Smart hospital residence and cultivation management and evaluation integrated system
By integrating artificial intelligence technology and big data analysis in the integrated system of smart hospital inpatient and training management evaluation, automatic and real-time student evaluation is achieved, the problems of inconsistent evaluation standards and subjective bias are solved, and the objectivity and accuracy of evaluation are improved. , improve the quality of residential training management and the learning and growth of students.
Patent Information
- Application Number
- CN202411518496.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-29
- Publication Date
- 2025-02-25
AI Technical Summary
The existing integrated system for in-house training management and evaluation of smart hospitals has inconsistent evaluation standards. The possible subjective biases and human error problems in traditional evaluation methods have limited the improvement of the quality of in-house training management and the learning and growth of residents.
Design an integrated system for inpatient and training management and evaluation of smart hospitals, integrate artificial intelligence technology and big data analysis, and realize automatic and real-time student evaluation through data collection modules, behavioral analysis modules and other modules to reduce subjective biases and human errors.
Improve the objectivity, accuracy and comprehensiveness of the assessment, enhance the objectivity of the teacher's rating of students, effectively avoid subjective bias and artificial errors in traditional evaluation methods, and improve the quality of residential training management and students' learning and growth .
Smart Images
Figure CN119515140A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the field of medical information technology, and in particular to an integrated system for resident training management and evaluation in a smart hospital. Background Art
[0002] With the rapid development of the medical industry and the continuous upgrading of patient needs, hospital informatization has become a key path to improve the quality and efficiency of medical services. As an indispensable part of the hospital informatization blueprint, the integrated system of resident training management and evaluation in smart hospitals is becoming increasingly important. The system aims to optimize the standardized training process of resident physicians through informatization, thereby ensuring the steady improvement of medical quality and reliable guarantee of medical safety.
[0003] However, the traditional resident training management model faces challenges such as information fragmentation and low management efficiency, and it is difficult to meet the needs of the modern medical system for efficient and accurate management. The introduction of the integrated resident training management and evaluation system for smart hospitals provides an innovative solution to this problem. By integrating advanced technology, it realizes the intelligence and standardization of the training process.
[0004] Although the existing integrated resident training management and evaluation system of smart hospitals has improved management efficiency to a certain extent, there are still problems such as inconsistent evaluation standards, subjective bias and human errors that may exist in traditional evaluation methods. This not only limits the further improvement of the quality of resident training management, but may also affect the learning and growth of residents and the overall optimization of medical service quality.
[0005] Therefore, solving the problems of inconsistent evaluation standards in the existing smart hospital resident training management and evaluation integrated system, as well as the subjective bias and human errors that may exist in traditional evaluation methods, has become an important issue that needs to be overcome in the current field. To solve this problem, it is necessary to design and introduce a new smart hospital resident training management and evaluation integrated system. Summary of the invention
[0006] To solve the above problems, the present invention provides an integrated system for resident training management and evaluation in a smart hospital. The system enhances the objectivity of the teaching teachers' scoring of resident training trainees and improves the accuracy and comprehensiveness of the evaluation. By integrating advanced artificial intelligence technology and big data analysis methods, the system can automatically and in real time monitor and evaluate the performance of trainees in clinical practice, effectively avoiding the subjective bias and human errors that may exist in traditional evaluation methods.
[0007] In order to achieve the above-mentioned object, the technical solution of the present invention is as follows: an integrated system for resident training management and evaluation in a smart hospital, comprising a data collection module, a trainee allocation module, an evaluation module, a behavior analysis module, a training formulation module and a smart hospital client module;
[0008] The data collection module is used to collect patient information and basic information of trainees; the data collection module includes a number of cameras arranged inside various departments of the hospital, and the cameras are used to capture the teaching practice behavior norms of trainees in the teaching practice to generate image data of the teaching practice behavior norms;
[0009] The trainee allocation module is used to use the trainee's basic information to allocate the trainees to various departments for rotation, and use the camera of the data acquisition module to recognize the trainee's face to generate the trainee's assessment information and the trainee rotation schedule;
[0010] The behavior analysis module is used to use the image data of the students' teaching practice behavior norms to score and evaluate the students' teaching practice behavior norms through the operation scoring algorithm, and generate a scoring report on the students' teaching practice behavior norms;
[0011] The training formulation module is used to automatically provide training content based on the trainees’ individual abilities and personal choices using the training resource pool;
[0012] The training formulation module also includes a feedback unit, which is used to collect students' teaching practice behavior standard score reports to automatically adjust the difficulty of training content and examination content;
[0013] The evaluation module is used to analyze data using student assessment information and student teaching practice behavior norms scoring reports to generate student personal evaluation reports;
[0014] The smart hospital client module is used to build a smart hospital client using the student rotation schedule, student personal evaluation report and student teaching practice behavior standard scoring report. Students use their personal accounts to log in to the smart hospital client.
[0015] Furthermore, the data collection module also includes a data encryption unit, which is used to encrypt the patient's medical record information.
[0016] Furthermore, the action scoring algorithm in the behavior analysis module includes the following steps:
[0017] Step 1: Input the image data of the students’ teaching practice behavior norms into the YOLOv5s model for training;
[0018] Among them, the YOLOv5s model extracts features from the image data of the students' teaching practice behavior norms through a convolutional neural network, and directly outputs the location and category information of the human body movements in the students' teaching practice behavior norms;
[0019] Step 2: Use the SimplePose human posture estimation algorithm to extract information about key points in the students' practical teaching behavior norms and generate key practical teaching behavior norm states;
[0020] The SimplePose human posture estimation algorithm uses a convolutional neural network to predict the key points of the human body in the image data of the students' teaching practice behavior norms; by analyzing the pixel information in the image, it identifies the key points of the human body and estimates the positions of the key points; the key point information is used to generate the key teaching practice behavior norms;
[0021] Step 3: Input the generated teaching practice behavior norm state into the ConvNeXt model to generate the feature vector, and extract the features through the convolution layer, normalization layer and activation function, and use the ConvNeXt model to output the probability distribution of the features;
[0022] Step 4: The behavior analysis module assigns a score to each key teaching practice behavior norm state based on the output probability distribution of the ConvNeXt model and the category with the highest probability, and obtains the student teaching practice behavior norm score report;
[0023] Among them, the ConvNeXt model uses a convolutional neural network to construct a deep network by stacking multiple convolutional layers, normalization layers and activation functions, gradually transforming and abstracting the input feature vector to extract a number of high-level features; during the training process, the ConvNeXt model maps high-level features to different key teaching practice behavior norms categories and outputs the probability distribution of each key teaching practice behavior norms category; in the evaluation stage, the ConvNeXt model assigns a score to each key teaching practice behavior norms state according to the output probability distribution.
[0024] Furthermore, the behavior analysis module includes a weight allocation unit, which is used to construct a judgment matrix by analyzing the personnel substructure, relative importance and influence degree of each operation factor in the trainees' teaching practice behavior norms; and calculate the weight of each operation factor, and allocate weights according to the importance of different indicators in the hospital.
[0025] Furthermore, the training resource pool in the training development module includes a large case resource library, a training course resource library, a simulated surgery resource library, and a bed information resource library.
[0026] Furthermore, trainee assessment information includes teaching standards, rotation status of each department, and compliance with management rate.
[0027] Furthermore, the smart hospital client module includes a student training management unit, which is used to manage students' personal information, teaching resources and attendance.
[0028] Furthermore, the smart hospital client includes a teacher client and a student client; the teacher client is used to log in using the teacher's personal account, and the student client is used to log in using the student's personal account.
[0029] Furthermore, the smart hospital client module also includes an interaction unit and a permission allocation unit;
[0030] Furthermore, the interactive unit is used to collect and process evaluation data of trainees from senior doctors, doctors of the same level, interns, nurses and patients. The evaluation data includes outpatient ability evaluation, bed management ability evaluation and trainee response evaluation to emergencies, and the evaluation data is sent to the evaluation module to generate a personal evaluation report; the authority allocation unit is used to allocate different permissions and access ranges to teacher clients and student clients.
[0031] Furthermore, the training development module also includes a bed allocation unit, which is used to allocate patient beds to trainees based on their rotation schedules, teaching practice behavior standard scoring reports, and trainees' personal assessment reports.
[0032] The above scheme has the following beneficial effects:
[0033] 1. Through the close cooperation of the data collection module and the behavior analysis module, the system can capture and analyze the behavioral norms of students in teaching practice in real time and generate detailed scoring reports. This not only helps students to understand their shortcomings in a timely manner and make self-improvements, but also provides an objective and comprehensive evaluation basis for teachers, so that they can guide students more targetedly and improve the efficiency and quality of training. At the same time, the training formulation module automatically adjusts the training content and difficulty according to the personal ability of the students and the scoring report, realizes personalized teaching, and further improves the training effect.
[0034] 2. In the present invention, the combination of the student allocation module and the evaluation module enables the hospital to manage the rotation plan and assessment information of the residents in a refined manner. Through the student rotation plan and personal evaluation report generated by the system, the hospital management can clearly understand the learning progress, clinical skills and comprehensive quality of each student, providing a scientific basis for the hospital's talent training and resource allocation. In addition, the system can also flexibly adjust the management strategy according to the actual needs of the hospital to improve the overall management efficiency of the hospital.
[0035] 3. In the present invention, the system enhances the objectivity of the teaching teachers' scoring of resident training students, and improves the accuracy and comprehensiveness of the evaluation; by integrating advanced artificial intelligence technology and big data analysis methods, the system can automatically and in real time monitor and evaluate the performance of students in clinical practice, effectively avoiding the subjective bias and human errors that may exist in traditional evaluation methods; at the same time, with the introduction of the smart hospital client module, teachers can view students' evaluation reports and scoring reports through the teacher client, and give timely teaching feedback and guidance; students can understand their own learning situation through the student client, obtain feedback on the quality of teaching services, and interact with teachers.
[0036] Additional aspects and advantages of the present invention will be given in part in the following description and in part will be obvious from the following description, or will be learned through practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS
[0037] Figure 1 This is a system structure diagram of the integrated system for resident training management and evaluation in a smart hospital in an embodiment of the present invention;
[0038] Figure 2 2 is a module structure diagram of the smart hospital client module in an embodiment of the present invention. DETAILED DESCRIPTION
[0039] The following is further described in detail through specific implementation methods:
[0040] Embodiment 1:
[0041] like Figure 1 and Figure 2 As shown: A smart hospital residency training management and evaluation integrated system, including a data collection module, an evaluation module, a behavior analysis module, a training formulation module and a smart hospital client module;
[0042] The data acquisition module is used to collect basic information of trainees and patient information; the data acquisition module includes several cameras arranged inside various departments of the hospital, and the cameras are used to capture the teaching practice behavior norms of trainees in teaching practice and generate image data of teaching practice behavior norms.
[0043] The data acquisition module also includes a data encryption unit, which is used to encrypt patient information.
[0044] The student allocation module is used to use the students' basic information to allocate students to various departments for rotation, use the camera of the data acquisition module to recognize the students' faces, generate student assessment information, and generate a student rotation schedule.
[0045] Trainee assessment information includes teaching standards, rotation status of each department and compliance rate.
[0046] The behavior analysis module is used to collect image data of students' teaching practice behavior norms using cameras, and to score and evaluate students' teaching practice behavior norms through an operation scoring algorithm to generate a scoring report on students' teaching practice behavior norms.
[0047] The action scoring algorithm in the behavior analysis module includes the following steps:
[0048] Step 1: Input the image data of students’ practical teaching behavior norms into the YOLOv5s model for training.
[0049] Among them, the YOLOv5s model uses a convolutional neural network to extract features from the image data of the students' practical teaching behavior norms, and directly outputs the location and category information of human body movements in the students' practical teaching behavior norms.
[0050] Step 2: Use the SimplePose human posture estimation algorithm to extract key point information from students’ practical teaching behavior norms and generate key practical teaching behavior norm states.
[0051] Among them, the SimplePose human posture estimation algorithm uses a convolutional neural network to predict the key point positions of the human body in the image data of the students' teaching practical behavior norms; by analyzing the pixel information in the image, it identifies the key points of the human body and estimates the positions of the key points; the key point information is used to generate the key teaching practical behavior norms status.
[0052] Step 3: Input the generated teaching practice behavior normative state into the feature vector of the ConvNeXt model, extract the features through the convolution layer, normalization layer and activation function, and use the ConvNeXt model to output the probability distribution of the features.
[0053] Step 4: The behavior analysis module assigns a score to each key teaching practice behavior norm state according to the output probability distribution of the ConvNeXt model and based on the category with the highest probability, and obtains the student teaching practice behavior norm score report.
[0054] Among them, the ConvNeXt model uses a convolutional neural network to build a deep network by stacking multiple convolutional layers, normalization layers and activation functions, and gradually transforms and abstracts the input feature vector to extract high-level features that are useful for classification tasks; during the training process, the ConvNeXt model maps high-level features to different key teaching practice behavior norms categories and outputs the probability distribution of each key teaching practice behavior norms category; in the evaluation stage, the ConvNeXt model will assign a score to each key teaching practice behavior norms state based on the output probability distribution.
[0055] The behavior analysis module includes a weight allocation unit, which is used to construct a judgment matrix by analyzing the personnel substructure, relative importance and influence degree of each operation factor in the trainees' teaching practice behavior norms; and calculate the weight of each operation factor, and allocate weights according to the importance of different indicators in the hospital.
[0056] The training formulation module is used to automatically provide training content and examination content based on the trainees' individual abilities and personal choices using the training resource pool.
[0057] The training development module also includes a feedback unit, which is used to collect students' teaching practice behavior standard scoring reports and test scores to automatically adjust the difficulty of training content and test content, and generate work efficiency data and learning efficiency data.
[0058] The evaluation module is used to perform data analysis using student assessment information and student teaching practice behavior standard scoring reports to generate student personal evaluation reports.
[0059] Among them, the training resource pool in the training development module includes large case resource library, training course resource library, simulated surgery resource library and bed information resource library.
[0060] For example, in a simulated surgery, a student performed a suturing operation for teaching practice. The camera captured the image data of his teaching practice behavior norms and input the image data of the teaching practice behavior norms into the behavior analysis module. The YOLOv5s model identified the use of suturing tools, and the SimplePose algorithm extracted the key point information of the student's hands and suturing parts, and constructed the key teaching practice behavior norms states. Then, the ConvNeXt model extracted features and scored these states, and finally generated the teaching practice behavior norms scoring report for the student.
[0061] The report pointed out that the student scored high in terms of stability and accuracy in the suturing process, but scored low in detail processing; at the same time, based on the student's personal assessment report and teaching practice behavior standard score report, the training development module automatically recommended relevant training content for him. Because the student lacked operational experience, the system provided him with more online learning courses and simulated surgery resources. At the same time, based on the deficiencies in his teaching practice behavior standard score report, the system also customized targeted exercises and examination content for him to help him improve his suturing skills.
[0062] The smart hospital client module is used to build a smart hospital client using the student rotation schedule, student personal evaluation report and student teaching practice behavior standard scoring report. Students use their personal accounts to log in to the smart hospital client.
[0063] The smart hospital client module also includes an interactive unit and a permission allocation unit; the teaching service quality feedback is displayed to the student client; the permission allocation unit is used to allocate different permissions and access ranges to the teacher client and the student client.
[0064] The training development module also includes a bed allocation unit, which is used to allocate patient beds to trainees based on their rotation schedule, teaching practice behavior standard scoring report and trainee personal assessment report.
[0065] For example, during a complete resident physician standardized training cycle, a trainee was assigned to the surgery department for rotation training. Through the camera in the data acquisition module, the system continuously captures the trainee's various behaviors in teaching practice, including surgical operations, ward rounds, and medical record writing. These image data are transmitted to the behavior analysis module in real time for detailed analysis and evaluation.
[0066] During the operation, the camera captured the entire process of the trainee's laparoscopic cholecystectomy. Through the YOLOv5s model, the system accurately identified the use of surgical instruments, including clamping, cutting, suturing and other actions, and located the trainee's position in front of the operating table. The SimplePose algorithm further extracted key point information of the trainee's hands and surgical site, such as the degree of bending of the fingers and the rotation angle of the wrist. This information was used to construct the key teaching practice behavior norms.
[0067] Subsequently, the ConvNeXt model extracted features and scored these states. The system generated a detailed teaching practice behavior standard scoring report based on the trainees' performance in terms of technique stability, operation accuracy, operation duration, and bleeding control during the operation. The report not only pointed out the trainees' strengths and weaknesses in surgical operations, but also provided suggestions for improvement and targeted training plans.
[0068] At the same time, the assessment module generates a comprehensive individual assessment report for students based on their assessment information and teaching practice behavior norms score report. This report covers the students’ learning attitude, clinical skills, teamwork, communication skills, and other aspects during the rotation, providing a strong basis for the students’ further development.
[0069] The training development module automatically recommends relevant training and examination content to students based on their personal assessment reports and teaching practice behavior norms scoring reports. In view of the students' deficiencies in surgical operations, the system provides them with online learning courses on laparoscopic surgery, simulated surgical resources, and expert guidance. In addition, the system also assigns them patient beds to be managed based on their rotation schedule and teaching practice behavior norms scoring reports, allowing them to further improve their clinical skills in practice.
[0070] In the smart hospital client module, students can log in to the smart hospital client through their personal account and view their rotation schedule, personal evaluation report and teaching practice behavior standard scoring report at any time. At the same time, the interactive unit provides the teacher client with the student's evaluation report and scoring report, which is convenient for teachers to understand the student's learning situation, provide feedback on the quality of teaching services, and communicate with students through the chat interactive platform. The permission allocation unit ensures that the teacher client and the student client have different permissions and access scopes, ensuring the security and privacy of information.
[0071] Example 2
[0072] The difference from the above embodiment is that the interactive unit is used to collect and process the evaluation data of the trainees from senior doctors, doctors of the same level, interns, nurses and patients, and send the evaluation data to the evaluation module to generate a personal evaluation report.
[0073] The evaluation data includes outpatient ability assessment, bed management ability assessment and trainee response assessment to emergencies.
[0074] In the standardized training of resident doctors, the 360-degree evaluation mechanism mainly evaluates the work behavior performance of residents from different angles, such as senior doctors, doctors of the same level, interns, nurses and patients who work closely with the residents, and encourages them to improve through analysis and feedback.
[0075] While implementing a 360-degree evaluation mechanism for resident physicians, residents should also conduct evaluations of their supervisors to promote the continuous improvement of the supervising physicians' teaching abilities.
[0076] The 360-degree evaluation mechanism can be carried out in various ways, such as paper quality scales, electronic scales, websites, apps and mini-programs. If there are a large number of people to be evaluated, we should make full use of information-based evaluation methods, which will be more conducive to the continuous promotion of this work.
[0077] Taking APP-based assessment as an example, after the assessment work is started, it should be implemented step by step.
[0078] The first step is to match the relationship. The administrator of the professional base or the rotation department needs to match the evaluation relationship for various types of personnel. Whether the evaluation relationship is correct has an important impact on the evaluation results. For training bases with conditions, it is recommended to adopt a management system that integrates rotation management and 360-degree evaluation mechanism. The evaluation relationship is determined when the instructor is assigned to the department, which is more conducive to simplifying the implementation process.
[0079] The second step is to notify the participants to fill in the form. The assessment administrator will notify the participants to log in to the system on their mobile phones and fill in the form within a certain period of time (recommended 3 to 5 days).
[0080] The third step is to supervise the completion. To ensure a certain completion rate, the assessment administrator must take various feasible ways to supervise.
[0081] In this embodiment, the 360-degree evaluation mechanism is used in the interactive unit of the smart hospital client. For example, when a trainee enters a department for rotation, the interactive unit will automatically match the trainee with evaluation subjects such as senior doctors, doctors of the same level, interns, nurses, and patients according to the trainee's rotation schedule and the department's personnel configuration, and set the corresponding evaluation authority and scope. When starting the evaluation, there is no need to manually match the evaluation relationship, which greatly improves the efficiency and accuracy of the evaluation work.
[0082] The interactive unit has built-in multiple assessment scales, including outpatient capacity assessment, bed management capacity assessment, and trainee response assessment to emergencies, etc. These scales can be selected and adjusted according to different assessment subjects and assessment purposes. At the same time, the content of the scale can also be customized according to the actual situation of the hospital and the training needs of the trainees to ensure the pertinence and effectiveness of the assessment.
[0083] In order to facilitate the evaluation subject to conduct evaluation anytime and anywhere, in this embodiment, the interactive unit links to the WeChat applet to fill in the scale. The evaluation subject only needs to log in to the WeChat applet within the specified time after receiving the evaluation notification, and can conduct an objective and fair evaluation of the trainee's work behavior performance based on his own observation and understanding.
[0084] After the assessment data is submitted, the assessment module will immediately summarize and analyze it to generate a personal assessment report for the trainee. These reports not only include the trainee's scores and rankings in each assessment dimension, but also provide detailed assessment feedback and suggestions to help trainees identify their strengths and weaknesses and develop targeted improvement plans. At the same time, the assessment data will also be used in the feedback unit of the training formulation module to automatically adjust the difficulty of the training content and examination content to achieve personalized and precise training.
[0085] In addition to evaluating trainees, the interactive unit also supports residents’ evaluation of their supervisors. This two-way evaluation mechanism not only helps improve the training effect of trainees, but also promotes the continuous improvement of the teaching ability of supervisors. Through regular evaluation and feedback, the hospital can promptly discover and solve problems and deficiencies in training, continuously optimize the training process and content, and improve the overall training quality and level.
[0086] The interactive unit of the smart hospital client has achieved a comprehensive, objective and fair evaluation of the work performance of resident trainees by introducing a 360-degree evaluation mechanism, providing strong training support and decision-making basis for the hospital. It also provides a comprehensive and objective evaluation of resident trainees, providing 360-degree support for the comprehensive growth of resident trainees.
[0087] Obviously, the above embodiments are merely examples for the purpose of clear explanation, and are not intended to limit the implementation methods. For those skilled in the art, other different forms of changes or modifications can be made based on the above description. It is not necessary and impossible to list all the implementation methods here. The obvious changes or modifications derived therefrom are still within the scope of protection of the invention.
Claims
1. An integrated system for resident training management and evaluation in smart hospitals, characterized by: It includes data collection module, student allocation module, assessment module, behavior analysis module, training formulation module and smart hospital client module; The data collection module is used to collect basic information of trainees and patient information; the data collection module includes a number of cameras arranged inside various departments of the hospital, and the cameras are used to capture the teaching practice behavior norms of trainees in the teaching practice to generate image data of the teaching practice behavior norms; The trainee allocation module is used to use the trainee's basic information to allocate the trainees to various departments for rotation, and use the camera of the data acquisition module to recognize the trainee's face to generate the trainee's assessment information and the trainee rotation schedule; The behavior analysis module is used to use the image data of the students' teaching practice behavior norms to score and evaluate the students' teaching practice behavior norms through the operation scoring algorithm, and generate a scoring report on the students' teaching practice behavior norms; The training formulation module is used to automatically provide training content based on the trainees’ individual abilities and personal choices using the training resource pool; The training formulation module also includes a feedback unit, which is used to collect students' teaching practice behavior standard score reports to automatically adjust the difficulty of training content and examination content; The evaluation module is used to analyze data using student assessment information and student teaching practice behavior norms scoring reports to generate student personal evaluation reports; The smart hospital client module is used to build a smart hospital client using the student rotation schedule, student personal evaluation report and student teaching practice behavior standard scoring report. Students use their personal accounts to log in to the smart hospital client.
2. The smart hospital residency training management and evaluation integrated system according to claim 1 is characterized in that: The data collection module also includes a data encryption unit, which is used to encrypt the patient's medical record information.
3. The smart hospital residency training management and evaluation integrated system according to claim 2 is characterized in that: The action scoring algorithm in the behavior analysis module includes the following steps: Step 1: Input the image data of the students’ teaching practice behavior norms into the YOLOv5s model for training; Among them, the YOLOv5s model extracts features from the image data of the students' teaching practice behavior norms through a convolutional neural network, and directly outputs the location and category information of the human body movements in the students' teaching practice behavior norms; Step 2: Use the SimplePose human posture estimation algorithm to extract information about key points in the students' practical teaching behavior norms and generate key practical teaching behavior norm states; The SimplePose human posture estimation algorithm uses a convolutional neural network to predict the key points of the human body in the image data of the students' teaching practice behavior norms; by analyzing the pixel information in the image, it identifies the key points of the human body and estimates the positions of the key points; the key point information is used to generate the key teaching practice behavior norms; Step 3: Input the generated teaching practice behavior norm state into the ConvNeXt model to generate the feature vector, and extract the features through the convolution layer, normalization layer and activation function, and use the ConvNeXt model to output the probability distribution of the features; Step 4: The behavior analysis module assigns a score to each key teaching practice behavior norm state based on the output probability distribution of the ConvNeXt model and the category with the highest probability, and obtains the student teaching practice behavior norm score report; Among them, the ConvNeXt model uses a convolutional neural network to construct a deep network by stacking multiple convolutional layers, normalization layers and activation functions, gradually transforming and abstracting the input feature vector to extract a number of high-level features; during the training process, the ConvNeXt model maps high-level features to different key teaching practice behavior norms categories and outputs the probability distribution of each key teaching practice behavior norms category; in the evaluation stage, the ConvNeXt model assigns a score to each key teaching practice behavior norms state according to the output probability distribution.
4. The smart hospital residency training management and evaluation integrated system according to claim 3 is characterized in that: The behavior analysis module includes a weight allocation unit, which is used to construct a judgment matrix by analyzing the personnel substructure, relative importance and influence degree of each operation factor in the trainees' teaching practice behavior norms; and calculate the weight of each operation factor, and allocate weights according to the importance of different indicators in the hospital.
5. The smart hospital residency training management and evaluation integrated system according to claim 4 is characterized in that: The training resource pool in the training development module includes large case resource library, training course resource library, simulated surgery resource library and bed information resource library.
6. The smart hospital residency training management and evaluation integrated system according to claim 5 is characterized in that: Trainee assessment information includes teaching standards, rotation status of each department and compliance rate.
7. The smart hospital residency training management and evaluation integrated system according to claim 6 is characterized in that: The smart hospital client module includes a student training management unit, which is used to manage students' personal information, teaching resources and attendance.
8. The smart hospital residency training management and evaluation integrated system according to claim 7 is characterized in that: The smart hospital client includes the teacher client and the student client; the teacher client is used to log in using the teacher's personal account, and the student client is used to log in using the student's personal account.
9. The smart hospital residency training management and evaluation integrated system according to claim 8 is characterized in that: The smart hospital client module also includes an interaction unit and a permission allocation unit; The interactive unit is used to collect and process the evaluation data of trainees from senior doctors, doctors of the same level, interns, nurses and patients. The evaluation data includes the evaluation of outpatient ability, the evaluation of bed management ability and the response of trainees to emergencies. The evaluation data is sent to the evaluation module to generate a personal evaluation report. The authority allocation unit is used to allocate different authorities and access scopes to the teacher client and the student client.
10. The smart hospital residency training management and evaluation integrated system according to claim 9 is characterized in that: The training development module also includes a bed allocation unit, which is used to allocate patient beds to trainees based on patient information, trainees' rotation schedules, teaching practice behavior code scoring reports, and trainees' personal assessment reports.