Minimally invasive surgery skill training system and evaluation method thereof

Abstract

The invention relates to the field of medical simulation training. The invention discloses a minimally invasive surgical skill training system and an evaluation method thereof, and the system comprises a recognition and collection module which is used for collecting doctor information and recognizing and extracting surgical features, a storage module which is used for storing the doctor surgical information and the surgical features, and a data processing and analysis module which is used for analyzing, processing and fitting the surgical features. The three-dimensional module is used for converting the analyzed, processed and fitted surgical features into a three-dimensional model; and the display simulation module is used for displaying the three-dimensional model. Multi-doctor operation characteristics are collected, data are separated according to time and success rate, a dynamic operation model is generated through fitting of a genetic algorithm and the like, and individuation and standardization are combined. The sensing module is used for collecting data, constructing a three-dimensional model and providing force feedback. Through dual-mode and dual-threshold grading evaluation intervention, success rate constraint is introduced to guarantee the quality of key steps, and the efficiency is optimized for time-limited surgery.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of medical simulation training, and particularly to a minimally invasive surgical skill training system and an evaluation method thereof. BACKGROUND

[0002] Minimally invasive surgery has become an important means of modern clinical treatment due to its advantages of small trauma and fast recovery. However, it has extremely high requirements for operation precision, instrument control and spatial perception ability, and standardized training of surgical skills is the key to ensuring surgical safety and efficacy. Traditional surgical training mainly relies on fixed standard procedures or single expert experience. On the one hand, the training standard is single and lagging. Existing systems are mostly based on pre-set standardized operation models, which are difficult to integrate the latest clinical practice and multiple expert experiences in real time. With the rapid development of surgical technology, the operation differences of different schools, surgical methods and complex cases are significant, and fixed standards cannot meet the needs of personalized training, resulting in a disconnection between the operation habits of students and actual clinical scenes. On the other hand, the data utilization and simulation accuracy are insufficient. Traditional training systems lack the ability to deeply mine massive surgical data, and cannot effectively extract surgical features of multiple doctors and intelligently reorganize them. In the operation simulation link, the collection and feedback of key parameters such as the position, speed and pressure of surgical instruments are not accurate enough, and it is difficult to restore the mechanical interaction and spatial operation experience in real surgery, so that students are difficult to form effective muscle memory and risk prediction ability in the virtual environment. In addition, the evaluation system is highly subjective and has poor adaptability. Existing evaluation methods mostly rely on subjective evaluation of instructors, lack quantitative analysis of parameters such as operation trajectory and pressure fluctuation, and cannot accurately locate the operation weaknesses of students. For different types of surgery, traditional systems are difficult to dynamically adjust training targets, resulting in insufficient training targeting, and students face the challenge of balancing operation risk and efficiency in real surgery. The existing technology has obvious defects in the multi-element fusion of surgical features, the accuracy of simulation training and the flexibility of the evaluation system. SUMMARY

[0003] In order to solve the defects of the prior art, the present application discloses a minimally invasive surgical skill training system and an evaluation method thereof, which can integrate massive clinical data, generate dynamic surgical models and provide real-time intelligent feedback.

[0004] The present application discloses a minimally invasive surgical skill training system and an evaluation method thereof, which comprises an identification and collection module, the identification and collection module is used for collecting doctor operation information and identifying and extracting surgical features;

[0005] A storage module is connected to the identification and collection module, and the storage module is used for storing the doctor operation information and the surgical features.

[0006] a data processing and analysis module, the data processing and analysis module being connected to the storage module, and the data processing and analysis module being configured to analyze and fit the surgical characteristics stored in the storage module;

[0007] a three-dimensional module, the three-dimensional module being connected to the data processing and analysis module, and the three-dimensional module being configured to convert the analyzed and fitted surgical characteristics into a three-dimensional model;

[0008] a display simulation module, the display simulation module being connected to the three-dimensional module, and the display simulation module being configured to display the three-dimensional model.

[0009] Further, the data processing and analysis module further comprises a characteristic separation and fitting module, the characteristic separation and fitting module being connected to the storage module, and the characteristic separation and fitting module being configured to separate the surgical characteristics of each of the plurality of doctors into a plurality of surgical characteristics, and then fit the surgical characteristics of each of the plurality of doctors according to a set condition to form a new surgical model.

[0010] Further, the surgical characteristics of each of the plurality of doctors are separated into a plurality of surgical characteristics, and the separation condition is divided into a time-oriented separation and a success rate-oriented separation.

[0011] The time-oriented separation separates the surgical characteristics by time nodes, and the success rate-oriented separation separates the surgical characteristics according to the success rate in the surgical process.

[0012] Further, the plurality of surgical characteristics of each of the plurality of doctors are fitted according to a set condition, and the set condition is at least one of a surgical genre, a surgical time, a surgical success rate, and a movement trajectory of a surgical knife in a surgical process.

[0013] Further, the fitting method is performed by an algorithm, and the algorithm includes at least one of a genetic algorithm, a simulated annealing algorithm, an ant colony algorithm, a particle swarm optimization algorithm, and deep reinforcement learning.

[0014] Further, the constraint condition of the set condition includes a success rate threshold, and after the plurality of surgical characteristics are fitted according to the set condition, the success rate of the entire surgical characteristics is greater than the success rate threshold, and the success rate threshold is set according to the surgical historical data.

[0015] Further, the recognition and collection module further comprises a surgical module and a force feedback module, the surgical module comprising a position display sub-module capable of determining the spatial position of the surgical module, a motion measurement sub-module for detecting the speed and trajectory of the surgical module, a plurality of pressure measurement sub-modules arranged at one end of the surgical module, and an optical detection sub-module for detecting the inclination angle of the surgical module, the surgical module being connected to the recognition and collection module for the recognition and collection module to collect surgical information, and the force feedback module being configured to provide mechanical feedback for the surgical module.

[0016] Further, S1: the recognition collection module collects a plurality of surgeon operation information and recognizes and extracts operation characteristics and stores them into a storage module;

[0017] S2: the data processing and analysis module analyzes and processes the operation characteristics and fitting, and the three-dimensional module converts the analyzed and processed operation characteristics into a three-dimensional model;

[0018] S3: the display simulation module displays the three-dimensional modeling, and the student selects a mode and then operates the operation module on the force feedback module through the display of the display simulation module;

[0019] S4: the motion measurement sub-module, the pressure measurement sub-module and the optical detection sub-module in the operation module transmit the motion trajectory, the operation pressure and the tilt angle of the student's operation to the display simulation module, and the data processing and analysis module analyzes and processes the motion trajectory, the operation pressure and the tilt angle of the student's operation in real time;

[0020] S5: when the motion trajectory and the operation pressure of the student are within the motion trajectory fluctuation threshold or the pressure threshold, it is judged that the student's operation is standard, and the student continues to operate, when the motion trajectory and the operation pressure of the student are outside the motion trajectory fluctuation threshold or the pressure threshold, it is judged that the student's operation is not standard, the student selects the examination mode to perform S6, and selects the learning mode to perform S7;

[0021] S6: examination mode, waiting for the student to complete the operation, recording the non-standard operation of the student in the whole operation process and scoring the whole operation of the student, and recording the non-standard operation, evaluating whether the student needs to strengthen the training or the ordinary training;

[0022] S7: learning mode, when the pressure and the motion trajectory exceed the pressure threshold or the motion trajectory fluctuation threshold, but do not exceed the pressure revision threshold or the motion trajectory fluctuation revision threshold, the student is not corrected, but S6 is performed, when the pressure and the motion trajectory exceed the pressure revision threshold or the motion trajectory fluctuation revision threshold, the display simulation module distinguishes and color labels the abnormal area of the pressure and the motion trajectory on the displayed three-dimensional model, the display simulation module is paused and prompts the student, and the student is given the option of continuing the training or performing the error training, the student continues the training to repeat S6, and the student performs the error training to perform S8;

[0023] S8: After the selection of the wrong training, the three-dimensional model displayed by the simulation module is returned to the state before the trainee's wrong operation, and a countdown prompt is given, the trainee re-performs the wrong training, at least several times, each time compared with the standard training and scored, and the average score is taken, when the average score is higher than or equal to the qualified score value, the trainee is judged to be qualified, when the average score is lower than the qualified score value, S8 is repeated, and the maximum number of repeated attempts is allowed, if the trainee still fails to meet the standard, S9 is forced to enter and the failure is recorded;

[0024] S9: The trainee's training is completed.

[0025] Further, S6 determines whether to select intensive training or ordinary training according to the time and frequency of exceeding the stress threshold or the motion trajectory fluctuation threshold during the trainee's training.

[0026] Further, in S7, the simulation module distinguishes and color-labels the abnormal stress and motion trajectory areas on the displayed three-dimensional model on the premise that:

[0027] When the stress and motion trajectory exceed the stress revision threshold or the motion trajectory fluctuation revision threshold, and the stress and motion trajectory return to within the stress revision threshold and the motion trajectory fluctuation revision threshold within the threshold time range, the trainee is prompted after completing a node feature;

[0028] When the stress and motion trajectory exceed the stress revision threshold or the motion trajectory fluctuation revision threshold for more than the threshold time, the trainee is directly prompted;

[0029] The threshold time is determined according to the type of training surgery and the specific steps of the surgery.

[0030] The beneficial effects of the present application are:

[0031] The application collects a plurality of surgeon operation characteristics, separates data based on time and success rate, and combines genetic algorithm and deep reinforcement learning to generate a dynamic operation model integrating multiple expert experiences, break the traditional standard limit, and realize the combination of personalized operation habit and standardized safety requirement. Real-time operation data are collected by position sensing and pressure measurement modules, a high-precision three-dimensional operation model is constructed, force feedback is provided, and a real operation scene is restored; through the test and learning dual mode, the operation specification is graded and intervened by combining the double threshold value basic threshold value and the revised threshold value, and the balance between autonomous learning and forced error correction is supported. The fitting process introduces a success rate threshold rigid constraint, sets a safety standard based on historical data, accurately calculates the overall success rate through a product formula, guarantees the quality of key steps, and for time-limited operation scenes, uses particle swarm optimization algorithm to reorganize operation steps, improves efficiency under time constraints, and balances safety and timeliness. The progressive training from novice to expert is supported, the operation intuition is cultivated through a loose threshold value, and the high-risk action is gradually tightened to a strict threshold value specification, the error retraining mechanism can be rolled back to the operation failure node, the operation stability is ensured through multiple comparison scores, the accidentalness of single score is avoided, and the special breakthrough of error-prone links is strengthened.

[0032] The application adopts time orientation + success rate orientation dual-dimension separation of operation characteristics, combines genetic algorithm and deep reinforcement learning, and introduces a success rate threshold rigid constraint. This combination accurately splits key steps through dual orientation separation, realizes the fusion of multiple expert experiences through multi-algorithm fitting, guarantees the safety bottom line through the success rate threshold, and finally solves the technical problem of conflict between personalized operation habit and standardized safety requirement. In the learning mode, if the operation exceeds the revised threshold value but returns to normal within the threshold time, the node is completed and a prompt is given; if the operation exceeds the revised threshold value and exceeds the threshold time, the operation is immediately paused and a prompt is given. This mechanism balances the fault tolerance space-autonomous learning and error correction efficiency-avoiding error accumulation through dynamic time judgment, and solves the problem of unreasonable timing of traditional evaluation and intervention. The operation module integrates real-time collection of multiple parameters such as position, speed, pressure and inclination angle, and provides real-time mechanical feedback through the force feedback module to form a closed loop of collection-feedback-adjustment. This combination accurately captures operation details through multiple parameters, and real-time restores the mechanical properties of tissues through force feedback, solving the technical problem of difficulty in muscle memory formation in virtual environment. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 A flowchart of the micro-invasive surgical skill training and evaluation method in the embodiments of the application.

[0034] Figure 2 An interface schematic diagram of the micro-invasive surgical skill training system in the embodiments of the application.

[0035] Figure 3Another interface schematic diagram of the minimally invasive surgical skill training system in the embodiments of the present application.

[0036] Figure 4 Another interface schematic diagram of the minimally invasive surgical skill training system in the embodiments of the present application.

[0037] Figure 5 Another interface schematic diagram of the minimally invasive surgical skill training system in the embodiments of the present application.

[0038] Figure 6 Another interface schematic diagram of the minimally invasive surgical skill training system in the embodiments of the present application.

[0039] Figure 7 Another interface schematic diagram of the minimally invasive surgical skill training system in the embodiments of the present application. DETAILED DESCRIPTION

[0040] In order to make the personnel in the art better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below.

[0041] The present application discloses a kind of minimally invasive surgical skill training system and its evaluation method, it is characterized in that, including identification acquisition module, identification acquisition module is used to gather doctor surgery information and identify extraction surgery feature.Storage module, storage module is connected identification acquisition module, storage module is used to store doctor surgery information and surgery feature.Data processing analysis module, data processing analysis module is connected to storage module, data processing analysis module is used to analyze and process and fit the surgery feature that storage module stores.Three-dimensional module, three-dimensional module connects data processing analysis module, three-dimensional module is used to convert the surgery feature of analysis processing and fitting into three-dimensional model.Display simulation module, display simulation module connects three-dimensional module, display simulation module is used to display three-dimensional model.

[0042] This invention extracts surgical features such as operating speed, trajectory, pressure, and angle from the surgeon through an identification and acquisition module. Combined with the fitting function of the data processing and analysis module, it transforms the chaotic raw data into structured and standardized model parameters. The data processing and analysis module uses fitting algorithms such as deep learning or statistical models to analyze the surgical features of different surgeons, such as the differences between experts and novices, generating an ideal surgical model that conforms to clinical standards, providing trainees with a comparable reference benchmark. The 3D module transforms the fitted surgical features into a high-precision 3D model, which can intuitively display the three-dimensional spatial relationships of the surgical operation, such as the positional relationship between instruments and organs, and the hierarchical sense of tissue structure, enhancing trainees' understanding of anatomical structures and surgical paths. The display and simulation module presents real-time changes in the 3D model, such as the real-time projection of the scalpel trajectory and the visualization of tissue stress and deformation, helping trainees intuitively identify operational deviations such as trajectory jitter and uneven pressure distribution, and quickly adjust their movements. The storage module permanently stores the surgeon's surgical information and feature data, forming a surgical feature database to provide data support for subsequent analysis. Based on the massive amount of stored data, the data processing and analysis module can continuously optimize the fitting algorithm, such as through machine learning, to improve model accuracy, ensuring that the 3D model always conforms to the latest clinical standards. The system can quickly generate corresponding 3D models by replacing or expanding the surgical feature data in the storage module, such as data on different surgical types and complex cases, supporting training in various surgical skills. A schematic diagram of the interface of the minimally invasive surgical skills training system of this invention is shown below. Figure 2 , Figure 3 , Figure 4 , Figure 5 , Figure 6 and Figure 7 As shown in the image.

[0043] As an implementation form, the recognition and collection module further comprises a surgery module and a force feedback module. The surgery module comprises a position display submodule capable of judging the spatial position of the surgery module, a motion measurement submodule for detecting the speed and trajectory of the surgery module, a plurality of pressure measurement submodules arranged at one end of the surgery module, and an optical detection submodule for detecting the inclination angle of the surgery module. The surgery module is connected to the recognition and collection module for the recognition and collection module to collect surgery information. The force feedback module is used to provide mechanical feedback for the surgery module. The position display submodule is used to display the specific position of the surgery module during the surgery process and monitor the collision between the surgery module and other objects in the environment, such as surrounding bones, organs, etc. The motion measurement submodule monitors the motion speed of the surgery module, judges whether the operation speed of the doctor is too fast or too slow, and analyzes the motion trajectory of the surgery module. The ideal surgery motion trajectory should be smooth and stable, avoiding large amplitude jitter or irregular polyline motion. The force feedback module is provided with a micro-electro-mechanical system (MEMS) structure. The MEMS structure is composed of a plurality of micro-structure units manufactured by MEMS technology. The micro-structure units are controlled by electrical signals to realize different degrees of stretching or contraction to adjust the friction force of the surface of the force feedback module.

[0044] As an implementation form, the minimally invasive surgery skill training system further comprises a feature separation and fitting module. The feature separation and fitting module is connected to the storage module. The feature separation and fitting module is used to separate the surgery features of each of the plurality of doctors extracted by the recognition and collection module into a plurality of surgery features, and then fit the surgery features of each of the plurality of doctors according to a set condition to form a new surgery model. The fitting is to fit different feature steps to form a new complete surgery step according to the type of the surgery.

[0045] As an implementation form, the surgery features of each of the plurality of doctors are separated into a plurality of surgery features. The separation condition is divided into time-oriented separation and success rate-oriented separation. The time-oriented separation separates the surgery features by time nodes. The success rate-oriented separation separates the surgery features according to the success rate in the surgery process.

[0046] As an implementation form, the plurality of surgery features of each of the plurality of doctors are fitted according to a set condition. The set condition is at least one of a surgery genre, a surgery time, a surgery success rate, and a motion trajectory of a surgical knife in the surgery.

[0047] The fitting method is performed by an algorithm. The algorithm comprises at least one of a genetic algorithm, a simulated annealing algorithm, an ant colony algorithm, a particle swarm optimization algorithm, and deep reinforcement learning.

[0048] The set condition of the surgery feature comprises the following dimensions: a surgery genre, a surgery time, a surgery success rate, and a motion trajectory of a surgical knife.

[0049] Core objective: After separating the surgical characteristics of different doctors through algorithms, generate a new surgical model based on the above conditions, ensuring that at least one of the following constraints is met: the success rate of the fitted surgical process ≥ success rate threshold, the surgical process time ≥ time threshold.

[0050] Specific scenarios and embodiments:

[0051] Scenario 1: Trajectory-guided fitting of surgical knife movement

[0052] Preconditions: The student has formed a specific knife usage habit, such as preferring "fast cutting" or "steady advancement", and needs to combine the student's trajectory with expert data to ensure that the success rate of the fitted trajectory ≥ success rate threshold.

[0053] Fitting process: Extract the student's surgical knife trajectory data such as speed, angle. Select similar surgical characteristics from the expert library as the student's trajectory, such as calculating trajectory similarity using dynamic time warping algorithm DTW. Calculate the success rate of the matched trajectory, such as speed deviation, pressure fluctuation, position deviation. If the success rate is lower than the threshold, adjust the trajectory parameters such as reduce the speed or adjust the angle, or reduce the proportion of surgical characteristics similar to the student's trajectory, and replace it with the proportion of surgical characteristics with high success rate, with the success rate of the surgery as the constraint condition.

[0054] Embodiment: Student A habit "fast cutting", but the historical success rate is lower than the threshold. The system selects the surgical characteristics of the "fast cutting school" from the expert library, but requires that the fitted trajectory meet the following conditions: speed fluctuation ≤ 15% of the standard speed, pressure fluctuation ≤ 2N, position deviation ≤ 3mm. The final fitted trajectory not only retains the student's fast habit, but also adjusts the angle and path smoothness to improve the success rate to above the threshold.

[0055] Scenario 2: Fitting guided by surgical school

[0056] Preconditions: The student needs to learn the operation mode of a specific school such as "minimally invasive school A" or "robot-assisted school B", and the operation style of different schools is significantly different such as instrument use, step sequence.

[0057] Fitting process: Filter the surgical characteristics in the expert library according to the school label such as "school A", extract the parameters specific to the school such as the average speed of the "suture step" in school A is 1.2mm / s, and in school B is 0.8mm / s. The weight of the school characteristics ≥ 80% to ensure that the fitting result conforms to the school style; the overall success rate ≥ success rate threshold, such as 95%.

[0058] Example: Student B needs to learn "Robot-assisted Approach B" for tumor resection. The system extracts features from expert data of Approach B, including: Cutting phase pressure threshold: 2.5N, Suturing phase angle limit: Tilt angle ≤ 15°. During fitting, the "Layered Cutting" step of Approach B is forced to be retained, while the suturing trajectory is optimized through algorithms to ensure a success rate ≥ 95%.

[0059] Scenario 3: Time-oriented fitting

[0060] Precondition: Students need to complete surgery within a specified time, such as emergency surgery to shorten time, and need to optimize operation efficiency under time constraints while ensuring success rate.

[0061] Fitting process: The surgical procedure is divided into time nodes, and an upper limit is set for each stage, such as "tumor resection phase ≤ 15 minutes", total time ≤ 120% of standard time; Each stage success rate ≥ success rate threshold, such as each node success rate ≥ 80%.

[0062] Example: Student C needs to complete cholecystectomy within 30 minutes. The system selects "fast completion" cases from the expert database, but requires: clamp stage pressure ≥ 3N; Cutting phase speed ≥ 1.5mm / s. Adjust the step order through particle swarm optimization algorithm, parallelize the suturing step, and finally complete in 28 minutes, success rate 92%.

[0063] Scenario 4: Success rate-oriented fitting

[0064] Precondition: Students need to ensure success rate in high-risk surgery such as pediatric heart surgery. Need to prioritize the operation path with the highest success rate, even if it takes longer.

[0065] Fitting process: Extract all possible paths from the expert database, calculate the overall success rate. Select the path with the highest success rate as the fitting target. The constraint condition requires the overall success rate ≥ success rate threshold, such as set 99%; Allow time extension, but not more than 150% of standard time.

[0066] Example: Student D performs pediatric heart valve repair surgery, and the system selects the path with the highest success rate, which includes: suturing phase speed ≤ 0.5mm / s; Pressure fluctuation ≤ 1N. Even if the total time is extended to 140% of the standard time, the success rate still reaches 99.5%, meeting the constraint condition.

[0067] Constraint of fitting conditions and previous example:

[0068] 1. Constraint conditions for surgical knife trajectory

[0069] Trajectory smoothness: Change in displacement between adjacent points ≤ 5mm / s (avoid sudden movements).

[0070] Direction consistency: Angle mutation ≤ 10° / s (reduce instrument jitter).

[0071] Real-time feedback: If the trajectory deviates from the set threshold, the force feedback module increases resistance.

[0072] 2. Constraints of surgical genre

[0073] Genre feature weight: Parameters specific to the genre, such as instrument holding angle, should account for more than 70% of the fitting weight.

[0074] Compatibility check: Different genre operation steps cannot conflict, such as "Genre A" needs to stop bleeding before suturing, and "Genre B" is the opposite.

[0075] 3. Time-oriented constraints

[0076] Upper limit of stage time: If a stage exceeds the time limit, it is forced to jump to an alternative path, such as the "threshold time" judgment in the example.

[0077] Resource allocation: When time is tight, resources are prioritized for critical steps.

[0078] 4. Success rate-oriented constraints

[0079] Dynamic threshold: Adjust the success rate threshold according to the surgical stage, such as a high-risk step threshold ≥ 95% and a low-risk step ≥ 85%.

[0080] Punishment mechanism: Violation of operation directly reduces the success rate of the current stage.

[0081] Algorithm selection and examples:

[0082] The algorithm includes at least one of genetic algorithm, simulated annealing algorithm, ant colony algorithm, particle swarm optimization algorithm, and deep reinforcement learning.

[0083] The algorithm can combine multiple optimization methods:

[0084] Genetic algorithm:

[0085] Scenario: Surgical genre fitting, such as the optimal features of mixed genres A and B.

[0086] Example: Generate new paths through crossover and mutation operations, and retain the individual with the highest success rate.

[0087] Deep reinforcement learning:

[0088] Scenario: Real-time adjustment of student trajectory, such as adjusting the next action based on real-time pressure feedback.

[0089] Example: When the student is operating, the DRL model dynamically adjusts the recommended path for the next step based on the current trajectory deviation.

[0090] Particle swarm optimization:

[0091] Scenario: Time-oriented path optimization, such as shortening the total time.

[0092] Embodiment: Particles represent different path combinations, and the path with the shortest time and success rate ≥ threshold is found through group search.

[0093] As an implementation, the constraint condition of the set condition includes a success rate threshold, and the success rate of the entire surgical feature during the surgery is greater than the success rate threshold after fitting the surgical feature according to the set condition, and the success rate threshold is set according to the surgical history data.

[0094] As an implementation, S1: The identification and collection module collects the surgical information of the doctors and extracts and stores the surgical features to the storage module;

[0095] S2: The data processing and analysis module analyzes and processes the surgical features and fitting, and the three-dimensional module converts the analyzed and processed surgical features into a three-dimensional model;

[0096] S3: The display simulation module displays the three-dimensional modeling, and the student selects the mode to operate the surgical module on the force feedback module through the display of the display simulation module;

[0097] S4: The motion measurement sub-module, pressure measurement sub-module, and optical detection sub-module in the surgical module transmit the motion trajectory, operation pressure, and inclination angle of the student's surgical operation to the display simulation module, and the data processing and analysis module analyzes and processes the motion trajectory, operation pressure, and inclination angle of the student's surgical operation in real time;

[0098] S5: When the student's motion trajectory and operation pressure are within the motion trajectory fluctuation threshold or the pressure threshold, it is judged that the student's operation is standard, and the student continues to operate. When the student's motion trajectory and operation pressure are outside the motion trajectory fluctuation threshold or the pressure threshold, it is judged that the student's operation is not standard. When the student selects the examination mode, S6 is performed, and when the student selects the learning mode, S7 is performed;

[0099] S6: Examination mode, wait for the student to complete the surgical operation, record the non-standard operation of the student during the entire surgical operation process and score the entire surgical operation of the student, and record the non-standard operation as an error. Evaluate whether the student needs to strengthen training or normal training;

[0100] S7: learning mode, when the pressure and movement trajectory exceeds the pressure threshold or movement trajectory fluctuation threshold, but also does not exceed the pressure revision threshold or movement trajectory fluctuation revision threshold, no correction is made to the trainee, but S6 is performed, when the pressure and movement trajectory exceeds the pressure revision threshold or movement trajectory fluctuation revision threshold, the simulation module distinguishes and color labels on the pressure and movement trajectory abnormal area of the displayed three-dimensional model, the simulation module is paused and prompts the trainee, and the trainee is given the option of continuing training or performing error training, the trainee continues training and repeats S6, and the trainee performs error training and performs S8;

[0101] S8: after selecting error training, the simulation module displays the three-dimensional model to return to the trainee's error operation before, and gives a countdown prompt, the trainee re-performs error training, at least several times, each time compared with standard training and scored, and finally the average score is taken, the average score is higher than or equal to the qualified score value, the trainee is judged to be qualified, the average score is lower than the qualified score value, S8 is repeated, and at most a certain number of repeated attempts are allowed, if still not up to standard, forced into S9 and record not passed;

[0102] S9: the trainee's training is over.

[0103] The pressure threshold or movement trajectory fluctuation threshold is the minimum required threshold of the operation specification, indicating that the trainee's error is within the acceptable range, while the pressure revision threshold or movement trajectory fluctuation revision threshold is the strict threshold that needs to be corrected, indicating that the trainee's error is very large and cannot continue training.

[0104] S1-S3: Through three-dimensional modeling and display simulation modules, a highly realistic surgical scene is constructed, providing an operating experience similar to the real surgical environment. Students can directly observe the details of the operation, improving spatial perception and operation accuracy. Reducing the experience gap caused by the lack of practical opportunities, especially suitable for the simulation of complex or high-risk surgeries. S4-S8 collects data in real time through motion, pressure, and angle sensors, analyzes it with the data processing analysis module, and provides immediate feedback. When the operation exceeds the threshold, the system quickly locates the error through color marking and pause prompts, avoiding the formation of incorrect muscle memory by the students. The learning mode allows students to adjust themselves when there is a slight deviation, and only triggers forced correction when there is a serious violation, balancing autonomous learning and forced regulation. The examination mode records operation data completely, generates objective scores and error reports, and provides a basis for subsequent improvement. S5-S8: According to the performance of the students' operation, the training intensity is dynamically allocated. Beginners can first practice under a relaxed threshold, and gradually challenge strict thresholds as they become proficient, conforming to the learning curve from basic to complex. Through the average score of multiple error training, the mastery of students on a specific operation is measured, avoiding the randomness of a single score. S1-S4: The storage module accumulates expert surgical feature data, and the data processing module generates a standard path through analysis, ensuring that the training content conforms to clinical standards. Quantify indicators such as operation success rate and trajectory smoothness into thresholds to reduce subjective evaluation bias. Through threshold constraints, avoid students' operations from exceeding the safe range, and reduce the risk of complications in real surgery. S6-S9 repeated training and modular process design improve learning efficiency. Students can repeat high-risk operations as many times as needed until they meet the requirements, shortening the skill acquisition period. S8-S9 forced repeated training ensures that students have a deep memory of incorrect operations, combined with the complete process examination mode, to strengthen knowledge retention. Through multiple error correction training and high-fidelity simulation, students form muscle memory and improve proficiency in real surgery.

[0105] As an implementation, S6 chooses to strengthen training or ordinary training according to the time and frequency of exceeding the pressure threshold or motion trajectory fluctuation threshold during the student's training process. This method determines the training intensity by analyzing the time and frequency of exceeding the threshold in the student's operation, accurately identifies the student's ability shortcomings and dynamically adjusts the training difficulty, providing ordinary training for occasional small errors and starting intensive training for high-frequency or long-time exceeding, effectively preventing the solidification of incorrect habits, while quantifying data to make scientific decisions, reduce subjective bias, improve training efficiency and safety, and ultimately help students efficiently master standard operations and reduce real surgery risks.

[0106] As an implementation, the premise for the simulation module in S7 to distinguish and color mark the abnormal area of pressure and motion trajectory of the displayed three-dimensional model is: when the pressure and motion trajectory exceed the pressure revision threshold or the motion trajectory fluctuation revision threshold, and the pressure and motion trajectory return to within the pressure revision threshold and the motion trajectory fluctuation revision threshold within the threshold time range, the student is prompted after completing a node feature; when the pressure and motion trajectory exceeds the pressure revision threshold or the motion trajectory fluctuation revision threshold for more than the threshold time, the student is directly prompted; the threshold time is determined according to the type of training surgery and the specific steps of surgery.

[0107] Embodiment: Surgical path planning for endoscopic surgery

[0108] Time node:

[0109] 1. Time node division

[0110] In a plurality of surgical procedures of doctors, surgical procedures with similar surgery times are divided into a category, and the entire surgical process of this category of surgery is divided into n stages according to time:

[0111] Stage division = [T0, T1, T2, …, T n ];

[0112] Wherein, T0=0, T n =total surgery time, and each stage i corresponds to a time interval [T i−1 , T i ].

[0113] For each stage i, the following feature vector is extracted:

[0114] X i =[time feature, position feature, speed feature, angle feature, pressure feature];

[0115] The specific definitions are as follows:

[0116] Time feature: the duration t i of stage i is T i -T i−1 .

[0117] Position feature: the real-time position p i of the instrument tip is (xi, yi, zi) (mm).

[0118] Speed feature: the real-time linear speed v i (mm / s) and angular speed ω i (° / s) of the instrument.

[0119] Angle feature: the real-time inclination angle θ i (°) and direction angle ϕi (°).

[0120] Pressure feature: Real-time pressure p i (N) and pressure fluctuation σ i (N).

[0121] 2. Stage success rate calculation

[0122] Define the success rate S of each stage i , based on expert data or historical surgery data, calculated by the following formula:

[0123] ;

[0124] : Speed deviation penalty term, α is the weight;

[0125] : Pressure deviation penalty term, β is the weight;

[0126] : Position deviation penalty term, γ is the weight;

[0127] Where:

[0128] Total operations: How many steps were performed in total in this time stage, defined by historical experience of the surgery;

[0129] Violations: The total number of errors, including pressure exceeding, angle errors, and excessive speed, in this time stage;

[0130] V i : Actual average linear speed in this time stage;

[0131] V 标准 : Standard speed under the surgical step in this time stage, defined by historical experience of the surgery;

[0132] α: Speed deviation weight, which indicates the influence of speed deviation on the success rate of the surgery in different stages. The speed deviation weight is different in different surgical stages, and different parts are affected differently by speed deviation. More fragile organs are more affected by speed deviation, which has a greater impact on the success rate of the surgery;

[0133] σ i : Actual size of the average pressure fluctuation in this time stage;

[0134] σ: Standard pressure fluctuation threshold size of pressure fluctuation in this time stage, defined by historical experience of the surgery;

[0135] β: Pressure fluctuation weight, which indicates the influence of pressure on the success rate of the surgery in different stages. The pressure weight is different in different surgical stages;

[0136] Position deviation: the maximum position deviation in the surgical step within this time phase;

[0137] Position threshold: the threshold of the actual position deviation allowed, defined by the historical experience of the surgery;

[0138] γ: the weight of position deviation, which represents the impact of position deviation on the success rate of surgery at different stages, and the weight of position deviation is different at different stages of surgery.

[0139] 3. Overall path success rate

[0140] Combine the success rates of all stages into the overall path success rate:

[0141] Or ;

[0142] According to the clinical needs, choose the product formula or the mean formula, or use both together.

[0143] Product formula:

[0144] ;

[0145] Multiply the success rates of all stages, and the final result is the product of the success rates of all stages.

[0146] Applicable scenarios:

[0147] All key steps must be successful: for example, if a step fails during surgery, resulting in the entire surgery failing or serious complications. At this time, even if other steps are successful, the overall path is considered a failure. The success rate of any stage below the threshold will significantly affect the overall result. Ensure that there are no loopholes in the entire process, and high-risk operations.

[0148] Example:

[0149] The surgery is divided into three nodes:

[0150] Node 1 success rate S1 = 0.9;

[0151] Node 2 success rate S2 = 0.8;

[0152] Node 3 success rate S3 = 0.7;

[0153] Then S 总 = 0.9 x 0.8 x 0.7 = 0.504;

[0154] Even if node 1 performs well, but the low success rate of node 3 still results in an overall success rate of only close to 50%.

[0155] Mean formula:

[0156] ;

[0157] Take the success rate of all stages and calculate the arithmetic mean. The final result is the average level of success rate of each stage.

[0158] Applicable scenarios:

[0159] High fault tolerance process, some steps can be compensated by subsequent operations after failure.

[0160] Optimization of resource allocation, need to balance overall efficiency, allow slight failure of some steps.

[0161] Clinical needs, focus on overall trends rather than extreme cases, such as routine physical examination or non-high-risk surgery.

[0162] Use the same node success rate calculation:

[0163] ;

[0164] At this time, the overall success rate is pulled up to 0.8, covering up the low success rate problem of node 3.

[0165] Selection basis:

[0166] Advantages of product formula:

[0167] High risk sensitivity, zero tolerance for failure of any stage, ensuring the quality of the whole process.

[0168] In line with clinical logic, failure of key steps in surgery will lead to irreversible consequences.

[0169] Risk of product formula:

[0170] Overly conservative, may underestimate the potential value of the overall path due to individual low success rate nodes.

[0171] Need to strictly meet the standard, require all nodes to achieve high success rate, extremely high requirements for doctors or system.

[0172] Advantages of mean formula:

[0173] Simple calculation, suitable for quick evaluation of overall trends, no need for complex calculation.

[0174] Strong fault tolerance, allows failure of some steps, suitable for early process optimization or non-high-risk scenarios.

[0175] Risk of mean formula:

[0176] Cover up problems, ignore the low success rate of key nodes.

[0177] Poor clinical applicability, in high-risk surgery, mean will mislead decision-making.

[0178] The specific formula to choose depends on the type of surgery.

[0179] 4. Optimal Path Selection

[0180] The goal is to select a set of stage paths {X1, X2, ..., X...} n This maximizes the overall success rate.

[0181] ;

[0182] Constraints:

[0183] Time continuity: T i =T i−1 +t i ;

[0184] Operational feasibility: Parameters at each stage must meet safety thresholds, including pressure p. i ≤p max σ i ≤σ and V i ≤V 标准 .

[0185] 5. Algorithm Steps

[0186] (1) Data preprocessing

[0187] Extract surgical data from m doctors from the storage module and divide it into n stages according to time.

[0188] For each stage i, collect the feature vectors {X} of all doctors. i1 X i2 , ..., X im}

[0189] (2) Construction of candidate set of stage path

[0190] For each stage i, generate a set of candidate paths.

[0191] C i ={Xi1,X i2 , ..., X im (or by generating extended candidates through mutation).

[0192] (3) Global path combination

[0193] Construct all possible path combinations P = C1 × C2 × ... × C n .

[0194] Calculate S for each path 总 Choose the path corresponding to the maximum value.

[0195] Surgical milestones:

[0196] 1. Division of surgical nodes

[0197] Surgical procedures with similar success rates among several doctors are grouped into one category. The entire surgical process of this category is then divided into n key nodes according to clinical operation steps:

[0198] ;

[0199] Each node N i For each key operation, the specific number of nodes is defined depending on the specific surgery.

[0200] 2. Node Feature Extraction

[0201] Time characteristics: Node N i Duration t i (Second).

[0202] Positional characteristics: The average position of the instrument tip is pi = (xi, yi, zi) (millimeters).

[0203] Velocity characteristics: Average linear velocity v of the instrument i (mm / s) and angular velocity ω i (° / s).

[0204] Angular characteristics: the average tilt angle θ of the instrument i (°) and direction angle ϕ i (°).

[0205] Pressure characteristics: mean pressure p i (N) and pressure fluctuation σ i (N).

[0206] Operational quality characteristics: Number of violations (such as vascular injury, tissue tearing, etc.).

[0207] For each order i, extract the following feature vectors:

[0208] X i =[Time characteristics, position characteristics, velocity characteristics, angle characteristics, pressure characteristics, operational quality characteristics];

[0209] The specific definitions are as follows:

[0210] Time characteristics: Node N i Duration t i (Second).

[0211] Positional characteristics: the average position p of the instrument tip i =(xi, yi, zi) (mm).

[0212] Velocity characteristics: Average linear velocity v of the instrumenti (mm / s) and angular velocity ω i (° / s).

[0213] Angular feature: average instrument tilt angle θ i (°) and direction angle ϕ i (°).

[0214] Pressure feature: average pressure p i (N) and pressure fluctuation σ i (N).

[0215] Operational quality feature: number of illegal operations (such as blood vessel injury, tissue tearing, etc.).

[0216] 3. Node success rate calculation

[0217] Define the success rate S of each node i , based on expert data or historical operation data, calculated by the following formula:

[0218] ;

[0219] : speed deviation penalty term, α is the weight;

[0220] : pressure deviation penalty term, β is the weight;

[0221] : position deviation penalty term, γ is the weight;

[0222] Where:

[0223] Total number of operations: how many steps were performed in this time phase of the operation, defined by historical experience of the operation;

[0224] Number of illegal operations: including the total number of errors in this time phase, such as excessive pressure, angle error, and excessive speed;

[0225] V i : actual average linear velocity in this node phase;

[0226] V 标准 : standard speed under the operation step in this node phase, defined by historical experience of the operation;

[0227] α: weight of speed deviation, weight of speed deviation represents the influence of speed deviation in different phases on the success rate of the operation, speed deviation weight is different in different operation phases, different parts are affected by speed deviation, more fragile organs are more affected by speed deviation, which has greater influence on the success rate of the operation;

[0228] σ iσ: the standard pressure fluctuation threshold size in this point phase, defined by historical experience of the surgery;

[0229] σ: the standard pressure fluctuation threshold size in this point phase, defined by historical experience of the surgery;

[0230] β: the weight of pressure fluctuation, the pressure weight represents the influence of pressure on the success rate of the surgery in different phases, and the pressure weight is different in different surgery phases;

[0231] Position deviation: the maximum position deviation in the surgery step in this point phase;

[0232] Position threshold: the threshold of the actual position deviation allowed, defined by historical experience of the surgery;

[0233] γ: the weight of position deviation, the weight of position deviation represents the influence of position deviation on the success rate of the surgery in different phases, and the weight of position deviation is different in different surgery phases.

[0234] 4. Overall path success rate

[0235] Combine the success rates of all nodes into the overall path success rate:

[0236] Or ;

[0237] According to the clinical needs, select the product formula or the weight formula, or use both.

[0238] The product formula is divided by time nodes;

[0239] Weight formula:

[0240] Multiply the success rate of each node by the weight coefficient w i , and then sum up the weight w i reflecting the importance of the node.

[0241] The weight of the key node is higher, and the weight of the non-key node is lower. Whether a node is key is set according to different surgeries.

[0242] Clinical application scenarios:

[0243] Allow slight mistakes in part of the secondary steps, but ensure that the key steps meet the standards.

[0244] Resource optimization, in gastric resection, the weight of tumor resection is much higher than that of suturing.

[0245] 5. Optimal path selection

[0246] The goal is to select a set of node paths {X1, X2, …, Xn} to maximize the overall success rate:

[0247] ;

[0248] Constraints:

[0249] Operation continuity, the position and angle of adjacent nodes need to be smoothly transitioned;

[0250] Time constraints, total time is set to constrain;

[0251] Operation feasibility: each stage parameter needs to meet the safety threshold, including pressure p i ≤p max , σ i ≤σ and V i ≤V 标准 .

[0252] The constraints also include the success rate of the total operation, which is set to be no less than the success rate threshold.

[0253] 6. Algorithm steps

[0254] (1) Data preprocessing

[0255] Extract the operation data of m doctors from the storage module, and extract the feature vector {X i1 , X i2 , …, X im} according to node division.

[0256] For each node N i , collect expert standard parameters including standard operation node time, total operation time and standard operation pressure, etc.

[0257] (2) Node path candidate set construction

[0258] For each node N i , generate candidate path set C i ={X i1 , X i2 , …, X im}.

[0259] (3) Global path combination

[0260] Construct all possible path combinations P=C1×C2×⋯×C n .

[0261] Calculate S 总 for each path, and select the path corresponding to the maximum value.

[0262] Where the time node divides the operation process more towards operation efficiency, and the operation node divides the operation process more towards success rate. The specific way for students to choose training is left to the students to decide.

[0263] Through the minimally invasive surgical skill training system and the evaluation method thereof, a minimally invasive surgical basic skill operation competition is carried out

[0264] Step 1: Clear the eligibility for participating in the competition

[0265] Age limit: 40 years old or younger.

[0266] Determination of unit and title.

[0267] Step 2: Division and organization of competition areas

[0268] Competition area setting: divided into four competition areas of east, north, south and west, and each competition area is responsible for being undertaken by a local alliance member unit.

[0269] Notice of registration: each competition area issues a notice of registration for the preliminary competition, and clearly indicates the time, place, registration method online / offline and materials ID card, title certificate, unit recommendation letter.

[0270] Step 3: Registration of players and qualification examination

[0271] Submission of materials: the players submit the registration materials according to the requirements of the competition area, including personal information, age certificate, unit qualification, etc.

[0272] Qualification examination: the competition area organizing committee examines the registration materials and confirms the qualification for participating in the competition.

[0273] Step 4: On-site operation examination of three basic skills

[0274] Project 1: Pinning board moving object

[0275] Operation steps:

[0276] Hold the separating forceps with the left hand and the forceps with the right hand, grab the six hanging pieces on the left side of the moving object board in turn, alternately transfer them in the air with the hands, place them on the right side of the pin column, and then move them back to the left side according to the same method.

[0277] The tip is always directed downward, regardless of color / sequence, and the timing starts when the instrument is touched, and the timing ends when the instrument is pulled out of the training box.

[0278] Scoring standard: only objective timing, full score 10 points:

[0279] ≤ 30 seconds: 10 points; 30-45 seconds: 8 points; 45-60 seconds: 6 points; > 60 seconds: 5 points.

[0280] Project 2: Cutting of specific shape

[0281] Operation steps:

[0282] Fix the edge of the paper sheet with the separating forceps, and cut the paper sheet along the central circle mark of the base, without pulling out the paper sheet or cutting from the edge.

[0283] The timing rules are the same as those of the pinning board moving object.

[0284] Scoring criteria:

[0285] Objective timing full score 10 points: ≤60 seconds 10 points; 60-90 seconds 8 points; 90-120 seconds 6 points; >120 seconds 5 points.

[0286] Subjective score full score 30 points: judges score from holding force, cutting accuracy, hand coordination, etc. 5 aspects.

[0287] Project 3: Suture and knot

[0288] Operation steps:

[0289] Hold the needle holder, separating forceps, and wire scissors to complete 2 times of simple interrupted suture of different incisions on the training module, 4 knots each time, which can contain 1 surgical knot + 2 single knots, the wire tail is reserved 0.5-1.0 cm, and the wire scissors are not retained in the box after cutting the wire.

[0290] Timing rules are the same as before.

[0291] Scoring criteria:

[0292] Objective timing full score 10 points: ≤150 seconds 10 points; 150-180 seconds 8 points; 180-210 seconds 6 points; >210 seconds 5 points.

[0293] Subjective score full score 40 points: judges score from holding angle, knotting technique, knot firmness, etc. 8 aspects.

[0294] Step 5: Preliminary score and promotion

[0295] Scoring rules:

[0296] Total score = nail plate moving objects + cutting timing + suture timing objective score + cutting subjective score + suture subjective score, after removing the same unit judge score, take the average score.

[0297] Promotion quota: A total of 80-100 people are selected from four competition areas to advance to the final, the quota will be adjusted by the organizing committee according to the registration situation.

[0298] Step 6: Final preparation

[0299] Material distribution: players receive 3 No. 1 suture needles with wire, and the wire tail is trimmed 2 minutes before the race.

[0300] Instrument confirmation: check the status of the needle holder, separating forceps, wire scissors, etc.

[0301] Step 7: Suture and knot examination of ex vivo specimens

[0302] Operation steps:

[0303] The real animal ex vivo small intestine incision is simply intermittently sutured, and 4 surgical knots + single knot or 4 single knots are tied, the thread tail is 0.5-1.0 cm, the operation time is 8 minutes, and the operation is stopped immediately if the time is exceeded.

[0304] Scoring criteria:

[0305] Objective score full score 20 points: 4 points for each qualified knot, a maximum of 5 knots 20 points.

[0306] Subjective score full score 80 points: same as the subjective scoring dimension of the suture and knot tying in the preliminary round, focusing on tissue protection and clinical practicality.

[0307] Same score processing: those with more qualified knots are given priority.

[0308] Step 8: Final round

[0309] Promotion rule: according to the final total score ranking, the top 10-15 enter the additional competition.

[0310] Step 9: Virtual endoscopic surgery assessment

[0311] Operation selection: select 1 from appendectomy, cholecystectomy, and oophorectomy.

[0312] Operation rule: according to the animation / PPT steps demonstrated on site, complete the operation on the minimally invasive surgical skill training system of the invention, and automatically score based on the real operation points by the minimally invasive surgical skill training system.

[0313] Step 10: Comprehensive score and ranking

[0314] Additional competition weight: the additional competition results are used as the basis for final ranking, and the specific weight is announced by the organizing committee before the competition.

[0315] Step 11: On-site award ceremony

[0316] Award setting: first, second and third prizes and excellent awards are set, and certificates and trophies are awarded. The specific number of awards will be announced before the competition.

[0317] Expert comments: industry experts summarize the event, and comment on the operation highlights and improvement direction.

[0318] Step 12: Event summary

[0319] Feedback collection: players and judges fill out satisfaction questionnaires to optimize the next event.

[0320] Data archiving: save competition videos, score sheets and other data for future review.

[0321] It should be understood that for those of ordinary skill in the art, improvements or changes can be made according to the above description, and all such improvements and changes shall fall within the scope of the appended claims of the present invention.

Claims

1. A minimally invasive surgical skills training system, characterized in that, include: The identification and acquisition module is used to collect surgical information from doctors and identify and extract surgical features. The storage module is connected to the recognition and acquisition module and is used to store the doctor's surgical information and surgical characteristics. The data processing and analysis module is connected to the storage module. The data processing and analysis module is used to analyze and fit the surgical features stored in the storage module. The 3D module connects to the data processing and analysis module. The 3D module is used to convert the analyzed and fitted surgical features into a 3D model. The display simulation module connects to the 3D module and is used to display 3D models.

2. The minimally invasive surgical skills training system according to claim 1, characterized in that: The data processing and analysis module also includes a feature segmentation and fitting module, which is connected to the storage module. The feature segmentation and fitting module is used to segment the surgical features of each doctor from several doctors extracted by the recognition and acquisition module into several surgical features. Then, the surgical features of each doctor from several doctors are fitted according to the set conditions to form a new surgical model.

3. The minimally invasive surgical skills training system according to claim 2, characterized in that: The surgical characteristics of each doctor among a number of doctors are divided into several surgical characteristics. The separation conditions are divided into time-oriented separation and success rate-oriented separation. Time-oriented segmentation separates surgical features based on time nodes, while success rate-oriented segmentation is guided by the success rate during the surgical process.

4. The minimally invasive surgical skills training system according to claim 3, characterized in that: Several surgical characteristics of each doctor from a group of doctors are fitted according to set conditions, which are at least one of surgical school, operation time, operation success rate, and the movement trajectory of the scalpel during operation.

5. The minimally invasive surgical skills training system according to claim 4, characterized in that: The fitting method is performed by an algorithm, which includes at least one of the following: genetic algorithm, simulated annealing algorithm, ant colony algorithm, particle swarm optimization algorithm, and deep reinforcement learning.

6. The minimally invasive surgical skills training system according to claim 5, characterized in that: The constraints set include a success rate threshold. After fitting several surgical features according to the set conditions, the success rate of the entire surgical feature is greater than the success rate threshold. The success rate threshold is set based on historical surgical data.

7. The minimally invasive surgical skills training system according to claim 1, characterized in that: The identification and acquisition module also includes a surgical module and a force feedback module. The surgical module includes a position display submodule that can determine the spatial position of the surgical module, a motion measurement submodule that detects the speed and trajectory of the surgical module, several pressure measurement submodules set at one end of the surgical module, and an optical detection submodule for detecting the tilt angle of the surgical module. The surgical module is connected to the identification and acquisition module for the identification and acquisition module to collect surgical information. The force feedback module is used to provide mechanical feedback to the surgical module.

8. A method for assessing minimally invasive surgical skills, used in any of the minimally invasive surgical skills training systems described in claims 1-7, characterized in that, include: S1: The identification and acquisition module collects surgical information from several doctors and identifies and extracts surgical features, storing them in the storage module; S2: The data processing and analysis module analyzes and fits the surgical features, and the 3D module converts the analyzed and fitted surgical features into a 3D model; S3: The display simulation module displays the 3D model. After the trainee selects a mode, the operation surgery module performs surgical operations on the force feedback module through the display simulation module. S4: The motion measurement submodule, pressure measurement submodule and optical detection submodule in the surgical module transmit the motion trajectory, operating pressure and tilt angle of the trainee's surgical operation to the display simulation module. The data processing and analysis module performs real-time analysis and processing of the motion trajectory, operating pressure and tilt angle of the trainee's surgical operation. S5: When the student's movement trajectory and operating pressure are within the movement trajectory fluctuation threshold or pressure threshold, the student's operation is judged to be standard and the student continues to operate. When the student's movement trajectory and operating pressure are outside the movement trajectory fluctuation threshold or pressure threshold, the student's operation is judged to be non-standard. If the student selects the exam mode, proceed to S6; if the student selects the learning mode, proceed to S7. S6: Examination mode. Wait for the trainee to complete the surgical procedure, record any non-standard operations during the entire surgical procedure, score the trainee's entire surgical procedure, record any non-standard operations, and assess whether the trainee needs intensive training or regular training. S7: Learning mode. When the pressure and movement trajectory exceed the pressure threshold or movement trajectory fluctuation threshold, but do not exceed the pressure revision threshold or movement trajectory fluctuation revision threshold, no correction is made to the learner. Instead, S6 is performed. When the pressure and movement trajectory exceed the pressure revision threshold or movement trajectory fluctuation revision threshold, the simulation module displays color-coded markings on the abnormal areas of pressure and movement trajectory in the displayed 3D model. The simulation module pauses and prompts the learner, giving them the option to continue training or perform error training. If the learner continues training, S6 is repeated. If the learner performs error training, S8 is performed. S8: After selecting the error training, the simulation module displays the 3D model before the student's erroneous operation and provides a countdown prompt. The student re-trains against the error, at least several times, comparing and scoring each time with the standard training. The final average score is taken. If the average score is higher than or equal to the passing score, the student is considered qualified. If the average score is lower than the passing score, S8 is repeated. A maximum of several repetitions are allowed. If the student still fails, S9 is forced and the failure is recorded. S9: Trainee training has ended.

9. The method for assessing minimally invasive surgical skills according to claim 8, characterized in that: The choice between intensive and regular training for S6 is determined by the time and frequency during which the trainee exceeds the stress threshold or the movement trajectory fluctuation threshold.

10. A method for assessing minimally invasive surgical skills according to claim 8, characterized in that: The premise for the S7 display simulation module to differentiate and color-code abnormal areas of pressure and motion trajectory in the displayed 3D model is: When the pressure and motion trajectory exceed the pressure revision threshold or the motion trajectory fluctuation revision threshold, and the pressure and motion trajectory return to within the pressure revision threshold and the motion trajectory fluctuation revision threshold within the threshold time range, the student will be prompted after completing a node feature; When the pressure and movement trajectory exceed the pressure revision threshold or the movement trajectory fluctuation revision threshold exceeds the threshold time, the student will be prompted directly. The threshold time is determined based on the type of surgery being trained and the specific steps involved.

Images