Simulation teaching and management system based on real medical record medical image file of radiology department

Abstract

The invention provides a simulation teaching and management system based on a radiology department real medical record medical image file, and the system comprises a file classification module which is used for collecting and classifying the radiology department real medical record medical image file; the sample extraction module is used for extracting a plurality of training samples from each type of real medical record medical image file of the radiology department; the model training module is used for training each training sample according to a decision tree algorithm; the model combination module is used for combining a plurality of classifiers to obtain a simulation teaching and management model when the training results all reach the standard; the scheme making module is used for inputting the learning labels of the students into the simulation teaching and management model to make a simulation teaching and management scheme; and the learning data evaluation module is used for recording the learning data of the students completing the simulation teaching and management schemes and evaluating the learning conditions of the students. According to the invention, a personalized teaching scheme is customized for each student, and the learning effect is automatically evaluated.

Description

Technical Field

[0001] The present application relates to the technical field of intelligent teaching and management of medical imaging archives. Specifically, the present application relates to a simulation teaching and management system based on real medical imaging archives of the radiology department. Background Art

[0002] PACS (Picture Archiving and Communication System), also known as image storage and transmission system, is a comprehensive application system for collecting, storing, managing, diagnosing and processing digital medical image information generated by digital medical equipment in hospitals.

[0003] The cases in the PACS system are all based on the real world, with complete clinical and imaging data. After being reviewed by the hospital's science and education department and the radiology teaching and research department and anonymized by the system, they are included in the teaching PACS system according to the disease system, teaching objectives and difficulty coefficient. The purpose of the teaching and research department's review is to ensure that all cases meet the requirements of the radiology training syllabus and are typical and / or rare cases with certain teaching purposes, which have a better teaching effect than general clinical cases. Students have their own PACS accounts, and in addition to completing their daily work, they complete the teaching plans formulated by the PACS system. However, most of the existing PACS systems are clinical work systems, and it is impossible to formulate teaching plans for each student according to the syllabus, disease system or difficulty classification, and it is impossible to meet the needs of students of different natures and levels.

[0004] For example, in the technical proposal with patent application number 202311182717.8, the internship function is integrated into the original PACS formal system, an internship account is established and corresponding permissions are assigned. After the inspection data enters the system, it can be operated by formal doctors and interns at the same time. The interns log in to the PACS formal system through the internship account to practice report writing. The completed report is set with an internship label and stored separately. The formal doctor enters the PACS formal system, writes a formal report and sets the formal label to store it in the original path. The reviewing doctor reviews it in the PACS formal system through label distinction; however, the PACS system cannot automatically formulate suitable teaching plans for students according to different teaching personnel (interns, clinical trainees, radiology trainees, ultrasound trainees), the difficulty of teaching medical records, and the disease system has clear requirements, purposes and difficulties, and the level of intelligence is low.

[0005] In the technical solution with patent application number 202211268259.5, it obtains medical imaging data and generates teaching videos through the management end for scientific researchers to conduct scientific research and learners to study. The teaching videos are studied through the learning end, and scientific research is conducted for scientific researchers through the scientific research end. Thus, the medical imaging data in the hospital PACS database can be used as a reference for teaching and scientific research, which is easy to use and improves the learning efficiency of medical personnel. However, it cannot achieve the purpose of industry-standard training, and it is still impossible to customize personalized teaching plans for each student and meet the needs of different students. Summary of the invention

[0006] The main purpose of this application is to provide a simulation teaching and management system based on real medical imaging archives of the radiology department, so as to customize personalized teaching plans for each student by using the simulation teaching and management model, thereby realizing personalized customization of teaching content and methods.

[0007] In order to achieve the above invention objectives, the present application provides a simulation teaching and management system based on real medical imaging archives of the radiology department, including:

[0008] The file classification module is used to collect the real medical image files of the radiology department, classify the real medical image files of the radiology department, and obtain various types of real medical image files of the radiology department;

[0009] A sample extraction module, used to extract a plurality of training samples from each type of the real medical image archives of the radiology department, each of which includes a real medical image of the radiology department and a corresponding reference diagnosis result;

[0010] A model training module, used to train the multiple training samples of each type according to the decision tree algorithm to obtain multiple classifiers and corresponding training results;

[0011] A model combination module, used for combining a plurality of the classifiers to obtain a simulation teaching and management model when it is determined that the training results of each classifier meet the standards;

[0012] A program formulation module is used to obtain a student file of a student, determine a learning tag of the student based on the student file, input the learning tag of the student into the simulated teaching and management model, formulate a simulated teaching and management program for the student using the simulated teaching and management model, and send the simulated teaching and management program to the student;

[0013] The learning data evaluation module is used to record the learning data of the trainees completing the simulated teaching and management program, and evaluate the learning situation of the trainees based on the learning data.

[0014] Furthermore, the system also includes a diagnostic report evaluation module, which is used to:

[0015] Obtaining a preset number of diagnostic reports written by the trainee;

[0016] Using the simulation teaching and management model to evaluate each of the diagnostic reports, and obtain an evaluation value of each of the diagnostic reports;

[0017] Compare the evaluation value of each of the diagnostic reports with a preset evaluation value of the corresponding diagnostic report type, and calculate the proportion of diagnostic reports having the evaluation value greater than the preset evaluation value;

[0018] When the ratio is greater than a preset ratio, determining that the simulated teaching quality of the student meets the standard, and generating an evaluation report;

[0019] The evaluation report is randomly sent to the trainee's tutor for revision. After the tutor completes the revision, the evaluation report is sent to the teaching and research office for review and filing.

[0020] Furthermore, the system also includes a student assessment module, which is used to:

[0021] When the number of diagnostic reports written by the student reaches a preset number, determining the student's assessment target, student grade and practice record;

[0022] Randomly generate assessment questions according to the assessment objectives, student grades and practice records, and set completion time for the assessment questions;

[0023] Assess the students according to the assessment questions and completion time, and generate assessment results;

[0024] Adjust the student's simulation teaching and management plan according to the assessment results.

[0025] Furthermore, the system also includes a model updating module, which is used to:

[0026] Regularly collect the real medical image files of the radiology department to obtain the candidate real medical image files of the radiology department;

[0027] After classifying the candidate radiology department real case medical imaging files, they are input into the simulation teaching and management model, and the simulation teaching and management model is iteratively updated.

[0028] Preferably, the learning data evaluation module is also used for:

[0029] Performing data preprocessing on the learning data, and screening out data related to the learning behavior from the learning data after the data preprocessing to obtain target learning data;

[0030] Converting the target learning data into a transaction list, wherein each transaction in the transaction list includes a series of items;

[0031] Perform frequency statistics on the items in each of the transactions and arrange them in descending order of frequency;

[0032] Insert each of the transactions into a pre-constructed frequent pattern tree in order of frequency of the items, if the items in the transaction already exist in the frequent pattern tree, increase the count of the transaction, otherwise, create a new node;

[0033] Recursively mining frequent item sets from the frequent pattern tree until no frequent item sets can be mined from the frequent pattern tree;

[0034] Generate association rules from the frequent item sets, check the relationship between each subset of the frequent item sets and the entire item set, and calculate confidence and lift, wherein the confidence refers to the conditional probability that the subsequent item also appears in the transaction where the preceding item appears, and the lift refers to the ratio of the probability of the subsequent item appearing to the probability of the subsequent item appearing itself under the condition of the given preceding item;

[0035] The validity of the association rule is evaluated according to the confidence and the improvement. After the evaluation is passed, the learning data is analyzed according to the association rule, and a learning evaluation report of the student is generated.

[0036] Preferably, the data preprocessing method includes removing noise, filling missing values, and data normalization.

[0037] Preferably, the model combination module is also used for:

[0038] According to the training results of each classifier, the loss value of each classifier is calculated based on a preset cross entropy loss function;

[0039] Determine whether the loss value of each of the classifiers is lower than a preset loss value;

[0040] If yes, it is determined that the training results of each of the classifiers meet the standard;

[0041] If not, the classifiers whose training results do not meet the standards are screened out to obtain candidate classifiers;

[0042] Using stochastic gradient descent method to adjust parameters of the candidate classifier;

[0043] A plurality of the training samples of one type are randomly selected to retrain the candidate classifier after adjusting the parameters until the loss value of the candidate classifier is lower than the preset loss value, thereby obtaining a classifier whose training results meet the standards.

[0044] Preferably, the model combination module is also used for:

[0045] Randomly extract a training sample from the multiple training samples to obtain a target training sample;

[0046] Calculate, according to the cross entropy loss function, the gradient of each original parameter of the candidate classifier under the condition of the target training sample;

[0047] Calculate the difference between the gradient of each of the original parameters with respect to the loss value corresponding to the candidate classifier and the gradient of each of the corresponding original parameters on the target training sample to obtain a gradient difference;

[0048] Calculate the difference between the logarithm of the gradient difference and the logarithm of the reference gradient difference to obtain a target gradient difference, wherein the reference gradient difference is 0.18;

[0049] The original parameters of the candidate classifier are continuously adjusted in the negative gradient direction according to the target gradient difference.

[0050] Preferably, the preset cross entropy loss function includes:

[0051]

[0052] Wherein, N is the number of training samples, i is the i-th training sample, e is the base of the natural logarithm, k is the diagnosis result predicted by the i-th training sample, and y is i,k is the vector of the diagnosis result k corresponding to the i-th training sample, and the p i,k The probability that the diagnosis k predicted for the i-th training sample is the reference diagnosis.

[0053] Furthermore, the system also includes a probability calculation module, which is used to:

[0054] Vectorize the diagnosis result k predicted by the i-th training sample to obtain a first vector;

[0055] Querying the reference diagnosis result corresponding to the i-th training sample, and performing vectorization processing on the reference diagnosis result corresponding to the i-th training sample to obtain a second vector;

[0056] The Euclidean distance between the first vector and the second vector is calculated to obtain the probability that the diagnosis result k predicted by the i-th training sample is the reference diagnosis result.

[0057] The application provides a simulation teaching and management system based on the real medical image archives of the radiology department. The archive classification module classifies the real medical image archives of the radiology department through automation technology, which improves the efficiency and accuracy of classification; the sample extraction module automatically extracts training samples from the classified medical image archives, reduces the workload of manual sample selection, and ensures the representativeness and diversity of the samples; the model training module uses a decision tree algorithm to train the extracted training samples, generates multiple classifiers to process a large amount of data, and can learn complex patterns from the data; after confirming that the training results of each classifier meet the standards, the model combination module combines them into a comprehensive simulation teaching and management model, and the integration method can improve the generalization ability and prediction accuracy of the model; the program formulation module uses the simulation teaching and management model to customize personalized teaching plans for each student according to the student's learning archives and labels, and realizes the personalization of teaching content and methods to meet the needs of different students. The learning data evaluation module records the student's learning data, and evaluates the student's learning situation based on these data, provides feedback for teaching, and helps teachers and students understand the learning progress and effect. The entire system improves the efficiency of medical imaging teaching and management through automation and intelligent methods, reduces the need for manual operations, and improves the quality of decision-making through data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 This is a schematic diagram of the structure of a simulation teaching and management system based on real medical imaging archives of the radiology department according to an embodiment of the present application;

[0059] Figure 2 This is a structural diagram of a simulation teaching and management system based on real medical imaging archives of the radiology department according to another embodiment of the present application;

[0060] Figure 3 This is a structural diagram of a simulation teaching and management system based on real medical imaging archives of the radiology department in another embodiment of the present application;

[0061] Figure 4 This is a structural diagram of a simulation teaching and management system based on real medical imaging archives of the radiology department according to another embodiment of the present application.

[0062] The realization of the purpose, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. DETAILED DESCRIPTION

[0063] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.

[0064] refer to Figure 1 As shown, in one embodiment, the present application provides a simulation teaching and management system based on real medical imaging archives of the radiology department, the system comprising:

[0065] The file classification module 11 is used to collect the real medical image files of the radiology department, classify the real medical image files of the radiology department, and obtain various types of real medical image files of the radiology department;

[0066] The sample extraction module 12 is used to extract a plurality of training samples from each type of the real medical image archives of the radiology department, each of which includes a real medical image of the radiology department and a corresponding reference diagnosis result;

[0067] A model training module 13 is used to train the plurality of training samples of each type according to a decision tree algorithm to obtain a plurality of classifiers and corresponding training results;

[0068] A model combination module 14 is used to combine multiple classifiers to obtain a simulation teaching and management model when it is determined that the training results of each classifier meet the standards;

[0069] A program formulation module 15 is used to obtain a student profile of a student, determine a learning tag of the student based on the student profile, input the learning tag of the student into the simulated teaching and management model, formulate a simulated teaching and management program for the student using the simulated teaching and management model, and send the simulated teaching and management program to the student;

[0070] The learning data evaluation module 16 is used to record the learning data of the trainees completing the simulated teaching and management program, and evaluate the learning situation of the trainees based on the learning data.

[0071] In this embodiment, the file classification module 11 of the system collects real medical image files of medical records from the radiology department, such as CT, MRI, X-ray, etc. At the same time, the image files are automatically classified according to the characteristics of disease type, body part, etc., and the image files can also be automatically classified according to the examination plan, the disease system to which it belongs, the difficulty of assessment, the classification of assessors, etc., and all system cases are anonymized, errors are corrected, whether the cases are available or typical, the difficulty coefficient score (easy-medium-difficult), and the real-time update and maintenance of the medical records in the database are processed, and finally various types of real medical image files of the radiology department are obtained.

[0072] From each category, for trainees of different grades and natures, the system's sample extraction module 12 can extract multiple training samples according to the disease system and difficulty division mode. Each training sample includes a real medical image of the radiology department and the corresponding reference diagnosis result. The reference diagnosis result includes an evaluation of the report writing content, the diagnostic accuracy, the completeness of the differential diagnosis, etc., which are used to train machine learning models to identify and diagnose specific diseases.

[0073] The model training module 13 uses a decision tree algorithm to iteratively train multiple training samples of each type, and after the training is completed, multiple classifiers are generated. The decision tree algorithm is a machine learning algorithm that imitates the human decision-making process, which predicts the classification or results of unknown data by learning data features and decision rules. The decision tree displays the decision process in the form of a tree diagram, in which each internal node represents a test on a feature, each branch represents the result of the test, and each leaf node represents a decision result or classification. In addition, the decision tree algorithm builds a model by learning the relationship between image features and diagnostic results.

[0074] When the training results of all classifiers reach the predetermined accuracy standard, the model combination module 14 combines all classifiers into a comprehensive simulation teaching and management model, for example, assigning different weights to each classifier, the weights are set based on the performance or prediction accuracy of the classifier, and then all classifiers are spliced ​​according to the weights. The simulation teaching and management model is used to formulate simulation teaching and management plans for trainees based on real medical records.

[0075] The program formulation module 15 of the system obtains the student's profile information, including learning content, learning time, student level, etc., and determines the student's learning tags, such as learning style, knowledge level, etc. According to the learning tags, the system uses the simulation teaching and management model to customize a personalized simulation teaching and management plan for each student, and sends the simulation teaching and management plan to the student, full-time tutor and teaching management personnel. At the same time, the simulation teaching and management plan is preliminarily evaluated by AI to generate an AI preliminary evaluation report. The simulation teaching and management plan includes simulation cases, simulation videos, exercises and learning plans, etc.

[0076] The learning data evaluation module 16 of the system records the interaction and performance of the students in the simulated teaching activities, including learning data such as diagnosis accuracy and learning time, and uses the learning data to evaluate the learning situation of the students, generate evaluation results, and use the evaluation results to formulate a suitable learning plan for the students. The learning data is used to evaluate the learning progress and effect of the students in order to further optimize the teaching plan.

[0077] For example, if a radiology resident wants to improve his ability to diagnose lung nodules, the system first collects the hospital's lung CT image archives from the past few years and divides them into two categories: "benign nodules" and "malignant nodules". From the "malignant nodules" category, the system extracts 200 CT images and their diagnostic results as training samples. Using the decision tree algorithm, the system trains a classifier that can identify malignant lung nodules. This classifier is combined with other disease classifiers in the system to form a comprehensive lung disease diagnosis model. The system analyzes the resident's learning archive and finds that he needs to improve in identifying lung nodules. Therefore, the system develops a teaching plan for him that focuses on lung nodule identification. After completing a series of simulated diagnostic training, the system records the physician's diagnostic accuracy and learning time. The evaluation results show that the physician's diagnostic ability has been significantly improved. Through this technical solution, radiologists can improve their diagnostic skills through a learning environment that simulates real cases, and teaching managers can evaluate and optimize teaching results based on learning data.

[0078] The application provides a simulation teaching and management system based on the real medical image archives of the radiology department. The archive classification module classifies the real medical image archives of the radiology department through automation technology, which improves the efficiency and accuracy of classification; the sample extraction module automatically extracts training samples from the classified medical image archives, reduces the workload of manual sample selection, and ensures the representativeness and diversity of the samples; the model training module uses a decision tree algorithm to train the extracted training samples, generates multiple classifiers to process a large amount of data, and can learn complex patterns from the data; after confirming that the training results of each classifier meet the standards, the model combination module combines them into a comprehensive simulation teaching and management model, and the integration method can improve the generalization ability and prediction accuracy of the model; the program formulation module uses the simulation teaching and management model to customize personalized teaching plans for each student according to the student's learning archives and labels, and realizes the personalization of teaching content and methods to meet the needs of different students. The learning data evaluation module records the student's learning data, and evaluates the student's learning situation based on these data, provides feedback for teaching, and helps teachers and students understand the learning progress and effect. The entire system improves the efficiency of medical imaging teaching and management through automation and intelligent methods, reduces the need for manual operations, and improves the quality of decision-making through data analysis.

[0079] In one embodiment, reference Figure 2 As shown, the system further includes a diagnostic report evaluation module 17, which is specifically used for:

[0080] Obtaining a preset number of diagnostic reports written by the trainee;

[0081] Using the simulation teaching and management model to evaluate each of the diagnostic reports, and obtain an evaluation value of each of the diagnostic reports;

[0082] Compare the evaluation value of each of the diagnostic reports with a preset evaluation value of the corresponding diagnostic report type, and calculate the proportion of diagnostic reports having the evaluation value greater than the preset evaluation value;

[0083] When the ratio is greater than a preset ratio, determining that the simulated teaching quality of the student meets the standard, and generating an evaluation report;

[0084] The evaluation report is randomly sent to the trainee's tutor for revision. After the tutor completes the revision, the evaluation report is sent to the teaching and research office for review and filing.

[0085] This embodiment implements a student evaluation process based on a simulated teaching and management system, which aims to evaluate the student's diagnostic report writing ability and clinical thinking through an automated and standardized method. Specifically, the student writes a certain number of diagnostic reports in a simulated teaching environment, and the diagnostic reports are based on simulated cases or virtual patients. The system monitors the number of diagnostic reports of the student in real time. When the number of diagnostic reports written by the student meets the standard, the simulated teaching and management model is used to evaluate the content of each diagnostic report and generate an evaluation value; when the number of diagnostic reports written by the student does not meet the standard, the student is automatically reminded of the gap between the number of diagnostic reports and the target.

[0086] Determine the type of the diagnostic report, query the preset evaluation value corresponding to the diagnostic report based on the type of the diagnostic report, compare the evaluation value of each diagnostic report with the preset evaluation standard of the corresponding diagnostic report type to determine whether the student's performance reaches the expected level, and calculate the proportion of diagnostic reports with evaluation values ​​greater than the preset evaluation value.

[0087] If the proportion of students who meet the standard exceeds the preset proportion, the student's simulation teaching quality is considered to meet the standard, and a formal evaluation report is generated. The evaluation report is randomly sent to the student's full-time tutor for revision to ensure the accuracy and fairness of the evaluation. The revised evaluation report is sent to the teaching and research office for final review and archived as a formal record of the student's learning outcomes, and the full-time tutor will update the student's learning plan based on the actual situation.

[0088] For example, suppose a medical student is participating in a simulation teaching project with the goal of improving his ability to write diagnostic reports. If the student writes 10 diagnostic reports on different cases in a simulation environment, the simulation teaching and management model evaluates each report separately, gives a score from 0 to 100, and sets the preset evaluation value to 70 points, and the compliance rate to 70%. If the student has 8 reports with a score of more than 70 points, because the compliance rate exceeds the preset 70%, the system generates an evaluation report for the student, deeming the student's simulation teaching quality to be up to standard, and sends the evaluation report to the student's instructor, who revises the student's evaluation report and makes some suggestions for improvement. The revised evaluation report is sent to the teaching and research department for final review and archived as the student's formal learning achievement record.

[0089] This embodiment can ensure the accuracy, consistency and standardization of the evaluation by analyzing the proportion of the diagnostic report meeting the standards and using model evaluation, and also facilitate students to timely understand their performance and areas for improvement; at the same time, the quality and reliability of the evaluation report are ensured through the review of the tutor and the teaching and research department; the archiving of the evaluation report also helps to record and track the learning progress and achievements of the students.

[0090] In one embodiment, reference Figure 3 As shown, the system further includes a student assessment module 18, which can be specifically used for:

[0091] When the number of diagnostic reports written by the student reaches a preset number, determining the student's assessment target, student grade and practice record;

[0092] Randomly generate assessment questions according to the assessment objectives, student grades and practice records, and set completion time for the assessment questions;

[0093] Assess the students according to the assessment questions and completion time, and generate assessment results;

[0094] Adjust the student's simulation teaching and management plan according to the assessment results.

[0095] In this embodiment, after the number of diagnostic reports written by the trainee reaches a preset number, the system will determine the trainee's assessment goal, grade and practice record, and the assessment goal includes upgrade assessment, monthly assessment or year-end assessment, etc. The system randomly generates assessment questions based on the trainee's assessment goal, grade and practice record, aiming to evaluate the trainee's knowledge and skills in a specific field. At the same time, a completion time is set for each assessment question to ensure the fairness and challenge of the assessment.

[0096] Students complete all assessment questions within the specified time, and the system records the students' answers and the time taken. The system evaluates the students' answers and generates assessment results, including scores, ratings or other forms of feedback, which are also fed back to the students, full-time tutors and teaching management personnel.

[0097] Based on the assessment results, the system adjusts the trainees’ simulation teaching and management plans to strengthen their weaknesses and improve their overall performance.

[0098] For example, suppose an intern at a medical school is preparing for an internship assessment in internal medicine. The trainee has completed a certain number of diagnostic reports. The system determines the assessment content based on the trainee's goals (such as improving the ability to diagnose heart disease), grade (such as third year), and practice records. The system generates a series of case analysis questions about heart disease, including electrocardiogram interpretation, medical history analysis, and treatment plan design. The completion time for each question is set to 30 minutes, with a total assessment time of 90 minutes. The trainee answered all questions in a simulated exam environment and submitted the answers within the specified time. The system evaluated the trainee's answers, gave scores and feedback, and pointed out the trainee's shortcomings in electrocardiogram interpretation. Finally, based on the assessment results, the system recommended more simulation training for electrocardiogram interpretation to the trainee and arranged an online consultation meeting with a heart disease expert.

[0099] Through the above technical solutions, the system can generate assessment questions according to the specific situation of the students, ensuring the personalization and pertinence of the learning content. Students can also understand their performance in the assessment in a timely manner, which helps to quickly identify and improve deficiencies. In addition, the simulation teaching and management program model can adjust the teaching content and methods in a timely manner according to the performance of the students, thereby improving the teaching effect and the learning outcomes of the students. The dynamic adjustment of the teaching plan also helps students learn more effectively and improves the efficiency and quality of learning. Finally, through automated assessment and evaluation, the objectivity and accuracy of the assessment are also improved.

[0100] In one embodiment, reference Figure 4 As shown, the system further includes a model updating module 19, which can be specifically used for:

[0101] Regularly collect the real medical image files of the radiology department to obtain the candidate real medical image files of the radiology department;

[0102] After classifying the candidate radiology department real case medical imaging files, they are input into the simulation teaching and management model, and the simulation teaching and management model is iteratively updated.

[0103] This embodiment implements a radiology medical image archive processing and teaching update system based on a simulation teaching and management model. This system regularly collects and classifies real medical image archives of radiology, and then uses these archives to iteratively update the simulation teaching and management model to improve the teaching quality and effect. Specifically, the model update module 19 of the system regularly collects real medical image archives of radiology, which may include image data obtained by various imaging technologies (such as X-ray, CT, MRI, etc.). Then the collected archives are classified, which can be classified based on standards such as disease type, image characteristics, body parts, etc., and the classified medical image archives are input into the simulation teaching and management model to iteratively update the simulation teaching and management model using the newly collected archives, thereby ensuring that the model can reflect the latest medical knowledge and technological progress.

[0104] This embodiment can provide teaching cases that are closer to actual clinical situations and improve students' learning effects by continuously updating the simulation teaching model; iterative updates help improve the diagnostic accuracy of the model and enable it to better simulate the doctor's diagnostic decision-making process.

[0105] In one embodiment, the learning data evaluation module 16 may also be specifically used for:

[0106] Performing data preprocessing on the learning data, and screening out data related to the learning behavior from the learning data after the data preprocessing to obtain target learning data;

[0107] Converting the target learning data into a transaction list, wherein each transaction in the transaction list includes a series of items;

[0108] Perform frequency statistics on the items in each of the transactions and arrange them in descending order of frequency;

[0109] Insert each of the transactions into a pre-constructed frequent pattern tree in order of frequency of the items, if the items in the transaction already exist in the frequent pattern tree, increase the count of the transaction, otherwise, create a new node;

[0110] Recursively mining frequent item sets from the frequent pattern tree until no frequent item sets can be mined from the frequent pattern tree;

[0111] Generate association rules from the frequent item sets, check the relationship between each subset of the frequent item sets and the entire item set, and calculate confidence and lift, wherein the confidence refers to the conditional probability that the subsequent item also appears in the transaction where the preceding item appears, and the lift refers to the ratio of the probability of the subsequent item appearing to the probability of the subsequent item appearing itself under the condition of the given preceding item;

[0112] The validity of the association rule is evaluated according to the confidence and the improvement. After the evaluation is passed, the learning data is analyzed according to the association rule, and a learning evaluation report of the student is generated.

[0113] In this embodiment, the learning data may be preprocessed first, including removing noise, filling missing values, and normalizing data to ensure data quality. Data related to the learning behavior is screened out from the preprocessed learning data as target learning data for analysis. For example, the learning data is matched with features describing the learning behavior to determine which learning data is related to the learning behavior and is used as the target learning data.

[0114] The target learning data is converted into a transaction list. Each transaction represents a learning event and contains a series of items (such as learning time, learning content, learning outcomes, etc.). The frequency of the items in the transaction is counted and arranged in descending order of frequency to form a frequency order to identify the most frequently occurring items. Each transaction is inserted into the frequent pattern tree in the frequency order of the corresponding items. During the insertion process, if the item in the transaction already exists in the frequent pattern tree, the count of the transaction is increased; otherwise, a new node is created and the transaction is inserted in the new node. Among them, the frequent pattern tree is a data structure for mining frequent item sets. The frequent item set refers to the item set whose number of occurrences in the data set exceeds a preset threshold (i.e., the minimum support).

[0115] Recursively mine frequent item sets from the frequent pattern tree until no frequent item sets can be mined from the frequent pattern tree, generate association rules from the frequent item sets, and calculate the confidence and lift of each rule. The effectiveness of the association rules is evaluated based on the confidence and lift. The confidence represents the conditional probability that the subsequent item also appears in the transaction where the preceding item appears; the lift represents the ratio of the probability of the subsequent item appearing to the probability of the subsequent item appearing itself under the condition of the given preceding item.

[0116] Based on the evaluation results, a learning evaluation report for the students is generated, which includes the analysis of the students’ learning behavior and suggestions.

[0117] This embodiment can discover the intrinsic connection between learning behaviors, such as the relationship between certain learning habits and learning outcomes, through association rule mining, and can provide students with personalized learning suggestions and intervention measures based on the analysis results of the association rules.

[0118] For example, suppose there is an online learning platform that collects students' behavior data on the learning platform, including learning time, learning frequency, video types watched, test scores completed, etc. The system first screens out data related to learning behavior, such as learning time, video types watched, and test scores, and converts each student's learning behavior into a transaction list, each transaction containing a series of items. A frequent pattern tree is constructed, and transactions are inserted into the frequent pattern tree to form frequent item sets. Frequent item sets are mined from the frequent pattern tree, such as "watching teaching videos" and "completing tests" often appear together. Association rules are generated, such as "if students watch teaching videos, they are more likely to complete tests." At the same time, the confidence and lift of the association rules are calculated to evaluate the effectiveness of the rules. Based on the association rules, a learning evaluation report for the students is generated. The report points out the relationship between the students' learning habits and learning outcomes, and provides improvement suggestions to better understand the students' learning behavior and provide personalized learning suggestions for students. At the same time, it also provides a basis for the model to optimize teaching strategies.

[0119] Preferably, the data preprocessing method includes removing noise, filling missing values, and data normalization.

[0120] Data preprocessing is a crucial step in data analysis, which involves cleaning, converting and organizing raw data to improve data quality and ensure that the data is suitable for subsequent analysis.

[0121] In this embodiment, noise refers to errors or interference information in the data, such as random pixels in an image, errors in sensor readings, etc. Noise removal can be achieved by using filters (such as Gaussian filtering, median filtering) or more advanced techniques (such as wavelet transform), which are not specifically limited here.

[0122] Missing data can be filled in a variety of ways, such as using the mean, median, mode, or predictive models based on other variables to estimate the missing value.

[0123] Data normalization refers to scaling the data so that it falls into a specific small interval, such as [0,1], to avoid deviations in features of different dimensions during the analysis process. This normalization method includes minimum-maximum normalization and Z-score normalization.

[0124] This embodiment can improve the accuracy of data analysis through data preprocessing. At the same time, normalization and noise removal can reduce the time and resources required for data analysis and improve the efficiency of data analysis by the system.

[0125] In one embodiment, the model combination module 14 may also be specifically used for:

[0126] According to the training results of each classifier, the loss value of each classifier is calculated based on a preset cross entropy loss function;

[0127] Determine whether the loss value of each of the classifiers is lower than a preset loss value;

[0128] If yes, it is determined that the training results of each of the classifiers meet the standard;

[0129] If not, the classifiers whose training results do not meet the standards are screened out to obtain candidate classifiers;

[0130] Using stochastic gradient descent method to adjust parameters of the candidate classifier;

[0131] A plurality of the training samples of one type are randomly selected to retrain the candidate classifier after adjusting the parameters until the loss value of the candidate classifier is lower than the preset loss value, thereby obtaining a classifier whose training results meet the standards.

[0132] In this embodiment, a cross entropy loss function may be used to calculate the loss value of each classifier. The cross entropy loss function is a loss function used in classification tasks to measure the difference between the probability distribution predicted by the model and the probability distribution of the true label.

[0133] The loss value of each classifier is compared with the preset loss value to determine whether the model has achieved the expected performance standard. The preset loss value can be customized, such as 0.3.

[0134] If the performance of the classifier is not up to standard, the parameters of the candidate classifier that does not meet the standard are adjusted using stochastic gradient descent (SGD). SGD is an optimization algorithm that updates the parameters of the model by randomly selecting samples in the parameter space, thereby reducing the loss value of the model.

[0135] Then, multiple training samples of one type are randomly selected and used to train the classifier with adjusted parameters again until the loss value of the candidate classifier is also lower than the preset loss value.

[0136] For example, after training the model, the cross entropy loss value of the model on the validation set is first calculated. If the loss value of the model is higher than the set threshold, it means that the performance of the model needs to be improved. Then the SGD method is used to adjust the parameters of the model, and one or a small number of image samples are randomly selected each time to update the weights of the model. After several rounds of parameter adjustment, the model is trained again using specific training samples until the loss value of the model is lower than the preset loss value.

[0137] This embodiment can improve the prediction accuracy and generalization ability of the model through continuous parameter adjustment and retraining. At the same time, compared with calculating the gradient of the entire training set, the SGD method only uses one or a small part of the samples to update the parameters each time, which can greatly reduce the consumption of computing resources. In addition, the SGD method can also update the parameters faster, thereby speeding up the training speed of the model. By continuously updating the parameters during the training process, the SGD method also helps the model better adapt to the training data while avoiding overfitting.

[0138] In one embodiment, the model combination module 14 may also be specifically used for:

[0139] Randomly extract a training sample from the multiple training samples to obtain a target training sample;

[0140] Calculate, according to the cross entropy loss function, the gradient of each original parameter of the candidate classifier under the condition of the target training sample;

[0141] Calculate the difference between the gradient of each of the original parameters with respect to the loss value corresponding to the candidate classifier and the gradient of each of the corresponding original parameters on the target training sample to obtain a gradient difference;

[0142] Calculate the difference between the logarithm of the gradient difference and the logarithm of the reference gradient difference to obtain a target gradient difference, wherein the reference gradient difference is 0.18;

[0143] The original parameters of the candidate classifier are continuously adjusted in the negative gradient direction according to the target gradient difference.

[0144] The model combination module of this embodiment randomly selects a training sample from multiple training samples as the target training sample for the subsequent parameter update process. According to the cross entropy loss function, the gradient of each original parameter of the target training sample under the current model parameters is calculated. The gradient represents the direction in which the derivative of a multivariable function at a given point is the largest, which indicates the direction in which the function value increases the fastest. In machine learning, especially when training models, gradients are used to guide how to adjust model parameters (such as weights and biases) to minimize the loss function.

[0145] Calculate the gradient of the loss function with respect to each parameter, which indicates how the change of each parameter affects the value of the loss function under the current parameters. At the same time, calculate the difference between the gradient of each parameter and the gradient of the corresponding parameter on the target training sample to obtain the gradient difference to evaluate the difference between the current parameter update direction and the ideal update direction. Calculate the difference between the logarithm of the gradient difference and the logarithm of the preset reference gradient difference (for example, 0.18) to obtain the target gradient difference, so as to convert the gradient difference into a more refined form that is easier for the system to process.

[0146] According to the target gradient difference, the original parameters of the model are adjusted in the negative gradient direction. This negative gradient direction is the direction in which the loss function decreases fastest, so updating the parameters in this direction can effectively reduce the loss value of the model.

[0147] This embodiment can improve the performance of the model on the training data by continuously optimizing the model parameters, thereby improving the prediction accuracy of the model. A suitable parameter update strategy can accelerate the speed at which the model converges to the minimum loss value and reduce the training time. At the same time, by continuously adjusting the parameters during the training process, the overfitting of the model to the training data can be reduced and the generalization ability of the model can be improved.

[0148] Preferably, the preset cross entropy loss function includes:

[0149]

[0150] Wherein, N is the number of training samples, i is the i-th training sample, e is the base of the natural logarithm, k is the diagnosis result predicted by the i-th training sample, and y is i,k is the vector of the diagnosis result k corresponding to the i-th training sample, and the p i,k The probability that the diagnosis k predicted for the i-th training sample is the reference diagnosis.

[0151] In one embodiment, the system further includes a probability calculation module, which can be specifically used to:

[0152] Vectorize the diagnosis result k predicted by the i-th training sample to obtain a first vector;

[0153] Querying the reference diagnosis result corresponding to the i-th training sample, and performing vectorization processing on the reference diagnosis result corresponding to the i-th training sample to obtain a second vector;

[0154] The Euclidean distance between the first vector and the second vector is calculated to obtain the probability that the diagnosis result k predicted by the i-th training sample is the reference diagnosis result.

[0155] This embodiment converts the predicted diagnosis result k of the i-th training sample into a numerical vector. This process is called vectorization. The vectorization process is used to convert text into a machine-processable format, such as using the word2vec word vectorization algorithm to vectorize the text-based diagnosis result k to obtain a first vector.

[0156] The reference diagnosis result corresponding to the i-th training sample is queried, and the reference diagnosis result is also vectorized using the word2vec word vectorization algorithm to obtain a second vector. The reference diagnosis result can be provided by experts or clinical experience data as a standard for evaluating the prediction accuracy of the model.

[0157] The Euclidean distance between two vectors is calculated to measure the straight-line distance between two points in multidimensional space. Based on the calculated Euclidean distance, the similarity between the predicted diagnosis result and the reference diagnosis result is evaluated. The smaller the Euclidean distance, the closer the predicted diagnosis result is to the reference diagnosis result, and the higher the accuracy of the model diagnosis.

[0158] For example, suppose the model analyzes a lung CT image and predicts the result as "lung adenocarcinoma", and converts this diagnosis result into vector A. The expert reference diagnosis result of the query for this image is also "lung adenocarcinoma", and this result is converted into vector B. The Euclidean distance between vector A and vector B is calculated to obtain a value representing the similarity between the two. If the Euclidean distance is very small, close to zero, then it can be considered that the model's prediction result is very close to the reference diagnosis and the model's performance is very good. Through this process, the accuracy of the diagnostic report generated by the machine learning model can be automatically evaluated, and the model parameters can be adjusted according to the evaluation results to improve the accuracy of future diagnoses.

[0159] This embodiment can achieve the accuracy of the quantitative diagnosis report by comparing the predicted results and the reference results, which helps to improve the quality of diagnosis; at the same time, through continuous evaluation and feedback of the diagnosis results, the model can be continuously optimized to improve the accuracy of future diagnoses.

[0160] In summary, the application provides a simulation teaching and management system based on the real medical image archives of the radiology department. The archive classification module classifies the real medical image archives of the radiology department through automation technology, which improves the efficiency and accuracy of classification; the sample extraction module automatically extracts training samples from the classified medical image archives, reduces the workload of manual sample selection, and ensures the representativeness and diversity of the samples; the model training module uses a decision tree algorithm to train the extracted training samples to generate multiple classifiers to process a large amount of data and can learn complex patterns from the data; after confirming that the training results of each classifier meet the standards, the model combination module combines them into a comprehensive simulation teaching and management model. This integrated method can improve the generalization ability and prediction accuracy of the model; the program formulation module uses the simulation teaching and management model to customize personalized teaching plans for each student according to the student's learning archives and labels, and realizes the personalization of teaching content and methods to meet the needs of different students. The learning data evaluation module records the student's learning data, and evaluates the student's learning situation based on these data, provides feedback for teaching, and helps teachers and students understand the learning progress and effect. The entire system improves the efficiency of medical imaging teaching and management through automation and intelligent methods, reduces the need for manual operations, and improves the quality of decision-making through data analysis.

[0161] It should be noted that, in this article, the terms "include", "comprises" or any other variations thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, device, article or method. In the absence of further restrictions, an element defined by the sentence "includes a ..." does not exclude the existence of other identical elements in the process, device, article or method including the element.

[0162] The above description is only a preferred embodiment of the present application, and does not limit the patent scope of the present application. Any equivalent structure or equivalent process transformation made using the contents of the present application specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present application.

Claims

1. A simulation teaching and management system based on real medical imaging archives of radiology department, characterized by: include: The file classification module is used to collect the real medical image files of the radiology department, classify the real medical image files of the radiology department, and obtain various types of real medical image files of the radiology department; A sample extraction module, used to extract a plurality of training samples from each type of the real medical image archives of the radiology department, each of which includes a real medical image of the radiology department and a corresponding reference diagnosis result; A model training module, used to train the multiple training samples of each type according to the decision tree algorithm to obtain multiple classifiers and corresponding training results; A model combination module, used for combining a plurality of the classifiers to obtain a simulation teaching and management model when it is determined that the training results of each classifier meet the standards; A program formulation module is used to obtain a student file of a student, determine a learning tag of the student based on the student file, input the learning tag of the student into the simulated teaching and management model, formulate a simulated teaching and management program for the student using the simulated teaching and management model, and send the simulated teaching and management program to the student; The learning data evaluation module is used to record the learning data of the trainees completing the simulated teaching and management program, and evaluate the learning situation of the trainees based on the learning data.

2. The system according to claim 1, characterized in that Also included is a diagnostic report evaluation module for: Obtaining a preset number of diagnostic reports written by the trainees; evaluating each of the diagnostic reports using the simulated teaching and management model to obtain an evaluation value of each of the diagnostic reports; Compare the evaluation value of each of the diagnostic reports with a preset evaluation value of the corresponding diagnostic report type, and calculate the proportion of diagnostic reports having the evaluation value greater than the preset evaluation value; When the ratio is greater than a preset ratio, determining that the simulated teaching quality of the student meets the standard, and generating an evaluation report; The evaluation report is randomly sent to the trainee's tutor for revision. After the tutor completes the revision, the evaluation report is sent to the teaching and research office for review and filing.

3. The system according to claim 2, characterized in that It also includes a student assessment module for: When the number of diagnostic reports written by the student reaches a preset number, determining the student's assessment target, student grade and practice record; Randomly generate assessment questions according to the assessment objectives, student grades and practice records, and set completion time for the assessment questions; Assess the students according to the assessment questions and completion time, and generate assessment results; Adjust the student's simulation teaching and management plan according to the assessment results.

4. The system according to claim 1, characterized in that Also included is a model updating module for: Regularly collect the real medical image files of the radiology department to obtain the candidate real medical image files of the radiology department; After classifying the candidate radiology department real case medical imaging files, they are input into the simulation teaching and management model, and the simulation teaching and management model is iteratively updated.

5. The system according to claim 1, characterized in that The learning data evaluation module is also used to: Performing data preprocessing on the learning data, and screening out data related to the learning behavior from the learning data after the data preprocessing to obtain target learning data; Converting the target learning data into a transaction list, wherein each transaction in the transaction list includes a series of items; Perform frequency statistics on the items in each of the transactions and arrange them in descending order of frequency; Insert each of the transactions into a pre-constructed frequent pattern tree in order of frequency of the items, if the items in the transaction already exist in the frequent pattern tree, increase the count of the transaction, otherwise, create a new node; Recursively mining frequent item sets from the frequent pattern tree until no frequent item sets can be mined from the frequent pattern tree; Generate association rules from the frequent item sets, check the relationship between each subset of the frequent item sets and the entire item set, and calculate confidence and lift, wherein the confidence refers to the conditional probability that the subsequent item also appears in the transaction where the preceding item appears, and the lift refers to the ratio of the probability of the subsequent item appearing to the probability of the subsequent item appearing itself under the condition of the given preceding item; The validity of the association rule is evaluated according to the confidence and the improvement. After the evaluation is passed, the learning data is analyzed according to the association rule, and a learning evaluation report of the student is generated.

6. The system according to claim 5, characterized in that The data preprocessing methods include removing noise, filling missing values, and data normalization.

7. The system according to claim 1, characterized in that The model combination module is also used to: According to the training results of each classifier, the loss value of each classifier is calculated based on a preset cross entropy loss function; Determine whether the loss value of each of the classifiers is lower than a preset loss value; If yes, it is determined that the training results of each of the classifiers meet the standard; If not, the classifiers whose training results do not meet the standards are screened out to obtain candidate classifiers; Using stochastic gradient descent method to adjust parameters of the candidate classifier; A plurality of the training samples of one type are randomly selected to retrain the candidate classifier after adjusting the parameters until the loss value of the candidate classifier is lower than the preset loss value, thereby obtaining a classifier whose training results meet the standards.

8. The system according to claim 7, characterized in that The model combination module is also used to: Randomly extract a training sample from the multiple training samples to obtain a target training sample; Calculate, according to the cross entropy loss function, the gradient of each original parameter of the candidate classifier under the condition of the target training sample; Calculate the difference between the gradient of each of the original parameters with respect to the loss value corresponding to the candidate classifier and the gradient of each of the corresponding original parameters on the target training sample to obtain a gradient difference; Calculate the difference between the logarithm of the gradient difference and the logarithm of the reference gradient difference to obtain a target gradient difference, wherein the reference gradient difference is 0.18; The original parameters of the candidate classifier are continuously adjusted in the negative gradient direction according to the target gradient difference.

9. The system according to claim 7, characterized in that The preset cross entropy loss function includes: Wherein, N is the number of training samples, i is the i-th training sample, e is the base of the natural logarithm, k is the diagnosis result predicted by the i-th training sample, and y is i,k is the vector of the diagnosis result k corresponding to the i-th training sample, and the p i,k The probability that the diagnosis k predicted for the i-th training sample is the reference diagnosis.

10. The system according to claim 9, characterized in that Also included is a probability calculation module for: Vectorize the diagnosis result k predicted by the i-th training sample to obtain a first vector; Querying the reference diagnosis result corresponding to the i-th training sample, and performing vectorization processing on the reference diagnosis result corresponding to the i-th training sample to obtain a second vector; The Euclidean distance between the first vector and the second vector is calculated to obtain the probability that the diagnosis result k predicted by the i-th training sample is the reference diagnosis result.

Images