A total orthopedic platform surgical robot and a navigation method thereof

Abstract

The application discloses a full orthopedic platform surgical robot and a navigation method thereof, and belongs to the field of medical robots.The full orthopedic platform surgical robot comprises a core control system which is an embedded high-performance microprocessor, is used for coordinating the work processes among various components, and supports multi-task concurrent processing; a modular system which is installed at the end of the surgical robot; a high-precision positioning system which is used for tracking the position of a surgical instrument in space in real time; and an intelligent image fusion system which is used for fusing preoperative images with real-time images of a surgical site, planning and confirming a surgical path based on the fused image data.The full orthopedic platform surgical robot successfully overcomes a series of problems faced by traditional orthopedic surgery, significantly improves the safety, accuracy and efficiency of surgery, and has the advantages of the modular system, the accuracy of the high-precision positioning system, the application of the intelligent image fusion system, the intelligent features of the self-adaptive surgical planning and control system, and the full-process optimization brought by the surgical operation system.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of medical devices, and relates to a full-orthopedic-platform surgical robot and a navigation method thereof. BACKGROUND

[0002] Existing orthopedic surgery robots are mainly designed for specific types of orthopedic surgery, have limited application scope, and still have room for improvement in terms of surgical precision, flexibility and intelligent degree. In addition, the requirements in different orthopedic surgery scenarios are quite different, and a single-function surgical robot is difficult to meet the complex orthopedic surgery requirements. SUMMARY

[0003] The application is provided to overcome at least one deficiency of the prior art, and provides a full-orthopedic-platform surgical robot and a navigation method thereof.

[0004] In order to achieve the above-mentioned purpose, the application adopts the following technical scheme: a full-orthopedic-platform surgical robot, comprising

[0005] The core control system is an embedded high-performance microprocessor, which is used for coordinating the work flow among the components and supporting multi-task concurrent processing;

[0006] The modular system is installed at the end of the surgical robot;

[0007] The high-precision positioning system is used for real-time tracking of the position of the surgical instrument in space;

[0008] The intelligent image fusion system is used for fusing the preoperative image with the real-time image of the surgical site, planning and confirming the surgical path based on the fused image data;

[0009] The adaptive surgical planning and control system is used for constructing a three-dimensional model of the patient, planning a surgical path, and adjusting the contact force when the surgical instrument contacts the bone according to the real-time capture and quantification of the contact force and torque changes when the surgical instrument contacts the bone;

[0010] The surgical operation system supports diversified interaction, and further comprises a user interface for interaction.

[0011] Further, the modular system comprises a multi-degree-of-freedom mechanical arm, a multifunctional surgical terminal and a surgical instrument, the multi-degree-of-freedom mechanical arm is installed at the end of the surgical robot, the multifunctional surgical terminal is connected with the end of the multi-degree-of-freedom mechanical arm, the surgical instrument is detachably connected with the multifunctional surgical terminal, and the multifunctional surgical terminal has a quick replacement interface compatible with multiple surgical instruments and can be assembled with different types of surgical instruments.

[0012] Further, the high-precision positioning system comprises an optical tracking system and a magnetic navigation system, the optical tracking system is built in the surgical robot, the optical tracking system adopts an infrared or laser light source, the optical tracking system is provided with a reflective marker point, the reflective marker point is arranged on the surgical instrument, and the optical tracking system captures the reflective marker point on the surgical instrument.

[0013] Further, the magnetic navigation system comprises a magnetic sensitive element, the magnetic sensitive element is arranged on the surgical instrument, and the magnetic navigation system is combined with an external magnetic field generator and a sensor to position and track the position of the surgical instrument in a three-dimensional space.

[0014] Further, the adaptive surgical planning and control system comprises

[0015] A model establishing module, which is built in a deep learning algorithm, analyzes and learns preoperative patient data, processes and learns to construct a three-dimensional model capable of reflecting individual characteristics of the patient by using a neural network three-dimensional model;

[0016] A route adjustment module, which receives and processes feedback data of the surgical site in real time, adjusts the route of the surgical instrument in real time based on the real-time feedback;

[0017] A force-torque sensor array for capturing and quantifying the contact force and torque changes when the surgical instrument contacts the bone in real time;

[0018] An active impedance control module, which is built in an active impedance algorithm, adjusts the contact force when the surgical instrument contacts the bone according to the data fed back by the force-torque sensor array.

[0019] A navigation method of a full orthopedic platform surgical robot, comprising the following steps:

[0020] Step S1: Import the medical image data into the model establishing module to construct a three-dimensional model of the patient;

[0021] Step S2: The intelligent image fusion system creates a virtual surgical environment, and performs surgical simulation based on the three-dimensional model and the virtual surgical environment;

[0022] Step S3: Preoperative preparation of the surgical robot;

[0023] Step S4: The high-precision positioning system collects the position data of the surgical instrument in a three-dimensional space in real time, the position data is converted into a visual signal input into a user interface, and it is judged whether the surgical route conforms to the surgical planning, if yes, step S5 is executed, if not, the user interface displays an alarm, and the route adjustment module adjusts the route of the surgical instrument in real time according to the position data;

[0024] Step S5: The force-torque sensor array captures and quantizes the contact force and torque when the surgical instrument contacts the bone in real time, which is input into the impedance control module and the user interface, the impedance control module judges whether the contact force and torque exceed the preset threshold, if yes, the user interface displays an alarm, the core control system adjusts the driving torque of the multi-degree-of-freedom mechanical arm, if not, step S6 is executed;

[0025] Step S6: Determine whether the surgical instrument reaches the target anatomical position, if yes, execute step S7, if not, execute step S4;

[0026] Step S7: The surgical robot system automatically stores all key data of the surgical process, forms a detailed digital surgery report, and encrypts and saves it;

[0027] Step S8: Compare the digital surgery report with the surgical planning data, quantitatively evaluate the surgical precision, extract the indicators higher than the precision threshold and the corresponding surgical steps, form an optimized surgery report, and save it;

[0028] Step S9: End the step.

[0029] Further, the step S3 comprises the following steps:

[0030] Step S31: According to the surgical planning, place the total orthopedic platform surgical robot at the preset position in the operating room, and physically and electronically synchronize with the surgical bed and C-arm fluoroscopy device, select the surgical instrument matched with the surgical demand to install to the multifunctional surgical terminal, and ensure the sterile operation environment through sterilization packaging;

[0031] Step S32: Start the optical tracking system and the magnetic navigation system, monitor and correct the position and attitude of the surgical instrument in real time, collect the position data of the surgical instrument in the spatial position through the high-precision positioning system, adjust the route of the surgical instrument according to the position data, and verify whether the surgical instrument is consistent with the planned path in the whole surgical process;

[0032] Step S33: Start the surgical operation system, issue instructions through diversified interaction, and verify that the surgical robot performs fine operation actions according to the instructions.

[0033] In summary, the advantages of the present application are:

[0034] 1) The total orthopedic platform surgical robot of the present application successfully overcomes a series of problems faced by traditional orthopedic surgery by virtue of the advantages of its modular system, the accuracy of the high-precision positioning system, the application of the intelligent image fusion system, the intelligent features of the self-adaptive surgical planning and control system, and the whole-process optimization brought by the surgical operation system, significantly improving the safety, accuracy and efficiency of the surgery.

[0035] 2) The full orthopedic platform surgical robot of the present application covers the field of orthopedic surgery comprehensively, breaking the problem of single surgical robot function limitation.

[0036] 3) The modular system of the present application is convenient for expansion and upgrading, meeting the individualized needs of different orthopedic surgeries. BRIEF DESCRIPTION OF DRAWINGS

[0037] Figure 1 It is a schematic diagram of the full orthopedic platform surgical robot of the present application.

[0038] Figure 2 It is a flow chart of the navigation method of the full orthopedic platform surgical robot of the present application. DETAILED DESCRIPTION

[0039] The embodiments of the present application are described below through specific concrete examples, and those skilled in the art can easily understand other advantages and effects of the present application from the content disclosed in the specification. The present application can also be implemented or applied through other different specific embodiments, and various modifications or changes can be made to the details in the specification without departing from the spirit of the present application. It should be noted that the following examples and features in the examples can be combined with each other without conflict.

[0040] It should be noted that the diagrams provided in the following examples only illustrate the basic concept of the present application in a schematic manner, and the diagrams only show the components related to the present application, not the number, shape and size of the components when actually implemented. The actual implementation of each component may be a random change in type, number and proportion, and the layout pattern of the components may also be more complex.

[0041] All directional indications (such as up, down, left, right, front, back, transverse, longitudinal, etc.) in the embodiments of the present application are only used to explain the relative position relationship, motion condition, etc. between the components in a certain specific posture, and if the specific posture changes, the directional indications will also change accordingly.

[0042] Due to installation errors and other reasons, the parallel relationship referred to in the embodiments of the present application may actually be an approximate parallel relationship, and the vertical relationship may actually be an approximate vertical relationship.

[0043] Embodiment one:

[0044] As shown in Figure 1 A full orthopedic platform surgical robot, comprising.

[0045] The core control system, as the brain of the whole surgical robot, adopts an embedded high-performance microprocessor and carries a real-time operating system, is responsible for coordinating the work between components, and supports multi-task concurrent processing to ensure the smoothness and stability of the surgical process.

[0046] Multi-degree-of-freedom robot arm, with multiple degrees of freedom (such as 7-axis or higher), realizes omnidirectional flexible movement, and the multi-degree-of-freedom robot arm includes a high-torque-density motor and a precision harmonic reducer, which ensures high speed, high precision and smoothness during movement;

[0047] Multi-functional surgical terminal connected to the end of the multi-degree-of-freedom robot arm, the multi-functional surgical terminal has the ability to be compatible with multiple tools, and different types of surgical instruments can be assembled through quick replacement of the interface;

[0048] Surgical instrument, which is detachably connected to the multi-functional surgical terminal, ensures quick and targeted configuration for different orthopedic surgery needs, and the surgical instrument can use electric bone drills, micro-grinding heads, ultrasonic suction cutting knives, etc.

[0049] Multi-degree-of-freedom robot arm, multi-functional surgical terminal and surgical instrument realize a modular system,

[0050] High-precision positioning system for real-time tracking of the position of the surgical instrument in space;

[0051] Intelligent image fusion system for fusing preoperative images with real-time images of the surgical site, planning and confirming the surgical path based on the fused image data, ensuring that the surgical instrument accurately reaches the target position along the established path, reducing the risk of surgery and improving the success rate of surgery;

[0052] Preoperative images are two-dimensional or three-dimensional data collected by CT, MRI, DR and other imaging devices;

[0053] The intelligent image fusion system realizes the fusion of preoperative images and real-time images of the surgical site through advanced image registration algorithms such as Icp algorithm, B-spline deformation field registration, etc.

[0054] The intelligent image fusion system is based on high-performance graphics processor (GPU) accelerated computing, and uses computer-aided design (CAD) software to create a virtual surgical environment, which includes the patient's bone structure, soft tissue and key anatomical landmarks. Doctors can perform surgery simulation through an intuitive interface, including but not limited to planning the surgical path, determining the best approach (such as the least invasive path), setting the size, model and precise three-dimensional spatial position of the implant (such as pedicle screw, joint prosthesis) to the millimeter level, and simulating the surgical steps in advance to reduce uncertainty in actual surgery.

[0055] Adaptive surgery planning and control system for constructing a three-dimensional model of the patient, planning the surgical path, and intelligently adjusting the reaction force when the surgical instrument contacts the bone according to the real-time capture and quantification of the subtle force and torque changes when the surgical instrument contacts the bone.

[0056] The surgical operation system comprises a user interface, through which interaction is realized, further improving the convenience and accuracy of surgical operation, enabling doctors to quickly understand and master relevant data during surgery, thereby making accurate judgments and decisions.

[0057] The surgical operation system supports diversified interaction, such as touch screen gesture operation, enabling doctors to easily review data and adjust surgical parameters under sterile conditions through simple gestures; at the same time, the surgical operation system also supports voice control technology, and doctors can command surgical robots to operate through voice instructions, greatly reducing the operation threshold and improving the efficiency of human-machine cooperation during surgery. The humanized interaction design enables doctors to devote more energy to critical surgical decisions, achieving efficient and precise medical practice.

[0058] The high-precision positioning system comprises an optical tracking system and a magnetic navigation system, the optical tracking system is built into the surgical robot, the optical tracking system adopts infrared or laser light source, the optical tracking system is provided with a reflective marker point, the reflective marker point is arranged on the surgical instrument, the optical tracking system captures the reflective marker point on the surgical instrument, and the spatial position of the surgical instrument is tracked in real time, and the precision can reach sub-millimeter level.

[0059] The magnetic navigation system comprises a magnetic sensitive element, the magnetic sensitive element is arranged on the surgical instrument, the magnetic navigation system interacts with an external magnetic field generator and a sensor, and precise positioning and tracking of the surgical instrument in a three-dimensional space are realized.

[0060] The optical tracking system and the magnetic navigation system are complementary and integrated, so that even in a complex surgical environment, the surgical instrument can be accurately moved to the millimeter level in a three-dimensional space regardless of the angle or depth.

[0061] The adaptive surgical planning and control system comprises

[0062] The model establishment module is built-in with a deep learning algorithm, which analyzes and learns preoperative patient data, including but not limited to medical images, pathological reports, genetic information and other physiological indicators, processes and learns by using a complex neural network three-dimensional model, thereby constructing a three-dimensional model that can reflect the individual characteristics of the patient.

[0063] A route adjustment module receives and processes feedback data from the surgical site in real time, including but not limited to the position, angle, speed of the surgical instrument, and interaction with the surrounding tissue, etc. Based on these real-time feedback, the route adjustment module adjusts the route of the surgical instrument in real time to maximize the success probability of the surgery and minimize potential risks, such as reducing bleeding, reducing the possibility of nerve damage, etc. In this way, the robotic system can plan the precise surgical path and choose the method for each patient's specific situation, truly realizing the personalized surgical plan of "tailor-made";

[0064] A force-torque sensor array is used to capture and quantify the subtle force and torque changes when the surgical instrument contacts the bone, with an accuracy of nanometers or even piconewtons, thereby achieving fine perception of the surgical instrument-bone interaction behavior.

[0065] An active impedance control module with built-in active impedance algorithm can intelligently adjust the reaction force when the surgical instrument contacts the bone based on the data feedback from the force-torque sensor array, so that it always maintains appropriate contact force and rotational torque while maintaining stable operation, avoiding damage to the surface or internal tissue of the bone caused by excessive force. Specifically, once the contact force between the instrument and the bone is detected to exceed the preset threshold, the active impedance control module outputs instructions to the core control system to adjust the driving torque of the multi-degree-of-freedom robot arm, ensuring that the surgical instrument always maintains a safe and precise force range during contact with the bone.

[0066] The adaptive surgery planning and control system enables the total orthopedic platform surgical robot not only to provide the optimal solution in the surgery planning stage, but also to adapt and adjust in real time in the surgery implementation stage, thereby maximizing the precision, safety and effectiveness of the surgery, marking the entry of orthopedic surgery robot technology into a new intelligent era.

[0067] As shown in Figure 2 A navigation method for a total orthopedic platform surgical robot, comprising the following steps:

[0068] Step S1: The medical image data is imported into the model establishment module to construct a three-dimensional model of the patient;

[0069] The CT, MRI or X-ray medical image data of the patient is imported into the model establishment module and converted into a three-dimensional model to ensure high fidelity of the model and accurate restoration of the anatomical structure;

[0070] Step S2: An intelligent image fusion system creates a virtual surgical environment, and based on the three-dimensional model and the virtual surgical environment, a surgery simulation is performed;

[0071] The surgical simulation includes, but is not limited to, planning a surgical path, determining an optimal approach (such as a minimally invasive approach), setting the size, model and three-dimensional spatial position accurate to the millimeter level of an implant (such as a pedicle screw, joint prosthesis), and pre-simulating the surgical steps to reduce uncertainty in the actual surgery.

[0072] Step S3: preoperative surgical robot preparation;

[0073] Step S4: The high-precision positioning system collects the position data of the surgical instrument in the three-dimensional space in real time, and the position data is converted into a visual signal input to the user interface to determine whether the surgical route meets the surgical planning. If yes, execute step S5, if no, the user interface displays an alarm, and the route adjustment module adjusts the route of the surgical instrument in real time according to the position data;

[0074] Step S5: The force-torque sensor array captures and quantifies the contact force and torque when the surgical instrument contacts the bone in real time, which is input to the impedance control module and the user interface. The impedance control module determines whether the contact force and torque exceed the preset threshold. If yes, the user interface displays an alarm, and the core control system adjusts the driving torque of the multi-degree-of-freedom robot arm. If no, execute step S6;

[0075] Step S6: Determine whether the surgical instrument reaches the target anatomical position. If yes, execute step S7, if no, execute step S4;

[0076] Step S7: The surgical robot system automatically stores all key data of the surgical process, forms a detailed digital surgical report, and saves it in an encrypted manner;

[0077] The key data includes, but is not limited to, surgical video, instrument trajectory, force sensing data, patient physiological indicators, etc. Through the recording and saving of key data in step S7, not only can the surgical effect be evaluated and reviewed, but also can play a role in subsequent quality control, teaching training and scientific research;

[0078] Step S8: Compare the digital surgical report with the surgical plan data, quantitatively evaluate the surgical precision, extract the indicators and corresponding surgical steps that are higher than the precision threshold, form an optimized surgical report, and save it;

[0079] The optimized surgical report helps to continuously optimize the orthopedic surgery technology.

[0080] The surgical precision includes key indicators such as the position deviation of the implant, the size of the surgical incision, and the recovery time.

[0081] Step S9: End the step.

[0082] The establishment of the three-dimensional model in step S1 includes the following steps:

[0083] Step S11: Preprocessing of medical images;

[0084] The medical images are processed using wavelet transform or Non-local Means Denoising (NL-means) and the like.

[0085] Wavelet transform decomposes the image at different scales, removes high-frequency noise, and preserves image edge information. NL-means removes noise by comparing the similarity of pixel blocks in the image with the surrounding pixel blocks, and is suitable for preserving texture details.

[0086] The pixel value after denoising (\hat{f}_x) can be estimated by the following formula: [\hat{f}_x=\frac{\sum_{y\in\mathca l{N}(x)}w(x,y)f_y}{\sum_{y\i n\mathca l{N}(x)}w(x,y)}] where (f_y) is the pixel value in the neighborhood (\mathca l{N}(x)), and (w(x,y)) is the weight calculated according to the similarity of the pixel blocks.

[0087] Contrast enhancement processing is performed on the denoised image:

[0088] Histogram equalization or Contrast Limited Adaptive Histogram Equalization (CLAHE) is used.

[0089] CLAHE limits the range of contrast enhancement to avoid image distortion caused by excessive enhancement. It divides the image into multiple small regions (tiles), applies histogram equalization to each tile, and limits the upper limit of contrast enhancement for each tile.

[0090] Mutual correlation method, mutual information method or feature-based method (such as SIFT, SURF) are used for multi-modal or time series image registration:

[0091] Suppose two images are (I_1) and (I_2), the goal of registration is to find a transformation (T) (such as translation, rotation, scaling) that makes the registered images as similar as possible. When using mutual information as a similarity measure, the objective function can be expressed as: [\argmax_TM I(I_1,T(I_2))];

[0092] Step S12: Three-dimensional model construction;

[0093] Neural network 3D modeling: Before 3D reconstruction, the segmented 2D slices are first converted into voxel representation, i.e., continuous regions are divided into discrete small units in 3D space. This step is the basis for mapping 2D image information to 3D space.

[0094] Convolutional neural network (CNN) applied to 3D modeling: 3D convolutional layers are used to process voxel data, capturing spatial context information. For example, the 3D U-Net architecture can be extended to 3D space for more detailed volume segmentation and reconstruction.

[0095] 3D convolution operation can be regarded as a weighted sum operation in three dimensions, [V^{'}_{i jk}=\sum_{l=-k}^{k}\sum_{m=-k}^{k}\sum_{n=-k}^{k}W_{lmn}V_{(i+l)(j+m)(k+n)}] where (V^{'}_{i jk}) is the output voxel, (V_{i jk}) is the input voxel, and (W_{lmn}) is the weight of the convolution kernel.

[0096] Three-dimensional modeling includes surface reconstruction, surface smoothing, and volume consistency check:

[0097] Surface reconstruction uses the Marching Cubes algorithm: This algorithm traverses the voxel grid and applies a set of rules to each cube composed of 8 adjacent voxels to generate a smooth surface.

[0098] Surface smoothing uses Laplacian smoothing: a smoothing algorithm based on the Laplacian operator, which reduces noise and irregularities on the model surface while maintaining the overall shape.

[0099] [\vec{n}_{new}=\frac{\sum_{\text{neighbors}}\vec{n}_{ne i ghbor}}{|\sum_{\text{ne ighbors}}\vec{n}_{neighbor}|}] where (\vec{n}_{new}) is the normal vector of the smoothed vertex, and (\vec{n}_{neighbor}) is the normal vector of the adjacent vertex.

[0100] Volume consistency check: by calculating the volume change before and after the model, it is ensured that the actual volume of the anatomical structure is not significantly changed during the smoothing process. The volume calculation formula is simply: [V=\sum_{i}Voxe l Si ze^3] where (Voxe l Si ze) is the edge length of a single voxel, and the summation is for all voxels belonging to the model.

[0101] Step S13: Deep learning algorithm application;

[0102] Step S131: Convolutional layer feature extraction: Local features are extracted from the image using the convolutional layer of the convolutional neural network (CNN).

[0103] For each convolution kernel, its weight matrix (W) is convolved with the local region of the input image, and a bias term (b) is added to obtain a pixel value of the feature map: [F_{i j}=\sum_{m=0}^{M-1}\sum_{n=0}^{N-1}W_{mn}I_{(i+m)(j+n)}+b] where (F_{i j}) is an element of the feature map, (I) is the input image, and (M) and (N) are the dimensions of the convolution kernel.

[0104] Step S132: Dimensionality reduction by pooling layer;

[0105] The spatial dimension of the feature map is reduced by max-pooling or average-pooling while preserving important features, which can be simplified as: [P_{i j}=\max_{m,n}F_{(im)(jn)}] or [P_{i j}=\frac{1}{mn}\sum_{m,n}F_{(im)(j*n)}];

[0106] Step S133: Fully connected layer classification / regression;

[0107] After flattening the feature map, it is input to the fully connected layer for final feature vector extraction, and then the anatomical structure is classified or positioned. Assuming the output of the last fully connected layer is (Y), its calculation method is: [Y=\s i gma(W^TX+b)] where (X) is the output of the previous layer, (W) and (b) are the weight matrix and bias vector, and (\s i gma) is the activation function, such as ReLU.

[0108] Model training and validation;

[0109] Step S134: Loss function and optimizer;

[0110] The loss function is, for example, cross-entropy loss (for classification tasks) or mean squared error (for regression tasks). Taking cross-entropy as an example, the loss (L) can be represented as: [L=-\frac{1}{N}\sum_{i=1}^{N}\sum_{j=1}^{C}y_{i j}\log(p_{i j})] where (y_{i j}) is the true label, (p_{i j}) is the predicted probability, (N) is the number of samples, and (C) is the number of classes.

[0111] The optimizer (such as Adam) updates the model parameters according to the gradient descent method to minimize the loss function;

[0112] Step S135: Cross-validation;

[0113] K-fold cross-validation is used to evaluate model performance. The dataset is divided into K parts, and each time one part is taken as the test set and the rest as the training set. Repeat K times and take the average performance index (such as accuracy, recall, Dice coefficient, etc.).

[0114] Step S136: Early stopping method:

[0115] When the performance of the validation set no longer improves, the training is terminated in advance to avoid overfitting.

[0116] Regularization: To prevent overfitting, L1 or L2 regularization terms can be used, such as L2 regularization loss: [L_{reg}=\a lpha||W||^2_2] where (\a lpha) is the regularization strength, (||W||^2_2) is the L2 norm of the model parameters.

[0117] Step S14: Model interactive modification and confirmation:

[0118] Step S141: Use intuitive graphical user interface (GUI) to display three-dimensional models and allow doctors to interact through mouse, touch or professional equipment. The interface should provide rotation, scaling, cutting and other operations to allow doctors to view the model from multiple angles.

[0119] Step S142: Manual correction tools;

[0120] Provide a set of fine drawing and editing tools, such as a brush tool to add or delete incorrect parts of the model, and stretching and moving tools to adjust the position and shape of the anatomical structure, ensuring that the model is consistent with the actual anatomical structure.

[0121] Step S143: Marking and annotation;

[0122] Allow doctors to add markers or annotations to the model to point out areas that need special attention or modification suggestions. These information can be text, icons or color coding. Feedback loop

[0123] Step S144: Data labeling and collection:

[0124] Each modification by the doctor should be recorded and converted into labeled data. For example, if the doctor adds a missing bone segment to the model, the system should automatically or guide the doctor to accurately label the newly added structure.

[0125] Step S145: Online learning and model updating:

[0126] The doctor's corrections are input into the model as new data using an online learning mechanism, and the neural network is fine-tuned in real time or periodically. The process of updating the model parameters (W') can be achieved through the backpropagation algorithm, updating the gradient of the loss function (L) according to the corrected labeled data (D'): [DeltaW'= -eta nabla_WL(W',D')] where (eta) is the learning rate, and (nabla_WL) is the gradient of the loss function with respect to the weights (W).

[0127] Step S146: Performance evaluation and adjustment:

[0128] The performance of the updated model is periodically evaluated, and the data corrected by the doctor is used as the test set to calculate the accuracy, recall rate, and other indicators of the model, ensuring that the model is continuously optimized and not overfitting.

[0129] Step S147: Adaptive algorithm:

[0130] Adopting meta-learning or reinforcement learning methods, the model can more intelligently learn from the doctor's corrections, automatically adjust the learning strategy, and optimize the learning process for the doctor's different correction habits and preferences.

[0131] Through deep learning and neural network technology, this module realizes efficient conversion from two-dimensional images to three-dimensional models, improves the intelligent level of surgical robots, and provides strong technical support for precise medical treatment and personalized treatment of orthopedic surgery.

[0132] Step S3 includes the following steps:

[0133] Step S31: According to the surgical plan, the full orthopedic platform surgical robot is accurately set at the preset position in the operating room, and is physically and electronically synchronized with the operating bed, C-arm fluoroscopy device. Select the surgical instrument matching the surgical needs to install to the multifunctional surgical terminal, and ensure the sterile operation environment through sterilization packaging;

[0134] Step S32: Start the optical tracking system and the magnetic navigation system, monitor and correct the position and attitude of the surgical instrument in real time, collect the position data of the surgical instrument in space position through the high-precision positioning system, and adjust the route of the surgical instrument according to the position data, so as to verify whether the surgical instrument is consistent with the planned path during the whole surgical process;

[0135] Step S33: Start the surgical operation system, issue instructions through diversified interaction, and verify that the surgical robot performs fine operation actions according to the instructions.

[0136] Obviously, the described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor should belong to the protection scope of the present application.

Claims

1. A full-platform orthopedic surgical robot, characterized by: include The core control system is an embedded high-performance microprocessor that coordinates the workflow between components and supports multi-task concurrent processing; A modular system mounted on the end of a surgical robot; High-precision positioning system for real-time tracking of the position of surgical instruments in space; Intelligent image fusion system, used to fuse preoperative images with real-time images of the surgical site, and plan and confirm the surgical path based on the fused image data; Adaptive surgical planning and control system, used to construct a three-dimensional model of the patient, plan the surgical path, and adjust the contact force between the surgical instrument and the bone based on real-time capture and quantification of contact force and torque changes between the surgical instrument and the bone; The surgical operating system supports diversified interactions and also includes a user interface, through which interactions are carried out. The modular system includes a multi-degree-of-freedom robotic arm, a multi-function surgical terminal, and surgical instruments. The multi-degree-of-freedom robotic arm is mounted at the end of the surgical robot. The multi-function surgical terminal is connected to the end of the multi-degree-of-freedom robotic arm. The surgical instruments are detachably connected to the multi-function surgical terminal. The multi-function surgical terminal has a quick-change interface compatible with a variety of surgical instruments and can be equipped with different types of surgical instruments. The adaptive surgery planning and control system includes The model building module has a built-in deep learning algorithm that analyzes and learns preoperative patient data and uses a neural network 3D model to process and learn to construct a 3D model that can reflect the individual characteristics of the patient; Route adjustment module, which receives and processes feedback data from the surgical site in real time. Based on this real-time feedback, the route adjustment module adjusts the route of the surgical instrument in real time; A force-torque sensor array is used to capture and quantify in real time the changes in contact force and torque when surgical instruments come into contact with bones; The active impedance control module has a built-in active impedance algorithm that adjusts the contact force between the surgical instrument and the bone based on the data fed back by the force-torque sensor array.

2. The full orthopedic platform surgical robot according to claim 1, characterized in that: The high-precision positioning system includes an optical tracking system and a magnetic navigation system. The optical tracking system is built into the surgical robot. The optical tracking system uses an infrared or laser light source. The optical tracking system is provided with reflective marking points. The reflective marking points are set on the surgical instrument. The optical tracking system captures the reflective marking points on the surgical instrument.

3. The full orthopedic platform surgical robot according to claim 2, characterized in that: The magnetic navigation system includes a magnetic sensitive element, which is arranged on the surgical instrument. The magnetic navigation system is combined with an external magnetic field generator and a sensor to locate and track the position of the surgical instrument in three-dimensional space.

Images