Question and answer interaction virtual case development method for nurse inquiry training

Abstract

The invention provides a question and answer interaction virtual case development method for nurse inquiry training, which comprises an original case history module for constructing a virtual case text development framework based on a functional health form, a nursing inquiry teaching status and clinical nursing evaluation, and establishing an original case history; the initial response text module is used for determining the development sequence and inquiry elements of each unit, and constructing an initial response text with the guidance of cultivating clinical thinking ability, nursing science professional attainment and humanistic attainment; and the final response text module is used for constructing a final response text through the subject term and the subject term combination rule based on the initial response text. According to the method, all units of the virtual case text development framework are covered, and content comprehensiveness is guaranteed; an initial response text system is designed for cultivating clinical thinking ability, nursing science professional accomplishment and humanistic accomplishment, and educational guidance is achieved; on the basis of establishing a subject term library and designing a subject term combination rule, a response text is perfected, so that the response accuracy of the virtual patient is ensured.

Description

Technical Field

[0001] The present invention relates to the field of medical teaching technology, and more specifically, to a question-and-answer interactive virtual case development method for nursing staff consultation training. Background Art

[0002] Nursing interview, that is, the medical history collection of patients led by nurses, is an important means of nursing assessment and one of the core skills that clinical nurses need to master. Accurate medical history collection can provide 60% to 80% of the information for diagnosing diseases, among which 76% of diseases can be preliminarily diagnosed based on medical history alone. However, inappropriate medical interviews not only threaten professional integrity, but also lead to adverse medical outcomes, increase additional costs for patients and the medical and health system, and are an important cause of medical errors, accidents and lawsuits.

[0003] Interviewing is the key and difficult part of nursing teaching. Good interviewing ability requires a solid theoretical foundation and multiple practical training. Practical training includes clinical patient training. However, due to the typicality, safety and cooperation of real clinical patients, there is currently a lack of situations and opportunities for repeated and autonomous training for nursing staff. Moreover, real patients or trained standardized patients are willing to be interviewed multiple times by trainees, and the subjective differences in the answers given by different trainees are still difficult to eliminate, making it difficult to achieve objectivity and homogeneity in interview training.

[0004] Virtual patients simulated by computer programs play an important role in the consultation training of medical students. Virtual patients refer to a specific type of computer program that simulates real clinical scenarios. Users imitate the role of healthcare providers to obtain medical history, conduct physical examinations, and make diagnostic and treatment decisions. However, existing virtual patients have the following problems: ① It is difficult to achieve real-time human-computer interaction in natural language. Most products still allow users to select appropriate sentences from multiple inquiry sentences for consultation. Then the computer virtual patient gives corresponding answers, gradually guiding users to understand the condition of the virtual patient, thereby training the consultation ability. However, this form is very different from the real clinical consultation situation, and it is difficult to achieve good training results; ② Although there are some consultation training software for clinical medicine and traditional Chinese medicine, the consultation content of clinical medicine and traditional Chinese medicine is different from that of nursing consultation, and it cannot cover the key content of nursing assessment.

[0005] Therefore, in order to be able to fit and respond to the language of nursing consultations in real situations as closely as possible, and the content of the responses must cover the key contents of nursing, there is an urgent need for a question-and-answer interactive virtual case text development method for nursing staff consultation training to guide the development of virtual patient related products for nursing consultations. It can not only simulate real clinical consultation situations and meet the needs of nursing staff's independent consultation training, but also meet the evaluation thinking of clinical nursing staff, reflect the characteristics and key points of nursing consultations, and play a role in training nursing staff's clinical thinking ability, professional quality, and humanistic quality. Summary of the invention

[0006] The purpose of the present invention is to provide a question-answering interactive virtual case development method for nursing staff consultation training, so as to solve the problem that existing virtual patients are difficult to achieve real-time human-computer interaction in natural language and the consultation content cannot cover the key contents of nursing assessment.

[0007] In order to solve the above technical problems, the present invention provides a method for developing a virtual case with question-answer interaction for nurse consultation training, comprising: The original medical record module builds a virtual case text development framework based on functional health patterns, nursing consultation teaching status and clinical nursing evaluation to establish the original medical record; The initial response text module constructs the initial response text with the goal of cultivating clinical thinking ability, nursing professional literacy and humanistic literacy, and clarifies the development sequence and inquiry elements of each unit; A final response text module, based on the initial response text, constructs a final response text through subject words and subject word combination rules; The method for constructing the original medical record by the original medical record module is specifically to construct a virtual case text development framework based on functional health patterns, nursing consultation teaching status and clinical nursing evaluation, and then randomly generate diseases by the disease model, query relevant disease information in the disease database, and then generate basic information of the patient by the random model, and generate the original medical record in combination with the relevant disease information; The method for constructing the initial response text by the initial response text module is specifically to obtain all units and consultation elements of the original medical record, to cultivate the user's clinical thinking ability, nursing professional literacy and humanistic literacy, to extract questions from the standard question database according to a specific consultation order, and the initial response model generates a response statement according to the question, and integrates the question and the response statement to form a one-question-one-answer initial response text; The method for constructing the final response text by the final response text module is specifically as follows: based on the initial response text and the user's inquiry language, the subject word and subject word combination rule matching is performed; if the inquiry element corresponding to the initial response text is matched, the inquiry element model generates a response sentence; if the inquiry element corresponding to the initial response text is not matched, the non-inquiry element model generates a response sentence, and all response sentences are integrated to generate the final response text.

[0008] The original medical record module takes functional health patterns as the theoretical basis, combines the actual needs of current nursing consultation teaching and the latest standards of clinical nursing assessment, and constructs a comprehensive and practical framework for the development of virtual case texts. On this basis, an original medical record close to clinical practice and covering a wide range of disease types is established to lay the foundation for the subsequent construction of response texts; the initial response text module focuses on cultivating the clinical thinking ability, nursing professional literacy and humanistic care ability of nursing staff. By deeply analyzing the key elements in nursing consultation, an initial response text that is both in line with medical logic and rich in humanistic care is constructed, aiming to guide nursing staff to gradually establish and strengthen users' comprehensive analysis ability and communication skills in simulated consultations; based on the initial response text, the final response text module uses natural language processing technology to match the initial text with the subject word library and the designed subject word combination rules, and generates response sentences according to the user's consultation language to ensure the accuracy, fluency and pertinence of the response text, and improve its applicability and practicality in actual consultation scenarios. Finally, a set of virtual cases with question-answer interaction that is both in line with medical standards and highly interactive is formed, providing a training platform close to the real clinical environment for nursing staff.

[0009] In order to test the comprehensive ability of users in the nursing consultation link, a scientific and complete scoring system is needed to evaluate the user's consultation process, which also includes a scoring module. The scoring method of the scoring module is specifically based on the final response text, scoring criteria and scoring direction, and scoring is performed through a scoring model. The scoring is based on the matching degree between the user's consultation and the final response. Each response has a corresponding score, with the current medical history sub-unit accounting for 60% and the remaining units accounting for 40%. If the user asks about the inquiry elements in the case text, there will be an answer to this virtual case, and the corresponding score can be obtained. If the user's inquiry language does not involve the inquiry elements and does not reflect clinical thinking and professional thinking, although the virtual patient will give an answer, this inquiry will not be included in the score.

[0010] The scoring criteria are generated based on the consultation elements and the consultation sequence, and the scoring directions include clinical thinking ability, humanistic literacy, and nursing professional thinking ability. Clinical thinking ability aims to examine the logic and accuracy of the user's judgment of the condition; humanistic literacy focuses on evaluating the empathy, communication skills, and respect and understanding of the patient's emotions shown by the user during the consultation process; and nursing professional thinking ability focuses on testing the user's ability to use professional knowledge to analyze problems and formulate nursing plans. Through this scoring system, the user's nursing consultation level can be measured and improved from all aspects and angles.

[0011] In order to develop a virtual case text development framework that covers the important contents of the assessment of nursing consultation, the virtual case text development framework includes a starting consultation unit, a core medical history unit, a physiological-related health pattern unit, a psychological-related health pattern unit and an ending consultation unit.

[0012] Since accurate medical history collection can provide 60% to 80% of the information for diagnosing the disease, in order to obtain as much information as possible about the patient, the core medical history unit includes a general information sub-unit, a current medical history sub-unit, a past history sub-unit, a family history sub-unit, and a personal history sub-unit; the physiological-related health pattern unit includes a health perception and health management pattern sub-unit, a nutrition and metabolism pattern sub-unit, an excretion pattern sub-unit, an activity and exercise pattern sub-unit, and a sleep and rest pattern sub-unit; the psychosocial-related health pattern unit includes a perception and cognition pattern sub-unit, a stress and coping pattern sub-unit, a self-perception and self-concept pattern sub-unit, and a value and belief pattern sub-unit. Physiologically related health pattern units are important contents for obtaining information about patients in terms of nutrition, excretion, activity, sleep, health behaviors, etc. Among them, health perception and health management pattern describes the patient's perceived health status and health management behaviors, nutrition and metabolism pattern describes the food and fluid intake related to meeting the body's metabolic needs, excretion pattern describes a pattern of excretion function (including intestines, bladder and skin), activity and movement pattern describes the pattern of exercise, energy, endurance, leisure and entertainment for daily activities, and sleep and rest pattern describes a pattern of sleep, rest and relaxation. Since mental health, social adaptability and role relationships jointly affect an individual's mental health and social functioning, it is necessary to understand the patient's information on psychosocial health patterns. Cognitive and perceptual patterns describe a pattern of perception and cognition (such as taste, hearing, vision, smell, pain, learning style, etc.), stress and coping patterns describe a conventional coping pattern and the effectiveness of related strategies for improving stress resistance, self-perception and self-concept patterns describe a pattern of self-concept and self-perception (such as physical comfort, body image, etc.), and values ​​and beliefs patterns describe values ​​and beliefs (including spirituality) and goals used to guide choices and decisions.

[0013] Since the users of the virtual patient are professionals with nursing background, their consultations are orderly and purposeful. In order to fit the situation of nursing consultation and provide guided answers according to the user's questions, and cultivate the user's clinical thinking ability, nursing professional literacy and humanistic literacy, the initial response text module includes a consultation order submodule and an education guidance submodule. The consultation order submodule sorts each unit in the initial response text according to the consultation order that conforms to the clinical thinking of the nurse. The education guidance submodule is used to determine the consultation elements of each unit for the purpose of cultivating the user's clinical thinking ability, nursing professional literacy and humanistic literacy, extract questions from the standard question database according to the consultation elements and the consultation order, and then generate the response statement by the initial response model to generate the initial response text of one question and one answer. The consultation order submodule generates a variety of orderly consultation response sequences by combining the imported consultation order with the algorithm; the education guidance submodule extracts questions about clinical thinking ability, nursing professional literacy and humanistic literacy from the standard database according to the user's consultation language and consultation order through natural language processing technology, and generates the response statement by the initial response model to generate the initial response text. In addition, the education guidance sub-module can flexibly adjust the response strategy based on the user's real-time feedback to ensure the smoothness and effectiveness of the guidance process.

[0014] Since nursing consultation has a clear purpose, nursing consultation is generally orderly and the content of the consultation is relatively concentrated. The consultation sequence submodule determines the consultation sequence as follows: general information subunit → current medical history subunit → other medical history → physiological related health pattern unit → psychosocial related health pattern unit; the other medical history includes the past history subunit, the family history subunit and the personal history subunit. The inquiry of each content is structured and has specific consultation elements.

[0015] In order to improve the virtual case response text and ensure the accuracy of the virtual patient's response, the final response text module includes a subject word submodule and a subject word combination submodule; the subject word and main word combination rule matching in the construction of the final response text is specifically to establish a subject word library, and divide the subject words into general subject words, symptom subject words, supplementary subject words and special subject words according to different inquiry elements; establish subject word combination rules, and generate subject word combination rules according to the lexical structure by the statistical model. Link the user's inquiry language with the response according to different subject word combination rules. By identifying different types of subject words from the subject word library and matching different subject word combination rules, the corresponding response is obtained on the basis of the initial response text, and the initial response text is improved. The subject word library of the subject word submodule is obtained through large-sample user testing. According to the characteristics of different consultation units, the subject words are divided into general subject words, symptom subject words, supplementary subject words and special subject words. General subject words come from daily communication content and are mainly used in consultation units such as the start consultation unit, the patient's general information subunit, and the end consultation unit; symptom subject words are often used to describe symptoms, and mostly come from consultations related to the clinical manifestations of the disease, and are mainly used in the consultation unit of the current medical history; special subject words generally come from various functional health pattern consultations, other medical history (family history subunit, past history subunit, personal history subunit, menstrual and reproductive history) and other consultation units, reflecting the key content of nursing assessment; supplementary subject words are used to assist the core medical history and the consultation of physiological, psychosocial and related functional health patterns, making the consultation content more specific, so as to more accurately match the corresponding responses. When users ask medical questions, there may be many different ways of asking the same question element. The computer not only needs to identify the subject terms, but may also involve combinations of different subject terms. Therefore, a big data model is used to design subject term combination rules. After the computer identifies the relevant subject terms, it matches the correct answer through these different types of subject term combination rules.

[0016] In order to generate natural language text, understand the meaning of language text, and provide instant feedback, the initial response model, the inquiry element model, and the non-inquiry element model are all large language models. Large language models can generate natural language text and handle a variety of natural language tasks, such as text classification, question and answer, dialogue, etc.; they can also process and understand a variety of forms of information, including text and voice, to enhance their generalization capabilities; they can also introduce a "context learning" model training mechanism to trace back the content of the context and provide a more accurate response.

[0017] The implementation of medical consultation by clinical nursing staff is a comprehensive reflection of their own theoretical knowledge, communication skills, professional quality, and humanistic quality. The virtual patient used for nursing staff medical consultation training should be able to guide clinical nursing staff to inquire about the patient's condition through computer-designed responses, gradually clarify the disease diagnosis of the virtual case and the impact of the disease on its physical, psychological, and social health, and give play to the attributes of teaching and training. Therefore, the clinical thinking ability training is used to guide users to clarify the diagnosis of the disease during the user's inquiry about the demographic characteristics and clinical manifestations of the virtual patient; the nursing assessment focus training is used to guide users to clarify the nursing assessment focus and cultivate their nursing professional quality during the user's inquiry about the virtual patient's physiological-related health type and psychosocial-related health type; the humanistic quality training is used to cultivate users during the inquiry about the virtual patient's psychosocial-related health type. In the process of nursing consultation, clinical thinking ability refers to the ability of nursing staff to understand the patient's condition by questioning the patient. This requires that the content of the consultation is comprehensive, focused and can gradually point to a specific disease diagnosis; the focus of nursing professional literacy is that clinical nursing staff can evaluate patients from a nursing perspective, reflecting the key content of nursing evaluation (such as health management behaviors, nutrition, urination, defecation, daily living activities, physical activity, sleep, emotional responses, social relationships and social support, etc.), which includes the impact of the disease on the patient's physical, psychosocial health; humanistic literacy should be reflected in respect, understanding and empathy for patients during the nursing consultation process.

[0018] Compared with the prior art, the present invention is beneficial in that: 1. Comprehensiveness of content Based on the functional health pattern, a virtual case text development framework is constructed, and the original medical records of the virtual patient are designed. The original medical records cover all the units in the development framework. The construction of the development framework is to ensure the comprehensiveness of the content of the virtual patient's response, that is, the virtual patient text should be able to answer all the questions that the clinical nurses may ask during the consultation. The case text covers the comprehensive content of the clinical nurses' consultation, and is applicable to patients with different specialties and different diseases; 2. Educational guidance Design the initial response text of virtual cases with the goal of cultivating clinical thinking ability, nursing professional literacy and humanistic literacy, and clarify the development sequence, inquiry elements and training objectives of each inquiry unit in the initial response text; 3. Response accuracy On the basis of establishing a thesaurus and designing thesaurus combination rules, the virtual case response text is improved to ensure the accuracy of the virtual patient's response. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a structural diagram of a method for developing a virtual case with question-and-answer interaction for nursing staff consultation training according to the present invention.

[0020] Figure 2 This is an original medical record diagram of a virtual case development method for question-and-answer interaction for nursing staff consultation training according to the present invention.

[0021] Figure 3 It is a unit of the development framework of the present invention and a corresponding consultation element diagram.

[0022] Figure 4 and 5 For the purpose of initial response text training of the present invention and corresponding unit diagram.

[0023] Figure 6 This is a keyword type diagram of the present invention.

[0024] Figure 7 This is the subject word combination rule diagram of the present invention.

[0025] Figure 8 This is a process diagram of realizing human-computer interactive consultation based on the virtual case text of the nursing consultation of the present invention.

[0026] Fig. 9 This is a scoring standard diagram of the present invention.

[0027] Fig.10 This is a structural diagram of a question-and-answer interactive virtual case development system for nursing staff consultation training according to the present invention.

[0028] Fig.11 This is a structural diagram of the original medical record module of the present invention.

[0029] Fig.12 It is a structural diagram of the initial response text module of the present invention.

[0030] Fig.13 It is a structural diagram of the final response text module of the present invention.

[0031] Fig.14 It is a structural diagram of the scoring module of the present invention.

[0032] Fig.15 This is the home page image of the nursing consultation virtual patient software applet of the present invention.

[0033] Fig.16 It is the consultation and answering interface of the nursing consultation virtual patient software applet of the present invention. DETAILED DESCRIPTION

[0034] The method and system for selecting the site for connecting the drainage network proposed by the present invention are further described in detail below in combination with the accompanying drawings and specific embodiments. It should be noted that the accompanying drawings are all in a very simplified form and use non-precise proportions, which are only used to conveniently and clearly assist in explaining the purpose of the embodiments of the present invention. In addition, the structure shown in the accompanying drawings is often a part of the actual structure. In particular, the emphasis of each accompanying drawing is different, and sometimes different proportions are used.

[0035] It should be noted that the terms "first", "second", etc. in the specification, claims and drawings of the present invention are used to distinguish similar objects in order to describe the embodiments of the present invention, rather than to describe a specific order or sequence. It should be understood that the structures used in this way can be interchanged under appropriate circumstances. In addition, the terms "include" and "have" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0036] Example 1 like Figure 1 As shown, this embodiment provides a question-answer interactive virtual case development method for nursing staff consultation training, comprising: S100: Constructing original medical records, selecting specific diseases, and designing original medical records based on the nursing consultation virtual case development framework; S200: construct an initial response text, determine the training purpose and development order of each consultation unit, determine the consultation elements of each consultation unit, and design an initial response text for one question and one answer; S300: Construct the final response text, obtain as much inquiry language as possible through the user's initialization, match the keywords in the inquiry language with the corresponding subject words, and design the subject word combination rules of the inquiry elements to match the correct response; S400: Scoring: Scoring is performed based on the final response text, scoring criteria and scoring direction.

[0037] The specific method of constructing the original medical record is to build a virtual case text development framework based on functional health patterns, the current status of nursing consultation teaching and clinical nursing evaluation, and then randomly generate diseases based on common diseases in the department by the disease model, query relevant disease information in the disease database, and then generate the patient's basic information by the random model, and generate the original medical record in combination with the relevant disease information; the specific method of constructing the initial response text is to obtain all units and consultation elements of the original medical record, with the purpose of cultivating the user's clinical thinking ability, nursing professional literacy and humanistic literacy, and extract questions from the standard question database in accordance with the consultation order that conforms to the nurse's clinical thinking. The initial response model generates a response statement based on the question, and integrates the question and the response statement to form an initial response text with one question and one answer; the method for constructing the final response text is specifically to match the subject words and subject word combination rules based on the initial response text and the user's inquiry language. If the inquiry element corresponding to the initial response text is matched, the inquiry element model generates an answer statement; if the inquiry element corresponding to the initial response text is not matched, the non-inquiry element model generates an answer statement, and all the answer statements are integrated to generate the final response text; the scoring method is specifically to score based on the final response text, scoring criteria and scoring direction through the scoring model. The scoring is based on the matching degree between the user's inquiry and the final response. Each response has a corresponding score, with the current medical history subunit accounting for 60% and the remaining units accounting for 40%. If the user asks about the inquiry elements in the case text, there will be an answer to this virtual case and the corresponding score can be obtained. If the user's inquiry language does not involve the inquiry elements and does not reflect clinical thinking and professional thinking, although the virtual patient will give an answer, this inquiry will not be counted in the score.

[0038] In this embodiment, if Figure 2 As shown, the disease selected in the original medical record is acute pancreatitis, and the information of acute pancreatitis is searched from the disease database. The basic information of the patient is generated through a random model, and the general information subunit of the virtual patient is generated by combining the searched pancreatitis information, and the information of the current medical history subunit, past medical history subunit, family history subunit, personal history subunit, health perception and health management pattern subunit, nutrition and metabolism pattern subunit, excretion pattern subunit, activity and exercise pattern subunit, sleep and rest pattern subunit, perception and cognition pattern subunit, sex and reproduction pattern subunit, role and relationship pattern subunit, stress and coping pattern subunit, self-perception and self-concept pattern subunit and value and belief pattern subunit are generated to generate the original medical record.

[0039] In this embodiment, if Figure 3As shown in the figure, since the medical record text must cover all the contents of the clinical nursing staff's consultation and be applicable to patients with different specialties and different diseases, based on the analysis of the functional health pattern theory, the current teaching status of nursing consultation and clinical nursing evaluation were integrated to develop a framework for the virtual medical record text of nursing consultation. The original medical record of each virtual patient contains five units, namely, the start-up consultation unit, the core medical history unit, the physiological-related health pattern unit, the psychosocial-related health pattern unit and the end-of-consultation unit. The start-up consultation unit includes the consultation initiation sub-unit, the core medical history unit includes the general information sub-unit, the current medical history sub-unit, the past history sub-unit, the family history sub-unit and the personal history sub-unit, the physiological-related health pattern unit includes the health perception and health management pattern sub-unit, the nutrition and metabolism pattern sub-unit, the excretion pattern sub-unit, the activity and exercise pattern sub-unit, and the sleep and rest pattern sub-unit, and the psychosocial-related health pattern unit includes the perception and cognition pattern sub-unit, the role and relationship pattern sub-unit, the sex and reproduction pattern sub-unit, the stress and coping sub-unit, the self-perception and self-concept pattern sub-unit, and the value and belief pattern sub-unit. In addition, the Functional Health Pattern Theory is a nursing assessment theoretical framework with nursing characteristics. It discusses the key contents of nursing assessment from the physiological, psychological, social, and spiritual aspects, and guides nursing staff to conduct nursing assessment from a holistic perspective, making the content of the virtual case text development framework more comprehensive and complete.

[0040] In this embodiment, combined with Figure 4 and Figure 5 As shown, since the users that the virtual patient faces are professionals with a nursing background, their consultations are orderly and purposeful. In order to fit the actual nursing consultation situation and provide guided answers based on the users' questions, and to cultivate the users' clinical thinking ability, nursing professional qualities and humanistic qualities, an initial response text of one question and one answer is constructed. The construction of the initial response text also includes S210 determining the order of consultations and S220 providing educational guidance.

[0041] S210. Determine the specific sequence of medical consultations: Based on the imported medical consultation sequence that conforms to the clinical thinking of nurses, combined with algorithm technology, generate a variety of orderly medical consultation response sequences to match the user's medical consultations, laying a solid foundation for subsequent text construction; the order of medical consultations is general information sub-unit → current medical history sub-unit → other medical histories (past medical history sub-unit, family history sub-unit, personal history sub-unit, menstrual and reproductive history sub-unit, and there is no specific order for the consultation of these contents) → physiological-related health pattern unit (health perception and health management pattern sub-unit, nutrition and metabolism pattern sub-unit, excretion pattern sub-unit, activity and exercise pattern sub-unit, sleep and rest pattern sub-unit, and there is no specific order for the consultation of these contents) → psychosocial-related health pattern unit (perception and cognition pattern sub-unit, role and relationship pattern sub-unit, sex and reproduction pattern sub-unit, stress and coping pattern sub-unit, self-perception and self-concept pattern sub-unit, value and belief pattern sub-unit). Accurate medical history collection can provide 60% to 80% of the information for diagnosing diseases, and 76% of diseases can be preliminarily diagnosed based on medical history alone. Therefore, in order to obtain as much information as possible about the patient, it is necessary to ask about the core medical history immediately after asking about the patient's basic information to prevent the patient from being unable to receive assistance in an emergency; physiological-related health patterns cover the patient's information on nutrition, excretion, activity, sleep, health behaviors, etc., which can assist in the diagnosis of diseases. Among them, health perception and health management patterns describe the patient's perceived health status and health management behaviors, nutrition and metabolic patterns describe food and fluid intake related to meeting the body's metabolic needs, excretion patterns describe a pattern of excretion function (including intestines, bladder and skin), activity and movement patterns describe patterns of exercise, energy, endurance, leisure and entertainment for daily activities, and sleep and rest patterns describe a pattern of sleep, rest and relaxation. Since mental health, social adaptability and role relationships jointly affect an individual's mental health and social functioning, mastering information related to the patient's psychosocial health patterns is helpful to grasp the patient's current and future situation. Among them, cognitive and perceptual patterns describe a pattern of perception and cognition (such as taste, hearing, vision, smell, pain, learning style, etc.), stress and coping patterns describe a conventional coping pattern and the effectiveness of related strategies used to improve stress resistance, self-perception and self-concept patterns describe a pattern of self-concept and self-perception (such as physical comfort, body image, etc.), and values ​​and beliefs patterns describe values ​​and beliefs (including spirituality) and goals used to guide choices and decisions.

[0042] S220, conduct education guidance specifically, determine the elements of each unit for the purpose of cultivating the user's clinical thinking ability, nursing professional literacy and humanistic literacy, extract questions from the standard question database according to the elements and the order of the questions, and then generate the response sentence by the initial response model to generate the initial response text of one question and one answer. The purpose of guidance and training is achieved through the following methods: ① Through the questioning of the demographic characteristics and clinical manifestations of the virtual patient, guide the user to gradually clarify the diagnosis of the disease, that is, cultivate clinical thinking ability; demographic characteristics include the patient's general information, such as name, age, occupation and place of origin, and clinical manifestations include the patient's core medical history, such as current medical history, past history and family history; ② Through the questioning of the patient's physiological health type and psychosocial health type, guide the user to gradually clarify the focus of nursing assessment; physiological health type includes hobbies, allergy history and physical examination status; ③ Through the links of starting the questioning, ending the questioning and evaluating the psychosocial health type, train the user's humanistic literacy. Among them, clinical thinking ability training is used to guide users to clarify the diagnosis of the disease during the user's inquiry about the demographic characteristics and clinical manifestations of the virtual patient; nursing assessment focus training is used to guide users to clarify the nursing assessment focus during the user's inquiry about the virtual patient's physiological and psychosocial health patterns; humanistic literacy training is used to cultivate during the user's inquiry about the virtual patient's psychosocial health patterns. In the nursing consultation process, clinical thinking ability refers to the ability of nursing staff to understand the patient's condition through questioning the patient, which requires the content of the inquiry to be comprehensive, focused, and gradually point to a specific disease diagnosis; the focus of nursing professional literacy refers to the ability of clinical nursing staff to evaluate patients from the perspective of nursing, reflecting the key content of nursing assessment (such as health management behavior, nutrition, urination, defecation, daily living activities, physical activity, sleep, emotional response, social relations and social support, etc.), which includes the impact of the disease on the patient's physiological, psychosocial health; humanistic literacy should be reflected in the respect, understanding and empathy for patients during the nursing consultation process. The implementation of clinical nursing staff's consultation is a comprehensive reflection of their own theoretical knowledge, communication skills, professional literacy, and humanistic literacy. The virtual patient designed by computer can answer questions for nurses' consultation training, which can guide clinical nurses to ask questions about the patient's condition, gradually clarify the disease diagnosis of the virtual case and the impact of the disease on their physical, psychological and social health, and give full play to the attributes of teaching and training. In addition, the educational guidance submodule makes guiding responses based on the user's cognitive level and needs, combined with natural language processing technology, and can flexibly adjust the response strategy based on the user's real-time feedback to ensure the smoothness and effectiveness of the guidance process.

[0043] In order to improve the virtual case response text and ensure the accuracy of the virtual patient's response, the final response text is constructed, including S310 establishing a subject word library, S320 establishing subject word combination rules and S330 matching model: In this embodiment, if Figure 6 As shown, S310 establishes a subject word library, specifically, multiple subject words are obtained through large sample user testing, and subject words are divided into general subject words, symptom subject words, supplementary subject words and special subject words according to the characteristics of different consultation units. General subject words are used for daily communication content, and are often used in consultation units such as the start consultation unit, the patient's general information subunit, and the end consultation unit; symptom subject words are used to describe the symptoms of the disease, mostly from consultations related to the clinical manifestations of the disease, and are mainly used in the consultation unit of the current medical history subunit; special subject words are collected from various functional health type consultations, other medical history (family history, past history, personal history, menstrual and reproductive history) and other consultation units, and are used to reflect the key content of nursing assessment; supplementary subject words are used to assist the core medical history and the consultation of physiological, psychological and social related functional health types, so that the consultation content is more specific, so as to more accurately match the corresponding response. Since the consultation elements in different units involve different forms of consultation language, the computer can decompose the user's language into different keywords after recognizing it, but there will be a situation where there are multiple keywords for a certain same consultation element. For example, the query keywords for "age" in the medical inquiry element include "age, age, seniority, longevity", etc., and the subject word is to plan these different keywords into a unified search term "age". By planning multiple different keywords under a unified subject word, the system can well identify the subject words that match the keywords in the user's language, and thus make correct responses.

[0044] In this embodiment, combined with Figure 7 As shown, S320 establishes the subject word combination rules specifically. The subject word combination rules are based on the subject word type. A variety of subject word combination rules are designed through algorithms. Because when a user asks a question, there will be a variety of different ways of asking questions for the same question element. The system must not only identify the subject words in the question language, but also identify the type of subject words and the combination of different subject words, identify the corresponding question elements from the combination of these subject words, and match them to the correct answer; the keywords in the user's question language will be identified and matched to the corresponding subject words, and then matched to the corresponding answer through the combination rules based on the subject word type.

[0045] In this embodiment, combined with Figure 8As shown, the S330 matching model is specifically to match the input user's inquiry language and the initial response text with the subject words and main words combination rules. If the inquiry element corresponding to the initial response text is matched, the inquiry element model generates the response sentence; if the inquiry element corresponding to the initial response text is not matched, the non-inquiry element model generates the response sentence, and all the response sentences are integrated to generate the final response text. Since the initial response text has determined the training purpose, inquiry elements, initial inquiry language and response content of each inquiry unit, on this basis, for the inquiry elements of each inquiry unit, by supplementing the corresponding subject words and subject word combination rules, the keywords in the user's inquiry language can be identified and matched to the corresponding subject words in the subject word library, and then according to the different subject words and their combination rules in the user's inquiry language, the inquiry elements are identified and matched to the corresponding virtual patient response content. In addition, both the medical inquiry factor model and the non-medical inquiry factor model are big data models. The large language model can generate natural language text and handle a variety of natural language tasks, such as text classification, question and answer, and dialogue. It can also process and understand various forms of information, including text and voice, to enhance its generalization ability. It can also introduce the "context learning" model training mechanism to trace back the content of the context and provide more accurate responses.

[0046] In this embodiment, combined with Fig. 9 As shown in the figure, a set of scoring standards is formulated based on the elements and sequence of consultation, which is used to test the comprehensive ability of users in the nursing consultation process. The scoring directions include clinical thinking ability, humanistic literacy and nursing professional thinking ability. Clinical thinking ability aims to examine the logic and accuracy of users' judgment of the condition; humanistic literacy focuses on evaluating the empathy, communication skills and respect and understanding of patients' emotions shown by users during the consultation process; and nursing professional thinking ability focuses on testing the ability of users to use professional knowledge to analyze problems and formulate nursing plans. Through this scoring system, the user's nursing consultation level can be measured and improved from all aspects and angles. Based on the consultation elements and the sequence of consultation involved in the user's consultation process, the initial score is generated, and after calculation, the final score is generated, with a total score of 1000 points. The calculation process is: user initial score × 10% = user final score.

[0047] Example 2 like Fig.10 As shown, the present invention designs a question-answering interactive virtual medical record development system for nursing consultation training based on the above-mentioned question-answering interactive virtual medical record development method for nursing consultation training, including: The original medical record module builds the original medical record, selects a specific disease, and designs the original medical record according to the nursing consultation virtual case development framework; Initial response text module: construct the initial response text, determine the training objectives and development sequence of each consultation unit, determine the consultation elements of each consultation unit, and design the initial response text of one question and one answer; The final response text module constructs the final response text, obtains as much inquiry language as possible through the user's initialization, matches the keywords in the inquiry language with the corresponding subject words, and designs the subject word combination rules of the inquiry elements to match the correct response; The scoring module performs scoring based on the final response text, scoring criteria and scoring direction.

[0048] In this embodiment, if Fig.11 As shown in the figure, the original medical record module includes a text framework submodule and a disease generation submodule. The text framework submodule builds a virtual case text development framework based on functional health patterns, nursing consultation teaching status and clinical nursing evaluation. The virtual case text development framework includes a start-up consultation unit, a core medical history unit, a physiological-related health pattern unit, a psychological-related health pattern unit and an end-of-consultation unit. The core medical history unit includes a general information subunit, a current medical history subunit, a past medical history subunit, a family history subunit and a personal history subunit; the physiological-related health pattern unit includes a health perception and health management pattern subunit, a nutrition and metabolism pattern subunit, an excretion pattern subunit, an activity and exercise pattern subunit and a sleep and rest pattern subunit; the psychosocial-related health pattern unit includes a perception and cognition pattern subunit, a stress and coping pattern subunit, a self-perception and self-concept pattern subunit and a value and belief pattern subunit.

[0049] In the disease generation submodule, the disease model randomly generates diseases based on the common diseases in the department, queries the disease database for relevant disease information, and then generates the patient's basic information by the random model, and generates the original medical record in combination with the relevant disease information.

[0050] In this embodiment, if Fig.12As shown, the initial response text module includes a consultation sequence submodule and an education guidance submodule; the consultation sequence submodule generates a variety of ordered consultation response sequences based on the imported consultation sequence that conforms to the nurse's clinical thinking and combines algorithm technology to match the user's consultation and lay a solid foundation for subsequent text construction; the consultation sequence is general information subunit → current medical history subunit → other medical history (past history subunit, family history subunit, personal history subunit, menstrual and reproductive history subunit, and there is no specific order for the consultation of these contents) → physiological related health pattern unit (health perception and health management pattern subunit, nutrition and metabolism pattern subunit, excretion pattern subunit, activity and exercise pattern subunit, sleep and rest pattern, and there is no specific order for the consultation of these contents) → psychosocial related health pattern unit (perception and cognition pattern subunit, role and relationship pattern subunit, sex and reproduction pattern subunit, stress and coping subunit, self-perception and self-concept pattern subunit, value and belief pattern subunit).

[0051] Educational guidance submodule: The purpose of this module is to cultivate the user's clinical thinking ability, nursing professional quality and humanistic quality. The consultation elements of each unit are determined. According to the training direction, consultation elements and consultation sequence, questions are extracted from the standard question database. The initial response model generates response statements based on the questions, and the questions and response statements are integrated to form an initial response text of one question and one answer.

[0052] In this embodiment, if Fig.13 As shown in the figure, the final response text module includes a subject word submodule, a subject word combination submodule and a matching submodule; the subject word submodule obtains multiple subject words through large-sample user testing, and divides the subject words into general subject words, symptom subject words, supplementary subject words and special subject words according to the characteristics of different consultation units. General subject words are used for daily communication content, and are often used in consultation units such as the start consultation unit, the patient's general information subunit, and the end consultation unit; symptom subject words are used to describe the symptoms of the disease, and are mostly from consultations related to the clinical manifestations of the disease, and are mainly used in the consultation unit of the current medical history; special subject words are collected from various functional health pattern consultations, other medical history (family history, past history, personal history, menstrual and reproductive history) and other consultation units to reflect the key content of nursing assessment; supplementary subject words are used to assist the core medical history and the consultation of physiological, psychosocial and related functional health patterns, making the consultation content more specific, so as to more accurately match the corresponding response.

[0053] The subject word combination submodule designs a variety of subject word combination rules through algorithms; The matching submodule matches the input user inquiry language and the initial response text according to the combination rules of subject words and main words. If the inquiry elements corresponding to the initial response text are matched, the inquiry element model generates the response sentence; if the inquiry elements corresponding to the initial response text are not matched, the non-inquiry element model generates the response sentence, and all the response sentences are integrated to generate the final response text.

[0054] In this embodiment, if Fig.14 As shown, the scoring module includes a score generation submodule and a calculation submodule; the score generation submodule generates an initial score from the three directions of clinical thinking ability, humanistic literacy and nursing professional thinking ability, according to the inquiry elements and inquiry sequence of the final response text, by the scoring model; In the calculation submodule, the initial score is calculated through the formula: user initial score × 10% = user final score to generate the final score.

[0055] Example 3 like Fig.15 and 16 As shown, the present invention designs a nursing consultation virtual patient applet based on the above-mentioned question-and-answer interactive virtual medical record development method. The applet includes a home page, message and my interface. The home page contains home page, message and my icons below. Click the icon to jump to the corresponding interface. The content of the home page includes common symptom consultation content and common case consultation content. The common symptom consultation content includes symptoms of various diseases, such as cough, difficulty breathing, chest pain and abdominal pain, etc. The common case consultation content includes internal medicine, pediatrics, obstetrics and gynecology, and surgery. Different users can choose different cases for consultation.

[0056] The message interface of the nursing consultation virtual patient includes the interaction process between the virtual patient and the user. The content of the interaction is the content of the nursing consultation. The user diagnoses the virtual patient through the consultation and learns about the virtual patient's disease and condition. The mini program matches the subject words and subject word combination rules through the user's consultation language, identifies the consultation elements, matches the accurate answers according to the consultation elements, and scores the user's consultation process according to the consultation elements and the consultation order. The final scoring results are presented in my interface.

[0057] Obviously, the above embodiments of the present invention are merely examples for clearly illustrating the technical solution of the present invention, and are not intended to limit the specific implementation methods of the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the claims of the present invention shall be included in the protection scope of the claims of the present invention.

Claims

1. A question-answer interactive virtual case development method for nursing staff consultation training, characterized in that: include: The original medical record module builds a virtual case text development framework based on functional health patterns, nursing consultation teaching status and clinical nursing evaluation to establish the original medical record; The initial response text module clarifies the development order and inquiry elements of each unit, and constructs the initial response text with the goal of cultivating clinical thinking ability, nursing professional literacy and humanistic literacy; A final response text module, based on the initial response text, constructs a final response text through subject words and subject word combination rules; The method for constructing the original medical record by the original medical record module is specifically to construct a virtual case text development framework based on functional health patterns, nursing consultation teaching status and clinical nursing evaluation, and then randomly generate diseases by the disease model, query relevant disease information in the disease database, and then generate basic information of the patient by the random model, and generate the original medical record in combination with the relevant disease information; The method for constructing the initial response text by the initial response text module is specifically to obtain all units and consultation elements of the original medical record, to cultivate the user's clinical thinking ability, nursing professional literacy and humanistic literacy, to extract questions from the standard question database according to a specific consultation order, and the initial response model generates a response statement according to the question, and integrates the question and the response statement to form a one-question-one-answer initial response text; The method for constructing the final response text by the final response text module is specifically as follows: based on the initial response text and the user's inquiry language, the subject words and subject word combination rules are matched. If the inquiry elements corresponding to the initial response text are matched, the inquiry element model generates a response sentence; if the inquiry elements corresponding to the initial response text are not matched, the non-inquiry element model generates a response sentence, and all response sentences are integrated to generate the final response text.

2. A method for developing a virtual case for question-answer interaction for nurse consultation training according to claim 1, characterized in that: It also includes a scoring module, and the scoring method of the scoring module is specifically to score based on the final response text, scoring criteria and scoring direction through a scoring model.

3. A method for developing a virtual case with question-answer interaction for nurse consultation training according to claim 2, characterized in that: The scoring criteria are generated based on the elements of the medical consultation and the order of medical consultation, and the scoring directions include clinical thinking ability, humanistic literacy and nursing professional thinking ability.

4. A method for developing a virtual case with question-answer interaction for nurse consultation training according to claim 3, characterized in that: The virtual case text development framework includes a start-up consultation unit, a core medical history unit, a physiological-related health pattern unit, a psychological-related health pattern unit, and an end-of-consultation unit.

5. A method for developing a virtual case with question-answer interaction for nurse consultation training according to claim 4, characterized in that: The core medical history part includes a general information subunit, a current medical history subunit, a past history subunit, a family history subunit and a personal history subunit; the physiological-related health pattern unit includes a health perception and health management pattern subunit, a nutrition and metabolism pattern subunit, an excretion pattern subunit, an activity and exercise pattern subunit and a sleep and rest pattern subunit; the psychosocial-related health pattern unit includes a perception and cognition pattern subunit, a stress and coping pattern subunit, a self-perception and self-concept pattern subunit and a value and belief pattern subunit.

6. A method for developing a virtual case with question-answer interaction for nurse consultation training according to claim 5, characterized in that: The initial response text module includes a consultation sequence submodule and an education guidance submodule. The consultation sequence submodule is used to sort the units in the initial response text according to a specific consultation sequence. The education guidance submodule is used to determine the consultation elements of each unit with the purpose of cultivating the user's clinical thinking ability, nursing professional quality and humanistic quality, extract questions from the standard question database according to the consultation elements and consultation sequence, and then generate a response statement by the initial response model to generate an initial response text with one question and one answer.

7. A method for developing a virtual case with question-answer interaction for nurse consultation training according to claim 6, characterized in that: The order of medical interviews is: general information subunit → current medical history subunit → other medical histories → physiological-related health pattern unit → psychosocial-related health pattern unit; the other medical history includes the past history subunit, the family history subunit and the personal history subunit.

8. A method for developing a virtual case with question-answer interaction for nurse consultation training according to any one of claims 1 to 7, characterized in that: The final response text module includes a subject word submodule and a subject word combination submodule; The specific matching of the subject word and main word combination rule in constructing the final response text is: Establish a subject word library and divide the subject words into general subject words, symptom subject words, supplementary subject words and special subject words; Establish topic word combination rules, and generate topic word combination rules based on lexical structure by statistical model.

9. A virtual case development system for question-answer interaction for nursing staff consultation training according to any one of claims 1 to 8, characterized in that: The initial response model, the medical inquiry element model and the non-medical inquiry element model are all large language models.

10. A method for developing a virtual case with question-answer interaction for nurse consultation training according to any one of claims 1 to 8, characterized in that: The clinical thinking ability training is used to guide users to clarify the diagnosis of the disease during the process of users asking about the demographic characteristics and clinical manifestations of virtual patients; the nursing assessment focus training is used to guide users to clarify the nursing assessment focus and cultivate their nursing professional qualities during the process of users asking about the virtual patients' physiological-related health types and psychosocial-related health types; the humanistic literacy training is used to cultivate users during the questioning of virtual patients' psychosocial-related health types.

Images