A method and system for planning multi-branch paths at joint points of a virtual bronchoscope

The center line of the bronchial tree is extracted through a three-dimensional regional growth and refinement algorithm, and the branch path is planned using the splicing point algorithm, which solves the problem of complex multi-branch path planning in the virtual bronchoscope and realizes high-quality automatic roaming path planning.

CN112598639BActive Publication Date: 2024-11-15JINAN GUOKE MEDICAL TECH DEV CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202011525900.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2020-12-23
Publication Date
2024-11-15
Estimated Expiration
2040-12-23

AI Technical Summary

Technical Problem

Existing virtual bronchoscopy technology is difficult to effectively plan complex multi-branch paths, especially in the complex structure of the pulmonary bronchial tree, which cannot accurately screen out a single branch path, resulting in disordered automatic roaming paths.

Method used

The three-dimensional area growth algorithm is used to segment the bronchial tree, combine the three-dimensional refinement algorithm to extract the center line, and use the splicing point algorithm to plan a branch path in the complex branch path, and set the virtual camera and its focus position in a loop to realize automatic roaming.

Benefits of technology

It realizes high-quality planning of complex multi-branch paths and large curvature turning branch paths, meets the requirements of centrality, connectivity and singularity, solves the multi-branch path planning problem of virtual bronchoscopic automatic roaming, and can quickly and conveniently reach any branch end point. .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN112598639B_ABST
    Figure CN112598639B_ABST
Patent Text Reader

Abstract

The present invention relates to the technical field of splicing point multi-branch path planning, and specifically, to a splicing point multi-branch path planning method for a virtual bronchoscope, comprising the following steps: using a three-dimensional region growing algorithm to segment the acquired lung CT sequence image to obtain a bronchial tree. In the splicing point multi-branch path planning method and system for the virtual bronchoscope, the splicing point path planning algorithm can be used to achieve the planning effect of complex multi-branch paths and large curvature turning branch paths, and the high-quality centerline image that can meet the centrality, connectivity, and singularity requirements of the center path is an important prerequisite for using the splicing point algorithm for bronchial tree path planning, and can solve the multi-branch path planning problem in obtaining the automatic roaming of the virtual bronchoscope. The splicing point branch path planning algorithm is adopted, and after obtaining the complete centerline of the bronchial tree, the automatic roaming path to reach any branch end point can be quickly and conveniently planned.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of splicing point multi-branch path planning, and in particular to a splicing point multi-branch path planning method and system for a virtual bronchoscope. Background Art

[0002] Virtual endoscopy is a new non-invasive diagnostic method that uses computers to process three-dimensional medical images such as CT and MRI to generate three-dimensional dynamic views of specific anatomical structures inside the human body to simulate the inspection process and visual effects of a standard endoscope. It has unique advantages such as non-invasiveness, reusability, dynamic pathological analysis, and no inspection dead zones. It has broad application prospects, and many countries and research institutions have invested a lot of manpower and material resources in research.

[0003] Virtual bronchoscopy utilizes the characteristics of high-resolution lung CT to perform three-dimensional reconstruction of the trachea and plan the optimal surgical path, which can provide important reference for the early and accurate diagnosis of lung cancer and intraoperative guidance of bronchoscopy. Virtual bronchoscopy can overcome the complex environment of frequent airway branches, complete the detection of small bronchial branches without causing damage or missing detection, and overcome the difficulty of biopsy of peripheral lung lesions, playing a key role in the diagnosis and staging of lung cancer.

[0004] Developing a virtual bronchoscopy system requires the implementation of four key technologies: lung and airway tree image segmentation, three-dimensional reconstruction of segmented images, airway tree centerline extraction, and virtual roaming. For the patient's chest image data obtained from the clinic, the lungs and airway trees need to be segmented first, and the segmented lung and airway images are reconstructed using three-dimensional reconstruction technology to reconstruct the three-dimensional model of the lungs and airway tree. Then, a suitable path extraction algorithm is used for the extracted three-dimensional reconstructed model of the airway tree to obtain the three-dimensional skeleton centerline of the airway tree as the central path for automatic roaming. Finally, by setting relevant parameters such as the virtual camera position and focus, the virtual camera can move along the extracted central path, thereby realizing automatic roaming of the internal structure of the airway tree.

[0005] Automatic roaming is the core function of virtual endoscopes. The planning of automatic roaming paths is the key to realizing the automatic roaming function. Currently, the roaming paths of virtual endoscopes mostly use the center line of the three-dimensional image of the organ. The methods for extracting the center line include distance transformation method, topological refinement method, etc.

[0006] The topological refinement method is to peel off the surface voxels and internal voxels of the tissue layer by layer in an iterative process until the last layer, which is the center path. The three-dimensional refinement algorithm proposed by Lee et al. can quickly extract the complete centerline of the bronchus.

[0007] As the application of virtual endoscopy becomes more and more widespread, organs with multi-branch lumen structures such as lung bronchi have the same demand for virtual endoscopy technology. After the centerline of the entire bronchial tree is refined and extracted, the problem of complex multi-branch path planning cannot be solved.

[0008] The bronchial tree has a complex multi-branched structure. After the topological refinement method is used to obtain a complete bronchial tree centerline image, the centerline coordinates obtained along the height direction, i.e., the Z-axis direction, are relatively disordered. The automatic roaming path should be specific to a certain branch according to actual needs. Currently, there is a lack of path planning methods that can screen out a single branch. Summary of the invention

[0009] The purpose of the present invention is to provide a method and system for planning multi-branch paths of joint points of a virtual bronchoscope to solve the problems raised in the above-mentioned background technology.

[0010] In order to solve the above technical problems, one of the purposes of the present invention is to provide a method for planning multi-branch paths of splicing points of a virtual bronchoscope, comprising the following steps:

[0011] S1. Use the three-dimensional region growing algorithm to segment the bronchial tree from the acquired lung CT sequence images;

[0012] S2, extracting the centerline image from the segmented bronchial tree image using a three-dimensional thinning algorithm and obtaining its coordinate point set;

[0013] S3, using the splicing point algorithm to plan a branch path among the complex multiple branch paths of the bronchial tree centerline;

[0014] S4. Loop through the virtual camera and its focus position to achieve automatic roaming.

[0015] Selected, the principle steps of the splicing point algorithm in S3 are as follows:

[0016] S3.1, obtain the center line coordinate point set;

[0017] S3.2, select the starting point and end point of the path;

[0018] S3.3, traverse to find the adjacent point set of the decision point;

[0019] S3.4, traverse to find the point in the adjacent point set with the smallest distance to the end point;

[0020] S3.5, joining the smallest points into the path point set;

[0021] S3.5, determine whether the smallest point is equal to the end point;

[0022] S3.5, delete the decision point from the center line coordinate point set, and use the smallest point as the new decision point until the smallest point is equal to the end point;

[0023] S3.5. The path planning is completed after the stitching is completed.

[0024] Selected, in said S3.3, the Euclidean distance calculation formula between the determination point and the two points in the point set in the splicing point algorithm principle is as follows:

[0025]

[0026] Where: L1 is the Euclidean distance, P1 is the decision point, and S is the point set.

[0027] Selected, in said S3.3, the Euclidean distance calculation formula between the path endpoint and the adjacent point set in the splicing point algorithm principle is as follows:

[0028]

[0029] Where: M is the set of adjacent points, P r The end point of the path.

[0030] The smallest point in S3.4 is selected using a circular comparison method.

[0031] Preferably, the circular comparison method comprises the following steps:

[0032] S3.4.1. First, compare two values, and compare the smallest value with the other value until the smallest value among all distance values ​​is found;

[0033] S3.4.2. Determine the corresponding minimum adjacent point, take the minimum adjacent point as the second path point, and add the second path point to the planning coordinate point set to complete a splicing;

[0034] S3.4.3. Delete the first path point from the center line voxel coordinate point set, set the new center line voxel coordinate point set as the point set, take the second path point as the current judgment point, find the adjacent point of the second path point in the point set, and perform the subsequent splicing operation on the second path point until the coordinates of the splicing point are consistent with the coordinates of the end point, and the splicing is completed at the same time.

[0035] Compared with the prior art, the present invention has the following beneficial effects:

[0036] In the splicing point multi-branch path planning method and system of the virtual bronchoscope, after using the refinement algorithm to obtain the complete centerline of the bronchial tree, the splicing point path planning algorithm can be used to achieve the planning effect of complex multi-branch paths and large curvature turning branch paths. The high-quality centerline image that can meet the centrality, connectivity, and singularity requirements of the center path is an important prerequisite for using the splicing point algorithm for bronchial tree path planning, and can solve the multi-branch path planning problem in obtaining automatic navigation of the virtual bronchoscope. The splicing point branch path planning algorithm is adopted, which can quickly and conveniently plan an automatic navigation path to reach the end point of any branch after obtaining the complete centerline of the bronchial tree. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] Figure 1 A flowchart for realizing automatic navigation of virtual bronchoscopy;

[0038] Figure 2 This is a flowchart of the principle of the stitching point path planning algorithm for virtual bronchoscope;

[0039] Figure 3 This is the flow chart of the circular comparison method. DETAILED DESCRIPTION

[0040] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0041] Example 1

[0042] like Figure 1-3 As shown, the purpose of the present invention is to provide a method for planning multi-branch paths of joint points of a virtual bronchoscope, comprising the following steps:

[0043] S1. Use the three-dimensional region growing algorithm to segment the bronchial tree from the acquired lung CT sequence images;

[0044] S2, extracting the centerline image from the segmented bronchial tree image using a three-dimensional thinning algorithm and obtaining its coordinate point set;

[0045] S3, using the splicing point algorithm to plan a branch path among the complex multiple branch paths of the bronchial tree centerline;

[0046] S4. Loop through the virtual camera and its focus position to achieve automatic roaming.

[0047] The principle steps of the splicing point algorithm in S3 are as follows:

[0048] S3.1, obtain the center line coordinate point set;

[0049] S3.2, select the starting point and end point of the path;

[0050] S3.3, traverse to find the adjacent point set of the decision point;

[0051] S3.4, traverse to find the point in the adjacent point set with the smallest distance to the end point;

[0052] S3.5, joining the smallest points into the path point set;

[0053] S3.5, determine whether the smallest point is equal to the end point;

[0054] S3.5, delete the decision point from the center line coordinate point set, and use the smallest point as the new decision point until the smallest point is equal to the end point;

[0055] S3.5. The path planning is completed after the stitching is completed.

[0056] Among them, in S3.3, the Euclidean distance calculation formula between the determination point and the two points in the point set in the splicing point algorithm principle is as follows:

[0057]

[0058] Where: L1 is the Euclidean distance, P1 is the decision point, and S is the point set;

[0059] First, let S = {S1, ..., St} be the point set containing all the voxel coordinates of the center line, let the final planned path point set be P = {P1, ..., Pr}, select the automatic roaming starting point and end point on the center line of the bronchial tree as the first path point P1 and the last path point Pr; take P1 as the judgment point, traverse the center line point set S, and calculate the distance to the first path point less than or equal to For all points, let the coordinates of the decision point P1 be (P1a, P1b, P1c), and the coordinates of a point Si in the point set S be (Sia, Sib, Sic), because the adjacency of voxel points is divided into face-to-face, edge-to-edge, and point-to-point, with spacings of 1,

[0060] Among them, in S3.3, the Euclidean distance calculation formula between the path endpoint and the adjacent point set in the splicing point algorithm principle is as follows:

[0061]

[0062] Where: M is the set of adjacent points, P r is the end point of the path;

[0063] The point set S satisfies The points are the adjacent points of the decision point, and the set M = {M1,...,Mn} is the adjacent point set of the path point; traverse the adjacent point set M, and let the coordinates of the path end point Pr be (Pra,Prb,Prc), and the coordinates of a point Mi in the adjacent point set M be (Mia,Mib,Mic),

[0064] The smallest point in S3.4 is determined by a circular comparison method.

[0065] Wherein, the circular comparison method comprises the following steps:

[0066] S3.4.1. First, compare two values, and compare the smallest value with the other value until the smallest value among all distance values ​​is found;

[0067] S3.4.2. Determine the corresponding minimum adjacent point, take the minimum adjacent point as the second path point, and add the second path point to the planning coordinate point set to complete a splicing;

[0068] S3.4.3. Delete the first path point from the centerline voxel coordinate point set, set the new centerline voxel coordinate point set as the point set, take the second path point as the current decision point, find the adjacent point of the second path point in the point set, and perform the subsequent splicing operation on the second path point until the splicing point coordinates are the same as the end point coordinates, and the splicing is completed at the same time;

[0069] Let D = {D1, ... Dn} be the distance set between each adjacent point in the adjacent point set M and the end point Pr. Use the circular comparison method to find the minimum value Dmin in the distance set D. First, compare D1 and D2, and compare the smaller one with D3; then compare the smaller one with D4, until it is compared with Dn. Find the smallest value among all n distance values, and determine its corresponding adjacent point Mmin. Take Mmin as the second path point P2, and add P2 to the planning coordinate point set P. So far Complete one stitching; delete the first path point P1 from the centerline voxel coordinate point set S, set the new centerline voxel coordinate point set S', take the second path point P2 as the current judgment point, find the adjacent point of P2 in the point set S', and perform the subsequent stitching operation on P2, and so on. When the coordinates of the stitching point are the same as the coordinates of the end point Pr, the stitching is completed. The point set P = {P1,...,Pr} is the planned path point set. This method can complete the accurate path planning from the main trachea to the fifth-order bronchial branch.

[0070] It can solve the multi-branch path planning problem in obtaining automatic navigation of a virtual bronchoscope. This advantage is because the present invention adopts a splicing point branch path planning algorithm, which can quickly and conveniently plan an automatic navigation path to reach the end point of any branch after obtaining the complete center line of the bronchial tree.

[0071] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited by the above embodiments. The above embodiments and descriptions are only preferred examples of the present invention and are not intended to limit the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which fall within the scope of the present invention. The scope of protection of the present invention is defined by the attached claims and their equivalents.

Claims

1. A method for planning multi-branch paths at joint points of a virtual bronchoscope, characterized in that: The steps include: S1. Use the three-dimensional region growing algorithm to segment the bronchial tree from the acquired lung CT sequence images; S2, extracting the centerline image from the segmented bronchial tree image using a three-dimensional thinning algorithm and obtaining its coordinate point set; S3, using the splicing point algorithm to plan a branch path among the complex multiple branch paths of the bronchial tree centerline; S4, cyclically setting the position of the virtual camera and its focus to realize automatic roaming; The principle steps of the splicing point algorithm in S3 are as follows: S3.1, obtain the center line coordinate point set; S3.2, select the starting point and end point of the path; S3.3, traverse to find the adjacent point set of the decision point; S3.4, traverse to find the point in the adjacent point set with the smallest distance to the end point; S3.5, joining the smallest points into the path point set; S3.6, determine whether the smallest point is equal to the end point; S3.7, delete the decision point from the center line coordinate point set, and use the smallest point as the new decision point until the smallest point is equal to the end point; S3.8, the path planning is completed after the splicing is completed; In S3.3, the Euclidean distance calculation formula between the decision point and the center line coordinate point set is as follows: Where: L1 is the Euclidean distance, P1 is the decision point, and S is the point set; First, let S = {S1, ..., St} be the point set containing all the voxel coordinates of the center line, let the final planned path point set be P = {P1, ..., Pr}, select the automatic roaming starting point and end point on the center line of the bronchial tree as the first path point P1 and the last path point Pr; take P1 as the judgment point, traverse the center line point set S, and calculate the distance to the first path point less than or equal to For all points, let the coordinates of the decision point P1 be (P1a, P1b, P1c), the coordinates of a point Si in the point set S be (Sia, Sib, Sic), and the adjacency of voxel points is divided into face-to-face, edge-to-edge, and point-to-point, with spacings of 1, In S3.3, the Euclidean distance between the path endpoint and the adjacent point set is calculated as follows: Where: M is the set of adjacent points, P r is the end point of the path; The point set S satisfies The points are the adjacent points of the decision point, and the set M = {M1, ..., Mn} is the adjacent point set of the path point; traverse the adjacent point set M, and let the coordinates of the path end point Pr be (Pra, Prb, Prc), and the coordinates of a point Mi in the adjacent point set M be (Mia, Mib, Mic); The traversal in S3.4 to find the point with the smallest distance from the end point in the adjacent point set adopts a circular comparison method; The circular comparison method comprises the following steps: S3.4.

1. First, compare two values, and compare the smallest value with the other value until the smallest value among all distance values ​​is found; S3.4.

2. Determine the corresponding minimum adjacent point, take the minimum adjacent point as the second path point, and add the second path point to the planning coordinate point set to complete a splicing; S3.4.

3. Delete the first path point from the set of all voxel coordinate points of the center line, set the new set of voxel coordinate points of the center line as the point set, take the second path point as the current decision point, find the adjacent point of the second path point in the point set, and perform the subsequent splicing operation on the second path point until the splicing point coordinates are the same as the end point coordinates, and the splicing is completed; Assume that D = {D1, ... Dn} is the distance set between each adjacent point in the adjacent point set M and the end point Pr. Use the circular comparison method to find the minimum value Dmin in the distance set D. First, compare D1 and D2, and compare the smaller one with D3; then compare the smaller one with D4, until it is compared with Dn, find the smallest value among all n distance values, and determine its corresponding adjacent point Mmin. Mmin is used as the second path point P2, and P2 is added to the planned coordinate point set P, so that a splicing is completed; delete the first path point P1 from the center line voxel coordinate point set S, set the new center line voxel coordinate point set S', and use the second path point P2 as the current judgment point. Find the adjacent point of P2 in the point set S', and perform the subsequent splicing operation on P2, and so on. When the coordinates of the splicing point are the same as the coordinates of the end point Pr, the splicing is completed, and the point set P = {P1, ..., Pr} is the planned path point set.