Intelligent test paper composition method, medical examination system, equipment and medium

Abstract

The invention relates to the technical field of intelligent examination systems, in particular to an intelligent test paper composition method, a medical examination system, equipment and a medium. The method comprises the following steps: firstly, receiving department examination demand parameters, retrieving a test question set meeting conditions from a clinical case test question bank to form a candidate test question pool, and then intelligently screening candidate test questions by utilizing a preset clinical diagnosis and treatment dynamic model to obtain a preliminary test question combination; analyzing high-frequency clinical examination point features in hospital historical examination data, and performing optimization adjustment in combination with knowledge point distribution in the preliminary test question combination; and by evaluating the clinical difficulty balance degree and the diagnosis and treatment knowledge point coverage rate in real time, the test paper composition parameters are dynamically adjusted until preset standards are met. By introducing the clinical diagnosis and treatment dynamic model and historical examination data analysis, intelligent construction of clinical relevance between test questions and dynamic optimization of knowledge point distribution are realized, so that the generated test paper more accurately evaluates the actual diagnosis and treatment ability of medical personnel.

Description

Technical Field

[0001] This application relates to the technical field of intelligent examination systems, and in particular to intelligent test paper generation methods, medical examination systems, equipment, and media. Background Technology

[0002] With the rapid development of the medical industry and the continuous deepening of medical education, hospital examinations, as an important means of assessing the professional competence of medical personnel, have placed higher demands on the quality and effectiveness of these examinations. Especially in clinical practice, medical personnel need comprehensive abilities to quickly diagnose, accurately treat, and respond to emergencies, which presents new challenges to the design and organization of examination content.

[0003] Currently, mainstream intelligent test paper generation systems are mainly based on knowledge graphs and deep learning technologies. By establishing test question feature models and knowledge point association networks, they can automatically select and combine test questions. These systems can automatically construct test papers according to preset rules and ensure the completeness of knowledge point coverage.

[0004] However, the existing intelligent test paper generation system is too mechanical in its question selection and lacks consideration for the characteristics of clinical practice, thus failing to truly reflect the diagnostic and treatment capabilities of medical personnel; this situation needs further improvement. Summary of the Invention

[0005] To address the problems of existing intelligent test paper generation systems, such as overly mechanical question selection, lack of consideration for clinical practice characteristics, and inability to accurately reflect the diagnostic and treatment capabilities of medical personnel, this application provides an intelligent test paper generation method, medical examination system, equipment, and media, employing the following technical solution: Firstly, this application provides an intelligent test paper generation method, comprising the following steps: Receive hospital department assessment requirements parameters input by the user; Based on the assessment requirements parameters, a set of questions that meet the conditions is retrieved from the hospital's clinical case question bank to obtain a candidate question pool; Based on a preset dynamic model of clinical diagnosis and treatment, the candidate question pool is intelligently screened to obtain a preliminary question combination; The characteristics of high-frequency clinical test points are obtained from the hospital's historical examination data. Based on the distribution of clinical knowledge points in the preliminary test question combination and the characteristics of high-frequency clinical test points, an optimized test question combination is obtained. Based on the optimized question combination, the clinical difficulty balance and coverage of diagnostic and treatment knowledge points of the test paper are evaluated in real time, and the adjustment plan of the test paper parameters is determined until the preset standard is met.

[0006] By adopting the above technical solution, this application first receives the department's assessment requirements parameters, retrieves a set of qualified questions from the clinical case question bank to form a candidate question pool, and then uses a preset clinical diagnosis and treatment dynamic model to intelligently screen the candidate questions to obtain a preliminary question combination. Next, it analyzes the characteristics of high-frequency clinical test points in the hospital's historical examination data and optimizes and adjusts them in combination with the distribution of knowledge points in the preliminary question combination. By evaluating the balance of clinical difficulty and the coverage of diagnosis and treatment knowledge points in real time, it dynamically adjusts the test paper parameters until the preset standards are met. By introducing the clinical diagnosis and treatment dynamic model and historical examination data analysis, it realizes the intelligent construction of clinical correlation between questions and the dynamic optimization of knowledge point distribution, so that the generated test paper can more accurately assess the actual diagnosis and treatment ability of medical staff.

[0007] Optionally, high-frequency clinical test point characteristics are obtained from the hospital's historical examination data. Based on the distribution of clinical knowledge points in the preliminary test question combination and the high-frequency clinical test point characteristics, an optimized test question combination is obtained, specifically including the following steps: By analyzing the distribution of clinical diagnosis and treatment test points in historical exam data using a neural network model, we can identify test points for acute and critical illnesses, common diseases, and frequently occurring diseases, as well as their corresponding trends, and thus obtain the priority of clinical test points. Based on the priority of the clinical test points, the weight of the diagnostic and treatment knowledge points in the preliminary test question combination is adjusted to ensure that the proportion of test points related to acute and critical illnesses in the test paper is not lower than the preset threshold, thus obtaining the optimized test question combination.

[0008] By adopting the above technical solution, the rationality of the distribution of test points in the hospital examination system directly affects the effectiveness of the assessment. Existing technologies usually use preset weights to determine the distribution ratio of various test points. However, due to the constantly changing characteristics and development trends of the diagnosis and treatment of acute and critical illnesses, common diseases, and frequently occurring diseases in medical practice, fixed test point weight settings are difficult to adapt to such dynamic changes. At the same time, they cannot reflect the differences in the importance of different types of diseases in clinical work, resulting in a disconnect between the distribution of test points in the test paper and actual clinical needs. This application first uses a neural network model to deeply mine historical examination data, extracting features such as the frequency of occurrence, importance, and changing trends of test points for various diseases, and establishes a clinical test point priority system. Then, based on the obtained priority data, the weight allocation of various diagnostic and treatment knowledge points in the preliminary test question combination is dynamically adjusted, and a preset threshold is set to ensure that the proportion of test points related to acute and critical illnesses meets the needs of clinical practice. This achieves data-driven optimization of test point distribution, making the test paper structure more in line with clinical reality and the assessment effect more targeted.

[0009] Optionally, after obtaining the optimized question combination, the method further includes the following steps: Based on the clinical characteristics of different departments, a random perturbation factor was used to generate differentiated test paper versions, resulting in multiple versions of the test paper; The multiple versions of the exam paper were configured with intra-college network IP restrictions, random question order, and A / B version modes to obtain the final exam paper.

[0010] By adopting the above technical solution, this application first designs random perturbation factors based on the clinical characteristics of different departments, differentiates the optimized question combinations, and generates multiple versions of the test paper to ensure that the same test point is presented in different forms in different versions. Then, by setting intra-hospital network IP restrictions, it ensures that candidates can only take the exam in designated locations, and by combining random question order and A / B test modes, it constructs multiple anti-cheating barriers. Through the differentiated test paper generation driven by departmental characteristics and multi-dimensional examination environment control, the equivalence of the assessment content is guaranteed while effectively reducing the risk of cheating.

[0011] Optionally, the method may also include the following steps: The system acquires real-time heart rate, blood pressure, and facial expression data of test takers to obtain real-time physiological parameter information. Based on a preset stress index assessment model, the presentation speed, difficulty gradient, and number of concurrent tasks of clinical cases are dynamically adjusted according to the real-time physiological parameter information to simulate high-pressure medical scenarios. Based on the changing trends of the real-time physiological parameters and the examinee's performance, a personalized exam rhythm adjustment plan is generated, resulting in the final dynamic exam paper.

[0012] By adopting the above-mentioned technical solutions, in order to simulate the high-pressure environment of medical work, existing technologies mainly create exam pressure by setting fixed answering time limits and progressively increasing difficulty. However, since emergency rescue and other scenarios in medical practice are often accompanied by high tension and multiple concurrent tasks, and candidates' physiological reactions and adaptability to pressure vary, a fixed exam rhythm cannot realistically reproduce the clinical work state, nor can it accurately assess the candidates' actual performance under pressure. This results in a significant deviation between the assessment results and clinical practice ability. This application first collects candidates' heart rate, blood pressure, and facial expression data in real time through physiological sensing devices to obtain objective stress state indicators. Then, using a preset stress index assessment model, the presentation rhythm of clinical cases is dynamically adjusted according to real-time physiological parameters, including adjusting the case presentation speed, difficulty gradient, and number of concurrent tasks. Finally, by analyzing the changing trends of physiological parameters and combining them with the candidates' answering performance, a personalized exam rhythm adjustment plan is adaptively generated to achieve dynamic evolution of the exam paper. Through a physiological data-driven dynamic adjustment mechanism, not only is a deep integration of the exam environment with real medical scenarios achieved, but also an intelligent assessment system adapted to individual differences is established.

[0013] Optionally, based on a preset stress index assessment model, the presentation speed, difficulty gradient, and number of concurrent tasks of clinical cases are dynamically adjusted according to the real-time physiological parameter information, specifically including the following steps: The real-time physiological parameter information is compared with the preset medical scenario stress threshold to determine the candidate's current stress tolerance status. Based on the aforementioned stress tolerance status, the presentation interval of clinical cases and the time limit for answering each question are dynamically adjusted. Based on the aforementioned stress tolerance status, multiple clinical cases with competing treatment resources are retrieved from a question bank of different difficulty levels to construct a multi-patient concurrent treatment scenario; Based on the candidates' decisions regarding the order of handling concurrent clinical cases and resource allocation, the disease progression and complication triggering of subsequent cases are dynamically adjusted to obtain a dynamically adjusted combination of test questions.

[0014] By adopting the above technical solution, since patients' conditions often change dynamically and influence each other in actual medical work, and the allocation decisions of medical resources are directly related to the prognosis of multiple patients, simple parallel task execution cannot reflect the resource competition relationship between patients, nor can it reflect the chain reaction caused by decision delays. This application first compares the collected real-time physiological parameters with the preset medical scenario stress threshold to scientifically assess the candidate's current stress tolerance. Then, it dynamically adjusts the presentation rhythm of clinical cases according to the stress status, including the case interval time and the answering time limit. Next, it selects multiple clinical cases with competition for treatment resources from a question bank of different difficulty levels to construct a realistic multi-patient concurrent treatment scenario. Finally, based on the candidate's processing order of concurrent cases and resource allocation decisions, it triggers the corresponding disease development path and the occurrence of complications, realizing the dynamic evolution of the test questions. By introducing a resource competition mechanism and decision chain reaction, it not only restores the complexity of clinical work, but also realizes a three-dimensional assessment of the comprehensive decision-making ability of medical personnel.

[0015] Optionally, the method may also include the following steps: By identifying candidates' answering behavior through an anomaly detection mechanism, the duration and patterns of answering clinical case analysis questions are monitored to obtain early warning information on abnormal behavior; Based on the abnormal behavior warning information, the difficulty of subsequent clinical cases is dynamically adjusted or unexposed test questions are replaced from the backup case library to obtain the adjusted test paper parameters.

[0016] By adopting the above technical solution, this application first uses an anomaly detection mechanism to analyze candidates' answering behavior in real time, focusing on monitoring the distribution of answering time and changes in answering rhythm of clinical case analysis questions, and identifying possible abnormal patterns; then, based on the detected abnormal behavior warning information, it promptly adjusts the difficulty level of subsequent clinical cases, or retrieves unexposed test questions from a pre-prepared backup case library for replacement, and dynamically updates the test paper parameters; this improves the accuracy of identifying cheating behavior while achieving real-time optimization of test question content.

[0017] Optionally, the method may also include the following steps: Based on the RBAC access control of the hospital personnel management system, hierarchical management of the test paper generation parameters for different levels of examinations is carried out to obtain an access control scheme. According to the aforementioned access control scheme, the system interfaces with the hospital information system to obtain the candidate's department and professional title information, and generates a targeted clinical assessment test paper.

[0018] By adopting the above technical solution, this application first establishes a multi-level management system for the test paper generation parameters of different levels of examinations based on the RBAC access control mechanism of the hospital personnel management system, so as to realize refined access control of test paper resources; then, by connecting with the hospital information system, it automatically obtains the candidate's department affiliation and professional title level information, and generates clinical assessment test papers that match the job responsibilities based on these characteristic data; through refined access management and automated test paper generation, both the security of test paper resources and the accurate positioning of assessment content are ensured.

[0019] Secondly, this application provides a medical examination system that applies the aforementioned intelligent test paper generation method, including: The requirement parameter receiving module is used to receive the hospital department assessment requirement parameters input by the user; The candidate question pool acquisition module is used to retrieve a set of questions that meet the conditions from the hospital clinical case question bank according to the assessment requirement parameters, and obtain the candidate question pool. The preliminary question combination acquisition module is used to intelligently filter the candidate question pool based on a preset clinical diagnosis and treatment dynamic model to obtain a preliminary question combination; The question combination optimization module is used to obtain the characteristics of high-frequency clinical test points in the hospital's historical examination data, and to obtain the optimized question combination based on the distribution of clinical knowledge points in the preliminary question combination and the characteristics of the high-frequency clinical test points. The test paper parameter adjustment module is used to evaluate the balance of clinical difficulty and coverage of diagnostic and treatment knowledge points of the test paper in real time based on the optimized test paper combination, and determine the adjustment plan of the test paper parameters until the preset standard is met.

[0020] Thirdly, this application provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the above-described intelligent document generation method.

[0021] Fourthly, this application provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the steps of the above-described intelligent document generation method.

[0022] In summary, this application includes at least one of the following beneficial technical effects: 1. This application first receives the department's assessment requirements parameters, retrieves a set of qualified questions from the clinical case question bank to form a candidate question pool, and then uses a preset clinical diagnosis and treatment dynamic model to intelligently filter the candidate questions to obtain a preliminary question combination. Next, it analyzes the characteristics of high-frequency clinical test points in the hospital's historical examination data and optimizes and adjusts them in combination with the distribution of knowledge points in the preliminary question combination. By evaluating the balance of clinical difficulty and the coverage of diagnosis and treatment knowledge points in real time, it dynamically adjusts the test paper parameters until the preset standards are met. By introducing the clinical diagnosis and treatment dynamic model and historical examination data analysis, it realizes the intelligent construction of clinical correlation between questions and the dynamic optimization of knowledge point distribution, so that the generated test paper can more accurately assess the actual diagnosis and treatment ability of medical staff. 2. This application first utilizes a neural network model to deeply mine historical exam data, extracting features such as the frequency of occurrence, importance, and changing trends of various disease-related test points, and establishing a priority system for clinical test points. Then, based on the obtained priority data, it dynamically adjusts the weight distribution of various diagnostic and treatment knowledge points in the initial test question combination, and ensures that the proportion of test points related to acute and critical illnesses meets the needs of clinical practice by setting preset thresholds. This achieves data-driven optimization of test point distribution, making the test paper structure more in line with clinical reality and the assessment effect more targeted. 3. This application first collects real-time data on examinees' heart rate, blood pressure, and facial expressions using physiological sensing devices to obtain objective stress state indicators. Then, using a pre-set stress index assessment model, it dynamically adjusts the presentation pace of clinical cases based on real-time physiological parameters, including adjusting the case presentation speed, difficulty gradient, and number of concurrent tasks. Finally, by analyzing the changing trends of physiological parameters and combining them with the examinees' performance, it adaptively generates personalized exam pace adjustment plans, realizing the dynamic evolution of the exam paper. Through a physiological data-driven dynamic adjustment mechanism, it not only achieves a deep integration between the exam environment and real medical scenarios but also establishes an intelligent assessment system that adapts to individual differences. Attached Figure Description

[0023] Figure 1 This is a flowchart illustrating an intelligent document generation method according to an embodiment of this application; Figure 2 This is a flowchart illustrating step S140 in an intelligent document generation method according to an embodiment of this application. Figure 3 This is a schematic diagram of the anti-cheating process in an intelligent test paper generation method according to an embodiment of this application; Figure 4 This is a flowchart illustrating the personalized adjustment of test papers in an intelligent test paper generation method according to an embodiment of this application; Figure 5 This is a flowchart illustrating step S420 in an intelligent document generation method according to an embodiment of this application. Figure 6This is a flowchart illustrating the abnormal behavior warning process in an intelligent test paper generation method according to an embodiment of this application. Figure 7 This is a flowchart illustrating the access control process in an intelligent document generation method according to an embodiment of this application. Figure 8 This is a schematic diagram of a module of a medical examination system according to an embodiment of this application; Figure 9 This is an internal structural diagram of an electronic device according to an embodiment of this application. Detailed Implementation

[0024] The terminology used in the following embodiments of this application is for the purpose of describing particular embodiments only and is not intended to be limiting of this application. As used in the specification and appended claims of this application, the singular expressions “a,” “an,” “the,” “the,” “the,” and “this” are intended to include the plural expressions as well, unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this application refers to any or all possible combinations including one or more of the listed items.

[0025] Hereinafter, the terms "first" and "second" are used for descriptive purposes only and should not be construed as implying or suggesting relative importance or implicitly indicating the number of indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature, and in the description of the embodiments of this application, unless otherwise stated, "multiple" means two or more.

[0026] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0027] Firstly, this application provides an intelligent test paper generation method, referring to... Figure 1 It includes the following steps: S110: Receive the hospital department assessment requirements parameters input by the user.

[0028] In this embodiment, the hospital department assessment requirements parameters refer to the basic configuration information used to determine the scope and requirements of the test paper, including the target department, assessment type, exam duration, question structure, difficulty distribution, and knowledge point coverage requirements. The assessment types are divided into job entry exams, annual assessments, promotion exams, and periodic evaluations; the question structure includes the proportion of multiple-choice questions, clinical case analysis questions, and practical assessment questions; the difficulty distribution includes the proportion of easy, medium, and difficult questions; and the knowledge point coverage requirements specify the exact range of required and optional knowledge points.

[0029] Specifically, the system first receives the assessment requirements form input by the user through a web interface. The form contains multiple options and fill-in-the-blank fields. The options are in the form of drop-down menus, with preset lists of departments, assessment types, and question templates. The fill-in-the-blank fields receive specific duration and percentage values ​​through input boxes. The system validates the input parameters to ensure that the sum of the percentages is 100% and the duration is within a reasonable range. After successful validation, the parameters are organized into a standard format and stored in the exam configuration database.

[0030] S120. Based on the assessment requirements parameters, retrieve a set of questions that meet the conditions from the hospital's clinical case question bank to obtain a candidate question pool.

[0031] In this embodiment, the candidate question pool refers to the set of preliminary selection questions based on assessment requirements. Each question in the question bank includes basic attribute tags such as department classification, applicable exam type, difficulty level, knowledge points involved, and historical usage frequency. The system matches these tags with assessment requirement parameters to initially determine the range of questions that meet the requirements.

[0032] Specifically, the system establishes a question attribute index table, converts question tags into binary feature vectors, and quickly completes attribute matching through bitwise operations. First, a first round of filtering is performed based on department and exam type to obtain a subset of questions relevant to the department's specialty. Then, a second round of filtering is conducted according to question type and difficulty requirements to ensure a sufficient number of questions for each type. Finally, the coverage of knowledge points is checked, and questions related to essential exam points are prioritized for retention.

[0033] S130. Based on the preset dynamic model of clinical diagnosis and treatment, the candidate question pool is intelligently screened to obtain a preliminary question combination.

[0034] In this embodiment, the clinical diagnosis and treatment dynamic model is a knowledge graph that describes the key decision nodes in the disease diagnosis and treatment process, including logical connections across multiple dimensions such as symptom characteristics, examination plans, diagnostic approaches, treatment plans, and prognostic assessments.

[0035] Specifically, the system pre-builds a library of treatment pathway templates based on the characteristics of each department and specialty. Multiple-choice questions mainly test the understanding of key diagnostic and treatment points, and the system ensures that the questions cover typical diagnostic and treatment scenarios through template matching; case analysis questions focus on the clinical reasoning process, and the system analyzes the completeness of the diagnostic and treatment elements in the case text to ensure that the questions are set in accordance with the logic of clinical practice.

[0036] S140. Obtain the characteristics of high-frequency clinical test points from the hospital's historical examination data. Based on the distribution of clinical knowledge points and the characteristics of high-frequency clinical test points in the preliminary test question combination, obtain the optimized test question combination.

[0037] In this embodiment, high-frequency clinical test point features refer to the distribution patterns of important knowledge points and the characteristics of question combinations extracted by analyzing historical exam data. The system establishes a test point feature database, recording the frequency of occurrence, importance, and correlation of each knowledge point in different types of exams.

[0038] Specifically, the system uses text mining technology to extract knowledge point tags from historical test questions and establishes a knowledge point-test question association matrix. During optimization, it first ensures coverage of frequently tested points, then analyzes the patterns of combinations between knowledge points and adjusts the question combination structure. Simultaneously, it considers the timeliness of knowledge points, appropriately increasing the weight of frequently updated clinical guidelines.

[0039] S150. Based on the optimized test question combination, the clinical difficulty balance and coverage of diagnosis and treatment knowledge points of the test paper are evaluated in real time, and the adjustment plan of the test paper parameters is determined until the preset standard is met.

[0040] In this embodiment, clinical difficulty balance refers to the balance of the overall difficulty distribution of the test paper, and the coverage rate of diagnostic and treatment knowledge points refers to the degree to which the test questions cover the core knowledge of the target field.

[0041] Specifically, the system sets up two evaluation dimensions: difficulty balance and knowledge point coverage. Difficulty balance is assessed by calculating the difference in difficulty between adjacent questions and the smoothness of the cumulative difficulty curve; knowledge point coverage is calculated based on a weighted average of the importance of knowledge points. When the evaluation indicators fail to meet the standards, the system will replace some questions according to preset adjustment rules until the quality requirements are met.

[0042] In one embodiment, refer to Figure 2 In step S140, the high-frequency clinical test point characteristics in the hospital's historical examination data are obtained. Based on the distribution of clinical knowledge points and the high-frequency clinical test point characteristics in the preliminary test question combination, the optimized test question combination is obtained, which specifically includes the following steps: S141. Analyze the distribution of clinical diagnosis and treatment test points in historical exam data using a neural network model, identify test points for acute and critical illnesses, common diseases, and their corresponding trends, and obtain the priority of clinical test points.

[0043] In this embodiment, historical exam data refers to past exam paper information stored in the hospital's exam system, including question content, knowledge point annotations, usage scenarios, student score distribution, difficulty coefficient, and discrimination index. Clinical diagnosis and treatment exam points refer to key diagnostic and treatment knowledge points that need to be mastered in medical practice, categorized into three levels according to clinical importance: acute and critical illnesses, frequently occurring diseases, and common diseases. The changing trends of exam points reflect the impact of clinical guideline updates and treatment protocol iterations on the assessment content.

[0044] Specifically, the system first establishes a preprocessing workflow for exam data. Medical terms and keywords are extracted from the exam questions using text segmentation technology to build a knowledge point dictionary. Then, combined with the ICD disease coding system, the extracted keywords are mapped to standardized disease categories and treatment procedures. For changes in treatment protocols appearing in new clinical guidelines, timestamps are added to the knowledge point dictionary. The system employs a three-layer neural network structure: the input layer receives the feature vectors of the exam question text, the hidden layer identifies important test points through an attention mechanism, and the output layer predicts the category and importance of the test points. Model training uses labeled questions from historical exam data as samples, and network parameters are optimized through backpropagation.

[0045] S142. Based on the priority of clinical test points, adjust the weight of diagnosis and treatment knowledge points in the initial test question combination to ensure that the proportion of test points related to acute and critical illnesses in the test paper is not lower than the preset threshold, thus obtaining the optimized test question combination.

[0046] In this embodiment, the priority of clinical test points refers to the ranking of the importance of different types of test points in the exam paper. The system establishes a test point classification standard, classifying the identified test points according to clinical risk, incidence rate, and difficulty of diagnosis and treatment. Among them, acute and critical illness test points include clinical situations requiring urgent treatment, potentially life-threatening complications, and prevention of serious adverse reactions; common disease test points cover the diagnostic points, treatment options, and follow-up management principles for common diseases in the department; and common disease test points include basic theoretical knowledge, basic operating procedures, and general clinical manifestations.

[0047] Specifically, the system establishes a test point priority mapping table, setting basic weights for different categories of test points. Test points related to acute, critical, and severe illnesses have the highest basic weight to ensure that the proportion of such questions meets the requirements. During the adjustment process, the system first statistically analyzes the distribution of various test points in the initial question combinations and compares the statistical results with preset thresholds. When the proportion of acute, critical, and severe illness test points is insufficient, a replacement algorithm is used to select suitable questions for replacement. The replacement rules prioritize questions with low overlap in knowledge points to avoid insufficient knowledge point coverage after replacement. Simultaneously, the system analyzes the correlation between the question to be replaced and other questions through a question association matrix to ensure that the replacement does not affect the overall coherence of the exam paper.

[0048] In one embodiment, refer to Figure 3 After obtaining the optimized question combination in step S140, the method further includes the following steps: S310. Based on the clinical characteristics of different departments, a random perturbation factor is used to generate differentiated test paper versions, resulting in multiple versions of the test paper.

[0049] In this embodiment, the random perturbation factor refers to the set of rules used for question transformation, and the perturbation factors for different departments are designed according to the characteristics of that subject. Differentiated test paper versions refer to multiple equivalent but differently worded test papers generated after processing with the perturbation factor, maintaining consistency in the knowledge point structure and difficulty distribution. Intra-hospital network IP restriction refers to limiting access to the examination system to a specific range of IP addresses. Random question order refers to shuffling the order in which the test questions are presented. A / B test mode refers to adjacent candidates using different versions of the test paper.

[0050] Specifically, the system establishes a departmental perturbation rule base, setting transformation rules for different types of questions. Multiple-choice questions are primarily transformed by adjusting the order of options, changing numerical ranges, and replacing synonyms; case analysis questions are transformed by adjusting the order of medical history, replacing examination result values, and changing drug brand names. The system maintains a perturbation mapping table, recording interchangeable numerical ranges, synonym groups, and drug name groups to ensure semantic accuracy after transformation. The perturbation process uses random sampling to extract replacement elements from the mapping table, while rule validation ensures the logical consistency of the questions. Based on the number of examinees, 3-5 test paper versions are generated, maintaining a consistent knowledge point structure but differing in expression among versions.

[0051] S320. Set intra-hospital network IP restrictions, random question order, and A / B version modes for multiple versions of the exam paper to obtain the final exam paper.

[0052] In this embodiment, the system configures a database to record a dedicated IP address range for the examination, allowing only specific network segments within the institution to access the examination system. After the exam paper is generated, the system assigns a random number to each question and reorders them based on these random numbers to generate the final exam paper.

[0053] In one embodiment, refer to Figure 4 The method also includes the following steps: S410. Obtain the candidate's real-time heart rate, blood pressure, and facial expression data to obtain real-time physiological parameter information.

[0054] In this embodiment, real-time physiological parameter information refers to the physiological state data of the examinee collected by monitoring equipment. Heart rate data is acquired through a smart bracelet and recorded every 5 seconds; blood pressure data is measured using a non-invasive blood pressure monitor, and diastolic and systolic blood pressure are recorded every 3 minutes; facial expression data is captured by the examination room camera, identifying three characteristic parameters: furrowed brows, drooping corners of the mouth, and eye fatigue.

[0055] Specifically, the system establishes a physiological parameter acquisition platform, standardizing the data interfaces of different devices. Smart bracelets transmit heart rate data via Bluetooth, blood pressure monitors connect via USB, and facial expression recognition is based on video stream processing. The system maintains a device parameter table, recording the sampling frequency, data accuracy, and signal stability indicators for each device. When abnormal data is detected, the system activates backup devices or adjusts the sampling frequency. The collected raw data is preprocessed and stored in a time-series database, establishing a mapping relationship between examinee IDs and physiological parameters.

[0056] S420, based on a preset stress index assessment model, dynamically adjusts the presentation speed, difficulty gradient, and number of concurrent tasks of clinical cases according to real-time physiological parameter information to simulate high-pressure medical scenarios.

[0057] In this embodiment, the stress index assessment model is an assessment system based on weighted calculations of physiological parameters. The model sets baseline intervals and weight coefficients for different physiological indicators, and converts each indicator into a unified dimension through normalization. The dynamic adjustment mechanism includes presentation speed control, difficulty gradient adjustment, and concurrent task management to simulate stress scenarios in clinical work.

[0058] Specifically, the system establishes a stress index calculation rule table, assigning weights to different indicators based on their clinical significance. Heart rate variability has a weight coefficient of 0.4, blood pressure fluctuation has a weight of 0.3, and facial expression features have a weight of 0.3. The calculated stress index is divided into three intervals: low stress, moderate stress, and high stress. The system pre-sets adjustment strategies corresponding to these stress intervals, including question interval time, difficulty progression speed, and the number of concurrent tasks.

[0059] S430. Based on the changing trends of real-time physiological parameters and the candidate's performance, a personalized exam rhythm adjustment plan is generated to obtain the final dynamic exam paper.

[0060] In this embodiment, the exam pacing adjustment scheme refers to a test paper presentation strategy customized according to the individual differences of the examinees. The system records the examinees' performance under different pressure levels, including indicators such as answering speed, accuracy, and coherence of thought, and combines this with the trends of physiological parameter changes to generate a personalized pacing control scheme.

[0061] Specifically, the system constructs a framework for analyzing candidate performance and establishes a correlation table between stress level and answer performance. When abnormal fluctuations in physiological indicators are detected, the system queries the correlation table to obtain the candidate's optimal performance range. Adjustment plans prioritize stabilizing the candidate's state, maintaining stress levels within a reasonable range by dynamically adjusting question difficulty and intervals. The system sets limits on the adjustment range to ensure a smooth adjustment process. Simultaneously, the system records the adjustment effects, allowing for real-time evaluation and optimization of the adjustment strategies. The resulting dynamic exam paper assesses professional skills while also catering to individual differences, enhancing the scientific rigor of the assessment.

[0062] In one embodiment, refer to Figure 5 In step S420, based on a preset stress index assessment model, the presentation speed, difficulty gradient, and number of concurrent tasks of clinical cases are dynamically adjusted according to real-time physiological parameter information. Specifically, this includes the following steps: S421. Compare real-time physiological parameter information with preset medical scenario stress thresholds to determine the candidate's current stress tolerance status.

[0063] In this embodiment, the stress threshold in a medical setting refers to reference values ​​for physiological indicators set based on the characteristics of clinical work. The system divides the standard ranges of three types of indicators—heart rate, blood pressure, and facial expression—into four intervals: normal, mild stress, moderate stress, and severe stress. The stress tolerance status reflects the physiological load level of the examinee in the current examination environment.

[0064] Specifically, the system establishes a stress threshold comparison table, recording the indicator ranges for different stress levels. Heart rate indicators are based on heart rate variability analysis, calculating the standard deviation and coefficient of variation of continuous heart rate data; blood pressure indicators determine stress levels by calculating the degree of deviation from baseline blood pressure; facial expression features are extracted using image recognition to extract stress level characteristics. The system uses a sliding window approach to calculate the short-term trends of each indicator, combining multiple indicators with weighted averages to obtain the current stress state assessment result.

[0065] Furthermore, the system pre-establishes a basic information table for examinees, recording their age group, baseline blood pressure level, and baseline heart rate. Personalized stress threshold adjustment coefficients are set for different examinee groups, encompassing three dimensions: age group adjustment mechanism, baseline condition correction, and historical data reference. Age group adjustment differentiates assessments for examinees of different age groups by setting a correction ratio for the upper limit of the heart rate threshold and a tolerance coefficient for blood pressure fluctuations. Baseline condition correction establishes an individual baseline by recording the average physiological indicators before the exam, and calculates the deviation of real-time data using a personalized stress indicator deviation table, enabling dynamic adjustment of the assessment standards. The historical data reference mechanism identifies individual stress response characteristics by saving the physiological indicator change curves of examinees from previous exams, establishing personalized weighting coefficients that highlight key response indicators.

[0066] S422. Based on the stress tolerance level, dynamically adjust the presentation interval of clinical cases and the time limit for answering each question.

[0067] In this embodiment, the presentation interval refers to the transition time between adjacent questions, and the answering time limit refers to the maximum allowed answering time for a single question. The system dynamically adjusts these two time parameters based on the pressure tolerance status, ensuring the assessment effect while avoiding pressure overload.

[0068] Specifically, the system maintenance time parameter adjustment table sets a base time and adjustment coefficient for different pressure states. When an increase in pressure level is detected, the system automatically extends the presentation interval and response time limit; when the pressure level is within a controllable range, the original rhythm is maintained. The adjustment is gradual to avoid abrupt changes in time parameters.

[0069] S423. Based on the stress tolerance status, retrieve multiple clinical cases with competing treatment resources from a question bank of different difficulty levels to construct a multi-patient concurrent treatment scenario.

[0070] In this embodiment, the competition for medical resources refers to situations where multiple cases experience time or resource conflicts regarding the use of medical equipment, expert consultations, and operating room arrangements. The system determines the number of concurrent cases based on the stress level and constructs realistic multi-patient management scenarios by combining clinical cases of different difficulty levels.

[0071] Specifically, the system establishes a case resource association table to record the key medical resources involved in each case. When selecting concurrent cases, it prioritizes combinations with resource competition to create decision-making pressure. The system sets up basic case combination templates and dynamically increases or decreases the number of concurrent cases based on the pressure level.

[0072] S424. Based on the candidates' handling order and resource allocation decisions for concurrent clinical cases, the disease development and complication triggering of subsequent cases are dynamically adjusted to obtain the dynamically adjusted test question combination.

[0073] In this embodiment, the processing order refers to the priority ranking of concurrent cases by the examinee, and the resource allocation decision refers to the allocation plan under resource constraints. Based on the examinee's processing strategy, the system triggers the corresponding disease evolution path and evaluates the rationality of the decision.

[0074] Specifically, the system constructs a rule base for disease progression, pre-setting multiple evolutionary branches. After the candidate completes resource allocation decisions, the system selects the corresponding disease progression path based on the decision results. For example, delaying treatment of critically ill cases may trigger a branch indicating worsening of the condition, and improper resource allocation may lead to complications. Through this dynamic feedback mechanism, the system examines the candidate's clinical decision-making ability under pressure. The generated dynamic question combinations reflect both the correlation between cases and the continuous impact of decisions.

[0075] In one embodiment, refer to Figure 6 The method also includes the following steps: S610. Identify candidates' answering behavior through an anomaly detection mechanism, monitor the answering time and answering patterns of clinical case analysis questions, and obtain early warning information on abnormal behavior.

[0076] In this embodiment, the anomaly detection mechanism refers to identifying answering behavior characteristics that may pose a cheating risk through data analysis. Answering time refers to the actual time a candidate spends answering a single clinical case, and answering patterns refer to the behavioral patterns exhibited by the candidate during continuous answering. Abnormal behavior warning information includes risk assessment results across three dimensions: time anomalies, pattern anomalies, and content anomalies.

[0077] Specifically, the system establishes a database of answer behavior characteristics, recording the typical time distribution and answer patterns of normal clinical thinking processes. The time dimension monitors the time distribution for each question, setting reference time intervals based on question difficulty and knowledge point complexity; the pattern dimension analyzes changes in answer rhythm, identifying sudden abnormalities in answer speed; and the content dimension uses text similarity analysis to discover abnormally similar content between multiple answer sheets. The system employs a real-time monitoring mechanism, generating corresponding level of warning information when any dimension is detected to exceed the normal range.

[0078] S620. Based on the abnormal behavior warning information, dynamically adjust the difficulty of subsequent clinical cases or replace unexposed test questions from the backup case library to obtain the adjusted test paper parameters.

[0079] In this embodiment, the adjusted test paper parameters refer to a configuration scheme that dynamically adjusts the difficulty and source of test questions based on early warning information. The system sets corresponding adjustment strategies for different early warning levels to ensure the fairness and effectiveness of the assessment. The backup case library contains clinical cases with evenly distributed difficulty and complete coverage of knowledge points, which are used for dynamic replacement of test questions.

[0080] In one embodiment, refer to Figure 7 The method also includes the following steps: S710. Based on the RBAC access control of the hospital personnel management system, hierarchical management of the test paper generation parameters for different levels of examinations is carried out to obtain an access control scheme.

[0081] In this embodiment, RBAC (Role-Based Access Control) refers to a role-based access control mechanism that achieves secure management of test question resources through role definition and permission allocation. The hierarchical management of test paper generation parameters includes three levels: test question management permissions, review permissions, and test paper generation permissions, ensuring the security of sensitive test questions and high-level exam content. The permission management scheme specifies the access scope and operation permissions of different roles for test question resources.

[0082] Specifically, the system establishes a permission configuration table, defining the role types and permission items related to exam management. Role types include exam administrator, department exam manager, question-setting expert, and review expert; permission items include permission to edit, review, compile, and view exam questions. Corresponding permission thresholds are set for different levels of exams; for example, exam questions for the attending physician promotion exam require review by experts with the title of associate chief physician or higher. The system ensures that exam question resources are strictly managed according to prescribed procedures through permission inheritance and constraint mechanisms.

[0083] S720. Based on the access control scheme, connect with the hospital information system to obtain the candidate's department and professional title information, and generate targeted clinical assessment papers.

[0084] In this embodiment, the targeted clinical assessment test paper refers to the examination content customized according to the characteristics of the candidate's department and professional title requirements. The system obtains the candidate's basic information through the hospital information system interface and generates a test paper that matches the job responsibilities based on the professional requirements of different positions.

[0085] Specifically, the system constructs a job competency model library, recording the core competency requirements for different departments and professional title levels. Through data interfaces, it obtains candidates' department affiliation, professional title level, and specialization from the personnel system to create candidate profiles. When compiling test papers, it prioritizes clinical cases related to the candidates' professional backgrounds, setting the difficulty distribution and knowledge point coverage according to job requirements. The system also adjusts the case presentation format based on departmental characteristics; for example, surgical cases focus on surgical plan selection, while internal medicine cases emphasize medication decisions.

[0086] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0087] Secondly, this application provides a medical examination system. The medical examination system of this application will be described below in conjunction with the above-mentioned intelligent test paper generation method.

[0088] Reference Figure 8 A medical examination system, comprising: The requirement parameter receiving module is used to receive the hospital department assessment requirement parameters input by the user; The candidate question pool acquisition module is used to retrieve a set of questions that meet the conditions from the hospital's clinical case question bank according to the assessment requirements parameters, and obtain the candidate question pool. The preliminary question combination acquisition module is used to intelligently filter the candidate question pool based on a preset clinical diagnosis and treatment dynamic model to obtain a preliminary question combination; The question combination optimization module is used to obtain the characteristics of high-frequency clinical test points in the hospital's historical examination data, and to obtain the optimized question combination based on the distribution of clinical knowledge points and the characteristics of high-frequency clinical test points in the preliminary question combination; The test paper parameter adjustment module is used to evaluate the balance of clinical difficulty and coverage of diagnosis and treatment knowledge points of the test paper in real time based on the optimized test paper combination, and determine the adjustment plan of the test paper parameters until the preset standards are met.

[0089] In one embodiment, this application provides an electronic device, which may be a server, and its internal structure diagram may be as follows: Figure 9 As shown, this electronic device includes a processor, memory, and a network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The database stores data. The network interface communicates with external terminals via a network connection. When the computer program is executed by the processor, it implements an intelligent document generation method.

[0090] Those skilled in the art will understand that Figure 9 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the electronic device to which the present application is applied. The specific electronic device may include more or fewer components than shown in the figure, or combine certain components, or have different component arrangements.

[0091] In one embodiment, an electronic device is also provided, including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps in the above-described method embodiments.

[0092] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, or optical storage, etc. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM), etc.

[0093] The above are all preferred embodiments of this application, and are not intended to limit the scope of protection of this application. Therefore, all equivalent changes made in accordance with the structure, shape and principle of this application should be covered within the scope of protection of this application.

Claims

1. An intelligent test paper generation method, characterized in that, Includes the following steps: Receive user-input parameters for hospital department assessment requirements; Based on the assessment requirements parameters, a set of questions that meet the conditions is retrieved from the hospital's clinical case question bank to obtain a candidate question pool; Based on a preset dynamic model of clinical diagnosis and treatment, the candidate question pool is intelligently screened to obtain a preliminary question combination; The characteristics of high-frequency clinical test points are obtained from the hospital's historical examination data. Based on the distribution of clinical knowledge points in the preliminary test question combination and the characteristics of high-frequency clinical test points, an optimized test question combination is obtained. Based on the optimized question combination, the clinical difficulty balance and coverage of diagnostic and treatment knowledge points of the test paper are evaluated in real time, and the adjustment plan of the test paper parameters is determined until the preset standard is met.

2. The intelligent test paper generation method according to claim 1, characterized in that, The process involves obtaining high-frequency clinical test point characteristics from historical hospital exam data, and then, based on the distribution of clinical knowledge points in the initial test question combination and the high-frequency clinical test point characteristics, obtaining an optimized test question combination. This process specifically includes the following steps: By analyzing the distribution of clinical diagnosis and treatment test points in historical exam data using a neural network model, we can identify test points for acute and critical illnesses, common diseases, and frequently occurring diseases, as well as their corresponding trends, and thus obtain the priority of clinical test points. Based on the priority of the clinical test points, the weight of the diagnostic and treatment knowledge points in the preliminary test question combination is adjusted to ensure that the proportion of test points related to acute and critical illnesses in the test paper is not lower than the preset threshold, thus obtaining the optimized test question combination.

3. The intelligent test paper generation method according to claim 2, characterized in that, After obtaining the optimized question combination, the method further includes the following steps: Based on the clinical characteristics of different departments, a random perturbation factor was used to generate differentiated test paper versions, resulting in multiple versions of the test paper; The multiple versions of the exam paper were configured with intra-college network IP restrictions, random question order, and A / B version modes to obtain the final exam paper.

4. The intelligent test paper generation method according to claim 1, characterized in that, The method also includes the following steps: The system acquires real-time heart rate, blood pressure, and facial expression data of test takers to obtain real-time physiological parameter information. Based on a preset stress index assessment model, the presentation speed, difficulty gradient, and number of concurrent tasks of clinical cases are dynamically adjusted according to the real-time physiological parameter information to simulate high-pressure medical scenarios. Based on the changing trends of the real-time physiological parameters and the examinee's performance, a personalized exam rhythm adjustment plan is generated, resulting in the final dynamic exam paper.

5. The intelligent test paper generation method according to claim 4, characterized in that, Based on a preset stress index assessment model, the presentation speed, difficulty gradient, and number of concurrent tasks of clinical cases are dynamically adjusted according to the real-time physiological parameter information. Specifically, this includes the following steps: The real-time physiological parameters are compared with preset medical scenario stress thresholds to determine the candidate's current stress tolerance level. Based on the aforementioned stress tolerance status, the presentation interval of clinical cases and the time limit for answering each question are dynamically adjusted. Based on the aforementioned stress tolerance status, multiple clinical cases with competing treatment resources are retrieved from a question bank of different difficulty levels to construct a multi-patient concurrent treatment scenario; Based on the candidates' handling order and resource allocation decisions for concurrent clinical cases, the disease progression and complication triggering of subsequent cases are dynamically adjusted to obtain a dynamically adjusted test question combination.

6. The intelligent test paper generation method according to claim 1, characterized in that, The method also includes the following steps: An anomaly detection mechanism is used to identify candidates' answering behavior, monitor the answering time and answering patterns of clinical case analysis questions, and obtain early warning information on abnormal behavior. Based on the abnormal behavior warning information, the difficulty of subsequent clinical cases is dynamically adjusted or unexposed test questions are replaced from the backup case library to obtain the adjusted test paper parameters.

7. The intelligent test paper generation method according to claim 1, characterized in that, The method also includes the following steps: Based on the RBAC access control of the hospital personnel management system, hierarchical management of the test paper generation parameters for different levels of examinations is carried out to obtain an access control scheme. According to the aforementioned access control scheme, the system interfaces with the hospital information system to obtain the candidate's department and professional title information, and generates a targeted clinical assessment test paper.

8. A medical examination system, characterized in that, The intelligent paper generation method according to claims 1-7 includes: The requirement parameter receiving module is used to receive the hospital department assessment requirement parameters input by the user; The candidate question pool acquisition module is used to retrieve a set of questions that meet the conditions from the hospital clinical case question bank according to the assessment requirement parameters, and obtain the candidate question pool. The preliminary question combination acquisition module is used to intelligently filter the candidate question pool based on a preset clinical diagnosis and treatment dynamic model to obtain a preliminary question combination; The question combination optimization module is used to obtain the characteristics of high-frequency clinical test points in the hospital's historical examination data, and to obtain the optimized question combination based on the distribution of clinical knowledge points in the preliminary question combination and the characteristics of the high-frequency clinical test points. The test paper parameter adjustment module is used to evaluate the balance of clinical difficulty and coverage of diagnostic and treatment knowledge points of the test paper in real time based on the optimized test paper combination, and determine the adjustment plan of the test paper parameters until the preset standard is met.

9. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps of the intelligent volume-compilation method according to any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent paper-generating method according to any one of claims 1-7.

Images