An integrated puncture simulation system based on an autoinjector

Abstract

The present invention relates to the technical field of medical training, and particularly to a comprehensive puncture simulation system based on an auto-injector. The system includes a user training puncture simulation operation selection module, a puncture gyroscope deviation calibration module, a puncture training scenario path simulation generation module, and a puncture simulation path visualization module. It is possible to select the puncture scenario to be simulated, configure the puncture needle and operating instruments for the corresponding auto-injector, and simultaneously perform comprehensive puncture simulation operations on the selected puncture scenario to generate the simulation process of the selected puncture scenario; perform puncture gyroscope measurement processing and neural network regression deviation estimation on the simulation process of the selected puncture scenario, and simultaneously perform static deviation calibration calculation and puncture path simulation recognition analysis to generate a puncture simulation training path; perform puncture simulation visualization processing on the puncture simulation training path to execute the corresponding puncture scenario visualization teaching process. The present invention can significantly improve the effectiveness of puncture simulation training.

Description

Technical Field

[0001] The present invention relates to the technical field of medical training, and particularly to a comprehensive puncture simulation system based on an autoinjector. Background Art

[0002] As an innovation pioneer in the modern medical training field, the puncture simulation system integrates a highly simulated physical model and cutting-edge intelligent technology, creating a zero-risk and high-efficiency practical training platform for students, nurses, and doctors in the medical field. By precisely simulating the real tissue structure of various parts of the human body and the dynamic mechanical feedback during the puncture process, the system enables the trainer to feel as if they are in a real surgical scenario, thus significantly improving the accuracy, skill proficiency, and hand-eye coordination ability of the puncture operation. This not only widely covers various common puncture types such as the chest cavity, abdominal cavity, and lumbar spine, but also integrates multiple functional modes such as demonstration, actual combat training, and assessment to meet the personalized needs of learners at different levels. In addition, the simulation system also has a visual teaching function, which helps the trainer intuitively understand the operation details and improve the teaching effect by real-time displaying the puncture path and the surrounding tissue structure.

[0003] Particularly importantly, the puncture simulation system of the present invention also uses a gyroscope sensor as one of the core technologies. The gyroscope is responsible for identifying the angular rotation of the autoinjector during the operation, providing real-time and accurate operation feedback to the trainer. However, although the gyroscope sensor plays a crucial role in this application, its accuracy in actual operation largely depends on the performance of its inertial measurement unit, which is prone to large gyroscope measurement error terms and has noise characteristics, thus affecting the accuracy of the puncture path recognition simulation. Summary of the Invention

[0004] Based on this, it is necessary for the present invention to provide a comprehensive puncture simulation system based on an autoinjector to solve at least one of the above technical problems.

[0005] To achieve the above object, a comprehensive puncture simulation system based on an autoinjector includes the following modules:

[0006] A user training puncture simulation operation selection module, which is used to select the puncture scenario to be simulated through the trainer user interface, and configure the corresponding puncture needle and puncture injection depth operation instrument for the autoinjector according to the selected puncture scenario to obtain the selected simulated puncture autoinjector; use the selected simulated puncture autoinjector to perform a comprehensive puncture simulation operation on the corresponding selected puncture scenario to generate the simulation process of the selected puncture scenario;

[0007] A puncture gyroscope deviation calibration module is used to perform puncture gyroscope measurement processing on the selected puncture scenario simulation process by using the gyroscope sensor in the selected analog puncture auto-injector and upload it to the computer processing unit to obtain the gyroscope measurement value during the puncture simulation process; use the computer processing unit to perform neural network regression deviation estimation on the gyroscope measurement value during the puncture simulation process to obtain the deterministic part of the gyroscope regression deviation; perform static deviation calibration calculation on the gyroscope measurement value during the puncture simulation process based on the deterministic part of the gyroscope regression deviation to obtain the puncture process gyroscope deviation calibration value;

[0008] A puncture training scenario path simulation generation module is used to perform puncture path simulation recognition analysis on the selected puncture scenario simulation process based on the puncture process gyroscope deviation calibration value to generate the puncture simulation training path corresponding to the selected puncture scenario simulation process;

[0009] A puncture simulation path visualization module is used to perform puncture simulation visualization processing on the puncture simulation training path corresponding to the selected puncture scenario simulation process by using the trainer user interface to execute the corresponding puncture scenario visualization teaching process.

[0010] Furthermore, the user training puncture simulation operation selection module includes the following functions:

[0011] Select the puncture scenario to be simulated through the trainer user interface to obtain the selected simulated puncture scenario, where the selected simulated puncture scenario includes a thoracic puncture simulation scenario, a lumbar puncture simulation scenario, an abdominal puncture simulation scenario, and a bone marrow puncture simulation scenario;

[0012] Configure the corresponding puncture needle and puncture injection depth operating instrument for the corresponding auto-injector according to the selected simulated puncture scenario to obtain the selected simulated puncture auto-injector;

[0013] Set the puncture scenario environment parameters for the selected simulated puncture scenario to obtain the selected simulated puncture scenario environment parameter set; perform device performance test and optimization on the selected simulated puncture auto-injector based on the selected simulated puncture scenario environment parameter set to obtain the optimized auto-injector for simulated puncture equipment;

[0014] Design the simulated puncture operation process for the optimized auto-injector for simulated puncture equipment to generate a document of the steps for designing the simulated puncture operation process;

[0015] Perform a comprehensive puncture simulation operation on the selected simulated puncture scenario based on the document of the steps for designing the simulated puncture operation process to generate the selected puncture scenario simulation process.

[0016] Furthermore, the puncture gyroscope deviation calibration module includes the following functions:

[0017] Using the gyroscope sensor in the selected simulated puncture auto-injector to perform puncture gyroscope measurement processing on the selected puncture scenario simulation process to obtain the gyroscope measurement value during the puncture simulation process; uploading the gyroscope measurement value during the real-time measurement of the puncture simulation process on the gyroscope sensor to the computer processing unit through wireless transmission technology;

[0018] Using the computer processing unit to obtain the gyroscope coordinate system of the gyroscope sensor corresponding to the puncture simulation process, and performing measurement value coordinate transformation on the gyroscope measurement value during the puncture simulation process based on the gyroscope coordinate system to obtain the corresponding gyroscope measurement value in the gyroscope coordinate system;

[0019] Performing angular velocity vector statistical analysis on the corresponding gyroscope sensor in the selected puncture scenario simulation process based on the gyroscope coordinate system of the puncture simulation process to obtain the corresponding actual angular velocity vector of the gyroscope in the gyroscope coordinate system;

[0020] Performing gyroscope error modeling on the corresponding gyroscope sensor in the selected puncture scenario simulation process according to the gyroscope measurement value corresponding in the gyroscope coordinate system and the actual angular velocity vector of the gyroscope to generate a puncture simulation gyroscope error mathematical model;

[0021] By configuring a static condition for the gyroscope sensor and performing zero-order calibration processing on the gyroscope deviation part in the puncture simulation gyroscope error mathematical model based on the static condition to obtain the zero-order calibration part of the gyroscope deviation; setting up a virtual gyroscope array and the gyroscope sensor to form a gyroscope virtual module through a resistance element, where each gyroscope virtual module consists of orthogonal gyroscopes on three direction axes, and performing neural network regression assisted calibration estimation on the zero-order calibration part of the gyroscope deviation based on the gyroscope virtual module to obtain the deterministic part of the gyroscope regression deviation;

[0022] Performing static deviation calibration calculation on the gyroscope measurement value corresponding in the gyroscope coordinate system based on the deterministic part of the gyroscope regression deviation to obtain the gyroscope deviation calibration value during the puncture process.

[0023] Further, the puncture simulation gyroscope error mathematical model is specifically:

[0024] ;

[0025] In the formula, is the gyroscope measurement value corresponding in the gyroscope coordinate system , is the actual angular velocity vector of the gyroscope corresponding in the gyroscope coordinate system , is the matrix of gyroscope non-diagonal elements and scale factor errors, is the gyroscope deviation part, is zero-mean Gaussian white noise.

[0026] Further, the zero-order calibration process for the gyroscope bias part in the puncture simulation gyroscope error mathematical model based on the stationary condition includes:

[0027] Measure relevant parameters of the puncture simulation gyroscope error mathematical model when the gyroscope is stationary for a series of times, and obtain the corresponding angular velocity measurement vector and Gaussian measurement white noise under a series of stationary conditions of the gyroscope;

[0028] Reconstruct the error model according to the corresponding angular velocity measurement vector and Gaussian measurement white noise under a series of stationary conditions of the gyroscope to obtain the gyroscope error reconstruction model under the stationary condition:

[0029] ;

[0030] In the formula, is the gyroscope measurement value corresponding to the th stationary condition, is the angular velocity measurement vector corresponding to the th stationary condition, is the gyroscope estimated bias corresponding to the th stationary condition, is the Gaussian measurement white noise corresponding to the th stationary condition;

[0031] Assume that the Gaussian measurement white noise tends to zero mean through the zero-order calibration method, then the gyroscope measurement value correspondingly tends to zero, and take the expectation operator on both sides of the above formula for zero-order calibration processing to obtain the zero-order calibration part of the gyroscope bias:

[0032] ;

[0033] where is the zero-order calibration part of the gyroscope bias.

[0034] Further, the neural network regression assisted calibration estimation for the zero-order calibration part of the gyroscope bias based on the gyroscope virtual module includes:

[0035] Design a neural network regression assisted calibration method through the gyroscope virtual module, where the neural network regression assisted calibration method is an assisted calibration method for increasing the gyroscope input channel or an assisted calibration method for increasing the real and virtual gyroscope training data;

[0036] Based on the neural network regression assisted calibration method, perform assisted calibration estimation on the zero-order calibration part of the gyroscope bias to obtain the deterministic part of the gyroscope regression bias.

[0037] Further, the assistance calibration estimation for the zero-order calibration part of the gyroscope deviation by the method of increasing the gyroscope input channel for assistance calibration includes:

[0038] Design a convolutional neural network architecture through the method of increasing the gyroscope input channel for assistance calibration. The convolutional neural network architecture consists of a convolutional layer, a LeakyReLU activation function, and a max pooling layer, and input the features of each gyroscope module into the convolutional layer: , where is the number of gyroscope virtual modules, is the window size;

[0039] At the th convolutional layer, the network outputs , where is the window kernel size, is the th gyroscope deviation value at the window kernel, is the stride, is the th weight at the window kernel; According to the network output in the th convolutional layer, the LeakyReLU activation function can be defined as ;

[0040] Based on the network output in the th convolutional layer, perform pooling calculation in the max pooling layer , where is the pooling size, and flatten it into a two-dimensional tensor and input it to calculate the first fully connected layer , where is the weight of the first fully connected layer, is the bias of the first fully connected layer;

[0041] Repeat the calculation for the second fully connected layer connected to it by introducing the LeakyReLU activation function to obtain the final neural network regression output , where and are the weight and bias of the second fully connected layer respectively; Calculate the neural network regression loss through the mean square error loss function , where is the total number of training samples;

[0042] Through inputting the zero-order calibration part of the gyroscope deviation for assistance calibration estimation, obtain the deterministic part of the gyroscope regression deviation:

[0043] .

[0044] Further, the assistance calibration estimation for the zero-order calibration part of the gyroscope deviation by the method of assisting calibration based on increasing real and virtual gyroscope training data includes:

[0045] Collect the real gyroscope measurement data of the gyroscope virtual module , where , and respectively represent the real measurement readings on the orthogonal gyroscopes of the three-axis of the gyroscope virtual module;

[0046] Generate the virtual gyroscope measurement data of the gyroscope virtual module through simulation , where , and respectively represent the virtual measurement readings on the orthogonal gyroscopes of the three-axis of the gyroscope virtual module;

[0047] Merge the real gyroscope measurement data and the virtual gyroscope measurement data into a training set , and input it into the convolutional neural network for neural network regression training to output the gyroscope deviation values on the orthogonal gyroscopes of the three axes , where is the weight matrix of the rd layer, is the neural network activation value of the th layer, is the bias term of the th layer;

[0048] Calculate the neural network regression loss according to the gyroscope deviation values on the orthogonal gyroscopes of the three axes:

[0049] ;

[0050] where is the total number of training samples, is the index of the three orthogonal gyroscopes;

[0051] Update and optimize the weight matrix and the bias term through the backpropagation method:

[0052] ;

[0053] ;

[0054] where is the learning rate;

[0055] Through the input gyroscope deviation zero-order calibration part for assistance calibration estimation, obtain the deterministic part of the gyroscope regression deviation:

[0056] 。

[0057] Further, the puncture training scenario path simulation generation module includes the following functions:

[0058] Obtain the gyroscope calibration measurement values corresponding to different time points in the selected puncture scenario simulation process through the gyroscope deviation calibration values during the puncture process;

[0059] Based on the gyroscope calibration measurement values corresponding to different time points, construct a three-dimensional virtual space scenario for the corresponding selected puncture scenario simulation process to generate a gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process;

[0060] Conduct a puncture path dynamic simulation recognition analysis on the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to generate a puncture simulation training path corresponding to the selected puncture scenario simulation process.

[0061] Further, the puncture path dynamic simulation recognition analysis of the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process includes:

[0062] Mark the puncture key nodes on the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to obtain the gyroscope puncture key nodes corresponding to the selected puncture simulation process;

[0063] Conduct a physical geometry constraint analysis on the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to obtain the gyroscope puncture physical geometry constraint conditions corresponding to the selected puncture simulation process;

[0064] Based on the gyroscope puncture physical geometry constraint conditions corresponding to the selected puncture simulation process, conduct a puncture path preset tracking analysis on the corresponding gyroscope puncture key nodes to generate a gyroscope puncture constraint preset path corresponding to the selected puncture simulation process;

[0065] Conduct a puncture process mechanical simulation analysis on the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process based on the selected puncture scenario simulation process to generate a puncture mechanical reaction simulation field corresponding to the selected puncture simulation process;

[0066] According to the puncture mechanical reaction simulation field corresponding to the selected puncture simulation process, conduct a puncture path dynamic simulation recognition analysis on the gyroscope puncture constraint preset path corresponding to the selected puncture simulation process to generate a puncture simulation training path corresponding to the selected puncture scenario simulation process.

[0067] Advantages of the present invention:

[0068] The integrated puncture simulation system based on an autoinjector proposed by the present invention is generally composed of a user training puncture simulation operation selection module, a puncture gyroscope deviation calibration module, a puncture training scenario path simulation generation module, and a puncture simulation path visualization module. Compared with the prior art, the beneficial effects of this application are as follows: By using the trainer user interface to select the puncture scenario to be simulated, a personalized and targeted puncture simulation training experience is provided for the user. The key to this step lies in that it allows the trainer to select a specific puncture type according to their own needs, such as thoracic cavity, lumbar spine, abdominal cavity, and bone marrow puncture, etc. This flexibility can not only enhance the learner's sense of active participation but also ensure training in the required actual scenario, thereby improving the effectiveness and pertinence of learning. By configuring the corresponding puncture needle and puncture injection depth operating instrument for the autoinjector according to the selected simulated puncture scenario, the key lies in selecting the appropriate puncture needle and injection depth operating instrument for different puncture types. The effectiveness of this step is reflected in the accuracy and adaptability of the equipment. By using equipment specific to a particular scenario, trainees can experience a feeling closer to real operation in the simulation environment. The appropriate instrument configuration not only ensures the safety of training but also enables learners to master the characteristics and requirements of different puncture techniques, thereby enhancing the training effect. At the same time, by using the selected simulated puncture autoinjector to perform a comprehensive puncture simulation operation on the corresponding selected simulated puncture scenario, a comprehensive practical opportunity can be provided for trainees. The key to this step is reflected in providing a platform for trainees to combine theory with practice, allowing them to perform operation exercises in a real simulation scenario. Through comprehensive simulation, trainees can experience the application of different puncture techniques, further consolidating their puncture skills and knowledge, thereby providing puncture data support for the subsequent puncture gyroscope deviation calibration process. Secondly, by using the gyroscope sensor in the selected simulated puncture autoinjector to perform puncture gyroscope measurement processing on the selected puncture scenario simulation process, the dynamic information of the gyroscope during the puncture simulation process can be effectively captured. These gyroscope measurement values reflect the attitude changes of the autoinjector device during the puncture simulation process, laying a foundation for subsequent gyroscope deviation data analysis. The data is also uploaded to the computer processing unit in real time through wireless transmission technology, ensuring the timeliness and accuracy of data transmission. This real-time nature can ensure the immediate identification and recording of any abnormal situations during the puncture simulation process, thereby improving the safety and effectiveness of the puncture operation.By using a computer processing unit to perform neural network regression deviation estimation on the gyroscope measurement values during the puncture simulation process, the deviation part of the gyroscope can be effectively corrected. Additionally, by creating a virtual gyroscope array, the accuracy and stability of deviation calibration can be further enhanced. The use of the neural network regression method for calibration estimation enables the identification and correction of complex deviation patterns, thereby significantly improving the accuracy of gyroscope measurements. This step not only reduces the errors in the use of gyroscope devices but also enhances the repeatability and reliability of puncture simulations, resulting in more consistent results in practical applications. Furthermore, by performing static deviation calibration calculations on the corresponding gyroscope measurement values in the gyroscope coordinate system based on the deterministic part of the gyroscope regression deviation, the gyroscope deviation calibration values during the puncture process can be obtained. This ensures that the data and models accumulated in the previous steps can be effectively applied to the actual scenario. Through the precise correction of static deviation, not only can the accuracy of gyroscope measurements be improved, but also the safety of puncture operations can be enhanced, thereby effectively calibrating the error terms during the gyroscope measurement process. Then, based on the gyroscope deviation calibration values obtained from the previous calibration calculations during the puncture simulation process, a puncture path dynamic simulation identification analysis is performed on the corresponding selected puncture scenario simulation process. This process can identify and generate the optimal puncture path, providing scientific guidance for the medical training of trainees. By simulating and analyzing different paths, the best practices and potential challenges in puncture operations can be better understood, thus formulating a more reasonable training plan. This dynamic simulation can not only reflect the complexity of actual operations but also adjust the path in real time, providing personalized guidance according to different situations and patient conditions. The resulting puncture simulation training path can be used as a reference for the training of medical staff, providing them with clear operation steps and techniques, thereby improving the accuracy of puncture path recognition simulation. Finally, by using the trainer user interface to visualize the puncture simulation training path, this step greatly enhances the intuitiveness and interactivity of puncture simulation learning through graphical display. Through visualization technology, trainees can clearly see every detail during the puncture process, further understanding the key points and difficulties of the operation. This visual learning tool can not only attract the attention of trainees but also enhance their learning interest and participation. The visualization teaching process makes complex puncture techniques easier to understand and master, making the entire puncture simulation process more efficient and effective. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Other features, objects, and advantages of the present invention will become more apparent from the following detailed description of non-limiting embodiments read in conjunction with the accompanying drawings:

[0070] Figure 1 Schematic diagram of the modules of the integrated puncture simulation system based on an autoinjector according to the present invention;

[0071] Figure 2 For Figure 1 Schematic diagram of the functional flow of the user training puncture simulation operation selection module in

[0072] Figure 3 For Figure 1 Schematic diagram of the functional flow of the puncture gyroscope deviation calibration module in

[0073] Figure 4 Schematic diagram of the depth of the puncture simulation of the human chest part of the present invention;

[0074] Figure 5 Schematic diagram of the depth of the puncture simulation of the human lumbar spine part of the present invention;

[0075] Figure 6 Schematic diagram of the depth of the puncture simulation of the human abdominal cavity part of the present invention;

[0076] Figure 7 Schematic diagram of the depth of the puncture simulation of the human bone marrow part of the present invention. Detailed implementation manners

[0077] The technical system of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative efforts fall within the scope of protection of the present invention.

[0078] To achieve the above object, please refer to Figures 1 to 3 , the present invention provides a comprehensive puncture simulation system based on an auto-injector, and the system includes the following modules:

[0079] A user training puncture simulation operation selection module, configured to select a puncture scenario to be simulated through a trainer user interface, and configure a corresponding puncture needle and a puncture injection depth operating instrument for the auto-injector according to the selected puncture scenario to obtain a selected simulated puncture auto-injector; perform a comprehensive puncture simulation operation on the corresponding selected puncture scenario by using the selected simulated puncture auto-injector to generate a simulation process of the selected puncture scenario;

[0080] A puncture gyroscope deviation calibration module is used to perform puncture gyroscope measurement processing on the selected puncture scenario simulation process by using the gyroscope sensor in the selected analog puncture auto-injector and upload it to the computer processing unit to obtain the gyroscope measurement value during the puncture simulation process; use the computer processing unit to perform neural network regression deviation estimation on the gyroscope measurement value during the puncture simulation process to obtain the deterministic part of the gyroscope regression deviation; perform static deviation calibration calculation on the gyroscope measurement value during the puncture simulation process based on the deterministic part of the gyroscope regression deviation to obtain the puncture process gyroscope deviation calibration value.

[0081] A puncture training scenario path simulation generation module is used to perform puncture path simulation recognition analysis on the selected puncture scenario simulation process based on the puncture process gyroscope deviation calibration value to generate a puncture simulation training path corresponding to the selected puncture scenario simulation process.

[0082] A puncture simulation path visualization module is used to perform puncture simulation visualization processing on the puncture simulation training path corresponding to the selected puncture scenario simulation process by using the trainer user interface to execute the corresponding puncture scenario visualization teaching process.

[0083] In an embodiment of the present invention, please refer to Figure 1 As shown, it is a module schematic diagram of the integrated puncture simulation system based on an auto-injector of the present invention. In this example, the integrated puncture simulation system based on an auto-injector includes the following modules:

[0084] S1: A user training puncture simulation operation selection module is used to select the puncture scenario to be simulated through the trainer user interface, and configure the corresponding puncture needle and puncture injection depth operating instrument for the auto-injector according to the selected puncture scenario to obtain the selected analog puncture auto-injector; use the selected analog puncture auto-injector to perform a comprehensive puncture simulation operation on the corresponding selected puncture scenario to generate the selected puncture scenario simulation process.

[0085] In an embodiment of the present invention, by using the trainer user interface in the system to select the puncture scenario to be simulated, which includes thoracentesis (as shown in Figure 4 ), lumbar puncture (as shown in Figure 5 ), abdominal puncture (as shown in Figure 6 ), and bone marrow puncture (as shown in Figure 7As shown in the figure), etc. The user selects a scenario through an interactive interface, and the interface can display the detailed information and applicable instructions of each scenario in real time, so as to select the puncture scenario for the selected simulation training. By configuring the corresponding puncture injection depth operating instrument and puncture needle for the automatic syringe corresponding to the selected simulation puncture scenario according to the previously trained simulation puncture scenario, for each puncture scenario, determine the type and size of the puncture needle, as well as the parameters of the injection depth operating instrument. For example, a rubber tube puncture needle, a 5 ml syringe, and a 50 ml syringe are used for thoracic puncture; a rubber tube puncture needle, a 5 ml syringe, and a 50 ml syringe are also used for abdominal puncture; a lumbar puncture needle, a 5 ml syringe, a manometer tube, and a test tube are used for abdominal puncture; and a bone marrow puncture needle, a 5 ml syringe, and a 20 ml syringe are used for bone marrow puncture. This process uses a professional medical device configuration tool to ensure that all components meet medical standards and technical requirements. After configuration, the automatic syringe and the puncture instrument are assembled to obtain the selected simulation puncture automatic syringe. Then, through the use of the configured and assembled selected simulation puncture automatic syringe, a comprehensive simulation operation of puncturing the previously selected training selected simulation puncture scenario is carried out to execute the entire puncture process through the automatic syringe, real-time monitor various parameters and feedback information, and enhance the simulation experience through the use of virtual reality technology. Through the images and sounds generated by the simulation system, the operation is made more immersive and realistic. After completion, all data during the simulation process are recorded to provide a basis for subsequent analysis and evaluation, ensuring the effectiveness and scientific nature of the training, and finally recording and generating the simulation process of the selected puncture scenario.

[0086] Among them, the automatic syringe includes a housing and a push rod that is slidably fitted within the housing. The front end of the push rod is hermetically and slidably connected to the inner wall of the housing, so as to form an adjustable-length accommodation chamber between the front end face of the push rod and the front wall of the housing. An injection interface communicating with the accommodation chamber is provided at the front end of the housing, and the injection interface can be configured with a corresponding puncture needle. For example, the rubber tube puncture needle, lumbar puncture needle, and bone marrow puncture needle in the present invention. An opening is provided at the rear end of the housing for the push rod to be inserted. A resistance element is provided on the push rod along the length direction, and the resistance element is electrically connected to a detection circuit. In addition, the resistance element can form a gyroscope virtual module corresponding to the gyroscope sensor used in the present invention, which includes orthogonal gyroscopes on three direction axes, so that the gyroscope measurement values real-time monitored by the gyroscope virtual module can be transmitted to the computer processing unit electrically connected to the gyroscope sensor through the detection circuit.

[0087] S2: Puncture gyroscope deviation calibration module, which is used to perform puncture gyroscope measurement processing on the selected puncture scenario simulation process by using the gyroscope sensor in the selected simulated puncture auto-injector and upload it to the computer processing unit to obtain the gyroscope measurement value during the puncture simulation process; use the computer processing unit to perform neural network regression deviation estimation on the gyroscope measurement value during the puncture simulation process to obtain the deterministic part of the gyroscope regression deviation; perform static deviation calibration calculation on the gyroscope measurement value during the puncture simulation process based on the deterministic part of the gyroscope regression deviation to obtain the gyroscope deviation calibration value during the puncture process.

[0088] In the embodiment of the present invention, the gyroscope attitude change during the selected puncture scenario simulation process is monitored in real time by using the gyroscope sensor embedded in the previously configured selected simulated puncture auto-injector. During the puncture simulation process, the gyroscope sensor collects and records the gyroscope measurement values, including angular velocity and acceleration data, so as to obtain the gyroscope measurement value during the puncture simulation process. At the same time, the gyroscope measurement value during the puncture simulation process measured in real time on the gyroscope sensor is quickly uploaded to the computer processing unit electrically connected thereto by using wireless transmission technologies such as Bluetooth or Wi-Fi to ensure the integrity and timeliness of data transmission. By using the computer processing unit to receive the gyroscope measurement value during the puncture simulation process obtained by the previous real-time monitoring and convert it to the corresponding gyroscope coordinate system during the puncture simulation process, it is ensured that the data for subsequent analysis accurately reflects the actual motion state during the puncture process, and through combining the previously obtained gyroscope coordinate system, statistical analysis of the angular velocity vector of the corresponding gyroscope sensor in the selected puncture scenario simulation process is carried out, so as to calculate the actual angular velocity vector in the gyroscope coordinate system by aggregating the measurement data at each time point by using the statistical analysis tool. At the same time, by combining the previously analyzed gyroscope measurement value and the actual angular velocity vector of the gyroscope in the gyroscope coordinate system, a corresponding gyroscope error mathematical calculation model is established, and by using the least squares method or the Kalman filter algorithm, the collected data is fitted and modeled to identify the gyroscope system deviation and Gaussian white noise, forming a dedicated error model suitable for this puncture scenario, for example: . Then, by configuring corresponding static state conditions for it to perform zero-order calibration processing, that is, placing the gyroscope sensor in a static state, collecting the gyroscope measurement value in this state, extracting its deviation part, and at the same time placing the corresponding gyroscope sensor under multiple previously configured static state conditions for measurement to obtain a series of angular velocity measurement vectors corresponding to the gyroscope under the static conditions and Gaussian measurement white noise , and the obtained angular velocity measurement vector and Gaussian measurement white noise Input it into the previously constructed mathematical model of the error of the puncture simulation gyroscope for error model reconstruction to obtain the corresponding gyroscope error reconstruction model ; where is the gyroscope measurement value corresponding to the -th static condition, is the angular velocity measurement vector corresponding to the -th static condition, is the gyroscope estimation deviation corresponding to the -th static condition, is the Gaussian measurement white noise corresponding to the -th static condition; also, by using the zero-order calibration method, the Gaussian measurement white noise assumption is made to tend to zero mean. Since the gyroscope is in a static state, the gyroscope measurement value also correspondingly tends to zero, and the expectation operator is taken on both sides of the above gyroscope error reconstruction model for zero-order calibration processing, which is also approximately equal to the zero-order calibration part:

[0089] ;

[0090] Thus, the zero-order calibration part of the gyroscope deviation is obtained. Subsequently, a virtual gyroscope array including multiple orthogonal gyroscopes is simulated and constructed through the resistance elements configured on the automatic syringe, which forms a corresponding gyroscope virtual module with the actual gyroscope sensor, and a corresponding neural network regression environment is built by applying the gyroscope virtual module to improve the performance of gyroscope calibration. This environment utilizes a high-performance computing platform, combines the existing convolutional neural network model with the gyroscope sensor data, and uses the real gyroscope and the virtual gyroscope array to assist in calibration, that is, a single gyroscope module composed of three gyroscopes adopts the corresponding neural network regression calibration method, specifically the method of assisting calibration by increasing the input channels of the gyroscope: by designing a convolutional neural network architecture, which consists of a convolutional layer, a LeakyReLU activation function, and a max-pooling layer, and by inputting the features of each gyroscope module in the convolutional layer: , where is the number of gyroscope virtual modules, is the window size; in the -th convolutional layer, the network outputs , where is the window kernel size, is the gyroscope deviation value at the -th window kernel, is the stride, is the weight at the -th window kernel; according to the network output in the -th convolutional layer, the LeakyReLU activation function can be defined as ; based on the The network output in the convolutional layer is pooled in the max pooling layer , where is the pooling size, and it is flattened into a two-dimensional tensor to calculate the first fully connected layer , where is the weight of the first fully connected layer, is the bias of the first fully connected layer; the second fully connected layer connected to it is repeatedly calculated by introducing the LeakyReLU activation function to obtain the final neural network regression output , where and are the weight and bias of the second fully connected layer respectively; the neural network regression loss is calculated by the mean square error loss function , where is the total number of training samples; then the regression bias deterministic part of the method is obtained by assisting calibration estimation through the input gyroscope bias zero-order calibration part ; or an assisted calibration method that increases real and virtual gyroscope training data: through the real gyroscope measurement data of the gyroscope virtual module , where , and represent the real measurement readings on the orthogonal gyroscopes of the gyroscope virtual module in three direction axes respectively; the virtual gyroscope measurement data of the gyroscope virtual module is generated by simulation , where , and represent the virtual measurement readings on the orthogonal gyroscopes of the gyroscope virtual module in three direction axes respectively; the real gyroscope measurement data and the virtual gyroscope measurement data are combined into a training set , and it is input into the convolutional neural network for neural network regression training to obtain the gyroscope bias values on the orthogonal gyroscopes of the three direction axes , where is the weight matrix of the th layer, is the neural network activation value of the th layer, is the bias term of the th layer; the neural network regression loss is calculated according to the gyroscope bias values on the orthogonal gyroscopes of the three direction axes: ; where is the total number of training samples, is the index of the three orthogonal gyroscopes; the weight matrix and the bias term are updated and optimized by the backpropagation method:

[0091] ;

[0092] ;

[0093] wherein is the learning rate; assisted calibration estimation is performed through the input gyroscope deviation zero-order calibration part to obtain the deterministic part of the regression deviation of this method: , the regression calibration method of any of the above neural network models can continuously optimize the model parameters through supervised learning to reduce the error between the model prediction output and the actual gyroscope data, so as to regress and calculate the corresponding deterministic part of the gyroscope regression deviation. Finally, by combining the deterministic part of the gyroscope regression deviation obtained from the previous calibration estimation calibration calculation of the deviation of the gyroscope measurement value corresponding in the gyroscope coordinate system is performed to subtract the calculated deterministic deviation from the gyroscope measurement value: , and finally the gyroscope deviation calibration value during the puncture process is obtained.

[0094] S3: Puncture training scenario path simulation generation module, which is used to perform puncture path simulation recognition analysis on the selected puncture scenario simulation process based on the gyroscope deviation calibration value during the puncture process to generate the puncture simulation training path corresponding to the selected puncture scenario simulation process;

[0095] In an embodiment of the present invention, by analyzing the temporal variation of the gyroscope deviation calibration values obtained after real-time gyroscope deviation calibration during the previous puncture process, the measured calibration values of the gyroscope at each time point during the simulation process of the selected puncture scenario are analyzed and recorded. By combining the corresponding gyroscope calibration measurement values at different time points obtained from the previous analysis, a virtual space scenario of the corresponding selected puncture scenario simulation process is constructed using 3D modeling software. Based on the gyroscope measurement calibration data at different time points, a three-dimensional coordinate system is established, and an accurate three-dimensional model is drawn according to the dynamic changes of the puncture path, puncture angle, and time. The model includes the movement trajectory of the puncture needle and the morphology of the target tissue. These elements are integrated together through numerical calculations to form a three-dimensional virtual space model of the complete puncture simulation process, which intuitively reflects the changes in the puncture process at different time points. Then, by using dynamic simulation analysis tools (including 3D modeling software (such as Blender or Maya), finite element analysis software (such as ANSYS or COMSOL), and mechanical simulation software (such as MATLAB or Simulink)) to simulate and identify the puncture path of the previously constructed three-dimensional virtual space model, this process simulates the movement trajectory of the puncture needle through an algorithm and interacts with the physical properties of the target tissue in real time. By using computer vision technology and machine learning algorithms, the effectiveness and safety of different puncture paths are analyzed. The dynamic simulation tool can evaluate the possible risks during the puncture process, generate a visual training path map, and the output puncture simulation training path can provide a reference for actual puncture operations to ensure the accuracy and safety of the operation. Finally, a puncture simulation training path corresponding to the selected puncture scenario simulation process is simulated and generated.

[0096] S4: A puncture simulation path visualization module, which is used to perform puncture simulation visualization processing on the puncture simulation training path corresponding to the selected puncture scenario simulation process by using a trainer user interface, so as to execute the corresponding puncture scenario visualization teaching process.

[0097] In an embodiment of the present invention, the puncture simulation training path corresponding to the generated selected puncture scenario simulation process is visualized by using a trainer user interface. Through 3D visualization software, the simulation path is graphically displayed, and the model of the real scenario is superimposed on the path to ensure that the user can intuitively understand the puncture process. During the visualization process, the operator can view the details of each step through interface interaction and learn and train in combination with the corresponding teaching content. This visualization processing improves the learning efficiency, makes the teaching process of the puncture scenario more vivid and specific, and thus enhances the practical operation ability of the operator.

[0098] Further, as an embodiment of the present invention, with reference to Figure 2 shown, it is Figure 1Schematic diagram of the functional flow of the user training puncture simulation operation selection module. In this embodiment, the user training puncture simulation operation selection module includes the following functions:

[0099] S11: Select the puncture scenario to be simulated through the trainer user interface to obtain the selected simulated puncture scenario for training, where the selected simulated puncture scenario for training includes a thoracic puncture simulation scenario, a lumbar puncture simulation scenario, an abdominal puncture simulation scenario, and a bone marrow puncture simulation scenario;

[0100] In the embodiment of the present invention, the puncture scenario to be simulated is selected through the trainer user interface in the system, including thoracic puncture, lumbar puncture, abdominal puncture, bone marrow puncture, etc. The user selects the scenario through the interactive interface, and the interface can display the detailed information and applicable instructions of each scenario in real time. The system records the selected scenario and generates a corresponding configuration file to ensure that the specific parameters and requirements of the scenario can be accurately referenced in the subsequent steps, and finally the selected simulated puncture scenario for training is obtained.

[0101] S12: Configure the corresponding puncture needle and puncture injection depth operating instrument for the corresponding automatic syringe according to the selected simulated puncture scenario for training to obtain the selected simulated puncture automatic syringe;

[0102] In the embodiment of the present invention, the corresponding puncture injection depth operating instrument and puncture needle are configured for the corresponding automatic syringe according to the previously selected simulated puncture scenario. For each puncture scenario, the type and size of the puncture needle and the parameters of the injection depth operating instrument are determined. For example, a rubber tube puncture needle, a 5 ml syringe, and a 50 ml syringe are used for thoracic puncture; a rubber tube puncture needle, a 5 ml syringe, and a 50 ml syringe are also used for abdominal puncture; a lumbar puncture needle, a 5 ml syringe, a manometer tube, and a test tube are used for abdominal puncture; and a bone marrow puncture needle, a 5 ml syringe, and a 20 ml syringe are used for bone marrow puncture. This process uses a professional medical device configuration tool to ensure that all components meet medical standards and technical requirements. After configuration, the automatic syringe and the puncture instrument are assembled to form an automatic syringe specifically for the selected simulated puncture scenario, and finally the selected simulated puncture automatic syringe is obtained.

[0103] S13: Set the puncture scenario environment parameters for the selected simulated puncture scenario for training to obtain the selected simulated puncture scenario environment parameter set; based on the selected simulated puncture scenario environment parameter set, perform device performance test and optimization on the selected simulated puncture automatic syringe to obtain the optimized automatic syringe for the simulated puncture device;

[0104] In an embodiment of the present invention, environmental parameters of the previously selected training selected simulated puncture scenario are set to establish a scenario model by using environmental simulation software, and key environmental parameters related to puncture are input, such as the tissue density, blood vessel distribution, and organ position of a simulated person. After confirming that all environmental variable settings are correct through a debugging interface, an environmental parameter set is generated. These parameters will be used for subsequent device testing and simulation operations to ensure the realism and effectiveness of the simulation, thereby obtaining the environmental parameter set of the selected simulated puncture scenario. At the same time, the performance of the corresponding puncture auto-injector is tested and optimized by combining the environmental parameter set of the previously set selected simulated puncture scenario. By using a performance test platform to simulate different puncture conditions, by gradually increasing the puncture depth and adjusting the puncture speed, the reaction and output of the device are monitored. During this process, data is collected to evaluate the stability and accuracy of the device, and the parameters of the device are adjusted in a timely manner to ensure that an optimized simulated puncture device is finally obtained, meeting all safety and performance standards, and finally obtaining an optimized auto-injector for the simulated puncture device.

[0105] S14: Design the simulated puncture operation process for the optimized auto-injector of the simulated puncture device to generate a document of the steps of the simulated puncture operation process design;

[0106] In an embodiment of the present invention, the operation process of the optimized puncture auto-injector is designed by using a standardized operation process document template to record each operation step in detail, including the preparations before puncture, the key actions during puncture, and the treatment methods after puncture. Through multiple rehearsals and feedback, the operation process is gradually improved to ensure its clear logic and easy execution, and finally a document of the steps of the simulated puncture operation process design is generated.

[0107] S15: Perform a comprehensive simulated puncture operation on the training selected simulated puncture scenario based on the document of the steps of the simulated puncture operation process design to generate the simulation process of the selected puncture scenario.

[0108] In an embodiment of the present invention, a comprehensive simulated operation is performed on the selected simulated puncture scenario by combining the document of the steps of the simulated puncture operation process designed previously. The entire puncture process is executed by the auto-injector, and various parameters and feedback information are monitored in real time. By using virtual reality technology to enhance the simulation experience, through the images and sounds generated by the simulation system, the operation is made more immersive and realistic. After completion, all data during the simulation process is recorded to provide a basis for subsequent analysis and evaluation, ensuring the effectiveness and scientific nature of the training, and finally recording and generating the simulation process of the selected puncture scenario.

[0109] Further, as an embodiment of the present invention, refer to Figure 3 as shown, for Figure 1Schematic diagram of the functional process of the middle puncture gyroscope deviation calibration module. In this embodiment, the puncture gyroscope deviation calibration module includes the following functions:

[0110] S21: Use the gyroscope sensor in the selected simulated puncture automatic syringe to perform puncture gyroscope measurement processing on the selected puncture scenario simulation process to obtain the gyroscope measurement value during the puncture simulation process; upload the gyroscope measurement value during the puncture simulation process measured in real time on the gyroscope sensor to the computer processing unit through wireless transmission technology;

[0111] In the embodiment of the present invention, by using the gyroscope sensor embedded in the selected simulated puncture automatic syringe configured previously to monitor the gyroscope attitude change during the selected puncture scenario simulation process in real time. During the puncture simulation process, the gyroscope sensor collects and records the gyroscope measurement values, including angular velocity and acceleration data, so as to obtain the gyroscope measurement value during the puncture simulation process. At the same time, by adopting wireless transmission technologies such as Bluetooth or Wi-Fi, the gyroscope measurement value during the puncture simulation process measured in real time on the gyroscope sensor is quickly uploaded to the computer processing unit electrically connected thereto. This computer processing unit can receive and display the data in real time to ensure the integrity and timeliness of data transmission.

[0112] S22: Use the computer processing unit to obtain the gyroscope coordinate system corresponding to the puncture simulation process of the gyroscope sensor, and perform measurement value coordinate system conversion on the gyroscope measurement value during the puncture simulation process based on the gyroscope coordinate system to obtain the corresponding gyroscope measurement value in the gyroscope coordinate system;

[0113] In the embodiment of the present invention, by using the computer processing unit to receive the gyroscope measurement value during the puncture simulation process monitored in real time previously and convert it to the corresponding gyroscope coordinate system during the puncture simulation process. First, by defining the gyroscope coordinate system, ensure that the origin of the coordinate system coincides with the position of the gyroscope sensor. Then, through the mathematical conversion formula, perform coordinate system conversion on the measurement value during the puncture simulation process to obtain the corresponding measurement value in the gyroscope coordinate system. This process ensures that the data for subsequent analysis accurately reflects the actual motion state during the puncture process, and finally obtains the corresponding gyroscope measurement value in the gyroscope coordinate system.

[0114] S23: Perform angular velocity vector statistical analysis on the corresponding gyroscope sensor in the selected puncture scenario simulation process based on the gyroscope coordinate system of the puncture simulation process to obtain the corresponding actual angular velocity vector of the gyroscope in the gyroscope coordinate system;

[0115] In an embodiment of the present invention, by combining the gyroscope coordinate system of the previous obtained puncture simulation process, statistical analysis of the angular velocity vector of the corresponding gyroscope sensor in the selected puncture scenario simulation process is performed. By using statistical analysis tools, the measurement data at each time point is aggregated to calculate the actual angular velocity vector in the gyroscope coordinate system. This analysis process takes into account multiple factors, such as the sensitivity of the sensor and changes in the simulation environment, to ensure that the obtained angular velocity vector can accurately reflect the dynamic changes during the puncture operation, and finally obtain the actual angular velocity vector of the gyroscope corresponding to the gyroscope coordinate system.

[0116] S24: Perform gyroscope error modeling on the corresponding gyroscope sensor in the selected puncture scenario simulation process according to the gyroscope measurement value and the actual angular velocity vector of the gyroscope corresponding to the gyroscope coordinate system, so as to generate a puncture simulation gyroscope error mathematical model;

[0117] In an embodiment of the present invention, a corresponding gyroscope error mathematical calculation model is established by combining the gyroscope measurement value and the actual angular velocity vector of the gyroscope corresponding to the gyroscope coordinate system obtained from the previous analysis. By using the least squares method or the Kalman filter algorithm, the collected data is fitted and modeled to identify the gyroscope system deviation and Gaussian white noise. This modeling process will help the calibration work in the subsequent steps to ensure the measurement accuracy and reliability of the puncture simulation device under different conditions, form a dedicated error model suitable for this puncture scenario, and finally model and generate a puncture simulation gyroscope error mathematical model.

[0118] S25: Configure a stationary condition for the gyroscope sensor, and perform zero-order calibration processing on the gyroscope deviation part in the puncture simulation gyroscope error mathematical model to obtain the zero-order calibration part of the gyroscope deviation; Set up a virtual gyroscope array and the gyroscope sensor to form a gyroscope virtual module through a resistance element, where each gyroscope virtual module consists of orthogonal gyroscopes on three direction axes, and perform neural network regression assisted calibration estimation on the zero-order calibration part of the gyroscope deviation based on the gyroscope virtual module to obtain the deterministic part of the gyroscope regression deviation;

[0119] In an embodiment of the present invention, by configuring a corresponding stationary state condition for the gyroscope sensor on the auto-injector for zero-order calibration processing, that is, placing the gyroscope sensor in a stationary state, collecting the gyroscope measurement values in this state, extracting its deviation part, and at the same time placing the corresponding gyroscope sensor under multiple previously configured stationary state conditions for measurement to obtain the angular velocity measurement vector corresponding to a series of stationary conditions of the gyroscope and Gaussian measurement white noise and the obtained angular velocity measurement vector and Gaussian measurement white noise Input it into the previously constructed mathematical model of the error of the puncture simulation gyroscope for error model reconstruction to obtain the corresponding gyroscope error reconstruction model ; where is the gyroscope measurement value corresponding to the th static condition, is the angular velocity measurement vector corresponding to the th static condition, is the gyroscope estimation deviation corresponding to the th static condition, is the Gaussian measurement white noise corresponding to the th static condition; also, by using the zero-order calibration method, the Gaussian measurement white noise assumption is made to tend to zero mean. Since the gyroscope is in a static state, the gyroscope measurement value also correspondingly tends to zero, and the expectation operator is taken on both sides of the above gyroscope error reconstruction model for zero-order calibration processing, which is also approximately equal to the zero-order calibration part:

[0120] ;

[0121] Thus, the zero-order calibration part of the gyroscope deviation is obtained. Subsequently, a virtual gyroscope array including multiple orthogonal gyroscopes is simulated and constructed through the resistance elements configured on the automatic syringe, which forms a corresponding gyroscope virtual module with the actual gyroscope sensor, and a corresponding neural network regression environment is built by applying the gyroscope virtual module to improve the performance of gyroscope calibration. This environment utilizes a high-performance computing platform, combines the existing convolutional neural network model with the gyroscope sensor data, and uses the real gyroscope and the virtual gyroscope array to assist in calibration, that is, a single gyroscope module composed of three gyroscopes adopts the corresponding neural network regression calibration method, specifically the method of assisting calibration by increasing the input channels of the gyroscope: by designing a convolutional neural network architecture, which consists of a convolutional layer, a LeakyReLU activation function, and a max pooling layer, and by inputting the features of each gyroscope module in the convolutional layer: where is the number of gyroscope virtual modules, is the window size; in the rd convolutional layer, the network outputs where is the window kernel size, is the gyroscope deviation value at the th window kernel, is the stride, is the weight at the th window kernel; according to the network output in the th convolutional layer, the LeakyReLU activation function can be defined as ; based on the The network output in the convolutional layer is pooled in the max - pooling layer , where is the pooling size, and it is flattened into a two - dimensional tensor to calculate the first fully - connected layer , where is the weight of the first fully - connected layer, is the bias of the first fully - connected layer; the second fully - connected layer connected to it is repeatedly calculated by introducing the LeakyReLU activation function to obtain the final neural network regression output , where and are the weight and bias of the second fully - connected layer respectively; the neural network regression loss is calculated by the mean - squared error loss function , where is the total number of training samples; then, assisted calibration estimation is carried out through the input gyroscope bias zero - order calibration part to obtain the deterministic part of the gyroscope regression bias: ; or an assisted calibration method that increases real and virtual gyroscope training data: through the real gyroscope measurement data of the gyroscope virtual module , where , and represent the real measurement readings on the orthogonal gyroscopes of the gyroscope virtual module in three direction axes respectively; virtual gyroscope measurement data of the gyroscope virtual module is generated by simulation , where , and represent the virtual measurement readings on the orthogonal gyroscopes of the gyroscope virtual module in three direction axes respectively; the real gyroscope measurement data and the virtual gyroscope measurement data are combined into a training set , and it is input into the convolutional neural network for neural network regression training to obtain the gyroscope bias values on the orthogonal gyroscopes of the three direction axes , where is the weight matrix of the th layer, is the neural network activation value of the th layer, is the bias term of the th layer; the neural network regression loss is calculated according to the gyroscope bias values on the orthogonal gyroscopes of the three direction axes: ; where is the total number of training samples, is the index of the three orthogonal gyroscopes; the weight matrix and the bias term are updated and optimized by the back - propagation method:

[0122] ;

[0123] ;

[0124] wherein is the learning rate; assisted calibration estimation is performed through the input gyroscope bias zero-order calibration part to obtain the deterministic part of the gyroscope regression bias: , the regression calibration method of any of the above neural network models can continuously optimize the model parameters through supervised learning to reduce the error between the model prediction output and the actual gyroscope data, and finally obtain the corresponding deterministic part of the gyroscope regression bias through regression calculation.

[0125] S26: Perform static bias calibration calculation on the gyroscope measurement value corresponding to the gyroscope coordinate system based on the deterministic part of the gyroscope regression bias to obtain the gyroscope bias calibration value during the puncture process.

[0126] In the embodiment of the present invention, by combining the deterministic part of the gyroscope regression bias obtained through previous calibration estimation perform bias calibration calculation on the gyroscope measurement value corresponding to the gyroscope coordinate system to subtract the calculated deterministic bias from the gyroscope measurement value: , and finally obtain the gyroscope bias calibration value during the puncture process.

[0127] Furthermore, the specific mathematical model of the puncture simulation gyroscope error is as follows:

[0128] ;

[0129] In the formula, is the gyroscope measurement value corresponding to the gyroscope coordinate system , is the actual angular velocity vector of the gyroscope corresponding to the gyroscope coordinate system , is the matrix of gyroscope non-diagonal elements and scale factor errors, is the gyroscope bias part, is zero-mean Gaussian white noise.

[0130] The present invention obtains a puncture simulation gyroscope error mathematical model through the use of a specific mathematical model and verification, which is used to describe the gyroscope measurement deviation situation of the corresponding gyroscope sensor during the puncture simulation process in the selected puncture scenario simulation process. This mathematical model fully considers the gyroscope measurement value corresponding to the gyroscope coordinate system , the actual angular velocity vector of the gyroscope corresponding to the gyroscope coordinate system , and the matrix of gyroscope non-diagonal elements and scale factor errors , the gyroscope bias part , zero-mean Gaussian white noise The mutual correlation relationships among the above parameters constitute a functional relationship , This mathematical model can effectively correct the measured values of the gyroscope, thereby improving the measurement accuracy of the angular velocity during the puncture process. The bias part and noise modeling in the model can reduce measurement errors, making the obtained angular velocity vector more accurate. At the same time, this mathematical model provides a systematic way to analyze the error sources of the gyroscope, including off-diagonal elements and scale factor errors (matrix M), which can lead to inconsistent measured values in different directions. These effects can be quantified through the model. Gyroscope bias is often inevitable during actual use, and its influence can be reduced through calibration. Gaussian white noise can be statistically analyzed through noise modeling, thereby improving the stability of the measurement. Through the modeling and calibration of gyroscope errors, the robustness against external interference and internal noise can be greatly enhanced. In puncture simulation, the gyroscope can more accurately reflect the actual operation situation, thereby reducing the failure rate and improving safety. By establishing a mathematical model, the error sources and correction process are traceable, facilitating subsequent research and improvement. For future puncture processes or other related applications, corresponding optimizations can be made based on the previous model and data, thus providing a powerful tool for improving the measurement accuracy, robustness, and real-time performance of the puncture simulation process.

[0131] Furthermore, the puncture training scenario path simulation generation module includes the following functions:

[0132] Obtain the gyroscope calibration measurement values corresponding to different time points in the selected puncture scenario simulation process through the gyroscope bias calibration value of the puncture process

[0133] In the embodiment of the present invention, by analyzing the temporal variation of the gyroscope bias calibration value of the puncture process obtained after real-time gyroscope bias calibration, the calibration measurement values of the gyroscope at each time point in the selected puncture scenario simulation process are analyzed and recorded, and it is ensured that the measured values at each time point have been accurately calibrated, thereby eliminating the bias caused by equipment errors or environmental factors. By obtaining these calibration measurement values through temporal analysis, the dynamic changes of the puncture process can be better understood, and finally the gyroscope calibration measurement values corresponding to different time points in the selected puncture scenario simulation process are obtained.

[0134] Preferably, based on the gyroscope calibration measurement values corresponding to different time points, a three-dimensional virtual space scenario of the corresponding selected puncture scenario simulation process is constructed to generate a gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process

[0135] In the embodiments of the present invention, by combining the gyroscope calibration measurement values corresponding to different time points obtained from the previous analysis, a virtual space scene of the corresponding selected puncture scenario simulation process is constructed using three-dimensional modeling software. Based on the gyroscope measurement calibration data at different time points, a three-dimensional coordinate system is established, and an accurate three-dimensional model is drawn according to the dynamic changes of the puncture path, puncture angle, and time. The model includes the movement trajectory of the puncture needle and the morphology of the target tissue. These elements are integrated together through numerical calculation to form a three-dimensional virtual space model of the complete puncture simulation process. This model not only has high precision but also can intuitively reflect the changes in the puncture process at different time points. Finally, a gyroscopic puncture three-dimensional virtual space model corresponding to the selected puncture simulation process is generated through modeling.

[0136] Preferably, a puncture path dynamic simulation recognition analysis is performed on the gyroscopic puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to generate a puncture simulation training path corresponding to the selected puncture scenario simulation process.

[0137] In the embodiments of the present invention, by using dynamic simulation analysis tools (including three-dimensional modeling software (such as Blender or Maya), finite element analysis software (such as ANSYS or COMSOL), and mechanical simulation software (such as MATLAB or Simulink)) to perform simulation recognition and analysis of the puncture path on the previously constructed gyroscopic puncture three-dimensional virtual space model. This process simulates the movement trajectory of the puncture needle through an algorithm and interacts with the physical properties of the target tissue in real time. By using computer vision technology and machine learning algorithms, the effectiveness and safety of different puncture paths are analyzed. The dynamic simulation tool can evaluate the possible risks during the puncture process, generate a visual training path map, and the output puncture simulation training path can provide a reference basis for actual puncture operations to ensure the accuracy and safety of the operation. Finally, a puncture simulation training path corresponding to the selected puncture scenario simulation process is generated through simulation.

[0138] Further, the puncture path dynamic simulation recognition analysis of the gyroscopic puncture three-dimensional virtual space model corresponding to the selected puncture simulation process includes:

[0139] Label the puncture key nodes of the gyroscopic puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to obtain the gyroscopic puncture key nodes corresponding to the selected puncture simulation process;

[0140] In an embodiment of the present invention, during the selected puncture simulation process, first, a corresponding three-dimensional virtual space model of gyroscopic puncture is constructed, and using three-dimensional modeling software (such as Blender or Maya), after importing the model, key nodes are marked, where the key nodes include the starting point of the puncture, the puncture depth, and the boundary of the target tissue. By adding marks of key nodes to the previously constructed three-dimensional virtual space model of gyroscopic puncture, it is ensured that these key nodes have clear coordinate information, and the coordinates of each key node are recorded through a three-dimensional coordinate system, and finally, the key nodes of gyroscopic puncture corresponding to the selected puncture simulation process are obtained.

[0141] Preferably, a physical and geometric constraint analysis is performed on the three-dimensional virtual space model of gyroscopic puncture corresponding to the selected puncture simulation process to obtain the physical and geometric constraint conditions of gyroscopic puncture corresponding to the selected puncture simulation process;

[0142] In an embodiment of the present invention, through a physical and geometric constraint analysis of the three-dimensional virtual space model of gyroscopic puncture corresponding to the selected puncture simulation process obtained by previous modeling, by applying finite element analysis software (such as ANSYS or COMSOL), the mechanical properties of the model are simulated, and the material properties, boundary conditions, and applied forces are set. Through mesh generation, the physical and geometric constraint conditions of the model are obtained. This process ensures that the model can truly reflect the physical behavior during the puncture simulation and clarifies the interactions between various components, especially the physical and geometric stress distribution that may be encountered during the puncture process. Finally, the physical and geometric constraint conditions of gyroscopic puncture corresponding to the selected puncture simulation process are obtained.

[0143] Preferably, a puncture path preset tracking analysis is performed on the key nodes of gyroscopic puncture corresponding to the physical and geometric constraint conditions of gyroscopic puncture corresponding to the selected puncture simulation process to generate a preset puncture constraint path corresponding to the selected puncture simulation process;

[0144] In an embodiment of the present invention, through a puncture path preset tracking analysis of the key nodes of gyroscopic puncture corresponding to the physical and geometric constraint conditions of gyroscopic puncture corresponding to the selected puncture simulation process obtained by previous simulation constraint analysis, by using a path planning algorithm (such as the A* algorithm or Dijkstra algorithm), for the key nodes of gyroscopic puncture, an initial puncture path is generated under the corresponding physical and geometric constraint conditions. This path will consider the distance, direction, and constraint conditions between the key nodes to ensure the smooth progress of the puncture process. The generated preset puncture constraint path is recorded in digital form and provided for use in subsequent dynamic simulation processes. Finally, a preset puncture constraint path corresponding to the selected puncture simulation process is generated by tracking.

[0145] Preferably, based on the selected puncture scenario simulation process, a puncture process mechanical simulation analysis is performed on the gyroscopic puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to generate a puncture mechanical response simulation field corresponding to the selected puncture simulation process.

[0146] In an embodiment of the present invention, by combining the selected puncture scenario simulation process, a puncture process mechanical simulation analysis is performed on the gyroscopic puncture three-dimensional virtual space model generated by corresponding modeling. By using mechanical simulation software (such as MATLAB or Simulink), a puncture mechanical response simulation is implemented on the three-dimensional virtual space model. In this process, initial conditions and applied forces are set, and the puncture process is simulated through time steps to generate a mechanical response simulation field actually generated during the puncture scenario simulation process. This simulation not only visualizes the forces and responses required during puncture but also analyzes the mechanical performance at different depths, and finally, a puncture mechanical response simulation field corresponding to the selected puncture simulation process is generated through simulation.

[0147] Preferably, based on the puncture mechanical response simulation field corresponding to the selected puncture simulation process, a puncture path dynamic simulation identification analysis is performed on the gyroscopic puncture constraint preset path corresponding to the selected puncture simulation process to generate a puncture simulation training path corresponding to the selected puncture scenario simulation process.

[0148] In an embodiment of the present invention, by combining the puncture mechanical response simulation field corresponding to the selected puncture simulation process generated by the previous simulation, a dynamic simulation identification analysis is performed on the puncture constraint preset path. By using dynamic simulation software (such as Unity or Unreal Engine), the generated mechanical response is applied to the preset path, and the dynamic changes during the puncture simulation process are displayed through real-time simulation. This simulation analysis can identify potential path deviations and inadaptabilities and optimize the puncture simulation training path according to the simulation results to ensure the accuracy and effectiveness of the training effect. The generated optimal puncture training path will provide a reference basis for actual puncture training, and finally, a puncture simulation training path corresponding to the selected puncture scenario simulation process is generated through simulation.

[0149] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention will not be limited to these embodiments shown herein but will conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A comprehensive puncture simulation system based on an automatic syringe, characterized in that: The automatic injector comprises a shell, a gyro sensor and a push-pull rod slidably fitted in the shell, the front end of the push-pull rod is sealed and slidably connected to the inner wall of the shell, a length-adjustable accommodation chamber is provided between the front end surface of the push-pull rod and the front end wall of the shell, the front end of the shell is provided with an injection interface connected to the accommodation chamber, the injection interface is provided with a puncture needle, a resistor element and a gyro sensor are provided on the push-pull rod along the length direction, the resistor element and the gyro sensor are electrically connected to a computer processing unit, and the comprehensive puncture simulation system based on the automatic injector comprises the following modules: The user training puncture simulation operation selection module is used to select the puncture scene to be simulated through the trainer user interface, and configure the corresponding puncture needle and puncture injection depth operation instrument for the automatic injector according to the selected simulated puncture scene to obtain the selected simulated puncture automatic injector; use the selected simulated puncture automatic injector to perform a comprehensive puncture simulation operation on the corresponding selected simulated puncture scene to generate a simulation process of the selected puncture scene; The puncture gyroscope deviation calibration module is used to use the gyroscope sensor in the selected simulated puncture automatic syringe to perform puncture gyroscope measurement processing on the selected puncture scene simulation process and upload it to the computer processing unit to obtain the gyroscope measurement value of the puncture simulation process; use the computer processing unit to perform neural network regression deviation estimation on the gyroscope measurement value of the puncture simulation process to obtain the deterministic part of the gyroscope regression deviation; Based on the gyroscope regression bias deterministic part, a static bias calibration calculation is performed on the gyroscope measurement value of the puncture simulation process to obtain the gyroscope bias calibration value of the puncture process; wherein, the following functions are included: Using the gyro sensor in the selected simulated puncture automatic syringe to perform puncture gyro measurement processing on the selected puncture scene simulation process to obtain the gyro measurement value of the puncture simulation process; uploading the gyro measurement value of the puncture simulation process measured in real time on the gyro sensor to the computer processing unit through wireless transmission technology; Using a computer processing unit to obtain a gyro coordinate system of the gyro sensor corresponding to the puncture simulation process, and based on the gyro coordinate system, converting the gyro measurement value of the puncture simulation process into a measurement value coordinate system to obtain a corresponding gyro measurement value in the gyro coordinate system; Based on the gyro coordinate system of the puncture simulation process, a statistical analysis of the angular velocity vector of the gyro sensor corresponding to the selected puncture scene simulation process is performed to obtain the actual angular velocity vector of the gyro corresponding to the gyro coordinate system; Performing gyroscope error modeling on the corresponding gyroscope sensor in the selected puncture scenario simulation process according to the corresponding gyroscope measurement value in the gyroscope coordinate system and the actual angular velocity vector of the gyroscope to generate a puncture simulation gyroscope error mathematical model; The mathematical model of the puncture simulation gyroscope error is specifically: ； In the formula, In the gyroscope coordinate system The corresponding gyroscope measurements are: In the gyroscope coordinate system The corresponding actual angular velocity vector of the gyroscope is, is the matrix of gyroscope off-diagonal elements and scale factor errors, is the gyro bias part, is zero-mean Gaussian white noise; The zero-order calibration part of the gyroscope deviation is obtained by configuring a stationary condition for the gyroscope sensor and performing a zero-order calibration process on the gyroscope deviation part in the mathematical model of the puncture simulation gyroscope error based on the stationary condition; a virtual gyroscope array and a gyroscope sensor are set up through a resistor element to form a gyroscope virtual module, wherein each gyroscope virtual module is composed of orthogonal gyroscopes on three directional axes, and a neural network regression-assisted calibration estimation is performed on the zero-order calibration part of the gyroscope deviation based on the gyroscope virtual module to obtain a gyroscope regression deviation deterministic part; Based on the gyroscope regression bias deterministic part, the static bias calibration calculation is performed on the corresponding gyroscope measurement value in the gyroscope coordinate system to obtain the gyroscope bias calibration value during the puncture process; including: Based on the static condition, the mathematical model of the puncture simulation gyroscope error is used to measure the relevant parameters when the gyroscope is static for a series of times, and the corresponding angular velocity measurement vector and Gaussian measurement white noise under a series of static conditions of the gyroscope are obtained; The error model is reconstructed according to the angular velocity measurement vector corresponding to a series of stationary conditions of the gyroscope and the Gaussian measurement white noise to obtain the gyroscope error reconstruction model under stationary conditions: ； In the formula, For the The corresponding gyroscope measurements under substationary conditions, For the The corresponding angular velocity measurement vector under sub-stationary conditions is: For the The corresponding gyroscope estimated bias under substationary conditions is, For the The corresponding Gaussian measurement white noise under substationary conditions; By assuming that the Gaussian measurement white noise tends to zero mean through the zero-order calibration method, the gyroscope measurement value tends to zero accordingly, and the expected operator is taken on both sides of the above formula for zero-order calibration processing to obtain the zero-order calibration part of the gyroscope bias: ； in It is the zero-order calibration part of the gyroscope bias; A puncture training scene path simulation generation module is used to perform puncture path simulation identification analysis on the selected puncture scene simulation process based on the gyroscope deviation calibration value of the puncture process, so as to generate a puncture simulation training path corresponding to the selected puncture scene simulation process; The puncture simulation path visualization module is used to use the trainer user interface to perform puncture simulation visualization processing on the puncture simulation training path corresponding to the selected puncture scene simulation process, so as to execute the corresponding puncture scene visualization teaching process.

2. The comprehensive puncture simulation system based on automatic syringe according to claim 1 is characterized in that: The user training puncture simulation operation selection module includes the following functions: Select the puncture scene to be simulated through the trainer user interface to obtain the simulated puncture scene selected for training, wherein the simulated puncture scene selected for training includes a thoracic puncture simulation scene, a lumbar puncture simulation scene, an abdominal puncture simulation scene, and a bone marrow puncture simulation scene; According to the simulated puncture scenario selected for training, the corresponding automatic injector is equipped with a corresponding puncture needle and a puncture injection depth operation instrument to obtain the selected simulated puncture automatic injector; The puncture scene environment parameters are set for the selected simulated puncture scene for training to obtain the selected simulated puncture scene environment parameter set; the device performance test and optimization of the selected simulated puncture automatic injector is performed based on the selected simulated puncture scene environment parameter set to obtain the simulated puncture device optimized automatic injector; Design the simulated puncture operation process for the simulated puncture device optimized automatic syringe to generate the simulated puncture operation process design step document; Based on the simulated puncture operation process design step document, a comprehensive puncture simulation operation is performed on the simulated puncture scenario selected for training to generate a simulation process for the selected puncture scenario.

3. The comprehensive puncture simulation system based on automatic syringe according to claim 1 is characterized in that: The neural network regression-assisted calibration estimation of the zero-order calibration part of the gyroscope deviation based on the gyroscope virtual module includes: A neural network regression assisted calibration method is designed through a gyroscope virtual module, wherein the neural network regression assisted calibration method is a calibration method assisted by adding a gyroscope input channel or a calibration method assisted by adding real and virtual gyroscope training data; Based on the neural network regression assisted calibration method, the zero-order calibration part of the gyroscope deviation is assisted in calibration estimation to obtain the deterministic part of the gyroscope regression deviation.

4. The comprehensive puncture simulation system based on automatic syringe according to claim 3 is characterized in that: The method for adding a gyroscope input channel to assist in calibration estimation of the gyroscope deviation zero-order calibration part includes: A convolutional neural network architecture is designed by adding a gyroscope input channel to assist in the calibration method. The convolutional neural network architecture consists of a convolutional layer, a LeakyReLU activation function, and a maximum pooling layer, and each gyroscope module feature is input into the convolutional layer: ,in is the number of gyroscope virtual modules, is the window size; In the The network output in the convolutional layer ,in is the window kernel size, For the The gyroscope bias value at the window kernel, is the stride, For the The weight at the window kernel; according to the The network output in the convolutional layer can be defined as the LeakyReLU activation function ; Based on The network output in the convolutional layer is pooled in the maximum pooling layer. ,in is the pooling size and flattens its input into a 2D tensor to calculate the first fully connected layer ,in is the weight of the first fully connected layer, is the bias of the first fully connected layer; By introducing the LeakyReLU activation function, the second fully connected layer connected to it is repeatedly calculated to obtain the final neural network regression output ,in and are the weights and biases of the second fully connected layer respectively; the neural network regression loss is calculated by the mean square error loss function ,in is the total number of training samples; By inputting the zero-order calibration part of the gyroscope bias to assist in the calibration estimation, the deterministic part of the gyroscope regression bias is obtained: 。 5. The comprehensive puncture simulation system based on automatic syringe according to claim 3 is characterized in that: The method of adding real and virtual gyroscope training data to assist in calibration estimation of the gyroscope bias zero-order calibration part includes: Collecting real gyroscope measurements from the gyroscope virtual module ,in , and They represent the actual measurement readings of the orthogonal gyroscopes on the three directional axes within the gyroscope virtual module; Generate virtual gyroscope measurement data from the gyroscope virtual module through simulation ,in , and They represent the virtual measurement readings of the gyroscopes on the three orthogonal axes in the gyroscope virtual module; Combine the real gyroscope measurement data and the virtual gyroscope measurement data into a training set , and input it into the convolutional neural network for neural network regression training, so that the network outputs the gyroscope deviation values ​​on the three-axis orthogonal gyroscope ,in For the The weight matrix of the layer, For the The neural network activation value of the layer, For the The bias term of the layer; The neural network regression loss is calculated based on the gyroscope bias values ​​on the three orthogonal gyroscope axes: ； in is the total number of training samples, are the indices of the three orthogonal gyroscopes; Update and optimize the weight matrix and bias terms through the back-propagation method: ； ； in is the learning rate; By inputting the zero-order calibration part of the gyroscope bias to assist in the calibration estimation, the deterministic part of the gyroscope regression bias is obtained: 。 6. The comprehensive puncture simulation system based on automatic syringe according to claim 1 is characterized in that: The puncture training scenario path simulation generation module includes the following functions: Obtaining gyroscope calibration measurement values ​​corresponding to different time points of the selected puncture scenario simulation process through gyroscope deviation calibration values ​​during the puncture process; Based on the corresponding gyroscope calibration measurement values ​​at different time points, a three-dimensional virtual space scene is constructed for the corresponding selected puncture scene simulation process to generate a gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process; The gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process is subjected to dynamic simulation recognition analysis of the puncture path to generate a puncture simulation training path corresponding to the selected puncture scene simulation process.

7. The comprehensive puncture simulation system based on automatic syringe according to claim 6, characterized in that: The dynamic simulation and identification analysis of the puncture path of the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process includes: Annotating key puncture nodes of the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to obtain key puncture nodes of the gyroscope puncture corresponding to the selected puncture simulation process; Performing physical geometric constraint analysis on the gyroscopic puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to obtain the gyroscopic puncture physical geometric constraint conditions corresponding to the selected puncture simulation process; Based on the gyroscope puncture physical geometric constraint conditions corresponding to the selected puncture simulation process, a puncture path preset tracking analysis is performed on the corresponding gyroscope puncture key nodes to generate a gyroscope puncture constraint preset path corresponding to the selected puncture simulation process; Based on the selected puncture scene simulation process, a puncture process mechanical simulation analysis is performed on the gyroscope puncture three-dimensional virtual space model corresponding to the selected puncture simulation process to generate a puncture mechanical response simulation field corresponding to the selected puncture simulation process; According to the puncture mechanical response simulation field corresponding to the selected puncture simulation process, the puncture path dynamic simulation identification analysis is performed on the gyroscope puncture constraint preset path corresponding to the selected puncture simulation process to generate a puncture simulation training path corresponding to the selected puncture scene simulation process.

Images