Three-dimensional reconstruction method and system for vascular network structure, and storage medium

Abstract

Disclosed are a three-dimensional reconstruction method and system for a vascular network structure, and a storage medium. The method comprises: acquiring an image sequence of a target imaging area; recognizing a region position sequence and a diffusion feature sequence both corresponding to a magnetically driven micro-robot cluster in the image sequence; and reconstructing a three-dimensional vascular network structure image on the basis of the recognized region position sequence and a diffusion pattern sequence. In the method and the system disclosed in the embodiments, the position of each image in the image sequence when a magnetic field drives the magnetically driven micro-robot cluster to flow in blood vessels in a reverse direction or to actively diffuse to a blocking or narrow branch at a low flow velocity is acquired, so that comprehensive exploration of the vascular network is realized, thereby acquiring a complete vascular network structure image, and improving the accuracy of vascular network structure recognition.

Description

A method, system, and storage medium for three-dimensional reconstruction of vascular network structures Technical Field

[0001] This invention relates to the field of microscopic imaging technology, and in particular to a method, system, and storage medium for three-dimensional reconstruction of vascular network structures. Background Technology

[0002] Angiography is a method for imaging vascular structures. However, current angiography relies on the passive diffusion of contrast agents with blood and lymph. The range that can be achieved is highly dependent on the flow field within the blood vessels and lymphatic vessels. It cannot reverse the flow direction to enter upstream branches or diffuse into low-velocity embolism and stenosis branches. Therefore, it cannot achieve complete imaging of the vascular network structure.

[0003] Therefore, existing technologies need to be improved.

[0004] Summary of the Invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide users with a three-dimensional reconstruction method, system and storage medium for vascular network structure, overcoming the defect that the prior art cannot obtain complete imaging results when using angiography to image vascular network structure.

[0006] The technical solution adopted by this invention to solve the technical problem is as follows:

[0007] In a first aspect, this embodiment provides a method for three-dimensional reconstruction of a vascular network structure, comprising:

[0008] Acquire an image sequence of the target imaging region; wherein the image sequence contains multiple images of a cluster of magnetically driven microrobots;

[0009] Identify the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence;

[0010] A three-dimensional vascular network structure image is reconstructed based on the identified regional location sequence and diffusion morphology sequence.

[0011] Optionally, the region location sequence includes: multiple branch points and multiple branches, and the diffusion feature sequence includes: the branch direction and the optimal exploration direction corresponding to each branch;

[0012] The step of identifying the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence includes:

[0013] The location and diffusion characteristics of the magnetically driven microrobot cluster were isolated from each image in the image sequence.

[0014] Based on the location and diffusion characteristics of the magnetically driven microrobot cluster, determine multiple branch points, multiple branches, branch directions corresponding to the multiple branches, and the optimal exploration direction within the target imaging area.

[0015] Optionally, the step of separating the location and diffusion characteristics of the magnetically driven microrobot swarm from each image in the image sequence includes:

[0016] The location and diffusion characteristics of the cluster are separated by image binarization with dynamic thresholding and image difference method.

[0017] Optionally, the step of determining multiple branch points, multiple branches, branch directions, and optimal exploration directions of the magnetically driven microrobot cluster within the target imaging area based on the location and diffusion characteristics of the magnetically driven microrobot cluster includes:

[0018] Based on the intersection information contained in the area where the magnetically driven microrobot cluster is located, determine whether there are branch points in the imaging area;

[0019] If a branch point exists, the branch direction corresponding to the branch point is determined according to the vector connected to the branch point, and the optimal exploration direction is determined according to the branch direction;

[0020] If no branch point exists, the direction of movement of the magnetically driven microrobot cluster is determined based on its position. Sampling points are selected at fixed preset intervals to obtain the position coordinates of each sampling point and record the branch to which each sampling point belongs.

[0021] Optionally, after determining multiple branch points, multiple branches, branch directions corresponding to the multiple branches, and the optimal exploration direction within the target imaging area based on the location and diffusion characteristics of the magnetically driven microrobot cluster, the method further includes:

[0022] Determine whether the magnetically driven microrobot cluster has stopped spreading for an extended period, or whether it has spread to overlap with other branches, or whether it has spread beyond the area to be explored.

[0023] If this occurs, it indicates that the magnetically driven microrobot cluster has stopped spreading for an extended period of time, or has spread to overlap with other branches, or has spread to a branch outside the required exploration area, indicating that the exploration has been completed.

[0024] Continue to record the diffusion status information of all branches until all branches have been marked as explored.

[0025] Optionally, the step of reconstructing a three-dimensional vascular network structure image based on the identified region location sequence and diffusion morphology sequence includes:

[0026] Using data vectors to describe each branch point and branch, a tree map of the vascular network structure is constructed;

[0027] The tree map is updated based on the branch directions and optimal exploration direction corresponding to multiple branches, as well as the branch points and branches obtained through exploration.

[0028] A three-dimensional vascular network structure image was reconstructed based on the updated tree map.

[0029] Optionally, the step of using data vectors to describe each branch point and branch to construct a tree map of the vascular network structure includes:

[0030] Based on the collection time sequence of each branch point and branch, the branch points and branches are connected to obtain a tree map of the three-dimensional vascular network.

[0031] Optionally, the step of reconstructing the three-dimensional vascular network structure image based on the updated tree map includes:

[0032] For branches where both the starting point and the ending point are branch points, the direction from the starting point to the ending point is taken as the branch direction. The angle between the direction of the line connecting the last identified branch point to each scattered point and the corresponding branch direction of each branch is taken as the deviation angle. The connection order of the scattered points is determined by the deviation angle and the distance between each scattered point and the last identified branch point. The branch direction and structure are then restored based on the determined connection order of the scattered points.

[0033] For branches whose starting point is a branch point but whose ending point is unknown, the direction of the line connecting the last identified branch point within the branch is taken as the branch direction. The angle between the direction of the line connecting the last identified branch point within the branch to each scattered point and the branch direction corresponding to each branch is taken as the deviation angle. The connection order of the scattered points is determined by using the deviation angle and the distance between each scattered point and the last identified branch point. The branch direction and structure are then restored based on the determined connection order of the scattered points.

[0034] When all the scattered points in the tree map have been connected, or the distance between the remaining scattered points and the identified branch points exceeds the preset distance value, or the deviation angle between the remaining scattered points and the last identified branch point exceeds the preset angle value, and the last identified branch point coincides with the branch endpoint, then the reconstructed three-dimensional vascular network structure image is obtained.

[0035] Secondly, this embodiment provides a three-dimensional reconstruction system for vascular network structures, comprising:

[0036] An information acquisition module is used to acquire an image sequence of the target imaging area; wherein, the image sequence contains multiple images of a cluster of magnetically driven microrobots;

[0037] The information recognition module is used to identify the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence;

[0038] The network reconstruction module is used to reconstruct a three-dimensional vascular network structure image based on the identified regional location sequence and diffusion morphology sequence.

[0039] Thirdly, this embodiment also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the three-dimensional reconstruction method of the vascular network structure. Beneficial effects:

[0040] This embodiment discloses a method, system, and storage medium for three-dimensional reconstruction of vascular network structures. The method involves acquiring an image sequence of a target imaging region, wherein the image sequence contains images of a magnetically driven microrobot cluster. The method identifies the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence. Based on the identified region location sequence and diffusion pattern sequence, a three-dimensional vascular network structure image is reconstructed. The method and system disclosed in this embodiment acquire image sequences of the target imaging region, where each image represents a magnetically driven microrobot cluster actively diffusing within a blood vessel during countercurrent flow or at low flow rates in blocked or narrowed branches. By identifying the cluster's location information and diffusion pattern in the image sequence, a comprehensive exploration of the vascular network is achieved, resulting in a complete vascular network structure image and improving the accuracy of vascular network structure identification. Attached Figure Description

[0041] Figure 1 is a schematic diagram of clustered active angiography in an embodiment of the present invention;

[0042] Figure 2 is a flowchart of the three-dimensional reconstruction method of vascular network structure in an embodiment of the present invention;

[0043] Figure 3 is a schematic block diagram illustrating the principle of the three-dimensional reconstruction method in an embodiment of the present invention;

[0044] Figure 4 is a schematic diagram of the image processing process of the top view of the workspace in an embodiment of the present invention;

[0045] Figure 5 is a schematic diagram of the key steps in the active exploration process in this embodiment;

[0046] Figure 6 is a schematic diagram of the three-dimensional reconstruction algorithm of blood vessel structure in this embodiment;

[0047] Figure 7 is a schematic diagram of the principle of branch connection reconstruction in this embodiment;

[0048] Figure 8 is a block diagram of the three-dimensional reconstruction system in this embodiment. Detailed Implementation

[0049] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0050] Angiography is a method used to image vascular structures, often to analyze whether there are abnormalities in blood vessels. Therefore, the accuracy of the imaging results is crucial. In current angiography, as shown in Figure 1, contrast agents are typically passively diffused with blood and lymph. Therefore, the range that the contrast agent can reach is highly dependent on the flow field within the vascular network structure. When facing upstream branches that require flow against the current or low-velocity embolisms and stenotic branches, diffusion is impossible, thus preventing the achievement of complete vascular network imaging.

[0051] Currently, magnetically driven microrobot swarms hold promise as active tools for exploring vascular networks due to their ability to be remotely and precisely driven by magnetic fields. However, the exploration of how to use magnetic fields to drive these swarms to actively explore vascular network structures remains in the exploratory stage. Achieving efficient and precise guidance for these swarms to fully cover complex vascular network structures presents significant challenges.

[0052] To address the aforementioned challenge of efficiently and accurately guiding a cluster of magnetically driven microrobots to achieve full coverage of complex vascular network structures, this embodiment provides a method, system, and storage medium for three-dimensional reconstruction of vascular network structures. Utilizing a time-varying magnetic field, the microrobot cluster is driven to actively move within the vascular network structure, acquiring image sequences of the target imaging area during active exploration. Based on these image sequences, the regional location information and diffusion morphology characteristics of the magnetically driven microrobot cluster are identified, allowing for the reconstruction of a three-dimensional vascular network structure image. Since the magnetically driven microrobot cluster can enter upstream branches against the flow and diffuse into low-velocity, blocked branches, achieving full coverage of the vascular network structure, the method disclosed in this embodiment can significantly expand the imageable area in angiography, enabling accurate reconstruction of the three-dimensional vascular network structure.

[0053] The following description, in conjunction with the accompanying drawings provided in this embodiment, provides a more detailed explanation of a three-dimensional reconstruction method, system, and storage medium for a vascular network structure disclosed in this embodiment.

[0054] Firstly, this embodiment provides a three-dimensional reconstruction method for vascular network structures, as shown in Figure 2. The three-dimensional reconstruction method includes:

[0055] Step S1: Obtain an image sequence of the target imaging area; wherein the image sequence contains multiple images of a cluster of magnetically driven microrobots.

[0056] In this step, a cluster of magnetically driven microrobots is first injected into the vascular network structure as a contrast agent. Specifically, the magnetically driven microrobot cluster is a magnetically responsive, radiopaque material, thus enabling it to move under the drive of a time-varying magnetic field and allowing its location to be determined through image processing. In one embodiment, the magnetically driven microrobot cluster can be a cluster of paramagnetic nanoparticles. Because the microrobots in the magnetically driven microrobot cluster are only the size of nanoparticles, they can enter the narrow spaces within the vascular network structure to achieve full coverage exploration of the vascular network structure.

[0057] The time-varying magnetic field is generated by a magnetic field drive system, which can use electromagnetic coils or permanent magnets to produce the magnetic field, and its operating range can cover the required imaging area. Furthermore, by generating a magnetic field with magnetic flux density that varies with time, such as a rotating field or an oscillating field, the magnetically driven microrobot swarm can move in the direction driven by the magnetic field under the time-varying magnetic field.

[0058] Furthermore, the real-time position and morphology of the magnetically driven microrobot swarm can be acquired using a medical imaging system. This is achieved by capturing top and side views of the imaging area where the swarm is located in real time, thus obtaining image information of the magnetically driven microrobot swarm. Since the magnetically driven microrobot swarm dynamically diffuses within the vascular network structure under the drive of a time-varying magnetic field, this step allows for continuous imaging of the target imaging area, resulting in multiple image sequences containing the magnetically driven microrobot swarm within the vascular network structure. Specifically, to obtain the swarm's position information in three-dimensional coordinates, cameras positioned at the top and sides of the target imaging area can be used to capture images of the target imaging area, thereby obtaining position information of the swarm in three-dimensional coordinates.

[0059] Step S2: Identify the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence.

[0060] After obtaining the image sequence of the magnetically driven microrobot cluster in the vascular network structure during the duration period in step S1 above, the obtained image sequence is processed to identify the regional location information and diffusion morphology characteristics of the magnetically driven microrobot cluster in the vascular network structure. It can be inferred that the regional location information is also the regional location information of the vascular network structure, and its diffusion morphology characteristics are also the morphological characteristics of the vascular network structure.

[0061] Specifically, as shown in Figure 3, after capturing images of the target imaging area to obtain top and side views, the captured images are preprocessed to identify the regional location sequence and diffusion feature sequence, and to make exploration direction decisions and construct tree maps based on the regional location sequence and diffusion feature sequence.

[0062] The image preprocessing steps described above include: cropping the target imaging region from the captured top and side views to accurately identify information within the target imaging region. The top view is used to extract x-axis and y-axis information, and the side view is used to extract z-axis information. A tree map is constructed by combining the extracted x, y, and z-axis information, and the constructed tree map is updated in real-time based on images in the image sequence. Simultaneously, directional decisions are explored based on branch information in the images to obtain the optimal state decision. A driving magnetic field is generated based on the optimal state decision to control the movement of the magnetically driven microrobot cluster in the blood vessel. In one embodiment, the target imaging region is a portion containing blood vessels and lymphatic vessels. A magnetically driven microrobot cluster is injected into this region. As the magnetically driven microrobot cluster diffuses, this region can be adjusted according to the diffusion direction to accurately obtain the location and diffusion morphology characteristics of the magnetically driven microrobot cluster.

[0063] The region location sequence includes: multiple branch points and multiple branches, and the diffusion feature sequence includes: the branch direction and the optimal exploration direction corresponding to each branch.

[0064] Furthermore, the step of identifying the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence includes:

[0065] Step S21: Separate the location and diffusion characteristics of the magnetically driven microrobot cluster from each image in the image sequence.

[0066] The region and diffusion pattern of the magnetically driven microrobot cluster are first separated from the images in the image sequence to obtain the region and diffusion characteristics of the magnetically driven microrobot cluster. In one embodiment, the method for separating the magnetically driven microrobot cluster includes: separating the cluster region and pattern using a dynamic threshold image binarization method and an image difference method. Referring to Figure 5, the image is first processed using the image difference method to obtain the difference between the real-time image and the background image. Then, based on the obtained difference image, the difference image is binarized to obtain the separated region and diffusion characteristics of the magnetically driven microrobot cluster.

[0067] This step sequentially applies dynamic thresholding and image differencing to each image in the image sequence to separate the region containing the magnetically driven microrobot cluster from the extracted image corresponding to the cluster.

[0068] Furthermore, the location and diffusion characteristics of the magnetically driven microrobot cluster constitute the regional skeleton information of the separated magnetically driven microrobot cluster, which is the main structural information of the location of the magnetically driven microrobot. It contains not only the outline information of the location of the magnetically driven microrobot, but also the corresponding position information.

[0069] Step S22: Determine multiple branch points, multiple branches, branch directions corresponding to multiple branches, and optimal exploration directions within the target imaging area based on the location and diffusion characteristics of the magnetically driven microrobot cluster.

[0070] Based on whether there are intersections in the region corresponding to the structural information of the region, it is determined whether there are branch points in the current target imaging region, and the vascular network structure is described and the exploration progress is recorded according to the determined branch points.

[0071] Specifically, the steps of determining multiple branch points, multiple branches, branch directions, and optimal exploration directions of the magnetically driven microrobot cluster within the target imaging area based on the location and diffusion characteristics of the magnetically driven microrobot cluster include:

[0072] Step S211: Based on the intersection information contained in the area where the magnetically driven microrobot cluster is located, determine whether there are branch points in the imaging area.

[0073] Step S212: If there is a branch point, determine the branch direction corresponding to the branch point based on the vector connected to the branch point, and determine the optimal exploration direction based on the branch direction.

[0074] Step S213: If there is no branch point, determine the movement direction of the magnetically driven microrobot cluster based on its position, select sampling points at fixed preset intervals, obtain the position coordinates of each sampling point, and record the branch to which each sampling point belongs.

[0075] Referring to Figure 4, if the location and diffusion characteristics of the magnetically driven microrobot cluster constitute the cluster's regional skeleton information, and if this skeleton information contains intersection information (as shown in Figure 4b), then branch points are identified within the target imaging region. The branch directions are then extracted based on the vectors formed by the endpoints of the regional skeleton and the branch points, yielding potential explorable directions. Multiple unexplored branches are identified based on the determined branch points, and the unexplored branch with the highest cluster diffusion rate is selected for exploration. If no branch points are identified (as shown in Figure 4a), the movement direction of the cluster within the branch can be determined based on its historical diffusion location. This movement direction is then used to adjust the target imaging region for image processing, ensuring that the front end of the magnetically driven microrobot cluster diffusion remains within the target imaging region.

[0076] Based on the intersection information contained in the area where the magnetically driven microrobot cluster is located, after determining whether there are branch points in the imaging area, it is not only necessary to determine the branch direction and branch corresponding to each branch point and explore the optimal unexplored branch, but also to determine whether the exploration is completed based on the explored branch information. If the exploration is completed, the three-dimensional vascular network structure can be reconstructed based on the explored information.

[0077] Specifically, referring to Figure 6, after determining the multiple branch points, multiple branches, branch directions, and optimal exploration directions of the magnetically driven microrobot cluster within the target imaging area based on its location and diffusion characteristics, the method further includes:

[0078] Step S23: Determine whether the magnetically driven microrobot cluster has stopped spreading for a long time, or whether the magnetically driven microrobot cluster has spread to overlap with other branches, or whether the magnetically driven microrobot cluster has spread beyond the required exploration area.

[0079] Step S24: If the magnetically driven microrobot cluster stops spreading for a long time, or spreads to overlap with other branches or spreads outside the required exploration area, then mark the branch exploration as complete.

[0080] Step S25: Continuously record the diffusion status information of all branches until all branches have been marked as explored.

[0081] Once the optimal diffusion direction is determined, while controlling the magnetically driven microrobot swarm to diffuse along the optimal diffusion direction, it is necessary to simultaneously record the diffusion status information of the swarm, and determine whether the corresponding branch has been explored completely based on the recorded diffusion status information, and mark the explored branches. Once all branches of all nodes have been marked, the exploration process is complete.

[0082] During the aforementioned exploration of branches, data extraction and collection can be carried out simultaneously, or the nodes and branch information can be explored and collected first, and then three-dimensional image reconstruction can be performed based on the collected information. In addition, if there are no branch points in the target imaging area of ​​the image processing, sampling points are selected at fixed intervals on the skeleton of the cluster area, and the top and side views of the area are obtained through medical imaging. The coordinate information of the sampling points and the cavity diameter are calculated through image registration, and the branch to which the sampling point belongs is recorded. If there are branch points in the target imaging area of ​​the image processing, the coordinate information of the branch points and the two ends of each branch is recorded.

[0083] Step S3: Reconstruct a three-dimensional vascular network structure image based on the identified regional location sequence and diffusion morphology sequence.

[0084] Since the location information and diffusion morphology sequence of the cluster during the magnetic drive process are the location and structural morphology of the blood vessels, a three-dimensional image of the vascular network structure can be reconstructed based on the identified location information and diffusion morphology sequence.

[0085] Specifically, the step of reconstructing a three-dimensional vascular network structure image based on the identified region location sequence and diffusion morphology sequence includes:

[0086] Step S31: Use data vectors to describe each branch point and branch, and construct a tree map of the vascular network structure.

[0087] Based on the collection time sequence of each branch point and branch, the branch points and branches are connected to obtain a tree map of the three-dimensional vascular network.

[0088] Step S32: Update the tree map based on the branch directions and optimal exploration directions corresponding to multiple branches, as well as the branch points and branches obtained from the exploration.

[0089] Since the images in the image sequence obtained from the target imaging area are taken in a certain time sequence, by identifying each image in the image sequence, the regional location and diffusion characteristics corresponding to a series of magnetically driven microrobot clusters can be obtained. Therefore, in the continuous identification process, new branch points and branches will be identified. Therefore, the tree map is updated step by step according to the newly identified branch points, branch directions and optimal exploration directions, so as to obtain the final tree map.

[0090] Step S33: Reconstruct a three-dimensional vascular network structure image based on the updated tree map.

[0091] The vascular network structure image is reconstructed using information such as branch points, branches, and branch directions contained in the updated tree map.

[0092] Furthermore, the step of reconstructing the three-dimensional vascular network structure image based on the updated tree map includes:

[0093] For branches where both the starting and ending points are branch points, the direction from the starting point to the ending point is taken as the branch direction. The angle between the direction of the line connecting the last identified branch point to each scattered point and the corresponding branch direction of each branch is taken as the deviation angle. The connection order of the scattered points is determined by using the deviation angle and the distance between each scattered point and the last identified branch point. The branch direction and structure are then reconstructed based on the determined connection order of the scattered points.

[0094] For branches whose starting point is a branch point but whose ending point is unknown, the direction of the line connecting the last identified branch point within the branch is taken as the branch direction. The angle between the direction of the line connecting the last identified branch point within the branch (that is, the connected point closest to the unidentified scattered points in cases 1 and 2 in Figure 7) to each scattered point and the corresponding branch direction of each branch is taken as the deviation angle. The connection order of the scattered points is determined by using the deviation angle and the distance between each scattered point and the last identified branch point. The branch direction and structure are then reconstructed based on the determined connection order of the scattered points.

[0095] When all the scattered points in the tree map have been connected, or the distance between the remaining scattered points and the identified branch points exceeds the preset distance value, or the deviation angle between the remaining scattered points and the last identified branch point exceeds the preset angle value, and the last identified branch point coincides with the branch endpoint, then the reconstructed three-dimensional vascular network structure image is obtained.

[0096] Furthermore, when determining the branch direction, for branches in the tree map where both the starting and ending points are branch points, the line connecting the starting and ending points is taken as the branch direction. Referring to Figure 7, the angle between the line connecting the last branch point identified within the branch to the candidate scattered points and the trend direction is used as the deviation angle. An objective function is established using the deviation angle and distance as parameters. The smaller the deviation angle and the smaller the distance, the smaller the value of the objective function. The optimization process calculates the order of scattered point connections that minimizes the objective function. Scattered points are then connected according to this order to reconstruct the branch direction and structure.

[0097] In a tree map, for a branch with a starting point as a branch node and an unknown ending point, the direction of the line connecting the last two points within the branch is taken as the trend direction of the branch. The angle between the direction of the line connecting the last point within the branch to the candidate scattered points and the trend direction is taken as the deviation angle. Using the deviation angle and distance as parameters, an objective function is established. The smaller the deviation angle and the smaller the distance, the smaller the value of the objective function. The order of scattered points that minimizes the objective function is calculated through the optimization process. Scattered points are connected according to this order to restore the branch direction and structure.

[0098] When all scattered data points within a branch are connected sequentially, or when the remaining data points have too large a distance or deviation angle from the already connected points, or when the last connected point almost overlaps with the end point of the branch, the reconstruction of that branch is complete. Once all branches are connected, the reconstruction is finished, and the three-dimensional model of the vascular network can be viewed using drawing programs or software.

[0099] The method disclosed in this embodiment utilizes a magnetic field to remotely drive magnetically responsive imaging materials, enabling them to move controllably with the magnetic field, thereby actively expanding the imaging range and increasing the completeness of vascular network structure recognition. Furthermore, in this embodiment, the tree map is dynamically updated in real time based on branch information, and the tree map can also automatically determine the next direction of the magnetically responsive imaging materials, thereby achieving complete exploration and imaging of vascular network structures.

[0100] Secondly, this embodiment provides a three-dimensional reconstruction system for vascular network structures, as shown in Figure 8, including:

[0101] The information acquisition module 810 is used to acquire an image sequence of the target imaging area; wherein the image sequence contains multiple images of a cluster of magnetically driven microrobots; its function is as described in step S1.

[0102] The information recognition module 820 is used to identify the region location sequence and diffusion feature sequence corresponding to the magnetically driven microrobot cluster in the image sequence; its function is as described in step S2.

[0103] The network reconstruction module 830 is used to reconstruct a three-dimensional vascular network structure image based on the identified regional location sequence and diffusion morphology sequence, and its function is as described in step S3.

[0104] Furthermore, the regional location sequence includes: multiple branch points and multiple branches, and the diffusion feature sequence includes: the branch direction and the optimal exploration direction corresponding to each branch.

[0105] The information recognition module includes an information separation module and a feature recognition module.

[0106] The information separation module is used to separate the location and diffusion characteristics of the magnetically driven microrobot cluster from each image in the image sequence.

[0107] The feature recognition module is used to determine multiple branch points, multiple branches, branch directions corresponding to the multiple branches, and the optimal exploration direction within the target imaging area based on the location and diffusion characteristics of the magnetically driven microrobot cluster.

[0108] Furthermore, the information separation module utilizes dynamic threshold image binarization and image difference methods to separate the location and diffusion characteristics of the cluster.

[0109] Furthermore, the feature recognition module includes: a branch point judgment unit, an exploration direction determination unit, and an information recording unit.

[0110] The branch point determination unit is used to determine whether there is a branch point in the imaging area based on the intersection information contained in the area where the magnetically driven microrobot cluster is located.

[0111] The exploration direction determination unit is used to determine the branch direction corresponding to the branch point based on the vector connected to the branch point when a branch point exists, and to determine the optimal exploration direction based on the branch direction.

[0112] The information recording unit is used to determine the direction of movement of the magnetically driven microrobot cluster based on its position when there is no branch point, and to select sampling points at fixed preset intervals to obtain the position coordinates of each sampling point and record the branch to which each sampling point belongs.

[0113] Furthermore, the information recognition module also includes: an exploration and judgment module;

[0114] The exploration determination module is used to determine whether the magnetically driven microrobot cluster has stopped spreading for a long time, or whether the magnetically driven microrobot cluster has spread to overlap with other branches, or whether the magnetically driven microrobot cluster has spread beyond the required exploration area; if so, the exploration of the branch where the magnetically driven microrobot cluster has stopped spreading for a long time, or spread to overlap with other branches, or spread beyond the required exploration area is marked as complete; and the diffusion status information of all branches is continuously recorded until all branches have been marked as completed.

[0115] Furthermore, the network reconstruction module includes: a map building unit, a map updating unit, and an image reconstruction unit.

[0116] Map building units are used to describe each branch point and branch using data vectors to construct a tree-like map of the vascular network structure.

[0117] The map update unit is used to update the tree map based on the branch directions and optimal exploration directions corresponding to multiple branches, as well as the branch points and branches obtained through exploration.

[0118] The image reconstruction unit is used to reconstruct a three-dimensional vascular network structure image based on the updated tree map.

[0119] Furthermore, the map building unit includes: establishing a connection subunit;

[0120] The connection subunit is used to connect the branch points and branches according to the collection time sequence of each branch point and branch to obtain a tree map of the three-dimensional vascular network.

[0121] Furthermore, the image reconstruction unit includes: a first recognition subunit, a second recognition subunit, and a reconstruction subunit.

[0122] The first identification subunit is used to take the direction from the start point to the end point as the branch direction for branches where both the start point and the end point are branch points, and to take the angle between the direction of the line connecting the last identified branch point to each scattered point and the corresponding branch direction of each branch as the deviation angle. The connection order of the scattered points is determined by the deviation angle and the distance between each scattered point and the last identified branch point, and the branch direction and structure are restored according to the determined connection order of the scattered points.

[0123] The second identification subunit is used to take the direction of the line connecting the last identified branch point in the branch as the branch direction for a branch whose starting point is a branch point and whose ending point is unknown. The angle between the direction of the line connecting the last identified branch point in the branch to each scattered point and the branch direction corresponding to each branch is used as the deviation angle. The connection order of the scattered points is determined by the deviation angle and the distance between each scattered point and the last identified branch point. The branch direction and structure are restored according to the determined connection order of the scattered points.

[0124] The reconstruction subunit is used to obtain a reconstructed three-dimensional vascular network structure image when all scattered points in the tree map have been connected, or the distance between the remaining scattered points and the identified branch points exceeds a preset distance value, or the deviation angle between the remaining scattered points and the last identified branch point exceeds a preset angle value, and the last identified branch point coincides with the branch endpoint.

[0125] In addition, this embodiment also discloses a computer-readable storage medium, wherein the computer-readable storage medium stores one or more programs, which can be executed by one or more processors to implement the steps of the three-dimensional reconstruction method of the vascular network structure.

[0126] This embodiment discloses a three-dimensional reconstruction method, system, and storage medium for vascular network structures. The method involves acquiring an image sequence of the target imaging area where a magnetically driven microrobot cluster is located within the vascular network structure and driven by a time-varying magnetic field; identifying the regional location information and diffusion morphology characteristics of the magnetically driven microrobot cluster in the image sequence; and reconstructing a three-dimensional vascular network structure image based on the identified regional location information and diffusion morphology characteristics.

[0127] The method and system disclosed in this embodiment use a magnetic field to drive a cluster of magnetically driven microrobots to actively diffuse in reverse flow or low-flow-velocity blocked or narrowed branches within blood vessels, and identify the cluster's location information and diffusion morphology characteristics, thereby achieving a comprehensive exploration of the vascular network and obtaining a complete image of the vascular network structure, thus improving the accuracy of vascular network structure identification.

[0128] Other embodiments of the invention will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of the invention are indicated by the following claims.

[0129] The embodiments described above are merely illustrative of several implementation methods of this application, and while the descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the invention patent. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of this application, and these all fall within the protection scope of this application. Therefore, the protection scope of this patent application should be determined by the appended claims.

Claims

1. A method of three-dimensional reconstruction of a vascular network structure, characterized by, The method comprises the following steps: acquiring an image sequence of a target imaging region; wherein the image sequence contains multiple images of a magnetic-driven micro robot cluster; identifying a sequence of region positions and a sequence of diffusion characteristics corresponding to the magnetic-driven micro robot cluster in the image sequence; reconstructing a three-dimensional vascular network structure image according to the identified sequence of region positions and sequence of diffusion characteristics.

2. The method of three-dimensional reconstruction of a vascular network structure according to claim 1, characterized in that, The sequence of region positions comprises multiple branch points and multiple branches, and the sequence of diffusion characteristics comprises a branch direction and an optimal exploration direction corresponding to each branch. The step of identifying the sequence of region positions and the sequence of diffusion characteristics corresponding to the magnetic-driven micro robot cluster in the image sequence comprises: separating the region where the magnetic-driven micro robot cluster is located and the diffusion characteristics from each image in the image sequence; determining the multiple branch points, the multiple branches, the branch direction corresponding to each branch, and the optimal exploration direction in the target imaging region according to the region where the magnetic-driven micro robot cluster is located and the diffusion characteristics.

3. The method of three-dimensional reconstruction of a vascular network structure according to claim 2, characterized in that, The step of separating the region where the magnetic-driven micro robot cluster is located and the diffusion characteristics from each image in the image sequence comprises: separating the region where the magnetic-driven micro robot cluster is located and the diffusion characteristics by using an image binarization method with a dynamic threshold and an image difference method.

4. The method of three-dimensional reconstruction of a vascular network structure according to claim 2, wherein The step of determining the multiple branch points, the multiple branches, the branch direction corresponding to each branch, and the optimal exploration direction in the target imaging region according to the region where the magnetic-driven micro robot cluster is located and the diffusion characteristics comprises: determining whether there is a branch point in the imaging region according to the intersection information contained in the region where the magnetic-driven micro robot cluster is located; if there is a branch point, determining the branch direction corresponding to the branch point according to the connected vectors of the branch point, and determining the optimal exploration direction according to the branch direction; if there is no branch point, judging the motion direction of the magnetic-driven micro robot cluster according to the position of the magnetic-driven micro robot cluster, selecting sampling points at a fixed preset interval, obtaining the position coordinates of each sampling point, and recording the branch to which each sampling point belongs.

5. The method of three-dimensional reconstruction of a vascular network structure according to claim 4, characterized in that, After the step of determining the multiple branch points, the multiple branches, the branch direction corresponding to each branch, and the optimal exploration direction in the target imaging region according to the region where the magnetic-driven micro robot cluster is located and the diffusion characteristics, the method further comprises: determining whether the magnetic-driven micro robot cluster has stopped diffusing for a long time, or whether the magnetic-driven micro robot cluster has diffused to coincide with other branches, or whether the magnetic-driven micro robot cluster has diffused to outside the required exploration region; if so, marking that the magnetic-driven micro robot cluster has stopped diffusing for a long time, or has diffused to coincide with other branches, or has diffused to outside the required exploration region; continuously recording the state information of all branches until all branches have been marked as exploration completed.

6. The method of three-dimensional reconstruction of a vascular network structure according to claim 4, wherein, The step of reconstructing a three-dimensional vascular network structure image according to the identified sequence of region positions and sequence of diffusion characteristics comprises: using data vectors to describe each branch point and branch, and constructing a tree map of the vascular network structure; According to the branch direction corresponding to each branch and the optimal exploration direction, and each branch point and branch obtained by exploration, a tree map is updated; According to the updated tree map, a three-dimensional vascular network structure image is reconstructed.

7. The method of three-dimensional reconstruction of a vascular network structure according to claim 6, characterized in that, The step of using data vectors to describe each branch and branch, and constructing a tree map of the vascular network structure includes: According to the collection time sequence of each branch point and branch, the branch points and branches are connected to obtain a tree map of the three-dimensional vascular network.

8. The method of three-dimensional reconstruction of a vascular network structure according to claim 7, characterized in that, The step of reconstructing a three-dimensional vascular network structure image according to the updated tree map includes: For a branch with a branch point as the starting point and a branch point as the ending point, the direction from the starting point to the ending point is taken as the branch direction, and the included angle between the direction of the connection line between the last identified branch point in the branch and each scatter point and the branch direction corresponding to each branch is taken as the deviation angle. The connection sequence of the scatter points is determined by using the deviation angle and the distance between each scatter point and the last identified branch point, and the branch direction and structure are restored according to the determined connection sequence of the scatter points; For a branch with a branch point as the starting point and an unknown point as the ending point, the direction of the connection line of the last identified branch point in the branch is taken as the branch direction, and the included angle between the direction of the connection line between the last identified branch point in the branch and each scatter point and the branch direction corresponding to each branch is taken as the deviation angle. The connection sequence of the scatter points is determined by using the deviation angle and the distance between each scatter point and the last identified branch point, and the branch direction and structure are restored according to the determined connection sequence of the scatter points; When all the scatter points in the tree map have been connected, or the distance between the remaining scatter points and the identified branch point exceeds the preset distance value, or the corresponding deviation angle between the last identified branch point and the branch ending point exceeds the preset angle value, the last identified branch point coincides with the branch ending point, and then a three-dimensional vascular network structure image is obtained.

9. A three-dimensional reconstruction system of a vascular network structure, characterized by, It includes: An information acquisition module is configured to acquire an image sequence of a target imaging region, wherein the image sequence contains a plurality of magnetic driving micro robot cluster images; An information identification module is configured to identify the region positions corresponding to the magnetic driving micro robot clusters in the image sequence A sequence of region positions and a sequence of diffusion characteristics; A network reconstruction module is configured to reconstruct a three-dimensional vascular network structure image according to the identified sequence of region positions and sequence of diffusion morphologies.

10. A computer-readable storage medium, characterized in that, The computer readable storage medium stores one or more programs which can be executed by one or more processors to implement the steps of the three-dimensional reconstruction method of the vascular network structure according to any one of claims 1-8.

Images