A surgical robot navigation board and navigation method

Abstract

The present application discloses a surgical robot navigation board and a navigation method, which relate to the technical field of surgical robot navigation boards, and include: obtaining preoperative preparation data for puncture surgery; determining the initial position of the navigation board and establishing a unified coordinate system, obtaining patient skin deformation data, and calculating a first reference position of the navigation board based on a deformation model; obtaining a comprehensive deformation value based on the skin deformation data obtained in real time, calculating a real-time pressure adjustment value for each micro airbag, arranging the real-time pressure adjustment values ​​in descending order and adjusting the pressures of the corresponding micro airbags in sequence to obtain a second initial position; performing a depth calibration on the second initial position to obtain a second reference position; and adjusting the navigation board twice based on the patient's skin deformation data, thereby achieving the effect of precise fixation of the navigation board and precise positioning of the surgical robot, reducing navigation errors caused by skin deformation, and improving positioning accuracy.

Description

Technical Field

[0001] The present invention relates to the technical field of surgical robot navigation boards, and in particular to a surgical robot navigation board and a navigation method. Background Art

[0002] Traditional surgical robot navigation uses preoperative planning, combined with intraoperative X-ray image fusion matching measurement technology, and real-time optical image tracking technology. The use of this navigation technology requires the computer to have powerful computing power to achieve timely tracking of the movement of the target (object) and prevent the robot from lagging behind.

[0003] The Chinese invention patent with application number 202210333133.5 discloses a surgical robot navigation board, which has a central correction target point by distributing multiple correction targets on the navigation board body. By using the navigation board as an intermediate bridge, the stable tissue structure image obtained by intraoperative X-ray is matched with the preoperative MRI image and a unified coordinate system is established with the robot, and the unified coordinate system is used to guide the robot for positioning.

[0004] However, when a single integral navigation board is used for navigation positioning in the prior art, it will deform to varying degrees when close to the skin, and the error will accumulate at the center correction target, causing the navigation board to have a large positioning error, affecting the positioning accuracy. Reducing the positioning error of the navigation board and thereby improving the positioning accuracy has become a top priority for the surgical robot navigation board. Summary of the invention

[0005] The present application solves the problem of low positioning accuracy caused by positioning errors in the prior art by providing a surgical robot navigation board and a navigation method, thereby achieving the technical effect of reducing the positioning error of the navigation board and thereby improving the positioning accuracy.

[0006] The present application provides a surgical robot navigation method, which is applied to a navigation board, and the method comprises:

[0007] S100: obtaining preoperative preparation data for the puncture surgery through preoperative MRI image data, wherein the preoperative preparation data includes a percutaneous puncture point, a needle insertion angle, and a puncture depth to the percutaneous puncture point;

[0008] S200: determining the initial position of the navigation board according to the preoperative preparation data and establishing a unified coordinate system, acquiring the patient's skin deformation data, and calculating the first reference position of the navigation board based on the deformation model;

[0009] S300: Determine the coordinate positions of the central navigation panel and several auxiliary navigation panels according to the first reference position and mark them to form a three-dimensional coordinate matrix; obtain a deformation comprehensive value based on the skin deformation data acquired in real time, and if the deformation comprehensive value is greater than a preset deformation threshold, calculate the real-time pressure adjustment value of each micro airbag, arrange them in descending order according to the size of the real-time pressure adjustment value, and adjust the pressure of the corresponding micro airbag in turn to obtain a second initial position; perform depth calibration on the second initial position to obtain a second reference position;

[0010] S400: Fix the navigation board on the human body based on the second reference position, and under the guidance of the laser cross target on the working end of the surgical robot, align the laser cross with the central target and cross guide line of the central navigation board. The central target and cross guide line can be visualized under X-rays and overlap with the target tissue structure; calculate the numerical values ​​of the surgical entry point, entry angle and entry depth based on the unified coordinate system, and input the numerical values ​​into the surgical robot. After the operation is completed, an analysis report is generated.

[0011] Further, obtaining the patient's skin deformation data and calculating the first reference position of the navigation board based on the deformation model include:

[0012] Acquire the patient's skin surface data based on a three-dimensional scanning device, generate point cloud data, and then obtain skin deformation data; mesh the skin deformation data, input the processed skin deformation data into a deformation model, extract the skin deformation characteristics, calculate the error value according to the change of the deformation characteristics, and calculate the first reference position based on the error value and the initial position;

[0013] Wherein, the deformation characteristics include tension and curvature of the skin.

[0014] Furthermore, the processed skin deformation data is input into the deformation model, the deformation characteristics of the skin are extracted, and the error value is calculated according to the change of the deformation characteristics, including: based on the finite element analysis method, a collective model is established according to the processed skin deformation data, the collective model is divided into a finite number of units, each unit contains a number of nodes, and a finite element solver is used to solve the mechanical equation of each unit to obtain the displacement and stress results of the node, and then the tension and curvature changes of the skin are calculated, and the weighted average of multiple deformation characteristic changes is summed to obtain the error value.

[0015] Furthermore, the real-time pressure adjustment value of each micro airbag is calculated, including:

[0016] The total pressure value is calculated according to the deformation comprehensive value, deformation threshold and mapping formula, the corresponding pressure demand value is obtained based on the weight coefficient of each micro airbag, and the corresponding real-time pressure adjustment value is obtained according to the current pressure and pressure demand value of each micro airbag.

[0017] Furthermore, the mapping formula is as follows:

[0018]

[0019] in, is the total pressure value, is the comprehensive deformation value, is the deformation threshold, is the conversion coefficient, with a value range of [0, 1], which is used to measure the conversion relationship between deformation and pressure; the pressure demand value calculation formula is as follows:

[0020]

[0021] in, is the pressure requirement of the i-th micro airbag, is the weight coefficient of the i-th micro airbag.

[0022] Further, in step S300, depth calibration is performed on the second initial position to obtain a second reference position, including:

[0023] S310: acquiring the position of each correction target point at the first reference position and marking it as the first original position, acquiring the position of each correction target point at the second initial position and marking it as the second current position;

[0024] S320: calculating the position deviation of each correction target point according to the first original position and the second current position, and then calculating the overall offset of the navigation board, and selecting a number of correction target points as candidate target points according to the flotation mechanism;

[0025] S330: Perform depth calibration based on the candidate target point to obtain a second reference position.

[0026] Furthermore, the flotation mechanism is: based on the position deviation and overall offset of each correction target point, the comprehensive impact value of each correction target point is calculated, the comprehensive impact value of each correction target point is arranged in descending order according to size, and several correction target points are selected as candidate target points according to the extraction ratio.

[0027] Furthermore, the comprehensive impact value is obtained by comprehensively calculating the deformation value and contribution value of each correction target point, wherein the deformation value refers to the ratio of the position deviation of each correction target point to the average value of the position deviation; the contribution value refers to the product of the pre-set weight value of each correction target point and the ratio of the position deviation of each correction target point to the overall offset; the extraction ratio is set according to the ratio of the overall offset and the pre-set offset threshold. If the ratio is not less than 1, all correction targets are selected as candidate targets; if the ratio is less than 1, several correction targets are selected as candidate targets according to the ratio.

[0028] A surgical robot navigation board, used in the above-mentioned surgical robot navigation method, the navigation board comprises a central navigation board, a plurality of auxiliary navigation boards and a plurality of micro airbags, the central navigation board and the plurality of auxiliary navigation boards are connected by micro airbags; each micro airbag is connected to an air pump;

[0029] Several auxiliary navigation panels are distributed in a square shape, and the central navigation panel is located at the center of the auxiliary navigation panels; two adjacent auxiliary navigation panels are fixedly connected by micro airbags, and the central navigation panel is fixedly connected to the four nearest auxiliary navigation panels by micro airbags.

[0030] Both the central navigation panel and the auxiliary navigation panel are provided with a central target point, a group of cross guide lines and a plurality of correction target points; the center of the central target point coincides with the center of the cross guide lines.

[0031] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0032] By acquiring the patient's skin deformation data, adjusting the initial position according to the deformation characteristics to obtain the first reference position, acquiring the skin deformation data in real time, calculating the comprehensive deformation value and the real-time pressure adjustment value of each micro airbag, adjusting the pressure of the corresponding micro airbag in turn to obtain the second initial position, and performing depth calibration on the second initial position to obtain the second reference position; based on the patient's skin deformation data, the navigation board is adjusted twice, which achieves the effect of precise fixation of the navigation board and precise positioning of the working end of the surgical robot, further reducing the navigation error caused by skin deformation, and improving the positioning accuracy of the surgical robot. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] Figure 1 This is a schematic diagram of a surgical robot navigation method according to an embodiment of the present invention;

[0034] Figure 2 This is a schematic diagram of the overall structure of a surgical robot navigation panel in an embodiment of the present invention;

[0035] Figure numerals: 1. Central navigation panel; 2. Auxiliary navigation panel; 3. Central target point; 4. Correction target point; 5. Cross guide line; 6. Micro airbag. DETAILED DESCRIPTION

[0036] To facilitate the understanding of the present invention, the present application will be described more comprehensively below with reference to the relevant drawings; the drawings show preferred embodiments of the present invention, but the present invention can be implemented in many different forms and is not limited to the embodiments described herein; on the contrary, the purpose of providing these embodiments is to enable a more thorough and comprehensive understanding of the disclosed content of the present invention.

[0037] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by technicians in the technical field to which the present invention belongs; the terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention; the term "and / or" used herein includes any and all combinations of one or more related listed items.

[0038] Embodiment 1: Figure 1 As shown, a surgical robot navigation method is applied to a navigation board, the method comprising:

[0039] S100: Obtaining preoperative preparation data for the puncture surgery through preoperative MRI image data, wherein the preoperative preparation data includes a percutaneous puncture point, a needle insertion angle, and a puncture depth to the percutaneous puncture point.

[0040] In some embodiments, the puncture point information is obtained through intraoperative X-ray machine and preoperative MR or CT measurement, and the needle angle information and puncture depth information are obtained preoperatively through preoperative MR or CT measurement. The specific needs need to be obtained according to actual conditions, and this embodiment does not impose specific restrictions.

[0041] S200: Determine the initial position of the navigation board according to the preoperative preparation data and establish a unified coordinate system, obtain the patient's skin deformation data, and calculate the first reference position of the navigation board based on the deformation model.

[0042] In some embodiments, the initial position of the navigation board is determined based on the preoperative preparation data, specifically based on the center position of the percutaneous puncture point and the central navigation panel 1 of the navigation board to ensure position overlap. After determining the position of the central navigation panel 1, the positions of the remaining auxiliary navigation panels 2 are adaptively distributed.

[0043] In some embodiments, a unified coordinate system is established, which provides a unified metric for all points in the surgical area, so that data from different sources (such as MRI images, three-dimensional scanning data, navigation board position, etc.) can be compared and calculated in the same space. The positions of key elements such as the navigation board, puncture point, and skin deformation are accurately determined to ensure the accuracy and safety of the surgery. In this embodiment, the central target 3 of the central navigation board 1 is used as the origin of the coordinate system.

[0044] In some embodiments, the patient's skin deformation data is obtained, and the first reference position of the navigation board is calculated based on the deformation model, including: obtaining the patient's skin surface data based on a three-dimensional scanning device, generating point cloud data, and then obtaining skin deformation data; gridding the skin deformation data, inputting the processed skin deformation data into the deformation model, extracting the deformation characteristics of the skin, calculating the error value according to the change of the deformation characteristics, and calculating the first reference position based on the error value and the initial position; wherein the deformation characteristics include but are not limited to the tension and curvature of the skin.

[0045] Specifically, a three-dimensional scanning device is used to obtain the patient's skin surface data, generate point cloud data, and the point density is 100 points per square millimeter. The noise is removed by the point cloud registration algorithm to obtain the skin deformation data. It is specifically designed according to the actual patient site and is suitable for parts such as the back, chest or joints. Taking the back as an example, the skin deformation data is gridded, and the skin tension and curvature are calculated by the finite element analysis method through the pre-established deformation model to obtain the deformation characteristics of the back skin. The target position setting of the central navigation board and the auxiliary navigation board is the starting point of the whole process. These navigation boards are usually placed at key anatomical locations on the patient's back, such as the protruding point of the spine or the edge of the scapula. For example, for patients with scoliosis, the central navigation board may be placed at the maximum bending point, while the auxiliary navigation boards are distributed on both sides to capture the overall curvature change. According to the deformation characteristics of the back skin, the least squares method is used for curve fitting to obtain the error value. According to the initial position coordinates and error values ​​of each navigation board, the first reference position of the navigation board is determined.

[0046] The processed skin deformation data is input into the deformation model, the deformation characteristics of the skin are extracted, and the error value is calculated according to the change of the deformation characteristics, including: based on the finite element analysis method, a collective model is established according to the processed skin deformation data, the collective model is divided into a finite number of units, each unit contains a number of nodes, according to the mechanical properties of the skin, the elastic modulus, Poisson's ratio and other properties of the material are set, according to the stress conditions of the skin under different states, the corresponding boundary conditions and loads are applied, and the finite element solver is used to solve the mechanical equations of each unit to obtain the displacement, stress and strain results of the node, and according to the displacement and stress results of the node, the tension and curvature changes of the skin are calculated, and the weighted average of multiple deformation characteristic changes is summed to obtain the error value.

[0047] Specifically, tension is the pulling force generated by the skin in the direction of force. In finite element analysis, the tension distribution of the skin can be obtained by calculating the stress of the nodes; curvature is the degree of curvature of the skin surface. In finite element analysis, the surface shape of the skin can be obtained by calculating the displacement of the nodes, and then the curvature can be calculated.

[0048] The deformation model is obtained by pre-training, and the method is not specifically limited. In this embodiment, a nonlinear viscoelastic model is used as the basic mathematical model to describe skin deformation. The basic parameters are determined by the experimental data fitting method, including: relaxation time, initially set to 0.5 seconds, indicating the time required for the skin to return to its original shape after being subjected to force; elastic modulus, set to 10 MPa, indicating the stiffness of the skin within the elastic range; viscosity coefficient: set to 1 MPa·s, indicating the damping characteristics of the skin within the viscous range.

[0049] The constitutive equation of the skin is established based on the basic form and parameters of the nonlinear viscoelastic model. The constitutive equation uses an integral form to describe the stress-strain relationship of the skin during the force process. The skin surface mesh data of the patient's back is imported into the finite element analysis software. The viscoelastic properties of the material are set, that is, the basic parameters input into the constitutive equation. Boundary conditions and loads are applied to simulate the force conditions of the skin during surgery. For example, the displacement boundary conditions of the skin under the pressure of surgical instruments and the loads applied to the skin by surgical instruments are set. Run the finite element solver to solve the mechanical equations and obtain the displacement, stress, and strain results of the nodes.

[0050] The established viscoelastic model is verified using the skin deformation data from the actual surgery. The predictive ability of the model is evaluated, such as the error in calculating the skin deformation, the degree of agreement with the experimental results, etc. In this embodiment, the skin deformation error predicted by the model is controlled within 5%, which is in good agreement with the experimental results. According to the verification results, the model is adjusted and optimized as necessary. For example, parameters such as relaxation time, elastic modulus, and viscosity coefficient can be adjusted to improve the accuracy and robustness of the model.

[0051] The optimized viscoelastic model is applied to the surgical robot navigation method. During the operation, the patient's skin deformation data is obtained in real time and input into the viscoelastic model. According to the prediction results of the model, the position of the navigation board is dynamically adjusted. Preferably, the skin deformation data variation range is monitored in real time. If the curvature change rate of the skin deformation data is greater than a preset threshold (such as 0.5), the elastic coefficient and Poisson's ratio parameters of the deformation model are adjusted by the gradient descent method, and the error value is recalculated. According to the adjusted deformation model parameters, the database of the deformation model is updated, and a new model configuration file is generated. The updated deformation model is optimized by the random forest regression model in the machine learning algorithm, and the hyperparameters are adjusted by the cross-validation method to obtain the optimized deformation model. Through the optimized deformation model, the nonlinear equations are solved in combination with the Newton iteration method, and the precise error value of each navigation board is finally determined, with an accuracy of 0.1 mm. The first reference position of the central navigation board and the auxiliary navigation board is set according to the error value, and the coordinate marking data is generated by the interpolation algorithm, and the data format is a three-dimensional coordinate matrix.

[0052] This embodiment can automatically identify and adapt to patients of different body shapes, ages and skin conditions, improving the versatility and accuracy of the model. The precise position determined by the optimization model is the basis for the final placement of the navigation board. This process takes into account multiple factors, including skin deformation characteristics, anatomical structure and surgical needs. For example, for thoracic surgery, the model may recommend placing more navigation boards in the intercostal space to capture subtle deformations caused by breathing. The final generated coordinate marking data is the output of the entire process. These data contain not only the three-dimensional coordinates of the navigation boards, but also the relative position and orientation information between them. These precise coordinate data provide reliable reference points for the surgical navigation system, which helps to improve surgical accuracy, reduce operation time and patient radiation exposure

[0053] S300: Determine the coordinate positions of the central navigation panel 1 and several auxiliary navigation panels 2 according to the first reference position and mark them to form a three-dimensional coordinate matrix; obtain the comprehensive deformation value based on the skin deformation data acquired in real time, if the comprehensive deformation value is greater than the preset deformation threshold, calculate the real-time pressure adjustment value of each micro airbag 6, arrange them in descending order according to the size of the real-time pressure adjustment value and adjust the pressure of the corresponding micro airbag 6 in turn to obtain the second initial position; perform depth calibration on the second initial position to obtain the second reference position.

[0054] In some embodiments, a three-dimensional coordinate matrix is ​​a data structure used to represent the coordinates of multiple points in a three-dimensional space. It appears in the form of an array or a matrix, where each element (or row) represents the coordinates of a point in the three-dimensional space. These coordinates are expressed in the form of (x, y, z), where x, y, and z represent the horizontal coordinate, vertical coordinate, and vertical coordinate of the point in the three-dimensional space, respectively. In this space, each point can be uniquely determined by three values.

[0055] Specifically, first determine a reference position or origin, which is the starting point for all coordinate calculations. In this embodiment, the position of the central navigation panel 1 is used as the origin. For each navigation panel (including the central navigation panel 1 and the auxiliary navigation panel 2), it is necessary to calculate its coordinates in three-dimensional space based on the first reference position. The calculated coordinate data is organized into a matrix form. Each row represents the coordinates of a navigation panel, and the columns correspond to the x, y, and z coordinates, respectively. In actual situations, it may be necessary to use an interpolation algorithm to generate denser coordinate data points. The interpolation algorithm can estimate the coordinates of unknown points based on known coordinate points, thereby generating a more complete and refined coordinate matrix. It is also necessary to verify and correct the generated three-dimensional coordinate matrix to ensure the accuracy and reliability of the data. The specific needs should be set according to the actual situation, and this embodiment does not make specific restrictions.

[0056] In some embodiments, the initial adjustment point can also be set according to the coordinate position of the auxiliary navigation panel 2, and the auxiliary navigation panel 2 located at the edge corner in the three-dimensional coordinate matrix is ​​selected as the initial adjustment point. The auxiliary navigation panel 2 located at the "edge corner" refers to those auxiliary navigation panel 2 that are located at the four extremes or corners of the overall layout in terms of spatial position. Specifically, it is as follows: traverse the coordinates of all auxiliary navigation panels 2 in the three-dimensional coordinate matrix. Find the maximum and minimum points in the x, y, and z coordinate values. These points are the auxiliary navigation panels 2 located at the edge corners, and arbitrarily select an auxiliary navigation panel 2 as the initial adjustment point.

[0057] In some embodiments, real-time acquisition of skin deformation data is to use a three-dimensional scanning device or an optical tracking system to capture changes in the patient's skin surface in real time. The device is aimed at the patient's surgical area, and the appropriate scanning frequency and resolution are set to start real-time acquisition of three-dimensional data of the skin surface. The collected data is processed in real time through the device's own software or custom algorithms to generate point cloud data or mesh data to represent the current shape of the skin.

[0058] Based on the acquired real-time skin deformation data, deformation features are extracted and quantified. For example, the extracted tension features are calculated to obtain a numerical representation of the tension. For example, the average value of the tension can be calculated by summing all the tension data and then dividing by the number of data points. The extracted curvature features are calculated to obtain a numerical representation of the curvature. For example, the rate of change of the curvature can be calculated by comparing the curvature values ​​at different time points or spatial positions.

[0059] According to the influence of each deformation feature on the overall deformation degree, a weight is determined for each feature, and the quantified deformation features are combined using the weighted average or summation method to obtain a comprehensive deformation value representing the overall deformation degree.

[0060] In some embodiments, based on historical surgical data or clinical experiments, deformation data under different surgical types, patient body shapes and skin conditions are collected, and statistical analysis is performed on the collected data to determine the normal range and abnormal range of the deformation characteristics. According to the statistical analysis results, one or more deformation thresholds are set to determine whether the real-time deformation data exceeds the normal range. The specific needs need to be pre-set according to the actual situation, and this embodiment does not impose any specific restrictions.

[0061] In some embodiments, the real-time pressure adjustment value of each micro airbag 6 is calculated, including: calculating the total pressure value according to the deformation comprehensive value, the deformation threshold and the mapping formula, obtaining the corresponding pressure demand value based on the weight coefficient of each micro airbag 6, and obtaining the corresponding real-time pressure adjustment value according to the current pressure and the pressure demand value of each micro airbag 6; the mapping formula is as follows:

[0062]

[0063] in, is the total pressure value, is the comprehensive deformation value, is the deformation threshold, is a conversion coefficient with a value range of [0, 1], which is used to measure the conversion relationship between deformation and pressure. Its purpose is to obtain the corresponding pressure data based on the current deformation data. The conversion coefficient k can be used to convert the deformation data into pressure values. The specific needs need to be set according to the actual situation and experimental data, and can also be determined by experimental fitting methods. For example, different deformation comprehensive values ​​are set in the laboratory, and the corresponding pressure demand values ​​are measured. The deformation comprehensive value is used as the independent variable and the corresponding total pressure value is used as the dependent variable. A scatter plot is drawn, and then the mapping relationship between the deformation comprehensive value and the total pressure value is fitted by the linear regression method. In the fitting process, the conversion coefficient k will be determined as a parameter of the regression equation, and verified using new skin deformation data to check the accuracy and reliability of the fitting equation. According to the verification results, the fitting equation and the conversion coefficient k are adjusted and optimized as necessary.

[0064] In some embodiments, the pressure requirement value is calculated as follows:

[0065]

[0066] in, is the pressure requirement value of the i-th micro airbag 6, is the weight coefficient of the i-th micro airbag 6. According to the current pressure of each micro airbag 6 and the calculated pressure demand value, the real-time pressure adjustment value of each airbag can be calculated. The real-time pressure adjustment value is equal to the pressure demand value minus the current pressure value.

[0067] For example, there are three micro airbags 6A, B, and C, and their weight coefficients are 0.4, 0.3, and 0.3 respectively. The deformation comprehensive value is 1.5, the deformation threshold is 1.0, and the conversion coefficient is 2.0. The calculated total pressure value is 1.0; the pressure demand value of each micro airbag 6 is ; ; The current pressure value of each micro airbag 6 is initially set to 0.1, and the real-time pressure adjustment values ​​of each micro airbag 6 are 0.3, 0.2, and 0.2 respectively.

[0068] In some embodiments, each micro airbag 6 is adjusted according to the real-time pressure adjustment value of each micro airbag 6, and the pressure adjustment range is 10 to 50 kPa. The airbag pressure of the central navigation board and the auxiliary navigation board is dynamically adjusted to adapt to the skin deformation to obtain the second initial position; the second initial position is depth-calibrated, and the depth calibration is to perform secondary correction on the adjusted navigation board according to the correction target 4 on the navigation board, and obtain the second reference position according to the coordinate position of the navigation board after the depth calibration.

[0069] In the unified coordinate system, an auxiliary navigation panel 2 at an edge angle is selected as the starting point for adjustment, because the auxiliary navigation panel 2 at the edge angle is relatively independent, and its deformation has little effect on the position of the overall navigation panel. By monitoring the relative position changes between adjacent navigation panels and the real-time update of skin deformation data, the air pressure of the micro airbag 6 is adjusted one by one with the edge angle as the starting point, so that each navigation panel can better adapt to the dynamic deformation of the skin. For example, when the skin is greatly deformed, the airbag pressure can be increased to better fit the skin; when the skin is less deformed, the airbag pressure can be reduced to avoid excessive compression. By dynamically adjusting the position of the navigation panel and the airbag pressure, the positioning accuracy of the surgical robot and the success rate of the operation are improved.

[0070] S400: Fix the navigation board on the human body based on the second reference position, and under the guidance of the laser cross target on the working end of the surgical robot, align the laser cross with the central target 3 and the cross guide line 5 of the central navigation board 1. The central target 3 and the cross guide line 5 can be visualized under X-rays and overlap with the target tissue structure; calculate the numerical values ​​of the surgical entry point, entry angle and entry depth based on the unified coordinate system, and input the numerical values ​​into the surgical robot. After the operation is completed, an analysis report is generated.

[0071] In some embodiments, the target tissue structure is a target puncture site of the human body, and the specific operation is as follows: turn on the X-ray light source placed above the navigation board to image the navigation board and the target tissue structure on the DR imaging flat panel; perform navigation board image distortion correction to obtain the positional relationship between the navigation board and the robot; the robot is equipped with a ranging device with a ranging function, and the ranging device obtains the distance of any multiple positions on the navigation board, thereby obtaining the angular positional relationship between the robot and the navigation board on a unified plane. MR images are used to calibrate the edge and surface of the target tissue structure, and the distance between the target tissue structure and the human body surface is obtained through MR images, thereby obtaining the angular positional relationship between the navigation board and the target tissue structure on a unified plane. Calculate the values ​​of the surgical entry point, entry angle, and entry depth based on a unified coordinate system, and input the values ​​into the robot. In puncture surgery, when the navigation board is used to guide and correct the surgical robot, the surgical entry point is the percutaneous puncture point, the entry angle is the needle entry angle, and the entry depth is the puncture depth. Based on the determined position of the navigation board, using an X-ray machine and MR images to determine the final surgical entry point of the robot is a technology well known to those skilled in the art, and will not be described in detail in this embodiment.

[0072] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0073] The present application obtains the patient's skin deformation data, adjusts the initial position according to the deformation characteristics to obtain a first reference position, obtains the skin deformation data in real time, calculates the comprehensive deformation value and the real-time pressure adjustment value of each micro airbag 6, adjusts the pressure of the corresponding micro airbag 6 in turn to obtain a second initial position, and performs depth calibration on the second initial position to obtain a second reference position; the navigation board is adjusted twice based on the patient's skin deformation data, thereby achieving the effect of precise fixation of the navigation board and precise positioning of the working end of the surgical robot, further reducing the navigation error caused by skin deformation, and improving the positioning accuracy of the surgical robot.

[0074] Embodiment 2: In embodiment 1, depth calibration is performed to obtain the second reference position. This embodiment makes further improvements on the basis of the above content.

[0075] In step S300, depth calibration is performed on the second initial position to obtain a second reference position, including:

[0076] S310: acquiring the position of each correction target point 4 at the first reference position and marking it as the first original position, acquiring the position of each correction target point 4 at the second initial position and marking it as the second current position;

[0077] S320: calculating the position deviation of each correction target point 4 according to the first original position and the second current position, and then calculating the overall offset of the navigation board, and selecting a number of correction target points 4 as candidate target points according to the flotation mechanism;

[0078] S330: Perform depth calibration based on the candidate target point to obtain a second reference position.

[0079] In some embodiments, the adjustment of the navigation board refers to the displacement change of the navigation board relative to the initial position at the beginning of the operation due to factors such as skin deformation, tissue movement, or surgical mechanical operation. After determining the first reference position, the navigation board is photographed using an X-ray machine, and the positions of all correction target points 4 are recorded and marked as the first original position. After the pressure of the micro-airbag 6 is adjusted, the position of the navigation board is offset, and the navigation board is re-photographed, and the positions of all correction target points 4 are recorded again and marked as the second current position.

[0080] For each correction target point 4, the first original position and the second current position are compared to calculate the position deviation, which is calculated by the distance formula in three-dimensional space, as follows:

[0081]

[0082] in, is the position deviation, are the coordinates of the first original position, are the coordinates of the second current position.

[0083] In some embodiments, the flotation mechanism is as follows: based on the position deviation and overall offset of each correction target point 4, the comprehensive impact value of each correction target point 4 is calculated, the comprehensive impact value of each correction target point 4 is arranged in descending order according to size, and a number of correction targets 4 are selected as candidate targets according to the extraction ratio. The first original position and the second current position of each correction target point 4 are obtained, and the position deviation of each correction target point 4 is calculated using the coordinate difference formula. For example, the original position of target point A is (10, 20, 30), and the current position is (12, 22, 40), then its position deviation is 2.83. According to the position deviation of all correction targets 4, the average value of the position deviation is calculated using the average value formula. If the position deviation of target point B is 1.41 and the position deviation of target point C is 3.16, then the average value is 2.46. The deformation value is determined by the ratio of the position deviation of each correction target point 4 to the average value of the position deviation. For example, the deformation value of target point A is 2.83 / 2.46≈1.15. According to the deformation value of each correction target 4 and the preset contribution weight, for example, the weight coefficient is 0.7, the contribution value of target A is calculated to be 1.15×0.7=0.805. The deformation value and contribution value of each correction target 4 are weighted and comprehensively calculated. For example, the comprehensive weight is 60% for the deformation value and 40% for the contribution value, then the comprehensive impact value of target A is 1.15×0.6+0.805×0.4≈1.02. Arrange the comprehensive impact values ​​of all correction targets 4 in descending order from large to small.

[0084] In some embodiments, the comprehensive impact value is obtained by comprehensively calculating the deformation value and contribution value of each correction target point 4, wherein the deformation value refers to the ratio of the position deviation of each correction target point 4 to the average value of the position deviation; and the contribution value refers to the product of the weight value preset for each correction target point 4 and the ratio of the position deviation of each correction target point 4 to the overall offset. The deformation value and contribution value mentioned in this embodiment are both used to judge the actual importance of the correction target point 4, specifically including: collecting the position deviation data of each correction target point 4 through a sensor, for example, the deviation of target point A is 0.5 mm, the deviation of target point B is 0.8 mm, and the deviation of target point C is 1.2 mm, and calculating the average value of the position deviations of all targets to obtain an average deviation of 0.83 mm.

[0085] Then, the deformation value is calculated based on the ratio of the position deviation of each target point to the average deviation, for example, the deformation value of target point A is 0.5 / 0.83=0.6, the deformation value of target point B is 0.8 / 0.83=0.96, and the deformation value of target point C is 1.2 / 0.83=1.45. The preset weight value of each target point is obtained, for example, the weight of target point A is 0.3, the weight of target point B is 0.5, and the weight of target point C is 0.2, and the overall offset is calculated, for example, the overall offset is 1.0 mm.

[0086] The contribution value is then calculated based on the ratio of the position deviation of each target to the overall offset, combined with the preset weight value. For example, the contribution value of target A is 0.3*(0.5 / 1.0)=0.15. Similarly, the contribution value of target B is 0.4, and the contribution value of target C is 0.24.

[0087] The deformation value and contribution value are used for comprehensive calculation to obtain the comprehensive impact value of each target. For example, the comprehensive impact value of target A is 0.6+0.15=0.75, the comprehensive impact value of target B is 0.96+0.4=1.36, and the comprehensive impact value of target C is 1.45+0.24=1.69. The comprehensive impact value of each target is arranged in descending order according to size, such as target C (1.69), target B (1.36), and target A (0.75). According to the extraction ratio, for example, the first 50% of the targets are selected, and candidate targets are selected from the arranged targets, such as target C and target B. Finally, the selected candidate target data, such as the position deviation and comprehensive impact value of target C and target B, are used to optimize the correction plan or generate an adjustment report.

[0088] In some embodiments, the extraction ratio is set according to the ratio of the overall offset and the preset offset threshold. If the ratio is not less than 1, all the correction targets 4 are selected as candidate targets; if the ratio is less than 1, a number of correction targets 4 are selected as candidate targets according to the ratio. The overall offset is obtained, and the preset offset threshold is used as a reference value. For example, the overall offset is 1.5 mm and the offset threshold is 2 mm. The ratio of the overall offset to the offset threshold is calculated to be 0.75, and a preliminary extraction ratio is obtained, that is, the first 75% of the correction targets 4 are selected in order as candidate targets.

[0089] In some embodiments, a depth calibration is performed based on the candidate target to obtain a second reference position, specifically including: determining the calibration parameters of each candidate target according to the position deviation value, for example, fitting the deviation value by the least square method to obtain a calibration matrix A. Using the calibration parameters, the second initial position is adjusted to obtain a preliminary calibration position, for example, the coordinates of the second initial position are multiplied by the matrix A by matrix operation to obtain the adjusted coordinates (x1, y1, z1). According to the preliminary calibration position, the updated position information of each candidate target is obtained, for example, the target position is captured by a high-precision camera and marked as (x2, y2, z2). Using the updated position information, the secondary calibration deviation value of each correction target 4 is calculated, for example, the secondary deviation value is calculated by standard deviation analysis. According to the secondary calibration deviation value, the secondary calibration parameters of each candidate target are determined, for example, the calibration matrix is ​​optimized by the gradient descent method to obtain a matrix B. Using the secondary calibration parameters, the preliminary calibration position is depth calibrated to obtain a second reference position, for example, the coordinates of the preliminary calibration position are multiplied by the matrix B by matrix operation to obtain the final coordinates (x3, y3, z3).

[0090] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0091] The present application obtains the different positions of each correction target point 4, namely the first original position and the second current position, and calculates the overall offset of the navigation board by calculating the position deviation of each correction target point 4; calculates the comprehensive impact value of each correction target point 4, taking into account the deformation value and the contribution value; arranges the comprehensive impact values ​​in descending order of size, and selects candidate targets according to the extraction ratio; achieves the effect of accurately selecting effective targets, ensuring the accuracy of surgical positioning while improving the efficiency of surgical positioning.

[0092] Embodiment 3: This embodiment makes further improvements on the basis of the above embodiment.

[0093] The method further comprises: S510: obtaining an analysis report, presetting a positioning accuracy threshold, selecting the number and layout of candidate target points corresponding to a positioning accuracy greater than the positioning accuracy threshold, and performing statistical grouping, wherein each group includes a surgical identification number, the number and layout of candidate targets;

[0094] S520: Randomly select a group to perform fitness evaluation, and calculate the positioning error and positioning rate;

[0095] S530: According to the fitness evaluation result, a layout with a fitness greater than a first threshold is selected as a parent, a crossover operation is performed on the selected parent layout to generate a child layout; and a mutation operation is performed on the child layout;

[0096] S540: Repeat steps S520 and S530 until a preset number of iterations is reached or the fitness is greater than a second threshold; output the layout with the highest fitness as the optimal correction target point 4 layout;

[0097] The analysis report includes the number and layout of candidate targets selected for each surgery and the positioning accuracy of surgical entry points; and the first threshold is less than the second threshold.

[0098] In some embodiments, the number of candidate targets, layout and positioning accuracy data of surgical entry points for each surgery are extracted from the surgical analysis report. The layout refers to the position distribution of the correction target 4. For example, data with a positioning accuracy of 92%, a number of candidate targets of 5 and a layout of (10, 20, 30) are extracted. According to the extracted positioning accuracy data, the positioning accuracy threshold is set to 90%, and the number and layout of candidate targets with a positioning accuracy greater than 90% are screened out. The screened candidate target numbers and layouts are statistically grouped according to the surgical identification number, for example, the data with the surgical identification number 001 is grouped into {001, 5, (10, 20, 30)}. According to the statistical grouping results, the fitness function is used to calculate the fitness of each group of candidate target layouts, for example, the calculated fitness is 95. It is determined whether the fitness is greater than the first threshold 85. If the fitness is 95 or greater than 85, the group of candidate target layouts is recorded. If the fitness is less than 85, the candidate target layout is adjusted and the fitness is recalculated. The fitness calculation and layout adjustment steps are repeated until the preset number of iterations reaches 100 or the fitness is greater than the second threshold value 98. All recorded candidate target layouts are compared, and the layout with the highest fitness is selected as the optimal correction target 4 layout, for example, the layout with a fitness of 98 is selected.

[0099] In some embodiments, based on the selected layout, the positioning error of each candidate target point in the layout is calculated, for example, the distance error between each target point and the target position is calculated by the Euclidean distance formula, and the error range is 0.5 mm to 2.0 mm. According to the positioning error calculation result, the overall positioning error mean of the layout is statistically calculated, for example, the error values ​​of 10 target points are summed and divided by 10, and the mean is 1.2 mm. According to the number and distribution of candidate target points in the layout, the positioning rate of the layout is calculated, for example, according to the average distance between the target points and the moving speed of the surgical instrument, the calculation rate is 3 target points per second. The positioning error mean and the positioning rate are input into the fitness calculation model, and the fitness of the layout is converted, for example, the fitness is calculated using a weighted formula, and the fitness is calculated to be 0.85. If the fitness is greater than the first threshold value of 0.8 and less than the second threshold value of 0.9, the layout is marked as a candidate optimal layout. If the fitness is greater than the second threshold value of 0.9, the layout is directly determined to be the optimal correction target point 4 layout. The above process is repeated until the preset number of iterations of 100 times is reached or the optimal layout with a fitness greater than 0.9 is found.

[0100] In some embodiments, according to the fitness evaluation result, a layout with a fitness greater than a first threshold is selected from the statistical grouping as a parent layout set. For example, a layout with a fitness greater than 0.8 is selected as the parent, and the roulette selection algorithm is used to screen out the five parents with the highest fitness from the candidate layouts. The selected parent layout is cross-operated to generate a child layout, and a single-point crossover algorithm is used to cross-operate at the third site of the gene sequence to generate 10 child layouts. The gene sequence of the child layout is obtained, and the mutation site is determined by a random number generator. If the mutation site is located at the 1st to 2nd site of the gene sequence, a local mutation strategy is adopted, and a Gaussian mutation algorithm is used to perform a ±0.1 perturbation at the site. If the mutation site is located at the 4th to 10th site of the gene sequence, a global mutation strategy is adopted, and a uniform mutation algorithm is used to randomly replace the site with 0 or 1. According to the mutation strategy, the gene sequence of the child layout is randomly mutated to generate a mutated gene sequence. The mutated gene sequence is obtained, and the child layout is reconstructed by a decoding algorithm to obtain a mutated child layout. The fitness of the mutated offspring layout is evaluated, and the score of each layout is calculated using the fitness function. According to the fitness evaluation result, it is determined whether the mutated offspring layout meets the second threshold of 0.9. If it does, the layout is retained, otherwise it is eliminated.

[0101] The technical solutions in the above embodiments of the present application have at least the following technical effects or advantages:

[0102] The present application optimizes the layout of the correction target 4 through a genetic algorithm, thereby significantly improving the positioning accuracy of the surgical navigation system; enhancing the system efficiency, and the optimized layout reduces the time required for positioning and improves the positioning rate; and by setting thresholds and the number of iterations, the effect of automated optimization of the layout of the correction target 4 is achieved.

[0103] Embodiment 4: This embodiment also provides a surgical robot navigation board, such as Figure 2 As shown, the navigation board includes a central navigation panel 1, a plurality of auxiliary navigation panels 2 and a plurality of micro airbags 6, and the central navigation panel 1 and the plurality of auxiliary navigation panels 2 are connected by the micro airbags 6;

[0104] The auxiliary navigation panels 2 are arranged in a square shape, and the central navigation panel 1 is located at the center of the auxiliary navigation panels 2; two adjacent auxiliary navigation panels 2 are fixedly connected by micro airbags 6, and the central navigation panel 1 is fixedly connected to the four nearest auxiliary navigation panels 2 by micro airbags 6 on all sides;

[0105] Both the central navigation panel 1 and the auxiliary navigation panel 2 are provided with a central target point 3, a group of cross guide lines 5 and a plurality of correction target points 4; the center of the central target point 3 coincides with the center of the cross guide lines 5; each micro airbag 6 is connected to an air pump for inflating or deflating the micro airbag 6. Preferably, the micro airbag 6 is in the form of a sheet and is in the shape of an isosceles trapezoid.

[0106] In some embodiments, when the position of the navigation board is adjusted using the micro airbag 6, an air pump is used to inflate or evacuate the micro airbag 6 according to the calculated pressure adjustment value, so that the air pressure value of each micro airbag 6 reaches the required numerical range. When the air pressure of the micro airbag 6 changes, it will cause the auxiliary navigation panels 2 on the left and right sides to move, thereby achieving the effect of adjusting the navigation board.

[0107] The navigation board in this embodiment can be used for dorsal puncture positioning navigation or lateral puncture positioning navigation. The navigation board is made of metal or non-metal materials that can be clearly developed under X-ray and other imaging equipment. It is required to have sufficient strength and not easy to deform or be soft and adhere to the surface of the human body, and be fixed on the surface of the human body. According to the optimal layout obtained in the above embodiment, the number and distribution method of the corresponding correction target points are selected and arranged on the navigation board body, which is used to measure the distance parameters and direction parameters on the plane XY in the X-ray film. The navigation board image distortion correction and the target tissue structure image distortion correction under X-ray and other imaging equipment are performed through the distance parameters and direction parameters. The specific image distortion correction method is not specifically described here.

[0108] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A surgical robot navigation method, applied to a navigation board, characterized in that: The navigation panel comprises a central navigation panel (1), a plurality of auxiliary navigation panels (2) and a plurality of micro airbags (6); the plurality of auxiliary navigation panels (2) are arranged in a square shape, and the central navigation panel (1) is located at the center of the plurality of auxiliary navigation panels (2); the central navigation panel (1) and the plurality of auxiliary navigation panels (2) are connected by the micro airbags (6); The central navigation panel (1) and the auxiliary navigation panel (2) are both provided with a central target point (3), a group of cross guide lines (5) and a plurality of correction target points (4); The method comprises: S100: obtaining preoperative preparation data for the puncture surgery through preoperative MRI image data, wherein the preoperative preparation data includes a percutaneous puncture point, a needle insertion angle, and a puncture depth to the percutaneous puncture point; S200: determining the initial position of the navigation board according to the preoperative preparation data and establishing a unified coordinate system, acquiring the patient's skin deformation data, and calculating the first reference position of the navigation board based on the deformation model; S300: determining the coordinate positions of the central navigation panel (1) and a plurality of auxiliary navigation panels (2) based on the first reference position and marking them to form a three-dimensional coordinate matrix; obtaining a deformation comprehensive value based on the skin deformation data acquired in real time, and if the deformation comprehensive value is greater than a preset deformation threshold, calculating the real-time pressure adjustment value of each micro airbag (6), arranging them in descending order according to the size of the real-time pressure adjustment value and adjusting the pressure of the corresponding micro airbag (6) in turn to obtain a second initial position; performing depth calibration on the second initial position to obtain a second reference position; S400: The navigation board is fixed on the human body based on the second reference position, and under the indication of the laser cross target provided on the working end of the surgical robot, the laser cross is aligned with the central target (3) and the cross guide line (5) of the central navigation board (1), and the central target (3) and the cross guide line (5) can be visualized under X-rays and overlap with the target tissue structure; the numerical values ​​of the surgical entry point, entry angle and entry depth based on the unified coordinate system are calculated, and the numerical values ​​are input into the surgical robot, and an analysis report is generated after the operation is completed.

2. A surgical robot navigation method according to claim 1, characterized in that: Acquiring the patient's skin deformation data and calculating the first reference position of the navigation board based on the deformation model, including: Acquire the patient's skin surface data based on a three-dimensional scanning device, generate point cloud data, and then obtain skin deformation data; mesh the skin deformation data, input the processed skin deformation data into a deformation model, extract the skin deformation characteristics, calculate the error value according to the change of the deformation characteristics, and calculate the first reference position based on the error value and the initial position; Wherein, the deformation characteristics include tension and curvature of the skin.

3. A surgical robot navigation method as claimed in claim 2, characterized in that: The processed skin deformation data is input into the deformation model, the deformation characteristics of the skin are extracted, and the error value is calculated according to the change of the deformation characteristics, including: based on the finite element analysis method, a collective model is established according to the processed skin deformation data, the collective model is divided into a finite number of units, each unit contains a number of nodes, and a finite element solver is used to solve the mechanical equation of each unit to obtain the displacement and stress results of the node, and then the tension and curvature changes of the skin are calculated, and the weighted average of multiple deformation characteristic changes is summed to obtain the error value.

4. A surgical robot navigation method according to claim 1, characterized in that: Calculate the real-time pressure adjustment value of each micro airbag (6), including: A total pressure value is calculated based on the deformation comprehensive value, the deformation threshold and the mapping formula, a corresponding pressure demand value is obtained based on the weight coefficient of each micro airbag (6), and a corresponding real-time pressure adjustment value is obtained based on the current pressure and pressure demand value of each micro airbag (6).

5. A surgical robot navigation method as claimed in claim 4, characterized in that: The mapping formula is as follows: ； in, is the total pressure value, is the comprehensive deformation value, is the deformation threshold, is the conversion coefficient, with a value range of [0, 1], which is used to measure the conversion relationship between deformation and pressure; the pressure demand value calculation formula is as follows: ； in, is the pressure requirement of the i-th micro airbag (6), is the weight coefficient of the i-th micro airbag (6).

6. A surgical robot navigation method according to claim 1, characterized in that: In step S300, depth calibration is performed on the second initial position to obtain a second reference position, including: S310: acquiring the position of each correction target point (4) at the first reference position and marking it as the first original position, acquiring the position of each correction target point (4) at the second initial position and marking it as the second current position; S320: calculating the position deviation of each correction target point (4) according to the first original position and the second current position, and then calculating the overall offset of the navigation board, and selecting a plurality of correction target points (4) as candidate target points according to the flotation mechanism; S330: Perform depth calibration based on the candidate target point to obtain a second reference position.

7. A surgical robot navigation method as claimed in claim 6, characterized in that: The flotation mechanism is as follows: based on the position deviation and overall offset of each correction target point (4), the comprehensive impact value of each correction target point (4) is calculated, the comprehensive impact value of each correction target point (4) is arranged in descending order according to size, and a number of correction target points (4) are selected as candidate target points according to the extraction ratio.

8. A surgical robot navigation method as claimed in claim 7, characterized in that: The comprehensive impact value is obtained by comprehensively calculating the deformation value and contribution value of each correction target point (4), wherein the deformation value refers to the ratio of the position deviation of each correction target point (4) to the average value of the position deviation; the contribution value refers to the product of the preset weight value of each correction target point (4) and the ratio of the position deviation of each correction target point (4) to the overall offset; the extraction ratio is set according to the ratio of the overall offset and the preset offset threshold value, and if the ratio is not less than 1, all the correction target points (4) are selected as candidate target points; if the ratio is less than 1, a number of correction target points (4) are selected as candidate target points according to the ratio.

9. A surgical robot navigation method as claimed in claim 1, characterized in that: The method further comprises: S510: Obtain an analysis report, pre-set a positioning accuracy threshold, select the number and layout of candidate targets corresponding to a positioning accuracy greater than the positioning accuracy threshold, and perform statistical grouping, where each group includes a surgery identification number, the number and layout of candidate targets; S520: Randomly select a group to perform fitness evaluation, and calculate the positioning error and positioning rate; S530: According to the fitness evaluation result, a layout with a fitness greater than a first threshold is selected as a parent, a crossover operation is performed on the selected parent layout to generate a child layout; and a mutation operation is performed on the child layout; S540: Repeat steps S520 and S530 until a preset number of iterations is reached or the fitness is greater than a second threshold; output the layout with the highest fitness as the optimal correction target layout; The analysis report includes the number and layout of candidate targets selected for each surgery and the positioning accuracy of surgical entry points; and the first threshold is less than the second threshold.

10. A surgical robot navigation board, used in a surgical robot navigation method according to any one of claims 1 to 9, characterized in that: Each micro airbag (6) is connected to an air pump; two adjacent auxiliary navigation panels (2) are fixedly connected via the micro airbags (6); the central navigation panel (1) is fixedly connected to the four nearest auxiliary navigation panels (2) via the micro airbags (6); the center of the central target point (3) coincides with the center of the cross guide line (5).

Images