Organ deformation correction method and device, equipment and storage medium

Abstract

The invention discloses an organ deformation correction method, device and equipment and a storage medium, and relates to the technical field of deformation correction of surgical navigation systems.The method comprises the steps that firstly, a point cloud complementation module is provided, pre-operative organ geometric features can be effectively utilized to enhance intraoperative visual data, and secondly, the point cloud complementation module is provided for correcting the organ deformation; reliable deformation estimation is carried out with acceptable accuracy by means of an optimization control method. And finally, introducing a viscoelastic model for conforming to the viscoelastic characteristics of the organ, and improving the accuracy of deformation prediction. According to the method, the problem that organ deformation cannot be accurately predicted due to rare point cloud data in a laparoscope operation is solved, the point cloud data of the organ in the operation can be completed through a registration technology, and deformation estimation is carried out on a preoperative organ three-dimensional model in combination with an optimization control theory and a biomechanical model, so that the organ form in the operation is accurately matched.

Description

Technical Field

[0001] The present invention relates to the technical field of deformation correction of surgical navigation systems, and in particular to a method, device, equipment and storage medium for correcting organ deformation under laparoscopic guidance based on point cloud completion and biomechanical models. Background Art

[0002] In laparoscopic liver resection, augmented reality systems are often used to overlay virtual information (preoperative CT, MRI data) on intraoperative images to obtain detailed information about the patient's anatomical structure in real time. Accurately aligning preoperative images with the deformed liver state during surgery is a key challenge. The complex non-rigid deformation of the liver, caused by breathing, surgical manipulation, and physiological changes, requires advanced registration techniques that go beyond simple rigid or affine transformations. In addition, intraoperative data are local due to the limitations of the field of view. Therefore, a precise liver deformation correction that can cope with sparse visible surfaces is urgently needed in clinical applications.

[0003] To address these challenges, researchers have explored a variety of methods, including non-rigid point cloud registration, biomechanical simulation, machine learning and other techniques to improve the accuracy and robustness of intraoperative liver registration. However, these methods have different limitations.

[0004] First, non-rigid registration registers surfaces to surfaces and cannot estimate the deformation position within the liver.

[0005] Secondly, the biomechanical models used by most researchers play a key role in dealing with the complex deformation patterns of the liver and estimating the deformation inside the soft tissue. However, due to limited field of view, instrument interference and tissue occlusion in laparoscopic surgery, it is impossible to perform complete deformation estimation during surgery using only biomechanical models.

[0006] Finally, the use of deep learning methods can enhance the accuracy of the registration process by effectively utilizing high-quality sparse data, but the main disadvantage is that it relies on a large amount of high-quality data for training. Summary of the invention

[0007] In view of the above problems, the present invention provides an organ deformation correction method, device, equipment and storage medium for overcoming the above problems or at least partially solving the above problems.

[0008] The present invention provides the following scheme:

[0009] A method for correcting organ deformation, comprising:

[0010] Acquire a CT image obtained by scanning the organ before surgery, and use the CT image to perform three-dimensional reconstruction on the organ to obtain a three-dimensional model of the organ;

[0011] collecting a partial surface image of the organ during surgery through an endoscope, and processing the partial surface image to obtain an initial point cloud during surgery;

[0012] Using the intraoperative initial point cloud and the three-dimensional model, a non-rigid registration algorithm is used to complete the intraoperative point cloud to obtain an intraoperative completed point cloud;

[0013] Biomechanical analysis, applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ;

[0014] The deformation vector field of the biomechanical model is predicted by using the intraoperative completed point cloud using an optimization control method, so as to correct the deformation of the organ during surgery.

[0015] Preferably: using a non-rigid registration algorithm to complete the intraoperative point cloud to obtain the intraoperative completed point cloud includes:

[0016] Based on the R-tree, find the triangles and centroid coordinates of the projection points on the surface of the three-dimensional model corresponding to each point in the initial point cloud during the operation, so as to match the centroid of the triangle with the corresponding point in the initial point cloud during the operation as a set of reference points;

[0017] Part of the intraoperative surface and the preoperative surface are used as the template surface and the target surface respectively, and the template is deformed to fit the shape of the target while maintaining the topological structure of the template itself; a dense registration method is used to define an affine transformation matrix for each vertex of the three-dimensional model, and the affine transformation matrix of all points is obtained by fitting the cost function;

[0018] Based on the correspondence between the determined preoperative points and intraoperative points, the nearest neighbor search method is used on the boundary of the corresponding surface before the operation to find the neighboring points of the boundary triangle, and then expand outward to determine all the points in the position of the area to be completed;

[0019] The inverse of the affine transformation matrix is ​​used to inversely transform the pre-operative surface points to the intra-operative surface, so that all points in the position of the area to be completed are transformed to obtain the intra-operative completion point cloud.

[0020] Preferably, using the CT image to perform three-dimensional reconstruction on the organ to obtain a three-dimensional model of the organ includes:

[0021] Using an image segmentation algorithm in combination with the CT image to extract features to obtain an organ contour feature sequence;

[0022] The three-dimensional model is obtained by utilizing a three-dimensional reconstruction algorithm in combination with the contour feature sequence.

[0023] Preferably, the image segmentation algorithm includes any one of a grayscale separation algorithm and a level set algorithm.

[0024] Preferably, the organ includes a liver, and applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ includes:

[0025] Based on the finite element method of Burgers viscoelastic material, the Young's modulus is 4000Pa, the Poisson's ratio is 0.45, the bulk modulus is 7000kPa, and the shear modulus is 2000kPa.

[0026] Preferably, the objective function of the optimization control method is expressed by the following formula:

[0027]

[0028] Where: Represents the gradient value of the elastic model residual, p b represents the state variable, Represents the gradient of the difference between the predicted value and the deformed value with respect to the control variable.

[0029] Preferably: a unique optimization variable b is input, and a quasi-Newton optimization algorithm BFGS solver is used to iteratively calculate the objective function and the objective function gradient.

[0030] An organ deformation correction device, used to perform the above-mentioned organ deformation correction method, the device comprising:

[0031] An organ three-dimensional model construction unit is used to obtain a CT image obtained by scanning the organ before surgery, and to perform three-dimensional reconstruction of the organ using the CT image to obtain a three-dimensional model of the organ;

[0032] An initial point cloud determination unit, used for collecting a partial surface image of the organ during surgery through an endoscope, and processing the partial surface image to obtain an initial point cloud during surgery;

[0033] A point cloud completion unit, used to complete the intraoperative point cloud by using the intraoperative initial point cloud and the three-dimensional model using a non-rigid registration algorithm to obtain an intraoperative completed point cloud;

[0034] A biomechanical model building unit, used for biomechanical analysis, applies different loads and boundary conditions to the three-dimensional model to build a biomechanical model of the organ;

[0035] The optimization unit is used to predict the deformation vector field of the biomechanical model using the intraoperative completed point cloud by utilizing an optimization control method, so as to correct the deformation of the organ during the operation.

[0036] An organ deformation correction device, the device comprising a processor and a memory:

[0037] The memory is used to store program code and transmit the program code to the processor;

[0038] The processor is used to execute the above-mentioned organ deformation correction method according to the instructions in the program code.

[0039] A computer-readable storage medium is used to store program codes, and the program codes are used to execute the above-mentioned organ deformation correction method.

[0040] According to the specific embodiments provided by the present invention, the present invention discloses the following technical effects:

[0041] The embodiments of the present application provide an organ deformation correction method, device, equipment and storage medium. The method first provides a point cloud completion module, which can effectively use the preoperative organ geometric features to enhance the intraoperative visualization data. Secondly, reliable deformation estimation is performed with acceptable accuracy through the optimization control method. Finally, in order to meet the viscoelastic characteristics of the organ, a viscoelastic model is introduced to improve the accuracy of deformation prediction. The problem that the deformation of organs cannot be accurately predicted due to the scarcity of intraoperative point cloud data of laparoscopy is solved. The point cloud data of the intraoperative organ can be completed by the registration technology, and the deformation of the preoperative three-dimensional organ model can be estimated by combining the optimization control theory and the biomechanical model to accurately match the intraoperative organ morphology.

[0042] Of course, any product implementing the present invention does not necessarily need to achieve all of the advantages described above at the same time. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention, and for ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0044] Figure 1 is a flow chart of an organ deformation correction method provided by an embodiment of the present invention;

[0045] Figure 2 is a flow chart of a deformation correction method applied in laparoscopic liver surgery navigation provided by an embodiment of the present invention;

[0046] Figure 3 is a flow chart of an optimization control algorithm provided by an embodiment of the present invention;

[0047] Figure 4 is a schematic diagram of an organ deformation correction device provided by an embodiment of the present invention;

[0048] Figure 5Schematic diagram of an organ deformation correction device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] The technical scheme in the embodiment of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiment of the present invention. Obviously, the described embodiment is only a part of the embodiment of the present invention, not all of the embodiments. Based on the embodiment of the present invention, all other embodiments obtained by ordinary technicians in this field belong to the scope of protection of the present invention.

[0050] See also Figure 1 , is a method for correcting organ deformation provided by an embodiment of the present invention, such as Figure 1 As shown, the method may include:

[0051] S101: Obtain a CT image obtained by scanning an organ before surgery, and use the CT image to perform three-dimensional reconstruction on the organ to obtain a three-dimensional model of the organ; in specific implementation, the embodiment of the present application can provide that the three-dimensional reconstruction of the organ using the CT image to obtain the three-dimensional model of the organ includes:

[0052] Using an image segmentation algorithm in combination with the CT image to extract features to obtain an organ contour feature sequence;

[0053] The three-dimensional model is obtained by utilizing a three-dimensional reconstruction algorithm in combination with the contour feature sequence.

[0054] Furthermore, the image segmentation algorithm includes any one of a grayscale separation algorithm and a level set algorithm.

[0055] S102: collecting a partial surface image of the organ during the operation through an endoscope, and processing the partial surface image to obtain an initial point cloud during the operation;

[0056] S103: using the intraoperative initial point cloud and the three-dimensional model to complete the intraoperative point cloud using a non-rigid registration algorithm to obtain an intraoperative completed point cloud; in specific implementation, the embodiment of the present application can provide that the intraoperative point cloud is completed using a non-rigid registration algorithm to obtain an intraoperative completed point cloud, including:

[0057] Based on the R-tree, find the triangles and centroid coordinates of the projection points on the surface of the three-dimensional model corresponding to each point in the initial point cloud during the operation, so as to match the centroid of the triangle with the corresponding point in the initial point cloud during the operation as a set of reference points;

[0058] Part of the intraoperative surface and the preoperative surface are used as the template surface and the target surface respectively, and the template is deformed to fit the shape of the target while maintaining the topological structure of the template itself; a dense registration method is used to define an affine transformation matrix for each vertex of the three-dimensional model, and the affine transformation matrix of all points is obtained by fitting the cost function;

[0059] Based on the correspondence between the determined preoperative points and intraoperative points, the nearest neighbor search method is used on the boundary of the corresponding surface before the operation to find the neighboring points of the boundary triangle, and then expand outward to determine all the points in the position of the area to be completed;

[0060] The inverse of the affine transformation matrix is ​​used to inversely transform the pre-operative surface points to the intra-operative surface, so that all points in the position of the area to be completed are transformed to obtain the intra-operative completion point cloud.

[0061] S104: Biomechanical analysis, applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ; in specific implementation, the embodiment of the present application may provide that the organ includes a liver, and applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ includes:

[0062] Based on the finite element method of Burgers viscoelastic material, the Young's modulus is 4000Pa, the Poisson's ratio is 0.45, the bulk modulus is 7000kPa, and the shear modulus is 2000kPa.

[0063] S105: Using the optimization control method and the intraoperative completed point cloud to predict the deformation vector field of the biomechanical model, so as to correct the deformation of the organ during the operation.

[0064] The objective function of the optimization control method is expressed as follows:

[0065]

[0066] Where: Represents the gradient value of the elastic model residual, p b represents the state variable, Represents the gradient of the difference between the predicted value and the deformed value with respect to the control variable.

[0067] Furthermore, the only optimization variable b is input, and the objective function and the objective function gradient are iteratively calculated using the quasi-Newton optimization algorithm BFGS solver.

[0068] The organ deformation correction method provided in this application aims to solve the problem in existing research that liver deformation cannot be accurately predicted due to the scarcity of intraoperative point cloud data of laparoscopy. It mainly uses the registration technology to complete the point cloud data of the intraoperative organ, and combines the optimization control theory with the biomechanical model to estimate the deformation of the preoperative organ's three-dimensional model to accurately match the intraoperative organ morphology. First, a point cloud completion module is designed, which can effectively use the preoperative organ geometric features to enhance the intraoperative visualization data. Secondly, reliable deformation estimation is performed with acceptable accuracy through the optimization control method. Finally, in order to meet the viscoelastic characteristics of the organ, a viscoelastic model is introduced to improve the accuracy of deformation prediction.

[0069] The method provided in the embodiment of the present application is described in detail below using laparoscopic liver resection surgery as an example.

[0070] See also Figure 2 The deformation correction method in laparoscopic liver surgery navigation based on point cloud completion and biomechanical model provided in the embodiment of the present application mainly includes the following steps:

[0071] Firstly, the complete preoperative model is used to perform local completion of the intraoperative point cloud through non-rigid registration technology. Secondly, the preoperative three-dimensional model is constructed into a biomechanical model. Finally, the complete deformation vector field is predicted using the optimization control method.

[0072] 3D model construction:

[0073] First, a CT scan of the liver is performed. These images can clearly show the shape, size and surrounding tissues of the liver, providing basic data for subsequent modeling. Secondly, the image segmentation algorithm is a key technology for extracting liver features. In this step, grayscale separation algorithms, level set algorithms, etc. are needed to extract the contour feature sequence of the liver and separate the liver from other tissues. Then, the processed liver feature sequence is used for three-dimensional reconstruction, and the three-dimensional reconstruction technology can generate a three-dimensional model of the liver with accurate morphology and structure. The three-dimensional model constructed is a mesh model, which is a model that can use a series of polygons (usually triangles) of similar size and shape to approximate a three-dimensional object.

[0074] Point cloud completion:

[0075] First, find the correspondence between the intraoperative and preoperative points: collect partial surface images of the liver during the operation through endoscopy, and process the partial surface images to obtain the initial point cloud during the operation. Based on R-tree, find the triangle where the projection points of the preoperative surface corresponding to the points on the intraoperative partial surface are located and the centroid coordinates, so as to match the centroid of the triangle with the corresponding intraoperative points as a set of landmarks (reference points). R-tree is another form of B-tree developed into multidimensional space. It divides spatial objects by range. Each node corresponds to an area and a disk page. The disk page of non-leaf nodes stores the area range of all its child nodes, and the areas of all child nodes of non-leaf nodes fall within its area range; the disk page of leaf nodes stores the circumscribed rectangles of all spatial objects within its area range. There are upper and lower limits on the number of child nodes that each node can have. The lower limit ensures the effective use of disk space, and the upper limit ensures that each node corresponds to a disk page. When the insertion of a new node causes the space required by a node to be larger than a disk page, the node is divided into two (split). The R-tree is a dynamic index structure, that is, its query can be performed simultaneously with insertion or deletion, and there is no need to reorganize the tree structure regularly.

[0076] Set the affine transformation matrix: regard part of the intraoperative surface and the preoperative surface as the template surface and the target surface respectively, deform the template to fit the shape of the target while maintaining the template's own topological structure, use the dense registration method, define an affine transformation matrix for each vertex, and obtain the affine transformation matrix of all points through cost function fitting.

[0077] Selection of points to be completed before surgery: Based on the correspondence between the determined preoperative points and the intraoperative points, the nearest neighbor search method is used on the boundary of the corresponding surface before surgery to find the neighboring points of the boundary triangle, and then expand outward to determine the position of the area to be completed.

[0078] Inverse transformation of preoperative point cloud: The inverse of the affine transformation matrix is ​​used to inversely transform the preoperative surface points to the intraoperative surface. After the obtained points are transformed, the completed intraoperative point cloud will be obtained.

[0079] Constructing a biomechanical model of the liver:

[0080] Biomechanical analysis was performed to apply different loads and boundary conditions to the three-dimensional liver model, and its stress, strain and other mechanical parameters were calculated. Because the liver itself has viscoelasticity, the finite element method based on Burgers viscoelastic material was selected, and the Young's modulus was set to 4000Pa, the Poisson's ratio to 0.45, and the bulk modulus and shear modulus to 7000kPa and 2000kPa respectively.

[0081] Optimization control method:

[0082] like Figure 3As shown in the figure, first, at the beginning of the registration process, the biomechanical model is in the preoperative configuration and needs to be deformed to match the intraoperative point cloud. The intraoperative point cloud is only a partial target. After the point cloud completion step is completed, the intraoperative surface point cloud is represented by Γ = {y1,…,y m} means, definition The orthographic projection matrix on is

[0083]

[0084] Here the orthogonal projection operator returns a unique point.

[0085] Secondly, elastic registration can be regarded as an optimization problem between the intraoperative point cloud and the preoperative biomechanical model. We regard registration as a problem of minimizing the difference between the predicted value and the observed data. The optimization problem is:

[0086]

[0087] The constraint is F(u)=b.

[0088] Here B is the set of admissible controls and J is defined using a simple least squares term to make full use of the observed data:

[0089]

[0090] Finally, for the constrained optimization problem, we use the adjoint method to transform it into an unconstrained problem, where we define a state variable p b , then redefine the optimization problem as:

[0091]

[0092] Input the only optimization variable b, use the quasi-Newton optimization algorithm BFGS solver to iteratively calculate the objective function and the objective function gradient, which can achieve faster convergence speed and less memory requirements. The output result is the displacement field of all vertices of the liver.

[0093] In summary, the organ deformation correction method provided by the present application can achieve the purpose of correcting liver deformation on a small part of the surface during surgery with high accuracy.

[0094] See also Figure 4 , the embodiment of the present application can also provide an organ deformation correction device, such as Figure 4 As shown, the device for performing the above-mentioned organ deformation correction method may include:

[0095] The organ three-dimensional model construction unit 401 is used to obtain a CT image obtained by scanning the organ before surgery, and to perform three-dimensional reconstruction of the organ using the CT image to obtain a three-dimensional model of the organ;

[0096] An initial point cloud determining unit 402 is used to collect a partial surface image of the organ during surgery through an endoscope, and process the partial surface image to obtain an initial point cloud during surgery;

[0097] A point cloud completion unit 403 is used to complete the intraoperative point cloud by using the intraoperative initial point cloud and the three-dimensional model using a non-rigid registration algorithm to obtain an intraoperative completed point cloud;

[0098] A biomechanical model building unit 404 is used for biomechanical analysis, and applies different loads and boundary conditions to the three-dimensional model to build a biomechanical model of the organ;

[0099] The optimization unit 405 is used to predict the deformation vector field of the biomechanical model using the intraoperative completed point cloud by utilizing an optimization control method, so as to correct the deformation of the organ during the operation.

[0100] The embodiment of the present application may also provide an organ deformation correction device, the device comprising a processor and a memory:

[0101] The memory is used to store program code and transmit the program code to the processor;

[0102] The processor is used to execute the steps of the above-mentioned organ deformation correction method according to the instructions in the program code.

[0103] like Figure 5 As shown, an organ deformation correction device provided by an embodiment of the present application may include: a processor 10, a memory 11, a communication interface 12 and a communication bus 13. The processor 10, the memory 11 and the communication interface 12 communicate with each other through the communication bus 13.

[0104] In the embodiment of the present application, the processor 10 may be a central processing unit (CPU), an application specific integrated circuit, a digital signal processor, a field programmable gate array or other programmable logic devices.

[0105] The processor 10 may call a program stored in the memory 11. Specifically, the processor 10 may execute operations in an embodiment of the organ deformation correction method.

[0106] The memory 11 is used to store one or more programs, which may include program codes, and the program codes include computer operation instructions. In the embodiment of the present application, the memory 11 at least stores programs for implementing the following functions:

[0107] Acquire a CT image obtained by scanning the organ before surgery, and use the CT image to perform three-dimensional reconstruction on the organ to obtain a three-dimensional model of the organ;

[0108] collecting a partial surface image of the organ during surgery through an endoscope, and processing the partial surface image to obtain an initial point cloud during surgery;

[0109] Using the intraoperative initial point cloud and the three-dimensional model, a non-rigid registration algorithm is used to complete the intraoperative point cloud to obtain an intraoperative completed point cloud;

[0110] Biomechanical analysis, applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ;

[0111] The deformation vector field of the biomechanical model is predicted by using the intraoperative completed point cloud using an optimization control method, so as to correct the deformation of the organ during surgery.

[0112] In one possible implementation, the memory 11 may include a program storage area and a data storage area, wherein the program storage area can store an operating system and application programs required for at least one function (such as a file creation function, a data reading and writing function), etc.; the data storage area can store data created during use, such as initialization data, etc.

[0113] In addition, the memory 11 may include a high-speed random access memory, and may also include a non-volatile memory, such as at least one disk storage device or other volatile solid-state storage device.

[0114] The communication interface 12 may be an interface of a communication module, and is used to connect to other devices or systems.

[0115] Of course, it should be noted that Figure 5 The structure shown does not constitute a limitation on the organ deformation correction device in the embodiment of the present application. In actual applications, the organ deformation correction device may include Figure 5 More or fewer components than shown, or combinations of certain components.

[0116] The embodiment of the present application may also provide a computer-readable storage medium, wherein the computer-readable storage medium is used to store program code, and the program code is used to execute the steps of the above-mentioned organ deformation correction method.

[0117] It should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or device including the elements.

[0118] It can be known from the description of the above implementation methods that those skilled in the art can clearly understand that the present application can be implemented by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present application can be essentially or partly embodied in the form of a software product that contributes to the prior art. The computer software product can be stored in a storage medium such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in the various embodiments of the present application or certain parts of the embodiments.

[0119] Each embodiment in this specification is described in a progressive manner, and the same or similar parts between the embodiments can refer to each other, and each embodiment focuses on the differences from other embodiments. In particular, for the system or system embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can refer to the partial description of the method embodiment. The system and system embodiments described above are merely schematic, wherein the units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment. Ordinary technicians in this field can understand and implement it without creative work.

[0120] The above description is only a preferred embodiment of the present invention and is not intended to limit the protection scope of the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention are included in the protection scope of the present invention.

Claims

1. A method for correcting organ deformation, characterized in that: The method comprises: Acquire a CT image obtained by scanning the organ before surgery, and use the CT image to perform three-dimensional reconstruction on the organ to obtain a three-dimensional model of the organ; collecting a partial surface image of the organ during surgery through an endoscope, and processing the partial surface image to obtain an initial point cloud during surgery; Using the intraoperative initial point cloud and the three-dimensional model, a non-rigid registration algorithm is used to complete the intraoperative point cloud to obtain an intraoperative completed point cloud; Biomechanical analysis, applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ; The deformation vector field of the biomechanical model is predicted by using the intraoperative completed point cloud using an optimization control method, so as to correct the deformation of the organ during surgery.

2. The organ deformation correction method according to claim 1, characterized in that: The intraoperative point cloud is completed using a non-rigid registration algorithm to obtain the intraoperative point cloud, including: Based on the R-tree, find the triangles and centroid coordinates of the projection points on the surface of the three-dimensional model corresponding to each point in the initial point cloud during the operation, so as to match the centroid of the triangle with the corresponding point in the initial point cloud during the operation as a set of reference points; Part of the intraoperative surface and the preoperative surface are used as the template surface and the target surface respectively, and the template is deformed to fit the shape of the target while maintaining the topological structure of the template itself; a dense registration method is used to define an affine transformation matrix for each vertex of the three-dimensional model, and the affine transformation matrix of all points is obtained by fitting the cost function; Based on the correspondence between the determined preoperative points and intraoperative points, the nearest neighbor search method is used on the boundary of the corresponding surface before the operation to find the neighboring points of the boundary triangle, and then expand outward to determine all the points in the position of the area to be completed; The inverse of the affine transformation matrix is ​​used to inversely transform the pre-operative surface points to the intra-operative surface, so that all points in the position of the area to be completed are transformed to obtain the intra-operative completion point cloud.

3. The organ deformation correction method according to claim 1, characterized in that: Using the CT image to perform three-dimensional reconstruction on the organ to obtain a three-dimensional model of the organ includes: Using an image segmentation algorithm in combination with the CT image to extract features to obtain an organ contour feature sequence; The three-dimensional model is obtained by utilizing a three-dimensional reconstruction algorithm in combination with the contour feature sequence.

4. The organ deformation correction method according to claim 3, characterized in that: The image segmentation algorithm includes any one of a grayscale separation algorithm and a level set algorithm.

5. The organ deformation correction method according to claim 1, characterized in that: The organ includes a liver, and applying different loads and boundary conditions to the three-dimensional model to construct a biomechanical model of the organ includes: Based on the finite element method of Burgers viscoelastic material, the Young's modulus is 4000Pa, the Poisson's ratio is 0.45, the bulk modulus is 7000kPa, and the shear modulus is 2000kPa.

6. The organ deformation correction method according to claim 1, characterized in that: The objective function of the optimization control method is expressed as follows: Where: represents the gradient value of the elastic model residual, p b represents the state variable, Represents the gradient of the difference between the predicted value and the deformed value with respect to the control variable.

7. The organ deformation correction method according to claim 6, characterized in that: The only optimization variable b is input, and the quasi-Newton optimization algorithm BFGS solver is used to iteratively calculate the objective function and the objective function gradient.

8. An organ deformation correction device, characterized in that: Used to perform the organ deformation correction method according to any one of claims 1 to 7, the device comprising: An organ three-dimensional model construction unit is used to obtain a CT image obtained by scanning the organ before surgery, and to perform three-dimensional reconstruction of the organ using the CT image to obtain a three-dimensional model of the organ; An initial point cloud determination unit, used for collecting a partial surface image of the organ during surgery through an endoscope, and processing the partial surface image to obtain an initial point cloud during surgery; A point cloud completion unit, used to complete the intraoperative point cloud by using the intraoperative initial point cloud and the three-dimensional model using a non-rigid registration algorithm to obtain an intraoperative completed point cloud; A biomechanical model building unit, used for biomechanical analysis, applies different loads and boundary conditions to the three-dimensional model to build a biomechanical model of the organ; The optimization unit is used to predict the deformation vector field of the biomechanical model using the intraoperative completed point cloud by utilizing an optimization control method, so as to correct the deformation of the organ during the operation.

9. An organ deformation correction device, characterized in that: The device comprises a processor and a memory: The memory is used to store program code and transmit the program code to the processor; The processor is used to execute the organ deformation correction method according to any one of claims 1-7 according to the instructions in the program code.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium is used to store program codes, and the program codes are used to execute the organ deformation correction method according to any one of claims 1-7.

Images