Interventional operation planning method and system based on anatomical structure modeling

Abstract

The invention provides an interventional operation planning method and system based on anatomical structure modeling, and the method comprises the steps: carrying out the anatomical structure modeling of an interventional operation target focus and surrounding tissues and organs, and obtaining an anatomical structure three-dimensional model; based on the anatomical structure three-dimensional model, acquiring initial puncture entry candidate points of the interventional operation; screening the initial puncture entry candidate points of the interventional operation to obtain final puncture entry candidate points of the interventional operation; and carrying out joint visualization on the anatomical structure three-dimensional model, the plurality of final puncture entry candidate points of the interventional operation and the quantitative information of the puncture indexes of the puncture entry candidate points so as to provide puncture entry point simulation and reference for the interventional operation and complete interventional operation planning. According to the method, interpretable reference is provided for decision making of interventional operation planning based on multi-dimensional quantitative indexes and visual risk prompts, the pulmonary nodule interventional operation process can be optimized, and then the risk of the interventional operation is reduced.

Description

Technical Field

[0001] The present invention relates to the field of computer-assisted interventional technology, and in particular, to an interventional surgery planning method and system based on anatomical structure modeling, as well as a corresponding computer terminal and a computer-readable storage medium. Background Art

[0002] For patients with suspected lung cancer lesions in chest CT, further diagnostic examinations are crucial to determine the clinical outcome. On the one hand, it can provide timely and necessary treatment when the nodules are in the early stages of lung cancer, and on the other hand, it can avoid over-medicalization as much as possible when the nodules are considered benign. In the process of lung lesion management, obtaining a cytological or histological biopsy of the lesion is a follow-up diagnostic method for auxiliary imaging examination of lung lesions. "Biopsy" is the abbreviation of biopsy, which refers to the removal of living tissue samples or cell samples from the patient's body through surgery, clamping or puncture, etc. according to the needs of diagnosis and treatment, for pathologists to perform pathological tests. Biopsy and pathological tests are usually performed to gain a deeper understanding of whether the lesion / mass is cancer, the type of cancer, and the nature of the cancer, so as to guide the patient's subsequent overall treatment plan, including surgery, radiotherapy, and drug therapy. For small lesions or peripheral lung lesions, more advanced navigation and positioning systems and specific tools are needed to reach areas that cannot be explored by conventional bronchoscopes.

[0003] The whole lung diagnosis and treatment navigation technology based on bronchoscopic transparenchymal nodule access (BTPNA) can break through the tracheal space barrier. Through X-ray fluoroscopy and virtual navigation bronchoscope guidance, a pathway is directly established from the tracheal arm through the lung parenchyma to the lesion, which not only provides the possibility of diagnosing small lesions and peripheral lung lesions, but also builds a treatment platform. At the same time, the interventional diagnosis and treatment with less damage and side effects achieved through whole lung navigation solves the high risks of pneumothorax and bleeding during and after surgery caused by traditional thoracotomy or transthoracic lung needle aspiration biopsy, which can greatly reduce the patient's pain and facilitate the patient's recovery. The biggest challenge of BTPNA is to calculate the appropriate puncture entry point parameters on the tracheal wall by combining the patient's CT data and three-dimensional reconstruction results. The straight path from the puncture entry point through the lung parenchyma to the lung lesion needs to avoid the intrapulmonary blood vessels or ensure that the risk of puncturing the blood vessels is as small as possible. After calculating the optimal puncture entry point position and angle, a biopsy needle is used to puncture the tracheal wall and the puncture hole is expanded with a balloon. Then, with the assistance of C-arm X-ray fluoroscopy, a catheter with a blunt probe is gradually advanced into the patient's lung parenchyma to the lesion, thereby establishing a sheath working channel that can be used to deliver lesion biopsy or treatment devices.

[0004] Therefore, for the BTPNA procedure, the key issue of preoperative path planning of the bronchoscope is to calculate the parameters of the optimal puncture entry point, which can then assist in providing the optimal puncture entry point position, reach the puncture entry point through the virtual navigation guide line displayed by the endoscopic image, and perform the subsequent "last inch" operation.

[0005] The existing preoperative path planning of bronchoscopes usually has the following technical problems: First, when the extratracheal lesion is located between multiple airway branches, multiple possible navigation paths will be generated, and the navigation system needs to perform complex reachability assessments; second, the vascular environment between the lesion and the bronchus is complex, and coupled with the influence of respiratory movement and heartbeat, the safety of navigation puncture is difficult to guarantee. The navigation system needs to accurately identify blood vessels (even arteries and veins) and avoid puncturing blood vessels to achieve low risk requirements. The existing image-guided bronchoscope system is not yet mature in providing the best path guidance technology for navigation to extratracheal lesions before surgery.

[0006] After searching, we found:

[0007] The Chinese invention patent "Optimal Path Planning Method and Navigation System for Lung Puncture Surgery" with authorization announcement number CN107296645B provides an optimal path planning method for lung puncture surgery, which includes the following steps: S1. Quantitative analysis of the constraint conditions of the clinical standards of lung puncture surgery based on the patient's preoperative CT images, and screening out puncture surgery paths that do not meet the constraint conditions; S2. Quantitative analysis of the target conditions of the clinical standards of lung puncture surgery based on the patient's preoperative CT images, and establishing a sub-objective function based on path risk; S3. Establish an optimal path model based on the above-mentioned constraint conditions and target conditions to obtain the solution set of the optimal path. However, this method is different from the application scenario of the present invention, and there are also differences in technology. The following technical problems still exist:

[0008] The Chinese invention patent with the authorization announcement number CN107296645B is applied to percutaneous puncture surgery, that is, planning the path from the human skin to the tumor target point, while the present invention is aimed at the scenario of bronchoscopic interventional puncture diagnosis and treatment, that is, planning the path from the bronchus to the tumor target point. Compared with percutaneous puncture surgery, the risk of pneumothorax, hemothorax and other complications in transbronchial puncture is lower. The Chinese invention patent method with the authorization announcement number CN107296645B does not consider the obstacle avoidance constraints of the complex vascular environment, nor does it consider the influence of human respiratory movement and heartbeat. Therefore, the invention method is not applicable in the scenario of bronchoscopic interventional puncture diagnosis and treatment. Summary of the invention

[0009] In view of the above-mentioned deficiencies in the prior art, the present invention provides an interventional surgery planning method and system based on anatomical structure modeling, and also provides a corresponding computer terminal and computer-readable storage medium.

[0010] According to one aspect of the present invention, there is provided an interventional surgery planning method based on anatomical structure modeling, comprising:

[0011] Perform anatomical structure modeling on the target lesion of interventional surgery and its surrounding tissues and organs to obtain a three-dimensional anatomical structure model;

[0012] Based on the three-dimensional model of the anatomical structure, obtaining candidate puncture entry points for the initial interventional surgery;

[0013] Screening the initial candidate puncture entry points for the interventional surgery to obtain the final candidate puncture entry points for the interventional surgery;

[0014] The three-dimensional model of the anatomical structure and the quantitative information of the final puncture entry candidate points and puncture indicators of the multiple interventional surgeries are jointly visualized to provide puncture entry point simulation and reference for the interventional surgery and complete the interventional surgery planning.

[0015] Preferably, the acquiring of an initial candidate puncture entry point for an interventional surgery based on the three-dimensional model of the anatomical structure includes:

[0016] Resampling the three-dimensional model of the anatomical structure to remove anisotropy, thereby obtaining a three-dimensional model of the tracheal tree;

[0017] The 3D skeleton of the tracheal tree is extracted based on the connected 3D model of the tracheal tree, and the degree of connectivity DoC between each voxel on the skeleton and the neighboring voxels is calculated; if the DoC of the voxel is equal to 1, the voxel is the end point of the skeleton, that is, it represents the tip of the tracheal tree; if the DoC of the voxel is greater than 2, the voxel is the bifurcation point of the skeleton, that is, it represents the branch node of the tracheal tree; the end point and the bifurcation point are merged into the node of the skeleton;

[0018] All bifurcation points on the skeleton are removed, and connected domains are obtained for the remaining skeletons, each connected domain is identified as a branch, and the two ends of the branch are adjacent to different nodes, and the midpoint of each branch is extracted; the branches on the skeleton correspond one to one with the tracheal segments on the tracheal tree three-dimensional model;

[0019] A subset of the set consisting of all nodes and branch midpoints on the skeleton that is distributed within the region of interest specified by the target lesion is used as the initial puncture entry candidate point for the interventional surgery.

[0020] Preferably, the screening of the initial candidate puncture entry points for the interventional surgery to obtain the final candidate puncture entry points for the interventional surgery includes:

[0021] Providing a reachability index, and using the reachability index to screen out the puncture entry candidate points that do not meet the conditions among the initial puncture entry candidate points of the interventional surgery, to obtain a first group of puncture entry candidate points;

[0022] Providing a low-risk index, and using the low-risk index to screen out the puncture entry candidate points that do not meet the conditions in the first group of puncture entry candidate points to obtain final puncture entry candidate points;

[0023] in:

[0024] The reachability index is a hard index for measuring whether a candidate point on the tracheal wall can be used as a bronchoscope puncture point, and is used to evaluate the matching degree between bronchial tree parameters and bronchoscope and its instrument parameters;

[0025] The low-risk index is a soft index for measuring the risk of bronchoscopic puncture behavior. The index is evaluated when the reachability index is met, and is used to evaluate whether the straight channel from the puncture entry point through the lung parenchyma to the target lesion can avoid risk structures and the channel length is as short as possible.

[0026] Preferably, the method of using the reachability index to screen out the puncture entry candidate points that do not meet the conditions among the initial puncture entry candidate points of the interventional surgery includes:

[0027] The accessibility indicators include three types of information: anatomy, equipment and process; among which:

[0028] The anatomical information includes: defining the intratracheal navigation environment through the tracheal tree three-dimensional model, and providing the following local tracheal feature information using the tracheal tree three-dimensional model T={P, B, S}:

[0029] S={s1,s2,...,s N The inner diameter d of each tracheal segment in aw The inner diameter of the cavity is d aw The calculation is based on the shortest distance from the skeleton point to the trachea surface;

[0030] B={b1,b2,...,b N The bending angle θ between each branch and the adjacent branch in aw ;

[0031] The equipment information includes: the physical characteristics and mechanical properties of the bronchoscope and other related instruments and tools determine the passability, operability and final reachability of the bronchoscope in the tracheal tree, and provide the following parameters:

[0032] The outer diameter of the bronchoscope is d b ;

[0033] The maximum bending angle of the bronchoscope head hinge θ b , the bending angle θ b for passage through the endoluminal bends of the bronchus;

[0034] The length of the bronchoscope that can reach the body b , the length l b The extracorporeal loss caused by the operation needs to be subtracted;

[0035] The length of the puncture needle or transpulmonary tunnel sheath and other instruments t ;

[0036] The maximum bending angle of bronchoscope combined with other related instruments θ b-t , the bending angle θ b-t Provide diagnostic field of view for intraoperative EBUS guidance and puncture;

[0037] The process information includes: the conditions and factors considered for performing the expected tasks from the oral cavity to the puncture entry point and from the puncture entry point to the target lesion, and provides the following parameters:

[0038] B={b1,b2,...,b N The total length of the branches from the oral cavity to the puncture entry point is l mp ;

[0039] The angle θ between the line from the puncture entry point to the target lesion and the bronchial branch where the puncture entry point is located aw-p ;

[0040] The length from the puncture entry point to the target lesion l pn ;

[0041] The number of lung lobes n from the puncture entry point to the target lesion lob ;

[0042] Combining the mutual constraints between the three types of information, namely, anatomy, equipment and process, the constraints for selecting the location of the puncture entry point are established as follows:

[0043] d b <d aw ,θ b >θ aw ,l b > mp

[0044] The constraints for establishing a puncture path without piercing the lung fissure or pleura are:

[0045] θ b-t >θ aw-pn ,l t > pn ,nlob =1

[0046] According to the constraint conditions, a reachable pathway is obtained, and then a first group of candidate puncture entry points for interventional surgery are screened out.

[0047] Preferably, the step of using the low risk index to screen out the puncture entry candidate points that do not meet the conditions in the first group of puncture entry candidate points includes:

[0048] Establish regional risk fields around target lesions;

[0049] Using the expansion channel to simulate the puncture path, calculating the total risk value of the puncture path based on the regional wind scene, and quantifying the low risk index;

[0050] Combined with the quantified low-risk index, the first group of candidate puncture entry points is comprehensively evaluated and a plurality of final candidate puncture entry points are selected.

[0051] Preferably, the establishing of a regional risk field around the target lesion comprises:

[0052] Based on the voxel set Φ of the given 3D blood vessel segmentation mask, the Euclidean distance transform output D corresponding to the voxel point K = (x, y, z) in the target lesion area VOI is calculated as:

[0053]

[0054] Where V = (u, v, w) is the coordinate of the vascular voxel point;

[0055] Using the minimum and maximum normalization method, each D(K) in the target lesion area VOI is rescaled so that D(K)∈[0,1], and the risk field value F(K) corresponding to the voxel point K=(x,y,z) is 1-D(K).

[0056] Preferably, the puncture path is simulated by using the expansion channel, the total risk value of the puncture path is calculated based on the regional wind scene, and the low risk index is quantified, including:

[0057] The 3D Bresenham line algorithm is used to generate the voxel connection line L between the candidate puncture point and the center of the target lesion in the spatial coordinates. pn ;

[0058] The voxel connection line L is connected by morphological dilation method pn Expanded into a puncture path similar to the tracheal cylindrical passage The puncture path The total risk value Ψ in the risk field of the region is expressed as:

[0059] Ψ=∑F(K),

[0060] The total risk value Ψ is used to measure the total probability that each pixel of the puncture path body may collide with a blood vessel;

[0061] Set low-risk indicators, including: average risk value Indicators, arterial risk value Indicators and Venous Risk Values Indicators, the low-risk indicators are quantitatively calculated as follows:

[0062] Calculate the average risk value index: Where n is the puncture path The number of voxels in It is not affected by the puncture length and is used to measure the risk value of unit voxels on the puncture path;

[0063] Calculate arterial risk value index: Among them, Φ artery is the voxel set of the artery segmentation mask, It is used to measure the number of intersections between the puncture path and the arterial pixels;

[0064] Calculate the venous risk value index: Among them, Φ vein is the set of voxels of the vein segmentation mask, Used to measure the number of intersections between the puncture path and vein pixels.

[0065] Preferably, the step of combining the quantified low-risk index, comprehensively evaluating the first group of candidate puncture entry points and selecting a plurality of final candidate puncture entry points comprises:

[0066] Comprehensively evaluate multiple indicators of each candidate puncture entry point in the target lesion regional VOI, and establish the overall evaluation function E of low risk and accessibility:

[0067]

[0068] Among them, l pn and θ aw-p are the reachability indicators used to measure the puncture length and angle respectively; α, β, γ and δ are the hyperparameters for balancing various indicators. pn and θ aw- The hyperparameters γ and δ of the term are set according to the risk thresholds of the puncture path length and puncture angle;

[0069] A plurality of puncture entry candidate points whose total evaluation function values ​​are less than a set threshold are selected as the final puncture entry candidate points.

[0070] According to another aspect of the present invention, there is provided an interventional surgery planning method based on anatomical structure modeling, comprising:

[0071] The anatomical structure modeling module is used to perform anatomical structure modeling on the target lesion of the interventional surgery to obtain a three-dimensional model of the anatomical structure;

[0072] A puncture entry candidate point determination module, which obtains an initial puncture entry candidate point for interventional surgery based on the three-dimensional model of the anatomical structure;

[0073] A puncture entry candidate point screening module, which is used to screen the initial puncture entry candidate points of the interventional surgery to obtain the final puncture entry candidate points of the interventional surgery;

[0074] A visualization and human-computer interaction module is used to jointly visualize the anatomical structure three-dimensional model and multiple final puncture entry candidate points for the interventional surgery, so as to provide puncture entry point simulation and reference for the interventional surgery and complete the interventional surgery planning.

[0075] According to a third aspect of the present invention, there is provided a computer terminal comprising a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the processor, when executing the computer program, can be used to execute any one of the methods described above in the present invention, or to run the system described above in the present invention.

[0076] According to a fourth aspect of the present invention, there is provided a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, can be used to execute any of the methods described above in the present invention, or to run the system described above in the present invention.

[0077] Due to the adoption of the above technical solution, the present invention has at least one of the following beneficial effects compared with the prior art:

[0078] The interventional surgery planning method and system based on anatomical structure modeling provided by the present invention can assist in finding the optimal navigation path, puncture point and puncture angle through preoperative modeling, thereby improving the safety and success rate of interventional surgery, protecting the patient's postoperative tracheal function and thus improving the prognosis.

[0079] The interventional surgery planning method and system based on anatomical structure modeling provided by the present invention can intuitively and accurately visualize the three-dimensional spatial structure around the target lesion and puncture path and the evaluation index of each candidate puncture entry point, thereby providing an interpretable reference for precise preoperative planning and the formulation of individualized navigation plans.

[0080] The interventional surgery planning method and system based on anatomical structure modeling provided by the present invention can be applied to bronchoscopic surgery (BTPNA). It provides auxiliary explanatory reference for surgical planning decisions based on multi-dimensional quantitative indicators and visualized risk prompts, which helps to optimize the interventional surgery process and reduce the risk of interventional surgery. BRIEF DESCRIPTION OF THE DRAWINGS

[0081] Other features, objects and advantages of the present invention will become more apparent from the detailed description of non-limiting embodiments made with reference to the following drawings:

[0082] Figure 1 The figure is a workflow diagram of an interventional surgery planning method based on anatomical structure modeling in one embodiment of the present invention.

[0083] Figure 2 Schematic diagram of branch extraction of tracheal tree, tracheal segment and tracheal skeleton in a preferred embodiment of the present invention.

[0084] Figure 3 Schematic diagram of a strategy for screening candidate puncture entry points for interventional surgery in a preferred embodiment of the present invention.

[0085] Figure 4 This is a schematic diagram of the visualization results of the final six puncture points in a specific application example of the present invention.

[0086] Figure 5 This is a schematic diagram of the visualization results of the final three puncture points in another specific application example of the present invention.

[0087] Figure 6 Schematic diagram of component modules of an interventional surgery planning system based on anatomical structure modeling in one embodiment of the present invention. DETAILED DESCRIPTION

[0088] The following is a detailed description of the embodiments of the present invention: This embodiment is implemented on the premise of the technical solution of the present invention, and a detailed implementation method and a specific operation process are given. It should be pointed out that for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention.

[0089] The existing preoperative path planning technology of bronchoscopes usually fails to consider the vascular risks and the influence of respiratory movement and heartbeat during the puncture process, resulting in unsafe and unreasonable planned puncture paths. In response to the above problems, an embodiment of the present invention provides an interventional surgery planning method based on anatomical structure modeling. The method reconstructs lesions and other chest and lung anatomical structure models for a given patient's preoperative CT images and CT reports, obtains candidate puncture entry points for interventional surgery, and selects the best candidate points. It uses visualization and human-computer interaction technology to assist in planning preoperative decisions, which has clinical application value.

[0090] Specifically, Figure 1 As shown, the interventional surgery planning method based on anatomical structure modeling provided in this embodiment may include the following operations:

[0091] S1, anatomical structure modeling of the target lesion of interventional surgery and its surrounding tissues and organs to obtain a three-dimensional anatomical structure model;

[0092] S2, based on the 3D model of the anatomical structure, obtain the candidate entry point for the initial puncture of the interventional surgery;

[0093] S3, screening the initial puncture entry candidate points for the interventional surgery to obtain the final puncture entry candidate points for the interventional surgery;

[0094] S4, jointly visualizes the 3D model of the anatomical structure and the final candidate puncture entry points of multiple interventional surgeries and the quantitative information of their puncture indicators, to provide puncture entry point simulation and reference for interventional surgery and complete interventional surgery planning.

[0095] In order to realize anatomical structure modeling, in some preferred embodiments, the above S1 may further include the following operations:

[0096] For a given patient's preoperative CT images and CT reports, the three-dimensional coordinates of the target lesion of the interventional surgery are obtained, and the anatomical structure of the target lesion and multiple related tissues and organs is automatically reconstructed using targeted three-dimensional segmentation technology. The reconstructed anatomical structure three-dimensional model is used to intuitively display the location, size and shape of the target lesion, and can clearly present the spatial position relationship between the target lesion and the adjacent tissues. In a specific application example, for BTPNA interventional surgery, multiple related tissues and organs include: chest and lung tissues and organs; adjacent tissues include: pulmonary trachea, pulmonary arteries and veins, lung lobes (including pleura, lung fissures) and lymph nodes, etc.; the target lesion is a lung nodule; and the constructed anatomical structure three-dimensional model is a tracheal tree three-dimensional model.

[0097] In some preferred embodiments, Figure 2 As shown, the above S2 may further include the following operations:

[0098] S21, after resampling the anatomical structure three-dimensional model constructed in S1 to solve the anisotropy, a tracheal three-dimensional tree model is obtained; first, the tracheal tree 3D skeleton is extracted based on the connected tracheal tree three-dimensional model, and the degree of connectivity (DoC) between each voxel on the skeleton and the neighboring voxel is calculated. If the DoC of the voxel is equal to 1, then the voxel is the end point of the tracheal tree skeleton, that is, it represents the tip of the tracheal tree; if the DoC of the voxel is greater than 2, then the voxel is the bifurcation point of the tracheal tree skeleton, that is, it represents the branch node of the tracheal tree (the tracheal tree can derive multiple branches in any direction). The end point and the bifurcation point are merged into the node of the tracheal skeleton. In order to better analyze the structure of the tracheal tree, the branches of the tracheal tree skeleton are split. In the splitting process, first remove all bifurcation points on the skeleton, then find the connected domain of the remaining skeleton, each connected domain is identified as a branch, and the two ends of the branch are adjacent to different nodes. Extract the midpoint of each branch. Each tracheal segment on the tracheal tree has a branch on the tracheal tree skeleton corresponding to it.

[0099] S22, let the global tracheal tree structure be T = (P, B, S), where P = {p1, p2, ..., p M} is the set of all nodes (points) and midpoints (midpoints) of each branch on the tracheal tree skeleton, B = {b1, b2, ..., b N} is the set of all branches on the tracheal tree skeleton, S = {s1, s2, ..., s N} is the set of all tracheal segments on the tracheal tree, and the tracheal segments correspond to the branches of the tracheal skeleton one by one. The subset of the set P distributed within the specified VOI of the lesion is used as the initial puncture entry candidate point.

[0100] In some preferred embodiments, Figure 3 As shown, the above S3 may further include the following operations:

[0101] S31, firstly, screening out inappropriate puncture entry candidate points for interventional surgery based on the reachability index; wherein:

[0102] The reachability index is a hard indicator to measure whether a candidate point on the tracheal wall can be used as a bronchoscope puncture point. It is mainly used to evaluate the matching degree between bronchial tree parameters and bronchoscope and its instrument parameters. To achieve the arrival of all lung lesions, feasible intratracheal and extratracheal routes must be able to accommodate specific bronchoscopes and puncture instruments and have sufficient maneuverability throughout the route. At the same time, the lungs and trachea of ​​different patients differ in size and morphology. Therefore, under the existing medical equipment conditions of the hospital, it is necessary to first measure whether the available bronchoscope can reach the preset puncture entry point, and secondly verify whether the operable length and angle of the puncture instrument can reach the lesion.

[0103] S32, selecting the optimal candidate puncture entry point for the interventional surgery based on the low risk index; wherein:

[0104] The low-risk index is a soft index that measures the risk of bronchoscopic puncture. The soft index is evaluated when the hard index (i.e., the reachability index) is met. The low-risk index requires that the straight channel from the puncture entry point through the lung parenchyma to the lung lesion needs to avoid risky structures such as intrapulmonary blood vessels. At the same time, the length of the channel should be as small as possible.

[0105] In some preferred implementations, the above S31 may further include the following operations:

[0106] The constraints of three types of information, namely, anatomy, device, and procedure, are combined to screen out candidate puncture entry points that do not meet the reachability index conditions, including:

[0107] Anatomical information: The 3D segmentation structure of the tracheal tree defines the environment for intratracheal navigation. The 3D tracheal tree model T = {P, B, S} can provide the following local tracheal feature information (for clarity, the subscript index is removed below):

[0108] S={s1,s2,...,s N The luminal internal diameter of each tracheal segment in aw (Unit: mm), the inner diameter of each tracheal segment is calculated based on the shortest distance from the skeleton point to the tracheal surface;

[0109] B={b1,b2,...,b N The angle between each branch and the adjacent branch in aw (Unit: degree).

[0110] Equipment information: The physical characteristics and mechanical properties of bronchoscopes and other related instruments determine their passability, maneuverability, and ultimate reachability within the tracheal tree. Bronchoscopes can provide the following parameters:

[0111] The outer diameter of the bronchoscope is d b (Unit: mm);

[0112] Maximum bending angle of the bronchoscope head b (Unit: degrees), the bending angle is used to pass through the intraluminal bend of the bronchus;

[0113] The length of the bronchoscope that can reach the body b (Unit: mm), this length needs to take into account the external loss caused by the doctor or robot operation;

[0114] The length of the puncture needle or transpulmonary tunnel sheath and other instruments t (Unit: mm);

[0115] The maximum bending angle of bronchoscope combined with other related instruments θ b-t (Unit: degree), this bending angle provides a diagnostic field of view (FoV) for intraoperative EBUS guidance and puncture.

[0116] Process Information: Factors to consider in performing the intended tasks from the oral cavity to the puncture entry point and from the puncture entry point to the lesion:

[0117] B={b1,b2,...,b N The total length of the branches from the oral cavity to the puncture entry point is l mp (Unit: mm);

[0118] The angle θ between the line from the puncture entry point to the lesion and the bronchial branch where the puncture entry point is located aw-pn (Unit: degree);

[0119] The length from the puncture entry point to the lesion l pn (Unit: mm);

[0120] The number of lung lobes n from the puncture entry point to the lesion lob .

[0121] First, these parameter information constrain each other in a specific relationship to screen out accessible pathways. The bronchoscope must first be able to navigate along a complete path leading to the puncture entry point, that is, the outer diameter of the bronchoscope must be able to adapt to the minimum lumen diameter of all tracheal trees on the route, and can pass through all tracheal tree bends on the route through operation. At the same time, the available length of the bronchoscope itself needs to meet the travel length from the patient's mouth to the puncture entry point, so these restrictions bring three constraints to the location selection of the puncture entry point:

[0122] d b <d aw ,θ b >θ aw ,l b > mp

[0123] Secondly, considering the puncture entry point to reach the lesion, the bronchoscope combined with other related instruments must have sufficient flexibility and length to perform target positioning and the required puncture navigation tasks, and the established puncture pathway must not pierce the lung fissure or pleura to cause pneumothorax, which needs to meet the following requirements:

[0124] θ b-t >θ aw-pn ,l t > pn ,n lob =1.

[0126] In order to overcome the risk of bleeding caused by inaccurate puncture position and angle and respiratory movement and heartbeat, the selection method based on low-risk indicators of S32 is as follows: Figure 3 As shown, the following operations may be further included:

[0127] Establish a regional risk field around the lesion; use the expansion channel to simulate the puncture path, and quantify the risk index (puncture index) based on the expansion channel; combine the "low risk" and "reachability" indicators to comprehensively evaluate and select multiple optimal puncture entry candidate points. They include:

[0128] (1) Establishing a regional risk field around the lesion

[0129] Ideally, if the straight line L between the puncture entry point and the lesion does not intersect any blood vessels in the preoperative 3D model, the puncture path can be considered safe. However, in reality, if the path is close to a blood vessel, it is still considered to have a great risk. Therefore, for the quantification of the risk of vascular collision, the blood vessel cannot be used as a binary segmentation model alone, and the distance to the blood vessel must be included in the risk quantification to construct an elastic probabilistic risk model.

[0130] For this purpose, a regional risk field is proposed. First, in the lesion region VOI, the Euclidean distance transform is calculated based on the vascular segmentation mask. Given a voxel set Φ of a three-dimensional vascular segmentation mask (artery segmentation mask + vein segmentation mask), the Euclidean distance transform output D corresponding to the voxel point K = (x, y, z) in the lesion region VOI is calculated as follows:

[0131]

[0132] Where V = (u, v, w) is the coordinate of the vascular voxel point;

[0133] The distance transform calculates the minimum Euclidean distance from each voxel point in the lesion area to the nearest vascular voxel. If the voxel belongs to a vascular voxel, its distance transform value is 0.

[0134] Then, the minimum and maximum normalization method is used to rescale each D(K) in the lesion area VOI so that D(K)∈[0,1]. Since the risk value is higher when it is closer to the blood vessel, the risk field value F(K) corresponding to the voxel point K=(x,y,z) is 1-D(K).

[0135] (2) Use the expansion channel to simulate the puncture path and quantify the risk index based on the expansion channel

[0136] The three-dimensional anatomical structure reconstructed based on preoperative CT is fixed. During the operation, the anatomical structure of the patient's body will vary due to factors such as time, posture, and anesthesia status. In addition, due to the influence of breathing and heartbeat, human tissues and organs will change dynamically at all times. Therefore, the preoperative model information cannot be completely used as an absolute position reference. Therefore, the risk range can be widened by increasing the diameter of the puncture channel. This method converts the disturbance of structures such as arteries and veins to the expandable structure of the puncture channel. At the same time, considering the diameter of the puncture sheath, the 3DBresenham straight line algorithm is first used to generate the voxel connection L between the puncture entry candidate point and the center of the lesion in the spatial coordinates. pn Then, the morphological expansion method is used to expand the connection into a set of puncture channels (puncture paths) similar to the tracheal cylindrical passage. The structural element used for the dilation operation is a sphere with a radius of 2 unit pixels. The total risk value Ψ of this channel in the regional risk field is expressed as:

[0137] Ψ=∑F(K),

[0138] Among them, Ψ measures the sum of the probabilities that each pixel in the channel body may collide with a blood vessel.

[0139] The quantitative information of puncture indicators specifically refers to the various indicators in the total evaluation function of the puncture candidate point, including the average risk value, arterial risk value, venous risk value, puncture length and puncture angle. In order to fairly quantify and compare the channel risks of different puncture entry candidate points to nodules, three risk indicators are quantitatively calculated, such as Figure 3 The right part shows:

[0140] Calculate the average risk value index. n is The number of voxels in The risk value of each voxel on the puncture channel is measured without being affected by the puncture length.

[0141] Calculate arterial risk value index. Φ artery is the voxel set of the artery segmentation mask, The number of intersections between the puncture channel and the arterial pixels is measured.

[0142] Calculate the venous risk value Indicators. Similarly, Φ vein is the set of voxels of the vein segmentation mask, It measures the number of intersections between the puncture channel and vein pixels.

[0143] In order to obtain a puncture path that satisfies the low risk, short distance, and small angle indicators at the same time, multiple indicators of each puncture entry seed point (candidate point) in the lesion regional VOI are comprehensively evaluated, and the overall evaluation function combining low risk and reachability is defined as follows:

[0144]

[0145] Among them, l pn and θ aw-pn are the accessibility indicators for measuring the puncture length and angle, β, γ and δ are the hyperparameters for balancing various indicators. In a preferred embodiment, α=0.2, β=0.1, l pn and θ aw-pnThe hyperparameters γ and δ of the item are set according to the risk thresholds of the puncture path length and the puncture angle (preferably 20 mm and 90 degrees). Since the consequences of puncturing an artery are generally more serious than those of puncturing a vein, α is preferably twice β. Finally, multiple puncture entry points with a total evaluation function value less than the set threshold are taken as the optimal puncture entry points to provide candidate points for the final decision. When making decisions, you can also interactively see the positional relationship between the visualized anatomical structure and the puncture channel, simulate bronchoscope navigation and puncture operations, and at the same time, the risk information of the candidate points and the descriptions of the puncture length, angle information, etc. ensure the reliability and interpretability of the system point selection.

[0146] The effect of the interventional surgery planning method based on anatomical structure modeling provided by the above embodiment of the present invention is verified in conjunction with a verification example.

[0147] 1. Experimental conditions

[0148] This verification example assumes that the hospital equipment meets the reachability constraint, that is, the bronchus can reach the puncture point, and the puncture tool can reach the nodule from the puncture point. In the total evaluation score, the two reachability information (the length from the puncture entry point to the lesion l pn , the angle θ between the line from the puncture entry point to the lesion and the bronchial branch where the puncture entry point is located aw-pn ) as an additional risk factor, because as the length of the puncture path or the puncture angle increases, the stability of the instrument itself will be affected, and the uncertainty of the puncture process will also increase. All chest and lung anatomical structure reconstructions and candidate path indicator quantifications are inferred and calculated using dual Intel Xeon Gold 6226R CPUs and a single NVIDIA RTX A6000 GPU.

[0149] 2. Experimental Content

[0150] In order to verify the feasibility and interpretability of the interventional surgery planning method based on anatomical structure modeling provided by the above embodiment of the present invention, two patient cases were used for experimental verification. In the experiment, the clinician needs to cooperate with the path planning system to perform the following operations: 1) The doctor clicks to confirm the center position of the nodule to be punctured in the CT image. This operation not only provides accurate coordinates for nodule segmentation, but also determines the nodule to be punctured for the path planning algorithm. Especially in the case of multiple nodules in the patient, confirming the nodule to be punctured is an indispensable step. Clicking the center position of the nodule is to simulate the behavior of the end point of the puncture path entering the nodule. The puncture channel reaching the inside of the nodule is conducive to the full sampling and ablation of the nodule tissue. At the same time, setting the end point of the path planning at the center point of the nodule is conducive to fair comparison of paths from different directions. 2) The doctor determines the VOI search range of the puncture entry candidate point according to the nodule and the nearby bronchus. This operation requires setting the VOI side length parameter. The parameter defaults to 80 mm. If the nodule does not belong to a peripheral pulmonary nodule or is close to multiple bronchi, the parameter can be adjusted to narrow the search range and speed up the planning process. 3) After the candidate paths are generated, the doctor refers to the explanatory indicators and virtual visualization models of different candidate paths to decide the optimal navigation path. After this step is completed, the intraoperative process begins.

[0151] 1. Specific case 1

[0152] This case is from a 67-year-old male patient with pulmonary nodules. The patient's CT images were collected by Philips iCT 256 equipment, with a size of 1024×1024×400, an XY axis resolution of 0.3906×0.3906 mm, and a Z axis layer thickness of 1 mm. According to the CT report, the patient has a solid nodule in the lateral segment of the middle lobe, with a maximum diameter of about 17 mm, and lobulation and burr signs can be seen. Chronic inflammation of both lungs, emphysema, and multiple small bullae. The mediastinum is centered, and small lymph node shadows are seen in the mediastinum. Further examination is required.

[0153] like Figure 4 As shown in Figure 2, the nodule in this patient is close to multiple bronchi, but there are no bronchi directly connected to it. Note that one trachea is very close to the nodule (see Figure 4 Puncture path 5), this trachea has an advantage over other trachea in terms of the length of the puncture path (12.13 mm). Since this nodule does not belong to the peripheral pulmonary nodules and is close to multiple bronchi, the VOI search range of the puncture entry candidate point is reduced from VOI side length of 80 mm to 50 mm, and 20 puncture entry candidate points, i.e., 20 puncture candidate paths, are generated. Figure 4As shown, the path planning algorithm proposed in the present invention automatically calculated the six puncture paths with the lowest total evaluation scores, which took 446 seconds. Among them, puncture path 5 is only ranked fifth among the six puncture paths with the lowest total evaluation scores, and the four puncture paths with lower total evaluation scores than puncture path 5 are all longer than puncture path 5, and are even three times the length of puncture path 5. The main reason for this situation is that puncture path 5 intersects with the artery near the nodule, resulting in a higher average risk value (0.9002) and arterial risk value (16), as shown in Table 1. In this case, if the puncture entry point is selected according to the length of the puncture path, the patient will be at potential risk of massive bleeding. The average risk value (0.8355) and arterial risk value (5) of puncture path 4 are lower than those of puncture path 5, but the number of lobes passed by puncture path 4 is 2, which means that puncture path 4 has the risk of puncturing a pulmonary fissure. According to Figure 4 The left middle three-dimensional rendering and the 4-row CT image of the puncture path confirm that puncture path 4 passes through the trachea of ​​the right upper lobe, pierces the horizontal fissure and reaches the nodule of the right middle lobe. The cross-lobar puncture path is not advisable because the horizontal fissure is an extension of the pleura, and the rupture of the horizontal fissure also has the risk of pneumothorax or hemothorax. Due to the sparse trachea near the horizontal fissure of the lung, the risk field value near the puncture path is small, and the average risk value of puncture path 4 is less than the average risk value of puncture paths 1-3. Figure 4 The CT images of puncture paths 1-3 respectively show the process of the puncture path from the tracheal wall, into the lung parenchyma, and reaching the lung nodule. As shown in Table 1, none of the puncture paths 1-3 intersected the artery and vein, and the puncture length (31.50 mm) and puncture angle (34.31 degrees) of puncture path 1 were better.

[0154] Table 1. Quantitative results of evaluation indicators of the six optimal paths of Case 1 calculated by different planning methods

[0155]

[0156]

[0157] The experiment of specific case 1 measured the impact of using risk field and channel expansion on the recommendation of the optimal puncture path. As shown in Table 1, "naive calculation" means canceling the risk field and channel expansion, and "w / risk field" and "w / channel expansion" respectively apply two measures based on naive calculation. "Naive calculation" and "w / risk field" will expand the 3DBresenham line to a smaller scale (structural element: a sphere with a radius of 1 unit pixel) to simulate the puncture path. The total evaluation scores calculated by these two methods both mistakenly regard puncture path 5 as the optimal puncture path. Among them, the total evaluation score of puncture path 5 obtained by "naive calculation" is smaller than that of other puncture paths, and the difference is also greater; while "w / risk field" can alleviate this highly confident but wrong recommendation situation. Figure 4 The partial enlarged image in the middle shows the expanded simulation path. "w / channel expansion" reconstructs the puncture path into a thicker tunnel to allow it to interact more with the surrounding environment. At the same time, "w / channel expansion" is conducive to better leveraging the risk simulation advantages of "w / risk field", so "w / channel expansion" is crucial. Table 1 shows that "w / channel expansion" not only prevents the situation of seeking benefits and avoiding harm when calculating the total evaluation score, but also cooperates with "w / risk field" to achieve more objective risk calculation.

[0158] 2. Specific case 2

[0159] This case is from a 66-year-old male patient with lung nodules. The patient's CT images were collected by Philips Ingenuity equipment, with a size of 512×512×282, an XY axis resolution of 0.8203×0.8203 mm, and a Z axis layer thickness of 1 mm. According to the CT report, the patient has a solid nodule and a ground-glass nodule in the right upper lobe. The solid nodule has a maximum diameter of about 21 mm in the posterior segment of the right upper lobe, with clear boundaries, accompanied by adjacent pleural traction, and enhancement can be seen after enhancement, considering the possibility of malignant lesions; the ground-glass nodule is in the apical segment of the right upper lobe, with a maximum diameter of about 8 mm and clear boundaries. Multiple slightly enlarged lymph nodes were seen in the mediastinum, and the pleura on both sides was locally thickened, requiring further examination.

[0160] A path planning experiment was conducted on the more challenging small peripheral lung nodules in the specific case 2, such as Figure 5As shown in the figure, the nodule is located in the apical segment of the right upper lobe and is a certain distance away from any bronchus. Since CT is presented in the form of a section, it is difficult for doctors to mark the path on CT. At the same time, even if the pulmonary nodules, trachea, and arteries and veins are visualized using a three-dimensional reconstruction algorithm, it is difficult for doctors to find the optimal path from the intricate tubular forest. Since the nodule is a peripheral pulmonary nodule and is not close to any bronchus, the VOI search range for the puncture entry candidate point is set to the default side length of 80 mm. This VOI only generates 3 puncture entry candidate points. Therefore, the path planning algorithm proposed in this chapter only needs to calculate the indicators of these 3 candidate points and sort them according to the total evaluation score. These operations only take 46.36 seconds.

[0161] Table 2. Quantitative results of evaluation indicators of the three optimal paths in Case 2 calculated by the planning method proposed in this invention

[0162]

[0163] As shown in Table 2 and Figure 5 As shown, the three puncture paths corresponding to the three puncture entry candidate points all intersect with the blood vessels, and the volume of the intersection between puncture path 1 and the blood vessel is the smallest. Similar to case 1, the puncture length of puncture path 1 (33.97 mm) is longer than the puncture length of the subsequent puncture path 3 (26.50 mm), but because puncture path 3 intersects with the artery over a large area, the total evaluation score is much worse than that of puncture path 1. The automatic path planning algorithm proposed in the present invention can not only clearly display the accessibility and risk indicators of different candidate paths, but also intuitively show doctors the blood vessel segments that need to be paid attention to and avoided in the puncture path, thereby ensuring the clinical practicality and interpretability of the path planning algorithm.

[0164] Based on the same inventive concept, an embodiment of the present invention further provides an interventional surgery planning system based on anatomical structure modeling, which can be used to implement the interventional surgery planning method based on anatomical structure modeling of any of the above embodiments of the present invention.

[0165] Specifically, Figure 6 As shown, the interventional surgery planning system based on anatomical structure modeling provided in this embodiment may include: an anatomical structure modeling module, a puncture entry candidate point determination module, a puncture entry candidate point screening module, and a visualization and human-computer interaction module; wherein:

[0166] The anatomical structure modeling module is used to perform anatomical structure modeling on the target lesion of the interventional surgery to obtain a three-dimensional model of the anatomical structure;

[0167] The puncture entry candidate point determination module obtains the initial puncture entry candidate point for interventional surgery based on the 3D model of the anatomical structure;

[0168] The puncture entry candidate point screening module is used to screen the initial puncture entry candidate points for the interventional surgery to obtain the final puncture entry candidate points for the interventional surgery;

[0169] The visualization and human-computer interaction module is used to jointly visualize the 3D model of the anatomical structure and the final candidate puncture entry points of multiple interventional surgeries, to provide puncture entry point simulation and reference for interventional surgery, and to complete interventional surgery planning.

[0170] It should be noted that the steps in the method provided by the present invention can be implemented by using corresponding modules, devices, units, etc. in the system. Those skilled in the art can refer to the technical solution of the method to realize the composition of the system, that is, the embodiments in the method can be understood as preferred examples for constructing the system, which will not be elaborated here.

[0171] An embodiment of the present invention also provides a computer terminal, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the program, it can be used to execute any one of the interventional surgery planning methods based on anatomical structure modeling of the above-mentioned embodiments of the present invention, or, to run any one of the interventional surgery planning systems based on anatomical structure modeling of the above-mentioned embodiments of the present invention.

[0172] Optionally, the memory is used to store programs; the memory may include volatile memory (English: volatile memory), such as random-access memory (English: random-access memory, abbreviated: RAM), such as static random-access memory (English: static random-access memory, abbreviated: SRAM), double data rate synchronous dynamic random access memory (English: Double Data Rate Synchronous Dynamic Random Access Memory, abbreviated: DDR SDRAM), etc.; the memory may also include non-volatile memory (English: non-volatile memory), such as flash memory (English: flash memory). The memory is used to store computer programs (such as applications, functional modules, etc. that implement the above method), computer instructions, etc., and the above computer programs, computer instructions, etc. can be partitioned and stored in one or more memories. And the above computer programs, computer instructions, data, etc. can be called by the processor.

[0173] The above-mentioned computer programs, computer instructions, etc. may be stored in one or more memories in partitions, and the above-mentioned computer programs, computer instructions, data, etc. may be called by a processor.

[0174] The processor is used to execute the computer program stored in the memory to implement the various steps of the method or various modules of the system involved in the above embodiments. For details, please refer to the relevant descriptions in the above method and system embodiments.

[0175] The processor and the memory may be independent structures or integrated structures. When the processor and the memory are independent structures, the memory and the processor may be coupled and connected via a bus.

[0176] An embodiment of the present invention further provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, can be used to execute the interventional surgery planning method based on anatomical structure modeling of any one of the above-mentioned embodiments of the present invention, or to run the interventional surgery planning system based on anatomical structure modeling of any one of the above-mentioned embodiments of the present invention.

[0177] Among them, computer-readable media include computer storage media and communication media, wherein the communication media include any media that facilitates the transmission of computer programs from one place to another. The storage medium can be any available medium that can be accessed by a general or special-purpose computer. An exemplary storage medium is coupled to the processor so that the processor can read information from the storage medium and write information to the storage medium. Of course, the storage medium can also be an integral part of the processor. The processor and the storage medium can be located in an ASIC. In addition, the ASIC can be located in a user device. Of course, the processor and the storage medium can also be present in a communication device as discrete components.

[0178] The interventional surgery planning method and system based on anatomical structure modeling provided in the above-mentioned embodiments of the present invention can assist in finding the optimal navigation path, puncture point and puncture angle through preoperative modeling, thereby improving the safety and success rate of interventional surgery, protecting the patient's postoperative tracheal function and thus improving the prognosis; it can intuitively and accurately visualize the three-dimensional spatial structure around the target lesion and puncture path and the evaluation index of each puncture entry candidate point, thereby providing an interpretable reference for precise preoperative planning and the formulation of individualized navigation plans; it can be applied to bronchoscopic surgery (BTPNA), and provides auxiliary explanatory references for surgical planning decisions based on multi-dimensional quantitative indicators and visualized risk prompts, which helps to optimize the interventional surgery process and reduce the risk of interventional surgery.

[0179] All matters not covered in the above embodiments of the present invention are well known in the art.

[0180] The above describes the specific embodiments of the present invention. It should be understood that the present invention is not limited to the above specific embodiments, and those skilled in the art may make various modifications or variations within the scope of the claims, which do not affect the essence of the present invention.

Claims

1. An interventional surgery planning method based on anatomical structure modeling, characterized in that: include: Perform anatomical structure modeling on the target lesion of interventional surgery and its surrounding tissues and organs to obtain a three-dimensional anatomical structure model; Based on the three-dimensional model of the anatomical structure, obtaining candidate puncture entry points for the initial interventional surgery; Screening the initial candidate puncture entry points for the interventional surgery to obtain the final candidate puncture entry points for the interventional surgery; The three-dimensional model of the anatomical structure and the quantitative information of the final puncture entry candidate points and puncture indicators of the multiple interventional surgeries are jointly visualized to provide puncture entry point simulation and reference for the interventional surgery and complete the interventional surgery planning.

2. The interventional surgery planning method based on anatomical structure modeling according to claim 1, characterized in that: The anatomical structure modeling of the target lesion of the interventional surgery is performed to obtain a three-dimensional anatomical structure model, including: For a given preoperative CT image and CT report, obtain the three-dimensional coordinates of the target lesion for interventional surgery; The three-dimensional segmentation technology is used to automatically reconstruct the anatomical structure of the target lesion and multiple related tissues and organs to obtain a reconstructed three-dimensional model of the anatomical structure, which is used to describe the location, size and shape of the target lesion and the spatial position relationship between the target lesion and adjacent tissues.

3. The interventional surgery planning method based on anatomical structure modeling according to claim 1, characterized in that: The step of obtaining an initial candidate puncture entry point for an interventional surgery based on the three-dimensional model of the anatomical structure includes: Resampling the three-dimensional model of the anatomical structure to remove anisotropy, thereby obtaining a three-dimensional model of the tracheal tree; The 3D skeleton of the tracheal tree is extracted based on the connected 3D model of the tracheal tree, and the degree of connectivity DoC between each voxel on the skeleton and the neighboring voxels is calculated; if the DoC of the voxel is equal to 1, the voxel is the end point of the skeleton, that is, it represents the tip of the tracheal tree; if the DoC of the voxel is greater than 2, the voxel is the bifurcation point of the skeleton, that is, it represents the branch node of the tracheal tree; the end point and the bifurcation point are merged into the node of the skeleton; All bifurcation points on the skeleton are removed, and connected domains are obtained for the remaining skeletons, each connected domain is identified as a branch, and the two ends of the branch are adjacent to different nodes, and the midpoint of each branch is extracted; the branches on the skeleton correspond one to one with the tracheal segments on the tracheal tree three-dimensional model; A subset of the set consisting of all nodes and branch midpoints on the skeleton that is distributed within the region of interest specified by the target lesion is used as the initial puncture entry candidate point for the interventional surgery.

4. The interventional surgery planning method based on anatomical structure modeling according to claim 1, characterized in that: The screening of the initial puncture entry candidate points of the interventional surgery to obtain the final puncture entry candidate points of the interventional surgery includes: Providing a reachability index, and using the reachability index to screen out the puncture entry candidate points that do not meet the conditions among the initial puncture entry candidate points of the interventional surgery, to obtain a first group of puncture entry candidate points; providing a low risk indicator, and selecting the best puncture entry candidate point from the first group of puncture entry candidate points using the low risk indicator to obtain a final puncture entry candidate point; in: The reachability index is a hard index for measuring whether a candidate point on the tracheal wall can be used as a bronchoscope puncture point, and is used to evaluate the matching degree between bronchial tree parameters and bronchoscope and its instrument parameters; The low-risk index is a soft index for measuring the risk of bronchoscopic puncture behavior. The index is evaluated when the reachability index is met, and is used to evaluate whether the straight channel from the puncture entry point through the lung parenchyma to the target lesion can avoid risk structures and the channel length is as short as possible.

5. The interventional surgery planning method based on anatomical structure modeling according to claim 4, characterized in that: The method of using the reachability index to screen out the puncture entry candidate points that do not meet the conditions among the initial puncture entry candidate points of the interventional surgery includes: The accessibility indicators include three types of information: anatomy, equipment and process; among which: The anatomical information includes: defining the intratracheal navigation environment through the tracheal tree three-dimensional model, and providing the following local tracheal feature information using the tracheal tree three-dimensional model T={P, B, S}: S={s1,s2,...,s N The inner diameter d of each tracheal segment in aw The inner diameter of the cavity is d aw The calculation is based on the shortest distance from the skeleton point to the trachea surface; B={b1,b2,...,b N The bending angle θ between each branch and the adjacent branch in aw ; The equipment information includes: the physical characteristics and mechanical properties of the bronchoscope and other related instruments and tools determine the passability, operability and final reachability of the bronchoscope in the tracheal tree, and provide the following parameters: The outer diameter of the bronchoscope is d b ; The maximum bending angle of the bronchoscope head hinge θ b , the bending angle θ b Applied to the intraluminal bends through the bronchus; The length of the bronchoscope that can reach the body b , the length l b The extracorporeal loss caused by the operation needs to be subtracted; The length of the puncture needle or transpulmonary tunnel sheath and other instruments t ; The maximum bending angle of bronchoscope combined with other related instruments θ b-t , the bending angle θ b-t Provide diagnostic field of view for intraoperative EBUS guidance and puncture; The process information includes: the conditions and factors considered for performing the expected tasks from the oral cavity to the puncture entry point and from the puncture entry point to the target lesion, and provides the following parameters: B={b1,b2,...,b N The total length of the branches from the oral cavity to the puncture entry point is l mp ; The angle θ between the line from the puncture entry point to the target lesion and the bronchial branch where the puncture entry point is located aw-pn ; The length from the puncture entry point to the target lesion l pn ; The number of lung lobes n from the puncture entry point to the target lesion lob ; Combining the mutual constraints between the three types of information, namely, anatomy, equipment and process, the constraints for selecting the location of the puncture entry point are established as follows: d b <d aw ,i b >θ aw ,l b >l mp The constraints for establishing a puncture path without piercing the lung fissure or pleura are: i b-t >θ aw-pn ,l t >l pn ,n lob =1 According to the constraint conditions, a reachable pathway is obtained, and then a first group of candidate puncture entry points for interventional surgery are screened out.

6. The interventional surgery planning method based on anatomical structure modeling according to claim 4, characterized in that: The selecting the best puncture entry candidate point from the first group of puncture entry candidate points by using the low risk indicator includes: Establish regional risk fields around target lesions; Using the expansion channel to simulate the puncture path, calculating the total risk value of the puncture path based on the regional wind scene, and quantifying the low risk index; Combined with the quantified low-risk index, the first group of candidate puncture entry points is comprehensively evaluated and a plurality of optimal candidate puncture entry points are selected.

7. The interventional surgery planning method based on anatomical structure modeling according to claim 6, characterized in that: The step of establishing a regional risk field around the target lesion includes: Based on the voxel set Φ of the given 3D blood vessel segmentation mask, the Euclidean distance transform output D corresponding to the voxel point K = (x, y, z) in the target lesion area VOI is calculated as: Where V = (u, v, w) is the coordinate of the vascular voxel point; Using the minimum-maximum normalization method, each D(K) in the target lesion area VOI is rescaled so that D(K)∈[0,1], and the risk field value F(K) corresponding to the voxel point K=(x,y,z) is 1-D(K); The method of simulating the puncture path by using the expansion channel, calculating the total risk value of the puncture path based on the regional wind scene, and quantifying the low risk index includes: The 3D Bresenham line algorithm is used to generate the voxel connection line L between the candidate puncture point and the center of the target lesion in the spatial coordinates. pn ; The voxel connection line L is connected by morphological dilation method pn Expanded into a puncture path similar to the tracheal cylindrical passage The puncture path The total risk value Ψ in the risk field of the region is expressed as: The total risk value Ψ is used to measure the total probability that each pixel of the puncture path body may collide with the blood vessel; Set low-risk indicators, including: average risk value Υ avg Indicators, arterial risk value γ artery Indicators and Venous Risk Valuesγ vein Indicators, the low-risk indicators are quantitatively calculated as follows: Calculate the average risk value γ avg index: Where n is the puncture path The number of voxels in avg It is not affected by the puncture length and is used to measure the risk value of unit voxels on the puncture path; Calculate arterial risk value γ artery index: Among them, Φ artery is the voxel set of the artery segmentation mask, γ artery It is used to measure the number of intersections between the puncture path and the arterial pixels; Calculate the venous risk value γ vein index: Among them, Φ vein is the set of voxels of vein segmentation mask, γ vein It is used to measure the number of intersections between the puncture path and the vein pixels; The step of comprehensively evaluating the first group of puncture entry candidate points and selecting a plurality of optimal puncture entry candidate points in combination with the quantified low risk index comprises: Comprehensively evaluate multiple indicators of each candidate puncture entry point in the target lesion regional VOI, and establish the overall evaluation function E of low risk and accessibility: E=10c avg +ag artery +βγ vein +γl pn +δθ aw-pn , Among them, l pn and θ aw- are the reachability indicators for measuring the puncture length and angle respectively; α, β, γ and δ are the hyperparameters for balancing various indicators respectively, pn and θ aw-pn The hyperparameters γ and δ of the term are set according to the risk thresholds of the puncture path length and puncture angle; A plurality of puncture entry candidate points whose total evaluation function value is less than a set threshold are selected as the optimal puncture entry candidate points.

8. An interventional surgery planning method based on anatomical structure modeling, characterized in that: include: The anatomical structure modeling module is used to perform anatomical structure modeling on the target lesion of the interventional surgery to obtain a three-dimensional model of the anatomical structure; A puncture entry candidate point determination module, which obtains an initial puncture entry candidate point for interventional surgery based on the three-dimensional model of the anatomical structure; A puncture entry candidate point screening module, which is used to screen the initial puncture entry candidate points of the interventional surgery to obtain the final puncture entry candidate points of the interventional surgery; A visualization and human-computer interaction module is used to jointly visualize the anatomical structure three-dimensional model and multiple final puncture entry candidate points for the interventional surgery, so as to provide puncture entry point simulation and reference for the interventional surgery and complete the interventional surgery planning.

9. A computer terminal comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, it can be used to perform the method described in any one of claims 1 to 7, or to run the system described in claim 8.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, it can be used to perform the method described in any one of claims 1 to 7, or to run the system described in claim 8.

Images