said follow-up question using the large language model as the new follow-up question.
7. The method of claim 6, wherein the obtaining new follow-up questions from the question bank using the large language model comprises:
carrying out semantic understanding on the first answer by using the large language model to obtain a semantic result;
Selecting a new follow-up question with a preset association relation with the semantic result from a question library;
and/or said adapting, using said large language model, a description of a previous said follow-up question, comprising:
The location of at least some of the content in the previous said follow-up question is adjusted using the said large language model.
8. The method according to any one of claims 5 to 7, wherein in the case where the analysis results include the first analysis result and a second analysis result, and the second analysis result is a question containing the patient, the step of generating a second answer to the patient question using the large language model is preferentially performed as the new interactive content, and in the case where the interaction for the patient question is ended, the step of generating a new follow-up question using the large language model based on the first analysis result is performed again as the new interactive content;
And/or after said generating a second answer to said patient question using said large language model as said new interactive content, further comprising:
Updating the large language model with the patient questions and feedback content for the patient questions.
9. The method of any of claims 4 to 8, wherein the generating a follow-up question using the large language model comprises:
Acquiring disease information aimed at the follow-up visit by using the large language model, and searching out a problem set related to the disease information from a problem library;
And acquiring one follow-up question from the question set.
10. The method of any of claims 4 to 9, wherein the providing the interactive content to the patient and receiving feedback content from the patient for the interactive content comprises:
And providing the interactive content for the patient in a voice mode, and receiving feedback content of the patient aiming at the interactive content in a voice mode.
11. The method of claim 10, wherein said providing said interactive content to said patient in voice comprises:
Generating a first voice of the interactive content by using the large language model, and sending the first voice to an outbound device so that the outbound device transmits the first voice to the patient; or alternatively
Transmitting the interactive content to a voice processing device, so that the voice processing device generates the first voice for the interactive content and transmits the first voice to the patient through the outbound device;
The receiving, by voice, feedback content of the patient for the interactive content includes:
Receiving second voice fed back by the patient and received by the outbound device, and identifying the second voice by using the large language model to obtain feedback content of the patient aiming at the interactive content; or alternatively
And receiving feedback content of the interaction content, which is sent by the voice processing equipment and is fed back by the patient, of the interaction content, wherein the feedback content is obtained by the voice processing equipment through identifying second voice fed back by the patient and received by the outbound equipment.
12. A follow-up device, the follow-up device comprising:
the interaction module is used for carrying out follow-up interaction with the patient by utilizing the large language model to obtain feedback content of the patient in the follow-up interaction, wherein the follow-up interaction comprises multiple rounds of answers;
and the determining module is used for obtaining a follow-up result based on the feedback content.
13. An electronic device comprising a processor and a memory, the memory storing program instructions, the processor configured to execute the program instructions to implement the follow-up method of any of claims 1-11.
14. A computer readable storage medium for storing program instructions executable to implement the follow-up method of any one of claims 1-11.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311630883.XA CN118014078A (en) | 2023-11-30 | 2023-11-30 | Follow-up method and related device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311630883.XA CN118014078A (en) | 2023-11-30 | 2023-11-30 | Follow-up method and related device |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN118014078A true CN118014078A (en) | 2024-05-10 |
Family
ID=90952502
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311630883.XA Pending CN118014078A (en) | 2023-11-30 | 2023-11-30 | Follow-up method and related device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN118014078A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120220933A (en) * | 2025-05-29 | 2025-06-27 | 杭州惠每医疗科技有限公司 | Intelligent follow-up method and system based on large model |
-
2023
- 2023-11-30 CN CN202311630883.XA patent/CN118014078A/en active Pending
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN120220933A (en) * | 2025-05-29 | 2025-06-27 | 杭州惠每医疗科技有限公司 | Intelligent follow-up method and system based on large model |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11145390B2 (en) | Methods and systems for recommending filters to apply to clinical trial search results using machine learning techniques | |
| US20190018694A1 (en) | Virtual laboratory assistant platform | |
| WO2019080662A1 (en) | Information recommendation method, device and apparatus | |
| CN111046156B (en) | Reward data determination method, device and server | |
| Maguire et al. | Shaping the future of digital technology in health and social care | |
| Welvaars et al. | Implications of resampling data to address the class imbalance problem (IRCIP): an evaluation of impact on performance between classification algorithms in medical data | |
| CN110473636B (en) | Intelligent medical advice recommendation method and system based on deep learning | |
| CN109545394A (en) | A kind of way of inquisition and device | |
| CN109599176B (en) | Method and device for recommending inquiry skills and online auxiliary diagnosis system | |
| US20240105346A1 (en) | Optimizing messaging to service principals using machine learning | |
| CN120542540A (en) | Data processing method, device, equipment and medium based on context protocol | |
| Salvo et al. | Will the future of pharmacovigilance be more automated? | |
| KR102306953B1 (en) | Method of medical support based on voice recognition | |
| WO2021139223A1 (en) | Method and apparatus for interpretation of clustering model, computer device, and storage medium | |
| CN120596202A (en) | Interaction method, device, electronic device and computer-readable storage medium | |
| WO2025024554A1 (en) | Systems and methods for phenotyping using large language model prompting | |
| CN118014078A (en) | Follow-up method and related device | |
| CN112396325B (en) | Case processing method, case processing device, case processing medium, and electronic apparatus | |
| CN109524108A (en) | A kind of method for building up and device of inquiry model | |
| Wang et al. | Biomanufacturing harvest optimization with small data | |
| US20220013030A1 (en) | Extended service-providing system and method for providing artificial intelligence prediction results for extended education contents through api access interface server | |
| WO2025111558A1 (en) | Methods and systems for optimizing healthcare data management using generative artificial intelligence agents | |
| CN119396952A (en) | Method, device, equipment and storage medium for replying to inquiries | |
| CN113160914A (en) | Online inquiry method and device, electronic equipment and storage medium | |
| WO2021151329A1 (en) | Patient medication behavior intervention method and apparatus, and server and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |