Platform and system for inquiry assessment

Abstract

The invention discloses a platform and a system for inquiry assessment. A platform for interrogation assessment includes a computing resource and a large model. The large model is configured to receive cue input, where the cue input includes an interrogation conversation between the evaluated person and the virtual standardized patient and interrogation evaluation criteria. The large model is further configured to generate an interrogation evaluation output based on the cue input, where the interrogation evaluation output includes an interrogation evaluation parse associated with an interrogation evaluation criterion.

Description

Technical Field

[0001] The present invention generally relates to the evaluation of medical students during the teaching process, and in particular to the evaluation of medical students during the consultation process. Background Art

[0002] In the teaching and training process of medical students, the application of standardized patients (SP) has significantly improved the skills of medical students in collecting medical histories (for example, see Xue Laixi, Yang Changbo, Cao Shengping, et al. Analysis of the effect of the application of standardized patients (SP) on the collection of medical medical histories [J]. Health Must-Read 2020 Issue 16, Page 260, 2020).

[0003] Standardized patients are normal people or patients who can accurately express the actual clinical problems of patients after standardized and systematic training. Their patient attributes are essentially performed. At present, in the doctor-patient communication and clinical testing examinations, real standard patients first simulate typical symptoms and complaints, and then the candidates have a dialogue with the actors who play the standardized patients to complete the medical history collection, so as to examine the medical students' appearance, attitude, language and other doctor-patient communication skills during clinical consultation and questioning.

[0004] In the teaching process, standardized patient teaching cannot be easily applied to the academic learning schedule.

[0005] The teaching model of medical history collection thinking training system based on virtual standardized patient (VSP) in the teaching of medical history skills can significantly improve the medical students' medical history collection ability, thereby improving the teaching quality (for example, see Li Xiang, Chen Xiaoyun, Zhang Huiqun, et al. Research on the application of medical history collection thinking training system based on VSP in medical history skills teaching [J]. Chinese Medical Education Technology, 2023, 37(2): 186-190, 201. DOI: 10.13566 / j.cnki.cmet.cn61-1317 / g4.202302013.).

[0006] For teachers, there is a lack of convenient auxiliary tools in the teaching of virtual standardized patient medical history collection thinking training system. For academics, in the process of daily medical history collection exercises, it is impossible to effectively locate the problems in the exercises, which hinders the improvement of skills in medical history collection.

[0007] There is a need in this field for technology that facilitates medical history-taking exercises and assessments for both teachers and students during the teaching and training of medical students. Summary of the invention

[0008] The present invention is provided to facilitate teachers and students to perform medical history collection exercises and assessments during the teaching and training of medical students.

[0009] One aspect of the present invention provides a platform for medical consultation and evaluation, comprising: computing resources; and a large model, wherein the large model is configured to: receive a prompt word input, wherein the prompt word input comprises a medical consultation dialogue between a person being evaluated and a virtual standardized patient and medical consultation evaluation criteria; and generate a medical consultation and evaluation output based on the prompt word input, wherein the medical consultation and evaluation output comprises a medical consultation and evaluation analysis, and the medical consultation and evaluation analysis is associated with the medical consultation and evaluation criteria.

[0010] The platform as described above, wherein the consultation evaluation criteria include medical education evaluation criteria, and the prompt word input is at least partially based on processing the medical education evaluation criteria into prompt word instructions.

[0011] A platform as described in any of the above items, wherein the consultation evaluation criteria includes multiple consultation evaluation criteria, the multiple consultation evaluation criteria are respectively associated with multiple consultation evaluation sub-processes, the consultation evaluation output includes multiple consultation evaluation resolutions and multiple consultation evaluation scores respectively associated with the multiple consultation evaluation resolutions, and the large model is configured to: execute the multiple consultation evaluation sub-processes to extract the multiple consultation evaluation resolutions and the multiple consultation evaluation scores; and combine the multiple consultation evaluation resolutions and the multiple consultation evaluation scores to obtain the consultation evaluation output.

[0012] A platform as described in any of the above items, wherein the large model is further configured to: generate a deduction analysis and / or a total score based on the multiple consultation evaluation analysis and the multiple consultation evaluation scores.

[0013] A platform as described in any of the above items, wherein the large model is reinforced trained with multiple question-and-answer evaluation outputs and multiple question-and-answer evaluation parsings and question-and-answer evaluation scores for each question-and-answer evaluation output.

[0014] A platform as described in any of the above items, wherein the consultation dialogue is generated based on the patient's medical history and personality.

[0015] A platform as described in any of the above items, wherein the prompt word input is a first prompt word input, and the large model is further configured to: receive a second prompt word input, the second prompt word input including the patient's medical history and patient personality; and generate the consultation dialogue based on the second prompt word input and the questions asked by the person being evaluated, wherein the consultation dialogue includes the questions asked by the evaluator and the corresponding answers generated by the large model.

[0016] The platform as described in any of the above items further comprises: a user interface configured to receive input from the person being assessed and present output of the large model to the person being assessed.

[0017] A platform as described in any of the above items, wherein the platform is configured to operate in a practice mode, and in the practice mode, the input of the person being evaluated includes symptoms or disease classification and the patient's personality and character setting to generate the second prompt word input.

[0018] A platform as described in any of the above items, wherein the platform is configured to operate in an examination mode, and in the examination mode, the large model is configured to: receive the second prompt word input including randomly determined symptoms or disease types and the patient's personality.

[0019] A platform as described in any of the above items, wherein the patient's medical history is associated with a symptom or disease, and / or the patient's medical history includes a standard medical history.

[0020] A platform as described in any of the above items, wherein the patient's personality includes one of the following:

[0021] Normal type, negative type, impatient type, and pressure-applying type.

[0022] A platform as described in any of the above items, wherein the platform is a mobile platform.

[0023] Another aspect of the present invention provides a system for medical consultation and evaluation, comprising: a platform as described in any of the above items; and a web-based platform communicatively coupled to the platform, wherein the web-based platform is configured to analyze and manage the platform.

[0024] In the system as described above, the web platform is configured to perform at least one of the following: analyzing and presenting the consultation evaluation output and / or usage of the platform; providing a page for managing the medical records required by the platform; providing a page for managing the prompt word input received by the large model of the platform; providing a page for managing examination parameters.

[0025] The platform and system for medical consultation evaluation according to the present invention utilize a large model to provide a convenient means of practice and assessment for the medical students' medical consultation evaluation process, thereby promoting the cultivation, training and improvement of medical students in medical history collection. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Embodiments of the present invention are described with reference to the accompanying drawings.

[0027] Figure 1 is a block diagram of a platform for diagnostic evaluation according to some embodiments of the present invention.

[0028] Figure 2 is a schematic diagram of a medical consultation and evaluation process according to some embodiments of the present invention.

[0029] Figure 3 is a schematic diagram of a medical consultation and evaluation process according to some embodiments of the present invention.

[0030] Figure 4 is a schematic diagram of a medical consultation and evaluation process according to some embodiments of the present invention.

[0031] Figure 5 The diagram is a flowchart of a process based on a human-computer interaction perspective of a consultation and evaluation process according to some embodiments of the present invention.

[0032] Figure 6 is a flowchart of a process of a consultation and evaluation process based on a human-computer interaction perspective in a practice mode according to some embodiments of the present invention.

[0033] Figure 7 It is a flowchart of a process of a consultation and evaluation process based on a human-computer interaction perspective in an examination mode according to some embodiments of the present invention.

[0034] Figure 8 is a block diagram of a system for diagnostic evaluation according to some embodiments of the present invention. DETAILED DESCRIPTION

[0035] According to one aspect of the present invention, a platform for diagnostic evaluation is provided.

[0036] Figure 1 is a block diagram of a platform 100 for diagnostic evaluation according to some embodiments of the present invention.

[0037] The platform 100 may include a computing resource 110 and a large model 130. The large model 130 may call the computing resource 110 to complete its corresponding calculation, thereby realizing the corresponding function. As an example, the computing resource 110 may include one or more processors, one or more processor cores, one or more execution units, etc. As an example, the large model 130 may include a large language model (LLM).

[0038] The large model 130 may be configured to receive a prompt input. The prompt input may include a consultation dialogue between the person being evaluated (e.g., a medical student) and a virtual standardized patient and a consultation evaluation standard. The large model 130 may also be configured to generate a consultation evaluation output based on the prompt input. The consultation evaluation output may include a consultation evaluation analysis, which may be associated with the consultation evaluation standard.

[0039] Some embodiments use a large model to combine the medical consultation dialogue and medical consultation evaluation standards into a prompt word input for the large model, which can obtain a detailed medical consultation evaluation analysis of the medical consultation process, providing a convenient means for medical students to practice medical history collection and for teachers to assess this aspect, while also promoting the application of virtual standardized patients.

[0040] In some embodiments, the platform 100 may be a mobile terminal platform. As an example, the platform 100 may be a smart phone, a tablet computer, a laptop computer, a dedicated handheld device, etc. As an example, the mobile terminal platform may be implemented through an H5 page.

[0041] In some embodiments, the platform 100 may also be a fixed-end platform. As an example, the platform 100 may be a desktop computer, a dedicated fixed device, or the like.

[0042] The following will combine Figure 2 , Figure 3 Details of how the large model 130 generates the diagnostic assessment output are further described.

[0043] In some embodiments, the consultation dialogue between the person being evaluated and the virtual standardized patient can also be generated by the large model 130. Figure 3 , Figure 4 The details of the large model 130 generating the consultation dialogue are further described.

[0044] In some embodiments, the platform 100 may further include a user interface 150. The user interface 150 may be configured to receive input from the person being assessed and present the output of the large model to the person being assessed. As an example, the user interface 150 may include a display, a keyboard, a mouse, a touch screen, a microphone, a speaker, etc. Figure 5-Figure 7 The process of human-computer interaction performed by the person being evaluated using the interface 150 is described.

[0045] Figure 2 is a schematic diagram of a medical consultation and evaluation process according to some embodiments of the present invention. Figure 2 The process shown in Figure 1 The large model 130 in the example invokes the computing resource 110 for execution, but the scope of the present invention is not limited in this regard.

[0046] like Figure 2 As shown in FIG, the consultation dialogue 210 and the consultation evaluation criteria 220 can be combined into a prompt word input 230. The consultation dialogue 210 can include questions 211 asked by the person being evaluated and answers 212 of the virtual standardized patient. The consultation dialogue 210 can be generated by the person being evaluated talking to the same or different large models.

[0047] Large model 240 (which may be Figure 1The large model 130 in the example may receive a prompt word input 230 to generate a question evaluation output 250. The question evaluation output 250 may include a question evaluation analysis. The question evaluation analysis may be associated with a question evaluation criterion.

[0048] In some embodiments, the medical interview evaluation criteria may include medical education evaluation criteria. Accordingly, the prompt word input 230 may be obtained at least in part based on processing the medical education evaluation criteria into prompt word instructions. As an example, the medical interview evaluation analysis associated with the medical education evaluation criteria may include one or more of the following: 1. Whether the collection of the patient's basic information is comprehensive; 2. Whether the summary of the main complaint is correct and concise; 3. Whether the current medical history is centered around the main complaint and related identification inquiries; 4. Whether to inquire about the causes and accompanying symptoms of the main complaint symptoms; 5. Whether to inquire about diet, sleep, defecation, and weight changes; 6. Whether to inquire about medical treatment and examinations; 7. Whether the diagnosis and treatment process accurately describes the treatment effect; 8. Whether the past history is centered around the main complaint; 9. Whether the personal history is centered around the main complaint; 10. Whether the marriage and childbearing history are adjusted according to the patient's information; 11. Whether to inquire about family history based on the main complaint; 12. Whether the medical interview is organized and can grasp the key points; 13. Whether the medical interview can focus on the condition; 14. Whether the language of the medical interview is appropriate; 15. Whether the attitude of the medical interview is serious and whether the communication is polite,

[0049] 16. Whether the concept of loving and hurting is always present during the consultation.

[0050] Some embodiments combine medical education assessment standards into the prompt word input 230 of the large model 240 to obtain a detailed assessment analysis of the consultation process (rather than a simple score for the entire process), which helps medical students and teachers to accurately locate weak points in the learning and teaching process, thereby strengthening learning and practice in a targeted manner.

[0051] In some embodiments, when defining the input and output fields of the large model 240, the consultation evaluation criteria can be disassembled into multiple consultation evaluation sub-processes, so that the consultation evaluation output includes the consultation evaluation analysis and the consultation evaluation score associated with each sub-process. Accordingly, the large model 240 can be configured to execute multiple consultation evaluation sub-processes to extract multiple consultation evaluation analysis and multiple consultation evaluation scores. The large model 240 can also be configured to combine multiple consultation evaluation analysis and multiple consultation evaluation scores to obtain the consultation evaluation output.

[0052] In some embodiments, the large model 240 may also be configured to generate a deduction analysis based on multiple question-and-answer evaluation analysis and question-and-answer evaluation scores. In some embodiments, the large model 240 may also be configured to generate a total score based on multiple question-and-answer evaluation analysis and question-and-answer evaluation scores. In some embodiments, the large model 240 may also be configured to generate a deduction analysis and a total score based on multiple question-and-answer evaluation analysis and question-and-answer evaluation scores.

[0053] As an example, the format of the input and output fields of the large model 240 may be as follows:

[0054] {Extract 1 Parse

[0055] Extract 1 score

[0056] Extract 2 Parse

[0057] Extract 2 score

[0058] …

[0059] Deduction analysis

[0060] Total score}

[0061] Some embodiments can guide the thinking chain of the large model 240 and enhance its output effect by breaking down the medical consultation evaluation criteria into sub-processes, installing extraction analysis and extracting scores for each sub-process for combined output.

[0062] In some embodiments, the large model 240 may also be trained using multiple question-and-answer evaluation outputs and multiple question-and-answer evaluation parsing and question-and-answer evaluation scores of each question-and-answer evaluation output to enhance the evaluation effect.

[0063] Figure 3 is a schematic diagram of the consultation and evaluation process according to some embodiments of the present invention. Figure 3 In the diagnostic evaluation process shown in , the diagnostic dialogue 350 is generated by the big model 340 (ie, the same big model) that performs the diagnostic evaluation.

[0064] like Figure 3 As described in the above, the patient's medical history 310 and personality setting 320 can be combined into a prompt word input 330. In some embodiments, the person being evaluated can input a selection of the medical history 310 combined with the personality setting 320 to perform a medical interview and evaluation exercise. In some embodiments, the platform can also randomly generate the medical history 310 and the personality setting 320 to assess the person being evaluated.

[0065] By combining the patient's personality setting 320 with the prompt word input 330 of the large model 340, patients with various common clinical personalities can be effectively simulated, thereby increasing the authenticity of the consultation and evaluation process. At the same time, by using the large model 340, only the patient's personality setting 320 needs to be combined with its prompt word input 330 to simulate a standardized patient with a specific personality setting, which is very convenient to use.

[0066] In some embodiments, the patient's personality may include one of the following: 1. Normal type: more cooperative with the doctor; 2. Passive type: always worried about the condition; 3. Impatient type: not very cooperative in answering the doctor's questions, always giving brief answers; 4. Pressure-exerting type: putting the doctor under pressure during the consultation.

[0067] In some embodiments, medical records 310 may be associated with characteristic symptoms or diseases. Thus, the person being evaluated may be required to practice or take an exam on a specific symptom or disease, thereby enhancing the effect of teaching medical students.

[0068] In some embodiments, medical records 310 may include standard medical records. By including standard medical records with standardized formats / contents, large model 340 may better understand the medical record contents.

[0069] In some embodiments, a large amount of medical record data (eg, standard medical records) may be added to the training data of the big model 340 for fine-tuning to facilitate the big model 340's understanding of the medical record data.

[0070] The large model 340 may receive the prompt word input 330 , and based on the prompt word input, conduct a question-and-answer dialogue with the person being evaluated, thereby generating a consultation dialogue 350 .

[0071] Subsequently, the generated medical consultation dialogue 350 may be combined with the medical consultation evaluation criteria 360 to form a prompt word input 370. The large model 340 may receive the prompt word input 370 and generate a medical consultation evaluation output 380 based on the prompt word input. The process of combining the medical consultation dialogue 350 and the medical consultation evaluation criteria 360 into the prompt word input 370 may be combined with the above Figure 2 The process of combining the medical consultation dialogue 210 and the medical consultation evaluation criteria 220 into the prompt word input 230 is similar to the process described above; the process of the large model 340 receiving the prompt word input 370 and generating the medical consultation evaluation output 380 based on it can be combined with the above Figure 2 The process of the described large model 240 receiving the prompt word input 230 and generating the diagnostic evaluation output 250 based thereon is similar.

[0072] As an example, the content of prompt word input 330 may be as follows:

[0073]

[0074] The large model 340 generates a medical consultation dialogue 350 based on the above prompt word input and the questions asked by the person being evaluated (hereinafter referred to as "doctor") as follows:

[0075]

[0076]

[0077]

[0078] The prompt word input 370 may be as follows:

[0079]

[0080]

[0081]

[0082] As an example, the diagnostic evaluation output 380 generated by the large model 340 may be as follows:

[0083]

[0084] Figure 4 is a schematic diagram of a medical consultation and evaluation process according to some embodiments of the present invention.

[0085] like Figure 4 As described in the above, the patient's medical history 410 and personality setting 420 can be combined into a prompt word input 430. The large model 440 can generate a consultation dialogue 450 based on the prompt word input 430. The process of combining the patient's medical history 410 and personality setting 420 into the prompt word input 430 can be combined with the above Figure 3 The process of combining the patient's medical history 310 and personality 320 into the prompt word input 330 is similar to the process described above. The process of generating the consultation dialogue 450 based on the prompt word input 430 by the large model 440 can be combined with the above Figure 3 The process of the described large model 340 generating the consultation dialogue 350 based on the prompt word input 330 is similar.

[0086] Figure 5 is a flowchart of a process 500 based on a human-computer interaction perspective of a consultation and evaluation process according to some embodiments of the present invention. Figure 5 The human-computer interaction process shown in Figure 1 The present invention may be implemented by the user interface 150 of the platform 100 in FIG. 1 , but the scope of the present invention is not limited thereto.

[0087] After starting, process 500 may proceed to displaying a home page at block 510. As an example, the home page may include options for platform operation modes (eg, practice mode, exam mode). The assessee may select a corresponding platform operation mode.

[0088] Subsequently, process 500 may proceed to decision block 530 to determine whether the selected mode is a practice mode or an exam mode.

[0089] Subsequently, according to the determination result at determination block 530 , process 500 may proceed to block 550 : performing operations in a practice mode; or process 500 may proceed to block 570 : performing operations in a test mode.

[0090] Figure 6 FIG. 6 is a flowchart of a process 600 of a consultation and evaluation process in a practice mode based on a human-computer interaction perspective according to some embodiments of the present invention. The process 600 may be Figure 5 The specific implementation of the step at box 550 of process 500 in FIG. 5 is described below, but the scope of the present invention is not limited thereto.

[0091] In practice mode, the person being evaluated can input symptoms or disease classifications as well as the patient's personality and character through the human-computer interface to generate corresponding prompt word inputs, based on which the large model can produce a medical consultation dialogue.

[0092] At block 610 , the practice mode is enabled, and the platform then operates in the practice mode.

[0093] At block 620, the person being evaluated may input a symptom classification (or disease classification). The platform may present an exercise number based on the symptom classification input by the person being evaluated for further selection by the person being evaluated; or the platform may automatically assign an exercise number based on the symptom classification input by the person being evaluated.

[0094] At block 630, the person being evaluated may input the patient's personality. Optionally, other practice parameters may also be input by the person being evaluated. For example, the person being evaluated may input a countdown for the consultation process to better simulate an exam scenario.

[0095] At block 640, the person being evaluated can conduct a consultation process with the platform. This process can be combined with the above Figure 3 The process of generating the consultation dialogue 350 by the large model 340 described above, or combining with the above Figure 4 The process of generating the consultation dialogue 450 by the described large model 440 is similar, but the scope of the present invention is not limited thereto.

[0096] At determination block 650, the platform may determine whether the consultation process is over. As an example, the consultation process may end in response to the end of a countdown set by the person being evaluated. As an example, the consultation process may also end based on a consultation dialogue between the person being evaluated and the platform's large model indicating that the consultation process is complete (e.g., a dialogue between the person being evaluated and the virtual standardized patient indicates that the consultation process is over, e.g., the two say goodbye).

[0097] If it is determined at decision block 650 that the consultation process is over, process 600 proceeds to block 660 where the consultation process is over. Otherwise, the process returns to block 640 to continue the consultation process.

[0098] After the consultation process ends, process 600 may proceed to frame 670, where the platform displays the evaluation results to the person being evaluated. In some embodiments, the evaluation results may include a consultation evaluation analysis. In some embodiments, the evaluation results may include a consultation evaluation analysis and a consultation evaluation score for each of a plurality of consultation evaluation subprocesses. In some embodiments, the evaluation results may include a deduction analysis. In some embodiments, the evaluation results may include a total score.

[0099] Subsequently, process 600 may proceed to a determination block 680 to determine whether to practice the next question. In some embodiments, the determination may be based on the number of practice questions set by the person being evaluated before the practice mode begins. In some embodiments, after the current consultation process ends, it may be determined based on the person being evaluated's choice whether to continue practicing the next question.

[0100] If it is determined at decision block 680 that the next question is to be practiced, process 600 returns to block 640. Otherwise, process 600 continues to progress to block 690, where the platform rolls back. In some embodiments, the platform may roll back to the step at block 620 so that the person being evaluated continues to select other symptom categories to practice. In some embodiments, the platform may end the entire practice mode.

[0101] Although Figure 6 600 of the platform operating in the training mode is shown in FIG. 6 , but those skilled in the art will appreciate that the steps and the order in which they are performed are exemplary only. In some embodiments, the process 600 may be performed by omitting one or more steps therein. In some embodiments, one or more steps in the process 600 may be performed in an interchangeable order.

[0102] Figure 7 FIG. 7 is a flowchart of a process 700 of a consultation and evaluation process in an examination mode based on a human-computer interaction perspective according to some embodiments of the present invention. Process 700 may be Figure 5 The specific implementation of the step at box 570 of process 500 in FIG. 5 is described below, but the scope of the present invention is not limited thereto.

[0103] In the exam mode, the platform can randomly determine the symptoms or disease type and the patient's personality as part of the input (e.g., prompt word input) of the large model for the consultation dialogue with the person being evaluated. Usually, the exam also has a time limit, so in the exam mode, the platform also sets a countdown time for the exam (e.g., 10 minutes), and the consultation dialogue between the person being evaluated and the platform will be forced to end after the countdown ends.

[0104] At block 701 , exam mode is enabled.

[0105] Then, at decision block 702 , the platform may determine whether to conduct an exam or to manage exam records.

[0106] If it is determined at decision block 702 that an exam is to be conducted, process 700 may proceed to starting the exam at block 703 .

[0107] Subsequently, at block 704 , a consultation process may be conducted between the person being evaluated and the platform.

[0108] At judgment box 705, the platform may determine whether the countdown has ended. If so, process 700 may proceed to box 707 to end the consultation. If not, process 700 may proceed to box 706 to determine whether the consultation has ended. The determination at box 706 may be based on the completion of the consultation process between the person being evaluated and the platform before the countdown ends. If it is determined at judgment box 706 that the consultation process has ended, process 700 may continue to proceed to box 707 to end the consultation. Otherwise, process 700 may return to box 704 to continue the consultation process.

[0109] After the consultation is finished, process 700 may proceed to frame 708, where the platform displays the evaluation results to the person being evaluated. In some embodiments, the evaluation results may include a consultation evaluation analysis. In some embodiments, the evaluation results may include a consultation evaluation analysis and a consultation evaluation score for each of a plurality of consultation evaluation subprocesses. In some embodiments, the evaluation results may include a deduction analysis. In some embodiments, the evaluation results may include a total score.

[0110] Subsequently, process 700 may proceed to determine whether to take the test again at decision block 709. If yes, process 700 may return to block 701; otherwise, process 700 may end, that is, the operation in the test mode may end.

[0111] If the test record management is selected at the decision block 702, the process 700 may proceed to the test record at block 710. In this mode, the platform may record its usage and relevant statistics of the test results. For example, the number of times the platform is used, the time consumed, the pass distribution for each symptom, the overall pass distribution, etc. may be recorded.

[0112] Then, at judgment block 711, it may be determined whether to roll back. If so, process 700 rolls back to block 701. Otherwise, process 700 continues to proceed to check the record at block 712. As an example, the platform may present the recorded items to a teacher or an academic administrator.

[0113] Then, at decision block 713, it may be determined whether to roll back. If so, process 700 rolls back to block 710. Otherwise, process 700 may end, that is, the operation in the test mode may end.

[0114] According to another aspect of the present invention, a system for diagnostic evaluation is provided.

[0115] Figure 8 is a block diagram of a system 800 for diagnostic evaluation according to some embodiments of the present invention.

[0116] System 800 may include platform 810 and web platform 830. Platform 810 may be communicatively coupled to web platform 830 via network 850. As an example, platform 810 may be wiredly connected to web platform 830 via an Ethernet network. As an example, platform 810 may be wirelessly connected to web platform 830 via a wireless network (e.g., a WiFi network or a cellular network). Web platform 830 may be configured to analyze and manage platform 810.

[0117] Platform 810 may be a combination of the above Figure 1 The platform 100 described herein. In some embodiments, the platform 800 may be a mobile terminal platform. As an example, the platform 800 may be a smart phone, a tablet computer, a laptop computer, a dedicated handheld device, etc. In some embodiments, the platform 800 may also be a fixed terminal platform. As an example, the platform 800 may be a desktop computer, a dedicated fixed device, etc.

[0118] In some embodiments, the web platform 830 may include an analysis and presentation function 831. For example, the web platform 830 may present the index analysis of the test simulation to the teacher through a data screen. As an example, the presented content may include at least one of the following: the number of times the platform 810 is used, the time consumed, the pass distribution of each symptom, the overall pass distribution, etc.

[0119] In some embodiments, the web platform 830 may include a medical record management function 831. For example, the web platform 830 may provide a traversal management page to support the management of a standard medical record set.

[0120] In some embodiments, the web platform 830 may include a prompt word input management function 831. For example, the web platform 830 may provide a prompt word input management page to support the management of the patient's personality and diagnosis evaluation criteria.

[0121] In some embodiments, the web platform 830 may include a system settings management function 831. For example, the web platform 830 may provide a system settings management page to support the management of the passing score and the test duration.

[0122] Embodiments of the present invention have been described with reference to the accompanying drawings, the embodiments being illustrative and not restrictive.

Claims

1. A platform for consultation and evaluation, comprising: Computing resources; as well as A large model, wherein the large model is configured to: receiving a prompt word input, wherein the prompt word input includes a medical consultation dialogue between the person being evaluated and the virtual standardized patient and a medical consultation evaluation standard; and A diagnostic question evaluation output is generated based on the prompt word input, wherein the diagnostic question evaluation output includes a diagnostic question evaluation analysis, and the diagnostic question evaluation analysis is associated with the diagnostic question evaluation standard.

2. The platform of claim 1, wherein: The consultation evaluation criteria include medical education evaluation criteria, and the prompt word input is obtained at least in part based on processing the medical education evaluation criteria into prompt word instructions.

3. The platform of claim 1, wherein: The consultation evaluation criteria include a plurality of consultation evaluation criteria, the plurality of consultation evaluation criteria are respectively associated with a plurality of consultation evaluation sub-processes, the consultation evaluation output includes a plurality of consultation evaluation resolutions and a plurality of consultation evaluation scores respectively associated with the plurality of consultation evaluation resolutions, and The large model is configured as follows: Executing the plurality of diagnosis and evaluation sub-processes to extract the plurality of diagnosis and evaluation interpretations and the plurality of diagnosis and evaluation scores; as well as The plurality of diagnostic evaluation analyses and the plurality of diagnostic evaluation scores are combined to obtain the diagnostic evaluation output.

4. The platform of claim 3, wherein: The large model is also configured to: Based on the plurality of consultation evaluation analyses and the plurality of consultation evaluation scores, a deduction analysis and / or a total score is generated.

5. The platform of claim 3, wherein: The large model is intensively trained using multiple question-and-answer evaluation outputs and multiple question-and-answer evaluation parsings and question-and-answer evaluation scores for each question-and-answer evaluation output.

6. The platform of claim 1, wherein: The consultation dialogue is generated based on the patient's medical history and personality.

7. The platform of claim 6, wherein: The prompt word input is a first prompt word input, and The large model is also configured to: receiving a second prompt word input, wherein the second prompt word input includes the patient's medical history and the patient's personality; as well as The medical consultation dialogue is generated based on the second prompt word input and the questions asked by the person being evaluated, wherein the medical consultation dialogue includes the questions asked by the evaluator and the corresponding answers generated by the large model.

8. The platform of claim 7, further comprising: A user interface is configured to receive input from the person being assessed and present output of the large model to the person being assessed.

9. The platform of claim 8, wherein: The platform is configured to operate in a training mode, and In the practice mode, the input of the person being evaluated includes symptoms or disease classification and the patient's personality and character setting, so as to generate the second prompt word input.

10. The platform of claim 8, wherein: The platform is configured to operate in an exam mode, and In the exam mode, the large model is configured to: The second prompt word input including the randomly determined symptom or disease type and the patient's personality is received.

11. The platform of claim 6, wherein: The patient's medical history is associated with a symptom or disease, and / or the patient's medical history includes a standard medical history.

12. The platform of claim 6, wherein: The patient's personality Includes one of the following: Normal type, negative type, impatient type, and pressure-applying type.

13. The platform of claim 1, wherein: The platform is a mobile platform.

14. A system for medical consultation and evaluation, comprising: A platform as claimed in any one of claims 1 to 13; as well as A web-side platform is communicatively coupled to the platform, and the web-side platform is configured to analyze and manage the platform.

15. The system of claim 14, wherein: The web platform is configured to perform at least one of the following: Analyze and present the diagnostic assessment output and / or usage of the platform; Providing a page for managing the medical records required by the platform; Providing a page for managing the prompt word input received by the large model of the platform; Provides a page for managing exam parameters.

Images