A surgical simulation data analysis system and method based on PACS technology

Abstract

The present invention discloses a surgical simulation data analysis system and method based on PACS technology, which relates to the technical field of surgical simulation data analysis. The target surgical simulation data of the target medical staff is obtained from the surgical simulator, and the historical surgical simulation data of other medical staff is obtained from the central database; based on the surgical simulation data of the target medical staff, the probability of problems occurring in each link of the operation of the target medical staff is determined, and the link where the target medical staff has problems in the operation is determined; based on the link where the medical staff has problems in the operation, the medical image data required by the medical staff is determined and pre-downloaded from the central database to the terminal of the medical staff; the medical staff obtains the medical image data from the terminal to improve the proficiency in the operation; the data pre-download strategy is implemented to improve the overall work efficiency; and targeted medical image data pre-download is performed for each link of the operation to improve the effect of surgical simulation.

Description

Technical Field

[0001] The present invention relates to the technical field of surgical simulation data analysis, and specifically provides a surgical simulation data analysis system and method based on PACS technology. Background Art

[0002] The PACS system is a key technology for storing, managing, and transmitting medical image data in hospitals. It not only improves the storage and management efficiency of medical images but also provides technical support for surgical simulation and data analysis. In surgical simulation, the data generated by medical staff is multi-dimensional and of great value for improving surgical success rates, education, and research. The data generated by medical staff during surgical practice is crucial for improving surgical skills, evaluating surgical procedures, and optimizing surgical techniques. By obtaining the required medical image data from the PACS system, medical staff can perform surgical simulations more effectively and increase the success rate of actual surgeries. The PACS system stores all medical image data in a central database. As the data volume grows and user access requirements increase, the user experience is affected. Therefore, how to improve the efficiency of the PACS system and improve the user experience of medical staff has become an urgent problem to be solved. Summary of the Invention

[0003] The purpose of the present invention is to provide a surgical simulation data analysis system and method based on PACS technology to solve the problems raised in the above background art.

[0004] In one aspect of the present invention, a surgical simulation data analysis method based on PACS technology is provided, including:

[0005] S11, obtaining target surgical simulation data of a target medical staff from a surgical simulator and obtaining historical surgical simulation data of other medical staff from a central database;

[0006] S12, based on the surgical simulation data of the target medical staff, determining the probability of problems occurring in each link of the surgery for the target medical staff and determining the links in the surgery where the target medical staff has problems;

[0007] S13, based on the links in the surgery where the medical staff has problems, determining the medical image data required by the medical staff and pre-downloading it from the central database to the terminal of the medical staff;

[0008] S14, the medical staff obtains the medical image data from the terminal to improve the proficiency in the surgery.

[0009] In step S12, the determination of the links in the surgery where the target medical staff has problems further includes the following steps:

[0010] S21. Extract the input features of each surgical procedure from the target surgical simulation data and the historical surgical simulation data; annotate the j-th procedure during the historical surgical simulation. If the j-th procedure of the surgery is successfully completed, it is annotated as 0; otherwise, it is annotated as 1. Here, j is a positive integer between [1, m], and m is the number of procedures included in the surgery.

[0011] The annotation value reflects the probability of problems occurring in the procedure. For the historical surgical simulation data, whether there are problems in each procedure has been determined, so the result of the annotation value is 0 or 1. For the target surgical simulation data, an annotation of 1 indicates that there are problems, an annotation value of 0 indicates that no problems have been found for the time being, and a value between 0 and 1 indicates that there is a probability of problems occurring.

[0012] S22. For the j-th procedure of the surgery, train a binary classification neural network model. Divide the historical surgical simulation data into a training set and a test set. Use the input features of the j-th procedure in the training set as the input and the annotation value of the j-th procedure as the output to train m binary classification neural network models. Each binary classification neural network model corresponds to one procedure, and use the data in the test set to verify the binary classification neural network model.

[0013] The surgery includes m procedures, and there is a set of input features for each procedure. Train a binary classification model using the input features of the historical surgical simulation data, and use the obtained binary classification model to judge the input features of the target medical staff, so as to judge whether there are problems in each procedure of the target medical staff during the surgery. The input features include, but are not limited to, the duration of the procedure, the score of the procedure given by the surgical simulator or the supervisor, the number of times of triggering reminders, etc. This step obtains whether there will be problems in each procedure, and each procedure is judged separately directly without forming a whole.

[0014] S23. Input the input features of the target surgical simulation data in m procedures into the m binary classification neural network models respectively. Determine the procedures in which the target medical staff has problems during the surgery according to the output results. If the output result of the binary classification neural network model corresponding to the procedure is 1, there is a problem in the procedure; otherwise, there is no problem in the procedure.

[0015] In step S12, the step of determining the probability of problems occurring in each procedure of the target medical staff during the surgery further includes the following steps:

[0016] S31. Perform m unsupervised classifications on the input features of the target surgical simulation data in m procedures and the input features of the historical surgical simulation data in m procedures to determine the classification clusters to which the input features of the target surgical simulation data in all procedures belong. Let the average value of the annotation values of the historical surgical simulation data in the classification cluster be the annotation value of the target surgical simulation data in m procedures.

[0017] Step S31 is used to find historical surgical simulation data with a high similarity to the target medical staff's link. By taking the average value, the annotation values of the target surgical simulation data in m links are obtained, and the obtained annotation values are between 0 and 1. For the links determined to have problems in step S23, since it has been determined that there are problems, the annotation value is manually set to 1, and the annotation values of other parts remain unchanged.

[0018] S32. For the links where the target medical staff has problems during the operation, set the annotation value to 1, covering the result determined in step S31. Perform an unsupervised classification on the annotation values of the target surgical simulation data in m links and the annotation values of the historical surgical simulation data in m links to determine the target classification cluster to which the annotation values of the target surgical simulation data in m links belong. Let the historical surgical simulation data in the target classification cluster be the relevant simulation data, calculate the average value of the annotation values of the relevant simulation data in the j-th link, and obtain the probability P that the target medical staff has problems in the j-th link during the operation. 1j For the links where the target medical staff has problems during the operation, set the probability of having problems to 1.

[0019] Using the annotation values of all links, find historical surgical simulation data with a high similarity to the target medical staff for the entire operation to judge the probability that the target medical staff has problems in the operation links.

[0020] Due to the limited energy of the target medical staff and the limited storage space of the terminal, only part of the medical image data can be pre-downloaded to the medical staff's terminal. Set the data volume of the pre-downloaded medical image data, and the proportion of the medical image data in each link is the solution to be found. As a whole operation, any link is indispensable, so the contribution value compulsorily includes the probabilities of all links being completed.

[0021] In step S13, the step of determining the medical image data required by the medical staff and pre-downloading it from the central database to the medical staff's terminal further includes the following steps:

[0022] S41. Set the initial temperature T0, randomly select medical image data from m links as alternative medical image data, take the alternative medical image data as the initial solution, and determine the contribution value f0 corresponding to the initial solution. And take the initial solution as the current solution and the initial temperature as the current temperature.

[0023] S42. For the counting unit k = 1, 2,..., L, repeat steps S43 to S44; L is the number of cycles.

[0024] S43. By generating a perturbation based on the current solution, change the alternative medical image data, take the perturbed alternative medical image as the new solution, and determine the contribution value f corresponding to the new solution.

[0025] S44. Calculate the increment Δf of the contribution value brought by the new solution. If the increment Δf is less than 0, accept the new solution as the new current solution with a probability of 1. If the increment is not less than 0, accept the new solution as the new current solution with a probability, where T represents the current temperature;

[0026] S45. Lower the current temperature according to the temperature reduction scheme. If the current temperature is not less than the threshold, go to step S42; if the current temperature is less than the threshold, determine the alternative medical image data according to the current solution and pre-download the alternative medical image data to the terminal of the target medical staff.

[0027] In step S41, the contribution value f0 corresponding to the initial solution is generated by the following formula:

[0028] Let P 2j (0) represent the probability P of the target medical staff having problems in the j-th link of the operation after obtaining the pre-downloaded undisturbed alternative medical image data 1j after the change, then

[0029] According to the simulation results, the medical image data recommendation unit will recommend medical image data to the target medical staff, and determine the recommendation effect of the medical image data recommendation unit from the relevant simulation data as the coefficient E j , E j being 1 means that all the medical image data recommended by the medical image data recommendation unit are the medical image data required by the medical staff; for the same link, the higher the probability P 1j of the link having problems, the more medical image data the medical staff may need. Therefore, when the probability of the link having problems is 1, multiply the amount of medical image data required by P 1j to obtain the amount of medical image data required; D j (0) is the amount of data allocated to the link corresponding to the solution;

[0030] In step S43, the determination of the contribution value f corresponding to the new solution further includes the following steps:

[0031] Let P 2j (k) represent the probability P of the target medical staff having problems in the i-th link of the operation after obtaining the pre-downloaded alternative medical image data with the k-th disturbance 1j after the change, then

[0032] P 2j (0) is determined through the following steps:

[0033] Let the historical medical staff corresponding to the relevant simulation data be the relevant personnel, and determine the amount of medical imaging data C required by the target personnel in the j-th link from the relevant simulation data j Let the amount of medical imaging data in the j-th link in the alternative medical imaging data without perturbation be D j (0), then In the formula, E j is a coefficient;

[0034] P 2j (k) is determined through the following steps: Let the amount of medical imaging data in the j-th link in the alternative medical imaging data with the k-th perturbation be D j (k), then

[0035] C j is determined through the following steps: Determine the amount of medical imaging data required by the relevant personnel in the j-th link from the relevant simulation data, and take the average value μ of the amount of medical imaging data required by all relevant personnel in the j-th link j , and use μ j ×P 1j as the value of C j ; E j is determined according to the recommendation result of the medical imaging data recommendation unit, and is taken as the ratio of the total amount of data required by the relevant personnel in the data recommended by the medical imaging data recommendation unit to the total amount of data recommended by the medical imaging data recommendation unit

[0036] In another aspect of the present invention, there is provided a surgical simulation data analysis system based on PACS technology, including: a surgical simulator, a data storage module, a simulation data analysis module, and a pre-download module; the output end of the surgical simulator is connected to the input end of the data storage module, and is used to provide a virtual surgical environment and allow medical staff to practice surgical skills in the simulation environment; the output end of the data storage module is connected to the input end of the simulation data analysis module, and is used to store the surgical simulation data of medical staff and the medical imaging data generated during the actual surgical process; the output end of the simulation data analysis module is connected to the input end of the pre-download module, and based on the surgical simulation data of the target medical staff, determines the probability of problems occurring in each link of the surgery for the target medical staff, and determines the link where problems exist in the surgery for the target medical staff; the pre-download module is used to pre-fetch the medical imaging data that the target medical staff needs to access

[0037] The data storage module also includes a central database and a terminal; the central database is used to store all medical impact data and surgical simulation data, and to uniformly manage and maintain the data; the terminal is used to temporarily store the medical image data that the target medical staff needs to access. The simulation data analysis module also includes a problem detection unit, a labeling unit, an unsupervised classification unit, an optimization unit, and a medical image data recommendation unit; the problem detection unit is used to determine the link where the target medical staff has problems in the operation; the labeling unit is used to label the links in the operation; the unsupervised classification unit is used to determine the probability of problems in each link of the operation for the target medical staff; the optimization unit is used to determine the medical image data that needs to be downloaded to the terminal; the medical image recommendation unit recommends medical image data to the medical staff based on the surgical simulation data of the target medical staff. The optimization unit determines the medical image data that needs to be pre-downloaded to the terminal by simulated annealing, and selects medical image data from the central database based on the probability that no problems occur in all links of the operation.

[0038] Compared with the prior art, the beneficial effects achieved by the present invention are: by implementing a data pre-download strategy, the PACS system can provide faster data access speed, improve the experience of medical staff, and improve overall work efficiency; considering the time and energy of medical staff and the storage space limitations of the terminal, targeted medical imaging data pre-download is performed for each link of the operation, thereby improving the effect of surgical simulation and helping medical staff to better master surgical skills. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0040] Figure 1 It is a structural schematic diagram of a surgical simulation data analysis system based on PACS technology according to an embodiment of the present invention. DETAILED DESCRIPTION

[0041] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] In the embodiments of the present invention, please refer to Figure 1, provided is a surgical simulation data analysis system based on PACS technology, including: a surgical simulator, a data storage module, a simulation data analysis module, and a pre-download module; the output end of the surgical simulator is connected to the input end of the data storage module, which is used to provide a virtual surgical environment and allow medical staff to practice surgical skills in the simulation environment; the output end of the data storage module is connected to the input end of the simulation data analysis module, which is used to store the surgical simulation data of medical staff and the medical image data generated during the actual surgical process; the output end of the simulation data analysis module is connected to the input end of the pre-download module, and based on the surgical simulation data of the target medical staff, it determines the probability of problems occurring in each link of the operation for the target medical staff and determines the link where problems exist in the operation of the target medical staff; the pre-download module is used to pre-fetch the medical image data that the target medical staff needs to access. The data storage module further includes a central database and a terminal; the central database is used to store all medical impact data and surgical simulation data, and uniformly manage and maintain the data; the terminal is used to temporarily store the medical image data that the target medical staff needs to access. The simulation data analysis module further includes a problem detection unit, a labeling unit, an unsupervised classification unit, an optimization unit, and a medical image data recommendation unit; the problem detection unit is used to determine the link where problems exist in the operation of the target medical staff; the labeling unit is used to label the links in the operation; the unsupervised classification unit is used to determine the probability of problems occurring in each link of the operation for the target medical staff; the optimization unit is used to determine the medical image data that needs to be downloaded to the terminal; the medical image recommendation unit recommends medical image data for medical staff according to the surgical simulation data of the target medical staff. The optimization unit determines the medical image data that needs to be pre-downloaded to the terminal through the simulated annealing method, and selects medical image data from the central database based on the probability that no problems occur in all links of the operation.

[0043] In an embodiment of the present invention, provided is a surgical simulation data analysis method based on PACS technology, including:

[0044] S11, obtaining the target surgical simulation data of the target medical staff from the surgical simulator, and obtaining the historical surgical simulation data of other medical staff from the central database;

[0045] S12, based on the surgical simulation data of the target medical staff, determining the probability of problems occurring in each link of the operation for the target medical staff, and determining the link where problems exist in the operation of the target medical staff.

[0046] The determination of the link where problems exist in the operation of the target medical staff includes the following steps S21 to S23:

[0047] S21. Extract the input features of each surgical procedure from the target surgical simulation data and historical surgical simulation data; label the j-th procedure during the historical surgical simulation. If the j-th procedure of the surgery is successfully completed, label it as 0; otherwise, label it as 1. Here, j is a positive integer between [1, m], and m is the number of procedures included in the surgery.

[0048] S22. For the j-th procedure of the surgery, train a binary classification neural network model. Divide the historical surgical simulation data into a training set and a test set. Use the input features of the j-th procedure in the training set as the input and the labeled value of the j-th procedure as the output to train m binary classification neural network models. Each binary classification neural network model corresponds to one procedure, and verify the binary classification neural network model with the data in the test set.

[0049] S23. Input the input features of the target surgical simulation data in m procedures into the m binary classification neural network models respectively. Determine the procedures where the target medical staff have problems during the surgery according to the output results. If the output result of the binary classification neural network model corresponding to the procedure is 1, there is a problem with the procedure; otherwise, there is no problem with the procedure.

[0050] The training process of the binary classification neural network model is as follows:

[0051] 1. Load the historical surgical simulation data and standardize it, then divide it into a training set and a test set.

[0052] 2. Define the neural network architecture, specify the number of channels in the input layer as the number of input features, the number of channels in the output layer as 1, specify a neural network hidden layer and the number of neurons it contains. You can choose hidden layers such as LSTM and CNN. Add fully connected layers between the input layer and the neural network hidden layer, and between the neural network hidden layer and the output layer; add a softmax layer after the fully connected layer.

[0053] 3. Specify the training options, including the solver, number of training epochs, learning rate, and threshold.

[0054] 4. Train the binary classification neural network.

[0055] 5. Test the binary classification neural network model, classify the test data, and calculate the classification accuracy.

[0056] The steps for determining the probability of problems occurring in each procedure of the target medical staff during the surgery include the following steps S31 and S32:

[0057] S31. Perform unsupervised classification on the input features of the target surgical simulation data in m links and the input features of the historical surgical simulation data in m links for m times, determine the classification clusters to which the input features of the target surgical simulation data in all links belong, and use the average value of the annotation values of the historical surgical simulation data in the classification cluster as the annotation value of the target surgical simulation data in m links;

[0058] S32. For the links where the target medical staff have problems in the operation, set the annotation value to 1 to overwrite the result determined in step S31; perform unsupervised classification on the annotation values of the target surgical simulation data in m links and the annotation values of the historical surgical simulation data in m links once, determine the target classification cluster to which the annotation value of the target surgical simulation data in m links belongs, use the historical surgical simulation data in the target classification cluster as the relevant simulation data, calculate the average value of the annotation values of the relevant simulation data in the j-th link, and obtain the probability P that the target medical staff have problems in the j-th link of the operation. 1j For the links where the target medical staff have problems in the operation, set the probability of problems to 1.

[0059] S13. Based on the links where the medical staff have problems in the operation, determine the medical image data required by the medical staff and pre-download it from the central database to the terminal of the medical staff; including steps S41 to S45:

[0060] S41. Set the initial temperature T0, randomly select medical image data from m links as the alternative medical image data, use the alternative medical image data as the initial solution, determine the contribution value f0 corresponding to the initial solution; and use the initial solution as the current solution and the initial temperature as the current temperature.

[0061] Let P 2j (0) represents the probability P that the target medical staff have problems in the j-th link of the operation after obtaining the pre-downloaded alternative medical image data without perturbation. 1j The result after change, then

[0062] Let the historical medical staff corresponding to the relevant simulation data be the relevant personnel, and determine the amount C of medical image data required by the target personnel in the j-th link from the relevant simulation data. j Let the amount of medical image data in the j-th link of the alternative medical image data without perturbation be D j (0), then In the formula, E j is a coefficient;

[0063] S42. For the counting unit k = 1, 2,..., L, repeat steps S43 to S44; L is the number of cycles;

[0064] S43. By generating perturbations based on the current solution to change the alternative medical image data, taking the perturbed alternative medical image as the new solution, and determining the contribution value f corresponding to the new solution;

[0065] Let P 2j (k) denote the probability P that problems occur to the target medical staff in the i-th link of the operation after obtaining the alternative medical image data for generating the k-th perturbation for pre-downloading. 1j For the changed result, then

[0066] Let the amount of medical image data in the j-th link of the alternative medical image data for generating the k-th perturbation be D j (k), then

[0067] Determine the amount of medical image data required by relevant personnel in the j-th link from relevant simulation data, and take the average value μ of the amounts of medical image data required by all relevant personnel in the j-th link. j , and take μ j ×P 1j as the value of C j ; E j Determine according to the recommendation result of the medical image data recommendation unit, and take the ratio of the total amount of data required by relevant personnel in the data recommended by the medical image data recommendation unit to the total amount of data recommended by the medical image data recommendation unit;

[0068] S44. Calculate the increment Δf of the contribution value brought by the new solution. If the increment Δf is less than 0, accept the new solution as the new current solution with a probability of 1. If the increment is not less than 0, then accept the new solution as the new current solution with a probability, where T represents the current temperature;

[0069] S45. Lower the current temperature according to the temperature reduction scheme. If the current temperature is not less than the threshold, go to step S42; if the current temperature is less than the threshold, determine the alternative medical image data according to the current solution, and pre-download the alternative medical image data to the terminal of the target medical staff.

[0070] Optionally, use a linear temperature reduction curve to lower the temperature; when the initial temperature is set to 100, if a linear temperature reduction curve with a temperature reduction coefficient of 0.95 is used for temperature reduction, the temperature changes from 100 to 95 after the first iteration.

[0071] S14. The medical staff obtains medical image data from the terminal to improve the proficiency in the operation.

[0072] It should be noted that in this document, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the term "comprising", "including" or any other variant thereof is intended to cover non-exclusive inclusion, such that a process, method, article or device comprising a series of elements not only includes those elements but also includes other elements not expressly listed, or elements inherent to such process, method, article or device.

[0073] Finally, it should be noted that the above are only preferred embodiments of the present invention and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or perform equivalent replacements for some of the technical features. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A surgical simulation data analysis method based on PACS technology, characterized in that: The following steps are involved: S11, obtaining target surgical simulation data of target medical staff from the surgical simulator, and obtaining historical surgical simulation data of other medical staff from the central database; S12, based on the surgical simulation data of the target medical staff, determining the probability of problems occurring in various links of the surgery by the target medical staff, and determining the links where the target medical staff have problems during the surgery; S13, based on the problems encountered by the medical staff during the operation, determining the medical imaging data required by the medical staff and pre-downloading the medical imaging data from the central database to the terminal of the medical staff, including the following steps: S41, setting an initial temperature T0, randomly selecting medical image data from m links as candidate medical image data, taking the candidate medical image data as an initial solution, determining a contribution value f0 corresponding to the initial solution; and taking the initial solution as the current solution, and taking the initial temperature as the current temperature; S42, for counting units k=1, 2, ..., L, repeat steps S43 to S44; L is the number of cycles; S43, generating disturbances on the basis of the current solution to change the candidate medical image data, taking the disturbed candidate medical image as a new solution, and determining a contribution value f corresponding to the new solution; S44, calculate the increment Δf of the contribution value brought by the new solution. If the increment Δf is less than 0, the new solution is accepted as the new current solution with probability 1. If the increment is not less than 0, The probability of accepting the new solution as the new current solution is , where T represents the current temperature; S45, lowering the current temperature according to the cooling solution. If the current temperature is not less than the threshold, proceeding to step S42; if the current temperature is less than the threshold, determining candidate medical imaging data according to the current solution, and pre-downloading the candidate medical imaging data to the terminal of the target medical staff; In step S41, the contribution value f0 corresponding to the initial solution is generated by the following formula: Let P 2j (0) represents the probability P that the target medical staff will have problems in the jth step of the operation after obtaining the pre-downloaded non-disturbanced candidate medical image data. 1j The result after the change is In step S43, the step of determining the contribution value f corresponding to the new solution further includes the following steps: Let P 2j (k) represents the probability P that the target medical staff will have problems in the i-th link of the operation after obtaining the pre-downloaded candidate medical image data that produces the k-th disturbance. 1j The result after the change is S14, medical staff obtain medical imaging data from the terminal to improve their proficiency in surgery.

2. The method for analyzing surgical simulation data based on PACS technology according to claim 1, characterized in that: In step S12, the step of determining whether the target medical staff has problems during the operation also includes the following steps: S21, extracting input features of each link of the surgery from the target surgery simulation data and the historical surgery simulation data; marking the jth link of the surgery in the historical surgery simulation data, if the jth link of the surgery is successfully completed, it is marked as 0, otherwise it is marked as 1; where j is a positive integer between [1, m], and m is the number of links included in the surgery; S22, for the jth step of the surgery, train a binary classification neural network model, divide the historical surgery simulation data into a training set and a test set, use the input features of the jth step of the surgery in the training set as input, and the labeled value of the jth step as output, train m binary classification neural network models, each binary classification neural network model corresponds to one step, and verify the binary classification neural network model with the data of the test set; S23, input the input features of the target surgical simulation data in m links into m binary classification neural network models respectively, and determine the link where the target medical staff has problems during the operation based on the output results. If the output result of the binary classification neural network model corresponding to the link is 1, there is a problem in the link, otherwise there is no problem in the link.

3. The method for analyzing surgical simulation data based on PACS technology according to claim 2, characterized in that: In step S12, the process of determining the probability of problems occurring in various stages of the operation by the target medical staff further includes the following steps: S31, performing m unsupervised classifications on the input features of the target surgical simulation data at m stages and the input features of the historical surgical simulation data at m stages, determining the classification clusters to which the input features of the target surgical simulation data at all stages belong, and taking the average of the labeled values ​​of the historical surgical simulation data in the classification clusters as the labeled values ​​of the target surgical simulation data at m stages; S32, for the link where the target medical staff has problems during the operation, set the labeling value to 1, overwriting the result determined in step S31; perform an unsupervised classification on the labeling values ​​of the target surgical simulation data at m links and the labeling values ​​of the historical surgical simulation data at m links, determine the target classification cluster to which the labeling values ​​of the target surgical simulation data at m links belong, let the historical surgical simulation data in the target classification cluster be the relevant simulation data, calculate the average value of the labeling values ​​of the relevant simulation data at the jth link, and obtain the probability P that the target medical staff has problems at the jth link of the operation 1j ,For the links where the target medical staff have problems during the surgery, the probability of problems occurring is set to 1.

4. The method for analyzing surgical simulation data based on PACS technology according to claim 3, characterized in that: P 2j (0) Determine by the following steps: Let the historical medical staff corresponding to the relevant simulation data be the relevant personnel, and determine the amount of medical imaging data C required by the target personnel in the jth link from the relevant simulation data. j , let the amount of medical image data of the jth link in the candidate medical image data without disturbance be D j (0), then Where E j is the coefficient; P 2j (k) Determine by the following steps: Let the amount of medical image data of the jth link in the candidate medical image data that produces the kth disturbance be D j (k), then C j The determination is carried out through the following steps: determine the amount of medical imaging data required by the relevant personnel in the jth link from the relevant simulation data, and take the average value μ of the amount of medical imaging data required by all relevant personnel in the jth link j , μ j ×P 1j As C j The value of E j The determination is made based on the recommendation result of the medical imaging data recommendation unit, and the ratio of the total amount of data required by the relevant personnel in the data recommended by the medical imaging data recommendation unit to the total amount of data recommended by the medical imaging data recommendation unit is taken.

5. A surgical simulation data analysis system based on PACS technology, using a surgical simulation data analysis method based on PACS technology as claimed in any one of claims 1 to 4, characterized in that: include: A surgical simulator, a data storage module, a simulation data analysis module and a pre-download module; the output end of the surgical simulator is connected to the input end of the data storage module to provide a virtual surgical environment, allowing medical staff to practice surgical skills in a simulated environment; the output end of the data storage module is connected to the input end of the simulation data analysis module to store the surgical simulation data of the medical staff and the medical imaging data generated during the actual operation; the output end of the simulation data analysis module is connected to the input end of the pre-download module to determine the probability of problems occurring in various links of the operation of the target medical staff and the links where problems exist in the operation of the target medical staff based on the surgical simulation data of the target medical staff; the pre-download module is used to pre-fetch the medical imaging data that the target medical staff needs to access.

6. The surgical simulation data analysis system based on PACS technology according to claim 5, characterized in that: The data storage module also includes a central database and a terminal; the central database is used to store all medical impact data and surgical simulation data, and to manage and maintain the data in a unified manner; the terminal is used to temporarily store medical imaging data that target medical staff need to access.

7. The surgical simulation data analysis system based on PACS technology according to claim 5, characterized in that: The simulation data analysis module also includes a problem detection unit, a labeling unit, an unsupervised classification unit, an optimization unit and a medical image data recommendation unit; the problem detection unit is used to determine the links where problems exist in the operation of the target medical staff; the labeling unit is used to label the links in the operation; the unsupervised classification unit is used to determine the probability of problems occurring in each link of the operation of the target medical staff; the optimization unit is used to determine the medical image data that needs to be downloaded to the terminal; the medical image recommendation unit recommends medical image data to the medical staff based on the surgical simulation data of the target medical staff.

8. The surgical simulation data analysis system based on PACS technology according to claim 7, characterized in that: The optimization unit determines the medical imaging data that needs to be pre-downloaded to the terminal by means of simulated annealing, and selects the medical imaging data from the central database based on the probability that no problems occur in all links of the operation.

Images