WO2024001747A1 - 肺部血管模型的建立方法、装置及服务器 - Google Patents

肺部血管模型的建立方法、装置及服务器 Download PDF

Info

Publication number
WO2024001747A1
WO2024001747A1 PCT/CN2023/099704 CN2023099704W WO2024001747A1 WO 2024001747 A1 WO2024001747 A1 WO 2024001747A1 CN 2023099704 W CN2023099704 W CN 2023099704W WO 2024001747 A1 WO2024001747 A1 WO 2024001747A1
Authority
WO
WIPO (PCT)
Prior art keywords
contour
point
pixel point
starting pixel
subset
Prior art date
Application number
PCT/CN2023/099704
Other languages
English (en)
French (fr)
Inventor
成兴华
Original Assignee
上海市胸科医院
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 上海市胸科医院 filed Critical 上海市胸科医院
Publication of WO2024001747A1 publication Critical patent/WO2024001747A1/zh

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T19/00—Manipulating 3D models or images for computer graphics
    • G06T19/003—Navigation within 3D models or images
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T17/00—Three dimensional [3D] modelling, e.g. data description of 3D objects

Definitions

  • the present application relates to the field of medical technology, and in particular to a method, device and server for establishing a pulmonary blood vessel model.
  • determining the blood vessels associated with the lung lesion tissue is a more important goal.
  • the establishment of a pulmonary blood vessel model can guide the closure of blood vessels that supply blood to the lesion area during surgery.
  • related technologies propose that pulmonary blood vessel models can be established through threshold-based segmentation methods, region-based segmentation methods, and morphology-based segmentation methods.
  • the pulmonary blood vessel models established through the above solutions have low definition, and due to the The complex morphological structure makes the algorithm more difficult, and the probability of adhesion between blood vessels and surrounding tissues increases, thereby reducing the accuracy of establishing the pulmonary blood vessel model.
  • the purpose of this application is to provide a method, device and server for establishing a pulmonary blood vessel model, which can significantly improve the accuracy of establishing a pulmonary blood vessel model and reduce the adhesion phenomenon with surrounding tissues, thereby reducing the risk of pulmonary blood vessel model building. Simulation difficulty of blood vessel model.
  • embodiments of the present application provide a method for establishing a pulmonary blood vessel model, including:
  • At least one contour point subset is determined from the pixel points of the lung scan data until the contour point subset meets the preset boundary
  • the steps of the convergence condition include: for the first subset of contour points, determine the first-level starting pixel point from the pixel points included in the lung scan data, and extract the first feature value of the first-level starting pixel point, based on the first level starting pixel point.
  • the first eigenvalue of the first-level starting pixel point and the preset blood vessel constraint conditions are used to determine the first contour point subset from the pixel points of the lung scan data; for the contour point subset except the first contour point subset, based on the previous
  • the subset of contour points corresponding to the starting pixel point of the first level determines the starting pixel point of the level, and extracts the first feature value of the starting pixel point.
  • the subset of contour points corresponding to the starting pixel point is determined from the pixel points of the lung scan data.
  • the first feature value is used to characterize the normal direction of the first-level starting pixel point
  • the step of determining the first contour point subset from the pixel points of the lung scan data includes: based on the normal direction and the preset blood vessel constraint conditions , determine at least one contour point search direction; wherein, the preset vascular constraint conditions include bronchial companion conditions; for each contour point search direction, determine candidate contour pixel points from the pixel points of the lung scan data based on the contour point search direction , calculate the first distance value between the candidate contour pixel point and the first-level starting pixel point. If the first distance value is greater than or equal to the preset boundary threshold, it is determined that the candidate contour pixel point belongs to the first-level starting pixel point.
  • the step of determining at least one contour point search direction based on the normal direction and preset blood vessel constraints includes: determining bronchial pixel points from the pixels of the lung scan data, and determining the bronchial pixel point based on the bronchial pixel point. Points and preset blood vessel constraints adjust the normal direction to determine at least one contour point search direction.
  • the step of determining the starting pixel point of this level based on the subset of contour points corresponding to the starting pixel point of the previous level includes: extracting each of the subset of contour points corresponding to the starting pixel point of the previous level.
  • the second eigenvalue of the contour point wherein the second eigenvalue is used to characterize the tangent direction of the contour point; for each contour point, a candidate starting pixel is determined from the pixel points of the lung scan data based on the tangent direction of the contour point point; calculate the distance value between the candidate starting pixel point and the contour point; if the distance value is equal to the preset distance threshold, determine the candidate starting pixel point as the level starting pixel point.
  • the distance threshold is the search distance of the current contour point set to the next contour point set.
  • the step of determining the starting pixel point of this level based on the subset of contour points corresponding to the starting pixel point of the previous level further includes: determining the outline of the subset of contour points corresponding to the starting pixel point of the previous level. Radius; if the contour radius meets the preset blood vessel bifurcation condition, the starting pixel point is determined from the contour point subset corresponding to the previous level starting pixel point.
  • the step of satisfying the preset boundary convergence condition of the contour point subset includes: when the contour radius of the contour point subset corresponding to the starting pixel point of the level reaches the minimum radius threshold, determining the starting pixel of the level The subset of contour points corresponding to the point satisfies the preset boundary convergence condition; and/or, determine the lung fissure point set based on the lung scan data, and determine when the contour point set corresponding to the starting pixel point of this level reaches the point set at the lung fissure.
  • the subset of contour points corresponding to the starting pixel point of this level satisfies the preset boundary convergence conditions.
  • the lung scan data is medical digital imaging and communication image data scanned by CT equipment; or the lung scan data is three-dimensional data, and the lung scan data of different tissues have different CT values.
  • embodiments of the present application also provide a method for establishing a pulmonary blood vessel model, including: acquiring lung scan data of a target object, and determining a first-level starting pixel point; according to the first-level starting pixel point points, determine the contour point sub-set; determine the next-level contour point sub-set according to the contour radius of the contour point sub-set, the characteristic value of each point in the contour point sub-set and the preset distance threshold; the contour point sub-set satisfies the preset boundary convergence condition Stop the calculation when; establish a pulmonary blood vessel model based on each contour point subset.
  • embodiments of the present application also provide a device for establishing a pulmonary blood vessel model, including: an information acquisition module, is configured to obtain lung scan data of the target object; the data calculation module is configured to determine at least one contour point from the pixel points of the lung scan data based on the starting pixel point in the lung scan data and the preset blood vessel constraint conditions Set until the contour point sub-set satisfies the preset boundary convergence condition; the model building module is configured to establish a pulmonary blood vessel model according to each contour point sub-set.
  • the data calculation module is configured to: for the first subset of contour points, determine the first-level starting pixels from the pixels contained in the lung scan data, and extract the first-level starting pixels The first eigenvalue of the point, based on the first eigenvalue of the first-level starting pixel point and the preset blood vessel constraint conditions, determines the first contour point subset from the pixel points of the lung scan data; for except the first contour point subset For contour point subsets other than Feature values and preset blood vessel constraints are used to determine the subset of contour points corresponding to the starting pixel point from the pixel points of the lung scan data.
  • the first feature value is used to characterize the normal direction of the first-level starting pixel point.
  • the data calculation module is configured to: determine at least one contour point search direction according to the normal direction and the preset blood vessel constraint conditions; wherein, the preset The vascular constraint conditions include bronchial companion conditions; for each contour point search direction, candidate contour pixel points are determined from the pixels of the lung scan data based on the contour point search direction, and the candidate contour pixel points and the first-level starting pixels are calculated.
  • the first distance value between the points If the first distance value is greater than or equal to the preset boundary threshold, it is determined that the candidate contour pixel point belongs to the contour point subset corresponding to the first-level starting pixel point.
  • the data calculation module when performing the step of determining at least one contour point search direction according to the normal direction and the preset blood vessel constraint conditions, is configured to: from the pixel points of the lung scan data Determine the bronchial pixel points, adjust the normal direction according to the bronchial pixel points and preset blood vessel constraints, and determine at least one contour point search direction.
  • the data calculation module when performing the step of determining the starting pixel point of this level based on the subset of contour points corresponding to the starting pixel point of the previous level, is configured to: extract the starting pixel of the previous level The second eigenvalue of each contour point in the subset of contour points corresponding to the point; where the second eigenvalue is used to characterize the tangent direction of the contour point; for each contour point, scan data from the lungs based on the tangent direction of the contour point Determine the candidate starting pixel point among the pixels; calculate the distance value between the candidate starting pixel point and the contour point; if the distance value is equal to the preset distance threshold, determine the candidate starting pixel point as the level starting pixel point .
  • the data calculation module when performing the step of determining the starting pixel point of this level based on the subset of contour points corresponding to the starting pixel point of the previous level, is configured to: determine the starting pixel of the previous level Point corresponding outline ideas The contour radius of the set; if the contour radius meets the preset blood vessel bifurcation conditions, the starting pixel point is determined from the contour point subset corresponding to the previous level starting pixel point.
  • the data calculation module when performing the step of the contour point subset meeting the preset boundary convergence condition, is configured to: when the contour radius of the contour point subset corresponding to the starting pixel point of the level reaches the minimum radius When the threshold is reached, it is determined that the subset of contour points corresponding to the starting pixel point of this level meets the preset boundary convergence conditions; and/or, the set of lung fissure points is determined based on the lung scan data. When the set of contour points corresponding to the starting pixel point of this level reaches When collecting points at the lung fissure, it is determined that the subset of contour points corresponding to the starting pixel point of this level satisfies the preset boundary convergence condition.
  • embodiments of the present application further provide a server, including a processor and a memory.
  • the memory stores computer-executable instructions that can be executed by the processor.
  • the processor executes the computer-executable instructions to implement the first aspect or the second aspect. Any of the methods provided.
  • embodiments of the present application further provide a computer-readable storage medium.
  • the computer-readable storage medium stores computer-executable instructions.
  • the computer-executable instructions When the computer-executable instructions are called and executed by the processor, the computer-executable instructions prompt the processor to Methods for implementing any of the items provided by the first or second aspect.
  • the embodiments of this application provide a method, device and server for establishing a pulmonary blood vessel model
  • the method is used to obtain the lung scan data of the target object, and according to the starting pixel point in the lung scan data and the preset blood vessel constraint conditions, at least one contour point subset is determined from the pixel points of the lung scan data until the contour
  • the point set satisfies the preset boundary convergence conditions, and a pulmonary blood vessel model is established based on each contour point set.
  • This method can effectively distinguish blood vessels and perivascular tissues based on preset vascular constraints when establishing a pulmonary blood vessel model, which can significantly improve the accuracy of establishing a pulmonary blood vessel model and reduce adhesion with surrounding tissues, thereby reducing the risk of pulmonary vascular disease.
  • the simulation difficulty of the local blood vessel model is used to obtain the lung scan data of the target object, and according to the starting pixel point in the lung scan data and the preset blood vessel constraint conditions, at least one contour point subset is determined from the pixel points of the lung scan data until the contour
  • the point set satisfies the preset boundary convergence conditions, and
  • Figure 1 is a schematic flow chart of a method for establishing a pulmonary blood vessel model provided by an embodiment of the present application
  • Figure 2 is a schematic diagram of the tangent and normal directions of a blood vessel provided by an embodiment of the present application
  • Figure 3 is a schematic structural diagram of a blood vessel accompanying the bronchi provided by an embodiment of the present application.
  • Figure 4 is a schematic structural diagram of a lung fissure provided by an embodiment of the present application.
  • Figure 5 is a schematic flow chart of another method for establishing a pulmonary blood vessel model provided by an embodiment of the present application.
  • Figure 6 is a schematic structural diagram of a device for establishing a pulmonary blood vessel model provided by an embodiment of the present application.
  • Figure 7 is a schematic structural diagram of a server provided by an embodiment of the present application.
  • 602-information acquisition module 604-data calculation module; 606-model establishment module; 70-processor; 71-memory; 72-bus; 73-communication interface.
  • determining the blood vessels associated with the lung lesion tissue is a more important goal.
  • Establishing a pulmonary blood vessel model can guide the closure of blood vessels that supply blood to the lesion area during surgery, and pulmonary blood vessels are the human body's One of the most important tissues and organs with the most complex topological structure.
  • Manual segmentation of tissues and organs in images consumes a lot of energy and physical strength for doctors.
  • Threshold-based blood vessel segmentation methods are relatively simple to implement and mainly use local segmentation thresholds or local grayscales. With the segmentation strategy of structural analysis, the segmentation results are prone to blurred contours and mis-segmentation phenomena; the blood vessel segmentation method based on region growing is the most widely used, mainly based on the grayscale information of the blood vessel tissue in the image, and is prone to over-segmentation and hole phenomena; based on Morphological blood vessel segmentation methods mainly use operators to detect vascular tissue, such as SMDC connection cost operator, Canny operator, etc.
  • This type of method can better eliminate the interference of noise in the image and retain the details of blood vessel branches, but Due to the complex morphological structure of blood vessels, operator parameters are difficult to fix, and adhesion phenomena often occur in segmentation results.
  • blood vessel segmentation methods based on machine learning and spatial filtering, which have better segmentation results but higher algorithm complexity.
  • this application implements the method for establishing a pulmonary blood vessel model.
  • the accompanying conditions of the bronchi are added to the constraints of the algorithm, which can not only adjust the search path of the blood vessels in real time, but also avoid calculating the pixel points of the bronchi.
  • the interference of bronchi on pulmonary blood vessel segmentation is eliminated.
  • the anatomical and physiological characteristics of blood vessels that are different from other tissues are fully considered, so that the pulmonary blood vessels can be segmented in a targeted manner, which can significantly improve the accuracy of the pulmonary blood vessel model establishment. degree, and reduce the adhesion phenomenon with surrounding tissues, thereby reducing the simulation difficulty of the pulmonary blood vessel model.
  • the method mainly includes the following steps S102 to S106:
  • Step S102 Obtain the lung scan data of the target object.
  • the lung scan data is Digital Imaging and Communications in Medicine (DICOM) image data scanned by CT equipment.
  • DICOM Digital Imaging and Communications in Medicine
  • the lung scan data is three-dimensional data, and the lung scan data of different tissues have different CT values.
  • peripheral tissues such as blood vessels and lymph can be distinguished based on the different CT values.
  • the CT value of lymph is different from Blood vessels are relatively low, so the brightness of lymphatic CT images is lower than that of blood vessel CT images.
  • Step S104 Determine at least one contour point subset from the pixel points of the lung scan data according to the starting pixel point in the lung scan data and the preset blood vessel constraint conditions until the contour point subset satisfies the preset boundary convergence condition.
  • the outline point set can be understood as a set of pixel points in a section of blood vessels with the same outline radius.
  • the inner layer data in the lung scan data that is, the pixel points at the aortic blood vessels close to the heart are selected as the starting pixel points.
  • the constraint condition can be the bronchial companion condition, that is, the bronchi and blood vessels are in a parallel relationship, and the contour point set is searched according to the bronchial extension direction.
  • the boundary convergence condition can be the convergence condition for the pixel point and the convergence condition for the contour point set.
  • the boundary convergence condition is triggered when reaching the lung fissure, and/or the boundary convergence condition is triggered when the contour radius of the contour point set reaches the minimum radius threshold.
  • the contour point sub-set is a point set of contour points. The same group of contour point sub-sets at the vertical section The contour radii are the same.
  • the pulmonary airway and the pulmonary artery are concomitant, and there are no large blood vessels near the pulmonary fissure. Therefore, when the calculated boundary convergence condition is that the set of contour points reaches the set of points at the pulmonary fissure (such as, The calculation stops when the branch vessels of the pulmonary artery reach the boundary of the lung parenchyma, and when the pulmonary vein reaches the lung parenchyma and the lung fissures corresponding to the lung lobes), or the contour radius r reaches the minimum radius threshold r (min), the boundary convergence condition is reached.
  • Step S106 Establish a pulmonary blood vessel model based on each contour point subset.
  • the pulmonary blood vessel model is an intricate tree structure, containing approximately 23 levels of branches, with diameters ranging from 20um to 15mm.
  • the blood vessels in the pulmonary blood vessel model have the geometric characteristics of elongated, tubular, and tree-like distribution. Therefore, the pixels of the inner layer image in the lung scan data are selected as the starting point of the blood vessels, and are established from the inside to the outside.
  • Pulmonary blood vessel model for example, the starting pixel point for establishing the pulmonary blood vessel model is the point closest to the heart. There are multiple aortas in the pulmonary blood vessel model, so the starting pixel point can be multiple groups of pixel points.
  • the above-mentioned method for establishing a pulmonary blood vessel model provided by the embodiments of the present application can effectively distinguish blood vessels and perivascular tissues based on preset blood vessel constraint conditions when establishing a pulmonary blood vessel model, and can significantly improve the accuracy of establishing a pulmonary blood vessel model. And reduce the adhesion phenomenon with surrounding tissues, thereby reducing the simulation difficulty of the pulmonary blood vessel model.
  • the embodiment of the present application also provides an implementation method for determining a subset of contour points.
  • an implementation method for determining a subset of contour points please refer to the following (1) to (2):
  • the first subset of contour points determine the first-level starting pixel point from the pixels contained in the lung scan data, and extract the first feature value of the first-level starting pixel point, based on the first-level starting pixel point.
  • the first characteristic value of the pixel and the preset blood vessel Constraints are used to determine the first subset of contour points from the pixels of the lung scan data.
  • the feature value includes the first feature value, the second feature value and the coordinate value of the pixel point.
  • the first feature value is used to characterize the normal direction of the contour point, and the second feature value is used to characterize the tangent direction of the contour point.
  • the first-level starting pixel point is the point closest to the heart.
  • the first-level starting pixel point can be multiple groups of pixel points
  • the blood vessel constraint condition can be the bronchial companion condition, that is , the bronchi and blood vessels are in a parallel relationship.
  • the direction of the blood vessel can be adjusted according to the first feature value and the preset blood vessel constraint conditions. Since the bronchi and blood vessels accompany each other, in the process of finding pixel points in the contour point set, the constraint conditions are used to determine Whether the pixel point B (x, y, z) of the bronchus is within the eight-area range of the contour point P (x, y, z), adjust the direction of the normal e1 of the search path in real time, and search in the direction of the parallel bronchus.
  • the next-level starting pixel point determined based on the contour point subset corresponding to the previous-level starting pixel point can be any point at the cross-section of the same contour point subset.
  • the next-level starting pixel points of the two sub-vessels after the bifurcation are determined by the outline point set of the main blood vessel, can be the sub-vessel and the main blood vessel. Any pixel in the annular cross-section at the connection point set of blood vessel outline points.
  • the embodiment of the present application also provides an implementation method of determining the starting pixel point of the contour point sub-set except the first contour point sub-set. For details, please refer to the following (a) to (b):
  • the distance threshold is the search distance of the current contour point set to the next contour point set. The distance threshold is set to the initial distance d.
  • contour point set From any contour point in the current contour point set, search for vascular CT by extending in the tangential direction d. value of the pixel point, and use this pixel point as the initial contour point of the next contour point set.
  • the contour point set can be regarded as a whole, and the next-level contour is found by extending along the e1 direction of the current circular contour with an initial distance d. point set, and perform iterative calculations based on it.
  • the outline point set of blood vessels has an elongated, tubular structure, and the outline radius is the radius of the vertical section of the tubular outline point set.
  • the difference between blood vessels and other tissues is that the interval diameter of blood vessels does not change much, and the edge Symmetrical appearance, therefore, the contour radius of adjacent contour points gradually decreases within a certain range, thereby excluding the lymphatic group surrounding the blood vessel organizations, and other organizational structures.
  • the contour radius when the contour radius suddenly becomes larger, it indicates that the blood vessels begin to bifurcate, and multiple starting pixel points are generated at this time.
  • the embodiment of the present application also provides an implementation method of determining the search direction of the contour point subset based on the feature value and the accompanying condition. For details, please refer to the following (1) to (3):
  • a pixel point of a specific layer image is selected as the blood vessel starting point, assuming its value is V (x, y, z), and the Hessian matrix of this point and its eigenvalues e1 and e2 are calculated, where e1 represents the tangential direction of the blood vessel, and e2 represents the normal direction of the blood vessel.
  • the bending direction of the blood vessel is determined according to
  • the boundary threshold is a preset CT value threshold, and the CT value range of different tissues can be specified according to the boundary threshold, thereby distinguishing blood vessels from adjacent tissues.
  • the contour radius of the subset of contour points corresponding to the starting pixel point of the level reaches the minimum radius threshold, it is determined that the subset of contour points corresponding to the starting pixel point of the level satisfies the preset boundary convergence condition; and/or , determine the lung fissure point set based on the lung scan data.
  • the contour point set corresponding to the starting pixel point of this level reaches the point set at the lung fissure, it is determined that the contour point subset corresponding to the starting pixel point of this level meets the preset boundary convergence. condition.
  • the outline point set can be understood as the set of pixel points in a section of blood vessels with the same outline radius
  • the set of pulmonary fissure points is as shown in Figure 5.
  • the pulmonary airways and the pulmonary arteries are concomitant, and there are no large blood vessels near the pulmonary fissures. Therefore, when the calculated boundary convergence condition is that the set of contour points reaches Collection of points at lung fissures.
  • the calculation stops when (for example, the branch vessels of the pulmonary artery reach the boundary of the pulmonary parenchyma, and the pulmonary vein reaches the pulmonary parenchyma and the pulmonary fissures of the corresponding pulmonary lobes), or the contour radius r reaches the minimum radius threshold r (min), and the boundary convergence condition is reached at this time.
  • embodiments of the present application provide an application example of a method for establishing a pulmonary blood vessel model. See another pulmonary blood vessel model shown in Figure 5 A schematic flow chart of the establishment method. The method mainly includes the following steps S502 to S510:
  • Step S502 Obtain the lung scan data of the target object and determine the first-level starting pixel point.
  • the lung scan data is Digital Imaging and Communications in Medicine (DICOM) image data scanned by CT equipment.
  • DICOM Digital Imaging and Communications in Medicine
  • the lung scan data is three-dimensional data, and the lung scan data of different tissues have different CT values.
  • peripheral tissues such as blood vessels and lymph can be distinguished based on the different CT values.
  • the CT value of lymph is different from Blood vessels are relatively low, so the brightness of lymphatic CT images is lower than that of blood vessel CT images.
  • Step S504 Determine a subset of contour points based on the first-level starting pixel points.
  • the blood vessels in the pulmonary blood vessel model have the geometric characteristics of slender, tubular, and tree-like distribution. Therefore, the pixels of the inner image in the lung scan data are selected as the starting point of the blood vessels, and the pulmonary blood vessel model is established from the inside to the outside.
  • the starting pixel point for establishing the pulmonary blood vessel model is the point closest to the heart. There are multiple aortas in the pulmonary blood vessel model, so the starting pixel point can be multiple sets of starting pixel points.
  • Step S506 Determine the next-level contour point subset according to the contour radius of the contour point subset, the feature value of each point in the contour point subset, and the preset distance threshold.
  • the outline point set can be understood as the set of pixel points in a section of blood vessels with the same outline radius
  • Step S508 Stop calculation when the contour point subset meets the preset boundary convergence condition.
  • the pulmonary airway and the pulmonary artery are accompanied, and there are no large blood vessels near the pulmonary fissure. Therefore, when the calculated boundary convergence condition is that the contour point set reaches the point set at the pulmonary fissure (for example, the branch blood vessels of the pulmonary artery reach The calculation stops when the boundary of the lung parenchyma (the pulmonary vein reaches the lung parenchyma and the lung fissure corresponding to the lung lobe), or the contour radius r reaches the minimum radius threshold r (min), and the boundary convergence condition is reached at this time.
  • Step S510 Establish a pulmonary blood vessel model based on each contour point subset.
  • the pulmonary blood vessel model is an intricate tree structure, containing approximately 23 levels of branches, with diameters ranging from 20um to 15mm.
  • this application can effectively distinguish blood vessels and perivascular tissues based on preset vascular constraints when establishing a pulmonary blood vessel model, which can significantly improve the accuracy of establishing a pulmonary blood vessel model and reduce adhesion with surrounding tissues. phenomenon, thereby reducing the simulation difficulty of the pulmonary blood vessel model.
  • an embodiment of the present application provides a device for establishing a pulmonary blood vessel model. See Figure 6 for a schematic structural diagram of a device for establishing a pulmonary blood vessel model.
  • the device includes the following parts:
  • the information acquisition module 602 is configured to acquire lung scan data of the target subject
  • the data calculation module 604 is configured to determine at least one contour point subset from the pixel points of the lung scan data according to the starting pixel point in the lung scan data and the preset blood vessel constraint conditions, until the contour point subset meets the preset boundary convergence condition;
  • the model building module 606 is configured to build a pulmonary blood vessel model according to each contour point subset.
  • the above-mentioned data processing device provided by the embodiments of the present application can effectively distinguish blood vessels and perivascular tissues according to preset vascular constraints when establishing a pulmonary blood vessel model, which can significantly improve the accuracy of establishing a pulmonary blood vessel model and reduce the risk of interaction with surrounding tissues.
  • the adhesion phenomenon occurs, thereby reducing the simulation difficulty of the pulmonary blood vessel model.
  • the data calculation module 604 is also configured to: for the first subset of contour points, determine the first-level starting pixel points from the pixels included in the lung scan data, and extract the first-level starting pixel points.
  • the first eigenvalue of The starting pixel point of this level is determined based on the contour point subset corresponding to the starting pixel point of the previous level, and the first feature value of the starting pixel point is extracted.
  • Based on the first feature of the starting pixel point value and preset blood vessel constraint conditions and determine the subset of contour points corresponding to the starting pixel point from the pixel points of the lung scan data.
  • the first eigenvalue is used to characterize the normal direction of the first-level starting pixel point. Based on the first eigenvalue of the first-level starting pixel point and the preset blood vessel constraint conditions, the lungs are In the step of determining the first contour point subset among the pixels of the scan data, the above-mentioned data calculation module 604 is also configured to: determine at least one contour point search direction according to the normal direction and the preset blood vessel constraint conditions; wherein, the preset The vascular constraint conditions include bronchial companion conditions; for each contour point search direction, candidate contour pixel points are determined from the pixels of the lung scan data based on the contour point search direction, and the candidate contour pixel points and the first-level starting pixels are calculated. The first distance value between the points. If the first distance value is greater than or equal to the preset boundary threshold, it is determined that the candidate contour pixel point belongs to the contour point subset corresponding to the first-level starting pixel point.
  • the above-mentioned data calculation module 604 is also configured to: from the pixel points of the lung scan data Determine the bronchial pixel points, adjust the normal direction according to the bronchial pixel points and preset blood vessel constraints, and determine at least one contour point search direction.
  • the above-mentioned data calculation module 604 when performing the step of determining the starting pixel point of this level based on the subset of contour points corresponding to the starting pixel point of the previous level, is also configured to: extract the starting pixel of the previous level.
  • the above-mentioned data calculation module 604 is also configured to: determine the starting pixel of the previous level. The contour radius of the contour point subset corresponding to the point; if the contour radius meets the preset blood vessel bifurcation condition, the starting pixel point is determined from the contour point subset corresponding to the previous level starting pixel point.
  • the above-mentioned data calculation module 604 when performing the step of satisfying the preset boundary convergence condition for the contour point subset, is also configured to: when the contour radius of the contour point subset corresponding to the starting pixel point of the level reaches the minimum radius When the threshold is reached, it is determined that the subset of contour points corresponding to the starting pixel point of this level meets the preset boundary convergence conditions; and/or, the set of lung fissure points is determined based on the lung scan data. When the set of contour points corresponding to the starting pixel point of this level reaches When collecting points at the lung fissure, it is determined that the subset of contour points corresponding to the starting pixel point of this level satisfies the preset boundary convergence condition.
  • An embodiment of the present application provides a server.
  • the server includes a processor and a storage device; a computer program is stored on the storage device, and when the computer program is run by the processor, it executes any one of the above embodiments. the method described.
  • FIG. 7 is a schematic structural diagram of a server provided by an embodiment of the present application.
  • the server 100 includes: a processor 70, a memory 71, a bus 72 and a communication interface 73.
  • the processor 70, the communication interface 73 and the memory 71 pass through the bus 72 Connection;
  • the processor 70 is used to execute executable modules stored in the memory 71, such as a computer program.
  • the memory 71 may include high-speed random access memory (RAM, Random Access Memory), and may also include non-volatile memory (non-volatile memory), such as at least one disk memory.
  • RAM Random Access Memory
  • non-volatile memory such as at least one disk memory.
  • the communication connection between the system network element and at least one other network element is realized through at least one communication interface 73 (which can be wired or wireless), and the Internet, wide area network, local network, metropolitan area network, etc. can be used.
  • the bus 72 may be an ISA bus, a PCI bus, an EISA bus, etc.
  • the bus can be divided into address bus, data bus, control bus, etc. For ease of presentation, only one bidirectional arrow is used in Figure 7, but it does not mean that there is only one bus or one type of bus.
  • the memory 71 is used to store the program.
  • the processor 70 executes the program after receiving the execution instruction.
  • the method executed by the device for stream process definition disclosed in any of the embodiments of the present application can be applied to processing. in the processor 70, or implemented by the processor 70.
  • the processor 70 may be an integrated circuit chip with signal processing capabilities. During the implementation process, each step of the above method can be completed by instructions in the form of hardware integrated logic circuits or software in the processor 70 .
  • the above mentioned places The processor 70 may be a general-purpose processor, including a Central Processing Unit (CPU for short), a Network Processor (NP for short), etc.; it may also be a Digital Signal Processing (DSP for short), Application Specific Integrated Circuit (ASIC), off-the-shelf programmable gate array (Field-Programmable Gate Array, FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.
  • CPU Central Processing Unit
  • NP Network Processor
  • DSP Digital Signal Processing
  • ASIC Application Specific Integrated Circuit
  • FPGA off-the-shelf programmable gate array
  • FPGA field-Programmable Gate Array
  • a general-purpose processor may be a microprocessor or the processor may be any conventional processor, etc.
  • the steps of the method disclosed in conjunction with the embodiments of the present application can be directly implemented by a hardware decoding processor, or executed by a combination of hardware and software modules in the decoding processor.
  • the software module can be located in random access memory, flash memory, read-only memory, programmable read-only memory or electrically erasable programmable memory, registers and other mature storage media in this field.
  • the storage medium is located in the memory 71.
  • the processor 70 reads the information in the memory 71 and completes the steps of the above method in combination with its hardware.
  • the computer program product of the readable storage medium provided by the embodiment of the present application includes a computer-readable storage medium storing program code.
  • the instructions included in the program code can be used to execute the method described in the previous method embodiment. Specifically, Please refer to the foregoing method embodiments, which will not be described again here.
  • the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium.
  • the technical solution of the present application is essentially or contributes to the relevant technology or part of the technical solution can be embodied in the form of a software product.
  • the computer software product is stored in a storage medium and includes several The instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of this application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .
  • This application provides a method, device and server for establishing a pulmonary blood vessel model, wherein the method is used to obtain the target lung scan data of the target object, and according to the starting pixel point in the lung scan data and the preset blood vessel constraint conditions, at least one contour point subset is determined from the pixel points of the lung scan data until the contour point subset meets the preset boundary Based on the convergence conditions, the pulmonary blood vessel model is established based on each contour point subset.
  • This method can effectively distinguish blood vessels and perivascular tissues based on preset vascular constraints when establishing a pulmonary blood vessel model, which can significantly improve the accuracy of establishing a pulmonary blood vessel model and reduce adhesion with surrounding tissues, thereby reducing the risk of pulmonary vascular disease.
  • the simulation difficulty of the local blood vessel model is used to obtain the target lung scan data of the target object, and according to the starting pixel point in the lung scan data and the preset blood vessel constraint conditions, at least one contour point subset is determined from the pixel points of the lung scan data until
  • the method, device and server for establishing the pulmonary blood vessel model of the present application are reproducible and can be used in a variety of industrial applications.
  • the method, device and server for establishing a pulmonary blood vessel model in this application can be used in the field of medical technology.

Landscapes

  • Engineering & Computer Science (AREA)
  • Software Systems (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Computer Graphics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Hardware Design (AREA)
  • General Engineering & Computer Science (AREA)
  • Remote Sensing (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Geometry (AREA)
  • Apparatus For Radiation Diagnosis (AREA)
  • Magnetic Resonance Imaging Apparatus (AREA)

Abstract

本申请提供了一种肺部血管模型的建立方法，涉及医疗技术领域，包括：获取目标对象的肺部扫描数据；根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件；根据每个轮廓点子集合建立肺部血管模型。本申请可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。

Description

肺部血管模型的建立方法、装置及服务器
相关申请的交叉引用
本申请要求于2022年06月28日提交中国国家知识产权局的申请号为202210746710.3、名称为“肺部血管模型的建立方法、装置及服务器”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及医疗技术领域，尤其是涉及一种肺部血管模型的建立方法、装置及服务器。
背景技术
针对肺部手术的导航设备中，确定肺部病灶组织周围相关联的血管是比较重要的目标，建立肺部血管模型可以指导术中封闭给病灶区域供血的血管。目前，相关技术提出可以通过基于阈值的分割方法、基于区域的分割方法和基于形态学的分割方法等建立肺部血管模型，通过上述方案建立的肺部血管模型清晰度较低，且由于血管的形态结构复杂，从而导致算法难度较高，血管与周围组织发生粘连现象的概率提升，进而降低肺部血管模型建立的精确度。
发明内容
有鉴于此，本申请的目的在于提供一种肺部血管模型的建立方法、装置及服务器，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
第一方面，本申请实施例提供了一种肺部血管模型的建立方法，包括：
获取目标对象的肺部扫描数据；根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件；根据每个轮廓点子集合建立肺部血管模型。
在一种可选实施方式中，根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件的步骤，包括：对于首个轮廓点子集合，从肺部扫描数据包含的像素点中确定第一级起始像素点，并提取第一级起始像素点的第一特征值，基于第一级起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合；对于除首个轮廓点子集合之外的轮廓点子集合，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点，并提取该起始像素点的第一特征值，基于该起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定该起始像素点对应的轮廓点子集合。
在一种可选实施方式中，其中，第一特征值用于表征第一级起始像素点的法线方向， 基于第一级起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合的步骤，包括：根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向；其中，预设血管约束条件包括支气管伴行条件；对于每个轮廓点搜寻方向，基于该轮廓点搜寻方向从肺部扫描数据的像素点中确定候选轮廓像素点，计算候选轮廓像素点与第一级起始像素点之间的第一距离值，如果第一距离值大于或等于预设边界阈值，确定候选轮廓像素点属于第一级起始像素点对应的轮廓点子集合。
在一种可选实施方式中，根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向的步骤，包括：从肺部扫描数据的像素点中确定支气管像素点，并根据支气管像素点和预设血管约束条件对法线方向进行调整，确定至少一个轮廓点搜寻方向。
在一种可选实施方式中，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤，包括：提取前一级起始像素点对应的轮廓点子集合中每个轮廓点的第二特征值；其中，第二特征值用于表征轮廓点的切线方向；对于每个轮廓点，基于该轮廓点的切线方向从肺部扫描数据的像素点中确定候选起始像素点；计算候选起始像素点与该轮廓点之间的距离值；如果距离值等于预设距离阈值，确定该候选起始像素点为该级起始像素点。
在一种可选实施方式中，距离阈值为当前轮廓点集合对下一轮廓点集合的搜寻距离。
在一种可选实施方式中，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤，还包括：确定前一级起始像素点对应的轮廓点子集合的轮廓半径；如果轮廓半径满足预设血管分叉条件，从前一级起始像素点对应的轮廓点子集合中确定该起始像素点。
在一种可选实施方式中，轮廓点子集合满足预设边界收敛条件的步骤，包括：当该级起始像素点对应的轮廓点子集合的轮廓半径达到最小半径阈值时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件；和/或，根据肺部扫描数据确定肺裂隙点集合，当该级起始像素点对应的轮廓点集合到达肺裂隙处的点集合时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件。
在一种可选实施方式中，肺部扫描数据为CT设备扫描的医学数字成像和通信图像数据；或者肺部扫描数据为三维数据，不同组织的肺部扫描数据具有不同的CT值。
第二方面，本申请实施例还提供一种肺部血管模型的建立方法，包括：获取目标对象的肺部扫描数据，并确定第一级起始像素点；根据所述第一级起始像素点，确定轮廓点子集合；根据所述轮廓点子集合的轮廓半径、轮廓点子集合中各点的特征值及预设距离阈值确定下一级轮廓点子集合；所述轮廓点子集合满足预设边界收敛条件时停止计算；根据每个轮廓点子集合建立肺部血管模型。
第三方面，本申请实施例还提供一种肺部血管模型的建立装置，包括：信息获取模块， 被配置成获取目标对象的肺部扫描数据；数据计算模块，被配置成根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件；模型建立模块，被配置成根据每个轮廓点子集合建立肺部血管模型。
在一种可选实施方式中，在进行根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条的步骤时，数据计算模块被配置用于：对于首个轮廓点子集合，从肺部扫描数据包含的像素点中确定第一级起始像素点，并提取第一级起始像素点的第一特征值，基于第一级起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合；对于除首个轮廓点子集合之外的轮廓点子集合，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点，并提取该起始像素点的第一特征值，基于该起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定该起始像素点对应的轮廓点子集合。
在一种可选实施方式中，第一特征值用于表征第一级起始像素点的法线方向，在进行基于第一级起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合的步骤时，数据计算模块被配置用于：根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向；其中，预设血管约束条件包括支气管伴行条件；对于每个轮廓点搜寻方向，基于该轮廓点搜寻方向从肺部扫描数据的像素点中确定候选轮廓像素点，计算候选轮廓像素点与第一级起始像素点之间的第一距离值，如果第一距离值大于或等于预设边界阈值，确定候选轮廓像素点属于第一级起始像素点对应的轮廓点子集合。
在一种可选实施方式中，在进行根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向的步骤时，数据计算模块被配置用于：从肺部扫描数据的像素点中确定支气管像素点，并根据支气管像素点和预设血管约束条件对法线方向进行调整，确定至少一个轮廓点搜寻方向。
在一种可选实施方式中，在进行基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤时，数据计算模块被配置用于：提取前一级起始像素点对应的轮廓点子集合中每个轮廓点的第二特征值；其中，第二特征值用于表征轮廓点的切线方向；对于每个轮廓点，基于该轮廓点的切线方向从肺部扫描数据的像素点中确定候选起始像素点；计算候选起始像素点与该轮廓点之间的距离值；如果距离值等于预设距离阈值，确定该候选起始像素点为该级起始像素点。
在一种可选实施方式中，在进行基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤时，数据计算模块被配置用于：确定前一级起始像素点对应的轮廓点子 集合的轮廓半径；如果轮廓半径满足预设血管分叉条件，从前一级起始像素点对应的轮廓点子集合中确定该起始像素点。
在一种可选实施方式中，在进行轮廓点子集合满足预设边界收敛条件的步骤时，数据计算模块被配置用于：当该级起始像素点对应的轮廓点子集合的轮廓半径达到最小半径阈值时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件；和/或，根据肺部扫描数据确定肺裂隙点集合，当该级起始像素点对应的轮廓点集合到达肺裂隙处的点集合时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件。
第四方面，本申请实施例还提供一种服务器，包括处理器和存储器，存储器存储有能够被处理器执行的计算机可执行指令，处理器执行计算机可执行指令以实现第一方面或第二方面提供的任一项的方法。
第五方面，本申请实施例还提供一种计算机可读存储介质，计算机可读存储介质存储有计算机可执行指令，计算机可执行指令在被处理器调用和执行时，计算机可执行指令促使处理器实现第一方面或第二方面提供的任一项的方法。
本申请实施例带来了以下有益效果：
本申请实施例提供的一种肺部血管模型的建立方法、装置及服务器，
其中该方法用于获取目标对象的肺部扫描数据，并根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件，根据每个轮廓点子集合建立肺部血管模型。该方法可以在建立肺部血管模型时，根据预设血管约束条件有效区分血管与血管周围组织，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
本申请的其他特征和优点将在随后的说明书中阐述，并且，部分地从说明书中变得显而易见，或者通过实施本申请而了解。本申请的目的和其他优点在说明书、权利要求书以及附图中所特别指出的结构来实现和获得。
为使本申请的上述目的、特征和优点能更明显易懂，下文特举较佳实施例，并配合所附附图，作详细说明如下。
附图说明
为了更清楚地说明本申请具体实施方式或相关技术中的技术方案，下面将对具体实施方式或相关技术描述中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图是本申请的一些实施方式，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请实施例提供的一种肺部血管模型的建立方法的流程示意图；
图2为本申请实施例提供的一种血管切线、法线方向的示意图；
图3为本申请实施例提供的一种血管支气管伴行的结构示意图；
图4为本申请实施例提供的一种肺裂隙的结构示意图；
图5为本申请实施例提供的另一种肺部血管模型的建立方法的流程示意图；
图6为本申请实施例提供的一种肺部血管模型的建立装置的结构示意图；
图7为本申请实施例提供的一种服务器的结构示意图。
图中：602-信息获取模块；604-数据计算模块；606-模型建立模块；70-处理器；71-存储器；72-总线；73-通信接口。
具体实施方式
为使本申请实施例的目的、技术方案和优点更加清楚，下面将结合实施例对本申请的技术方案进行清楚、完整地描述，显然，所描述的实施例是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有做出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
目前，针对肺部手术的导航设备中，确定肺部病灶组织周围相关联的血管是比较重要的目标，建立肺部血管模型可以指导术中封闭给病灶区域供血的血管，而肺部血管是人体最为重要且拓扑结构最为复杂的组织器官之一，手动分割影像中的组织器官对医生精力、体力消耗极大，影像中模糊的血管边界、低的组织对比度和部分容积效应等影响，使得血管组织难以准确分割；相关肺血管分割方法主要包括基于阈值的分割方法、基于区域的分割方法和基于形态学的分割方法等，基于阈值的血管分割方法实现相对简单，主要采用局部分割阈值或局部灰度结构分析的分割策略，分割结果易出现轮廓模糊和误分割现象；基于区域生长的血管分割方法应用最为广泛，主要基于影像中血管组织的灰度信息进行判断，容易出现过度分割和空洞现象；基于形态学的血管分割方法，主要利用算子对血管组织进行探测，例如SMDC连接成本算子、Canny算子等，这类方法能较好地消除影像中噪声的干扰，保留血管分支的细节，但由于血管形态结构复杂，算子参数很难固定，分割结果中常有粘连现象发生；此外，还有基于机器学习、空间滤波的血管分割方法等，分割效果较好，但算法复杂度较高。
基于此，本申请实施提供的肺部血管模型的建立方法，在算法的约束条件中，加入了支气管的伴行条件，既可以实时调整血管的寻找路径，又可以避免把支气管的像素点计算在内，排除了支气管对肺血管分割的干扰，在计算的过程中，充分考虑了血管区别于其它组织的解剖生理特性，从而有针对性的分割肺血管，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
参见图1所示的一种肺部血管模型的建立方法的流程示意图，该方法主要包括以下步骤S102至步骤S106：
步骤S102，获取目标对象的肺部扫描数据。其中，肺部扫描数据为CT设备扫描的医学数字成像和通信(Digital Imaging and Communications in Medicine，DICOM)图像数据。
在一种实施方式中，肺部扫描数据为三维数据，不同组织的肺部扫描数据具有不同的CT值，示例性的，可以根据CT值的不同区分血管及淋巴等周边组织，淋巴CT值与血管相比较低，因此淋巴CT图像亮度低于血管CT图像。
步骤S104，根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件。其中，轮廓点集合可以理解为具有相同轮廓半径的一段血管中像素点的集合，选取肺部扫描数据中的内层数据，即靠近心脏的主动脉血管处的像素点为起始像素点，血管约束条件可以为支气管伴行条件，即，支气管与血管为并行关系，根据支气管延伸方向搜寻轮廓点集合，边界收敛条件可以为针对像素点的收敛条件和针对轮廓点集合的收敛条件，当像素点到达肺裂隙处时触发边界收敛条件，和/或当轮廓点集合的轮廓半径达到最小半径阈值时触发边界收敛条件，轮廓点子集合为轮廓点的点集，同一组轮廓点子集合竖直切面处的轮廓半径相同。
在一种实施方式中，肺气道与肺动脉是伴行的，且肺裂缝附近不存在大的血管，因此，当计算的边界收敛条件是轮廓点集合到达肺裂隙处的点集合时(诸如，肺动脉的分支血管达到肺实质的边界，肺静脉达到肺实质和对应肺叶的肺裂隙)时停止计算，或轮廓半径r到达最小半径阈值r(min)，此时达到边界收敛条件。
步骤S106，根据每个轮廓点子集合建立肺部血管模型。其中，肺部血管模型为错综复杂的树形结构，大概包含23级分支，管径在20um～15mm范围内变化。
在一种实施方式中，肺部血管模型中的血管具有细长、管状、树状分布的几何特征，因此，选取肺部扫描数据中内层图像的像素点作为血管起始点，由内向外建立肺部血管模型，示例性的，建立肺部血管模型的起始像素点为距离心脏最近的点，肺部血管模型中存在多跟主动脉，因此起始像素点可以为多组像素点。
本申请实施例提供的上述肺部血管模型的建立方法，可以在建立肺部血管模型时，根据预设血管约束条件有效区分血管与血管周围组织，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
本申请实施例还提供了一种确定轮廓点子集合的实施方式，具体的参见如下(1)至(2)：
(1)对于首个轮廓点子集合，从肺部扫描数据包含的像素点中确定第一级起始像素点，并提取第一级起始像素点的第一特征值，基于第一级起始像素点的第一特征值和预设血管 约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合。其中，特征值为包括第一特征值、第二特征值及像素点的坐标值，第一特征值用于表征轮廓点的法线方向，第二特征值用于表征轮廓点的切线方向，第一级起始像素点为距离心脏最近的点，肺部血管模型中存在多跟主动脉，因此第一级起始像素点可以为多组像素点，血管约束条件可以为支气管伴行条件，即，支气管与血管为并行关系。
在一种实施方式中，可以根据第一特征值和预设血管约束条件调整血管方向，由于支气管与血管是伴行的，因此在寻找轮廓点集合中像素点的过程中，利用约束条件是判断支气管的像素点B(x,y,z)是否在轮廓点P(x,y,z)的八领域范围内，实时调整寻找路径的法线e1方向，靠近并行的支气管方向寻找。
(2)对于除首个轮廓点子集合之外的轮廓点子集合，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点，并提取该起始像素点的第一特征值，基于该起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定该起始像素点对应的轮廓点子集合。其中，根据前一级起始像素点对应的轮廓点子集合确定的下一级起始像素点可以为同一轮廓点子集合截面处的任意一点。
在一种实施方式中，当一根血管发生分叉为两根血管时，由主血管的轮廓点集合确定的分叉后两个子血管的下一级起始像素点，可以为子血管与主血管轮廓点集合连接处的圆环形截面中的任一像素点。
本申请实施例还提供了一种确定除首个轮廓点子集合之外的轮廓点子集合的起始像素点的实施方式，具体的参见如下(a)至(b)：
(a)：提取前一级起始像素点对应的轮廓点子集合中每个轮廓点的第二特征值，对于每个轮廓点，基于该轮廓点的切线方向从肺部扫描数据的像素点中确定候选起始像素点，计算候选起始像素点与该轮廓点之间的距离值，如果距离值等于预设距离阈值，确定该候选起始像素点为该级起始像素点，其中，第二特征值用于表征轮廓点的切线方向。其中，距离阈值为当前轮廓点集合对下一轮廓点集合的搜寻距离，设定距离阈值为初始距离d，从当前轮廓点集合中任一轮廓点处，延切向方向延伸d搜寻符合血管CT值的像素点，并将该像素点作为下一轮廓点集合的初始轮廓点，可以将轮廓点集合看做一个整体，沿着当前圆形轮廓的e1方向延伸以初始距离d寻找下一级轮廓点集，并以此进行迭代计算。
(b)：确定前一级起始像素点对应的轮廓点子集合的轮廓半径，如果轮廓半径满足预设血管分叉条件，从前一级起始像素点对应的轮廓点子集合中确定该起始像素点。其中，血管的轮廓点集合具有细长、管状结构，轮廓半径为管状的轮廓点集合的竖直切面的半径，由于血管与其它组织的区别就是血管的区间直径不会发生太大变化，且边缘对称出现，因此，相邻轮廓点的轮廓半径按照一定范围内逐渐减小，由此排除围绕在血管周围的淋巴组 织，以及其他组织结构。
在一种实施方式中，当轮廓半径突然变大时表明血管开始分叉，此时产生多个起始像素点。
本申请实施例还提供了一种根据特征值及伴行条件确定轮廓点子集合搜寻方向的实施方式，具体的参见如下(1)至(3)：
(1)根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向；其中，第一特征值用于表征第一级起始像素点的法线方向，预设血管约束条件包括支气管伴行条件。其中，如图2所示为第一级起始像素点的法线方向，通过像素点的法线方向对确定轮廓点集合的初步搜寻方向，结合支气管伴行条件对初步搜寻方向进行进一步的限定，轮廓点集合最终的搜寻方向。
在一种实施方式中，选取特定层图像的像素点作为血管起始点，假设其值为V(x,y,z)，计算出该点的Hessian矩阵以及它的特征值e1和e2，其中e1表示血管的切向方向，e2表示血管的法线方向。
在一种实施方式中，根据|e1|>|e2|确定血管的弯曲方向。
(2)从肺部扫描数据的像素点中确定支气管像素点，并根据支气管像素点和预设血管约束条件对法线方向进行调整，确定至少一个轮廓点搜寻方向。其中，如图3所示，支气管与血管是伴行的，搜寻轮廓点集合中的轮廓点时，根据支气管的CT值，延靠近支气管方向调整搜寻方向。
(3)对于每个轮廓点搜寻方向，基于该轮廓点搜寻方向从肺部扫描数据的像素点中确定候选轮廓像素点，计算候选轮廓像素点与第一级起始像素点之间的第一距离值，如果第一距离值大于或等于预设边界阈值，确定候选轮廓像素点属于第一级起始像素点对应的轮廓点子集合。其中，边界阈值为预先设定的CT值阈值，可以根据边界阈值规定不同组织的CT值范围，从而区分血管与相邻的组织。
在一种实施方式中，延e2方向寻找起始点周围以r为扩散半径扩张的轮廓点集合P(r)，P(r)-P(r-1)>＝V(r)，从而得到的P(r)即为此点的轮廓集合点，其中，V(r)即为设定的边界阈值。
在一种实施方式中，当该级起始像素点对应的轮廓点子集合的轮廓半径达到最小半径阈值时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件；和/或，根据肺部扫描数据确定肺裂隙点集合，当该级起始像素点对应的轮廓点集合到达肺裂隙处的点集合时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件。其中，轮廓点集合可以理解为具有相同轮廓半径的一段血管中像素点的集合，
在一种实施方式中，肺裂隙点集合如图5所示，肺气道与肺动脉是伴行的，且肺裂缝附近不存在大的血管，因此，当计算的边界收敛条件是轮廓点集合到达肺裂隙处的点集合 时(诸如，肺动脉的分支血管达到肺实质的边界，肺静脉达到肺实质和对应肺叶的肺裂隙)时停止计算，或轮廓半径r到达最小半径阈值r(min)，此时达到边界收敛条件。
为便于对上述实施例提供的肺部血管模型的建立方法进行理解，本申请实施例提供了一种肺部血管模型的建立方法的应用示例，参见图5所示的另一种肺部血管模型的建立方法的流程示意图，该方法主要包括以下步骤S502至步骤S510：
步骤S502：获取目标对象的肺部扫描数据，并确定第一级起始像素点。其中，肺部扫描数据为CT设备扫描的医学数字成像和通信(Digital Imaging and Communications in Medicine，DICOM)图像数据。
在一种实施方式中，肺部扫描数据为三维数据，不同组织的肺部扫描数据具有不同的CT值，示例性的，可以根据CT值的不同区分血管及淋巴等周边组织，淋巴CT值与血管相比较低，因此淋巴CT图像亮度低于血管CT图像。
步骤S504：根据第一级起始像素点，确定轮廓点子集合。其中，肺部血管模型中的血管具有细长、管状、树状分布的几何特征，因此，选取肺部扫描数据中内层图像的像素点作为血管起始点，由内向外建立肺部血管模型，示例性的，建立肺部血管模型的起始像素点为距离心脏最近的点，肺部血管模型中存在多跟主动脉，因此起始像素点可以为多组起始像素点。
步骤S506：根据轮廓点子集合的轮廓半径、轮廓点子集合中各点的特征值及预设距离阈值确定下一级轮廓点子集合。其中，轮廓点集合可以理解为具有相同轮廓半径的一段血管中像素点的集合，
步骤S508：轮廓点子集合满足预设边界收敛条件时停止计算。其中，肺气道与肺动脉是伴行的，且肺裂缝附近不存在大的血管，因此，当计算的边界收敛条件是轮廓点集合到达肺裂隙处的点集合时(诸如，肺动脉的分支血管达到肺实质的边界，肺静脉达到肺实质和对应肺叶的肺裂隙)时停止计算，或轮廓半径r到达最小半径阈值r(min)，此时达到边界收敛条件。
步骤S510：根据每个轮廓点子集合建立肺部血管模型。其中，肺部血管模型为错综复杂的树形结构，大概包含23级分支，管径在20um～15mm范围内变化。
综上所述，本申请可以在建立肺部血管模型时，根据预设血管约束条件有效区分血管与血管周围组织，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
对于前述实施例提供的肺部血管模型的建立方法，本申请实施例提供了一种肺部血管模型的建立装置，参见图6所示的一种肺部血管模型的建立装置的结构示意图，该装置包括以下部分：
信息获取模块602，被配置成获取目标对象的肺部扫描数据；
数据计算模块604，被配置成根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件；
模型建立模块606，被配置成根据每个轮廓点子集合建立肺部血管模型。
本申请实施例提供的上述数据处理装置可以在建立肺部血管模型时，根据预设血管约束条件有效区分血管与血管周围组织，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
一种实施方式中，在进行根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条的步骤时，数据计算模块604还被配置用于：对于首个轮廓点子集合，从肺部扫描数据包含的像素点中确定第一级起始像素点，并提取第一级起始像素点的第一特征值，基于第一级起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合；对于除首个轮廓点子集合之外的轮廓点子集合，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点，并提取该起始像素点的第一特征值，基于该起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定该起始像素点对应的轮廓点子集合。
一种实施方式中，第一特征值用于表征第一级起始像素点的法线方向，在进行基于第一级起始像素点的第一特征值和预设血管约束条件，从肺部扫描数据的像素点中确定首个轮廓点子集合的步骤时，上述数据计算模块604还被配置用于：根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向；其中，预设血管约束条件包括支气管伴行条件；对于每个轮廓点搜寻方向，基于该轮廓点搜寻方向从肺部扫描数据的像素点中确定候选轮廓像素点，计算候选轮廓像素点与第一级起始像素点之间的第一距离值，如果第一距离值大于或等于预设边界阈值，确定候选轮廓像素点属于第一级起始像素点对应的轮廓点子集合。
一种实施方式中，在进行根据法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向的步骤时，上述数据计算模块604还被配置用于：从肺部扫描数据的像素点中确定支气管像素点，并根据支气管像素点和预设血管约束条件对法线方向进行调整，确定至少一个轮廓点搜寻方向。
一种实施方式中，在进行基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤时，上述数据计算模块604还被配置用于：提取前一级起始像素点对应的轮廓点子集合中每个轮廓点的第二特征值；其中，第二特征值用于表征轮廓点的切线方向；对 于每个轮廓点，基于该轮廓点的切线方向从肺部扫描数据的像素点中确定候选起始像素点；计算候选起始像素点与该轮廓点之间的距离值；如果距离值等于预设距离阈值，确定该候选起始像素点为该级起始像素点。
一种实施方式中，在进行基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤时，上述数据计算模块604还被配置用于：确定前一级起始像素点对应的轮廓点子集合的轮廓半径；如果轮廓半径满足预设血管分叉条件，从前一级起始像素点对应的轮廓点子集合中确定该起始像素点。
一种实施方式中，在进行轮廓点子集合满足预设边界收敛条件的步骤时，上述数据计算模块604还被配置用于：当该级起始像素点对应的轮廓点子集合的轮廓半径达到最小半径阈值时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件；和/或，根据肺部扫描数据确定肺裂隙点集合，当该级起始像素点对应的轮廓点集合到达肺裂隙处的点集合时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件。
本申请实施例所提供的装置，其实现原理及产生的技术效果和前述方法实施例相同，为简要描述，装置实施例部分未提及之处，可参考前述方法实施例中相应内容。
本申请实施例提供了一种服务器，具体的，该服务器包括处理器和存储装置；存储装置上存储有计算机程序，计算机程序在被所述处理器运行时执行如上所述实施方式的任一项所述的方法。
图7为本申请实施例提供的一种服务器的结构示意图，该服务器100包括：处理器70，存储器71，总线72和通信接口73，所述处理器70、通信接口73和存储器71通过总线72连接；处理器70用于执行存储器71中存储的可执行模块，例如计算机程序。
其中，存储器71可能包含高速随机存取存储器(RAM，Random Access Memory)，也可能还包括非不稳定的存储器(non-volatile memory)，例如至少一个磁盘存储器。通过至少一个通信接口73(可以是有线或者无线)实现该系统网元与至少一个其他网元之间的通信连接，可以使用互联网，广域网，本地网，城域网等。
总线72可以是ISA总线、PCI总线或EISA总线等。所述总线可以分为地址总线、数据总线、控制总线等。为便于表示，图7中仅用一个双向箭头表示，但并不表示仅有一根总线或一种类型的总线。
其中，存储器71用于存储程序，所述处理器70在接收到执行指令后，执行所述程序，前述本申请实施例任一实施例揭示的流过程定义的装置所执行的方法可以应用于处理器70中，或者由处理器70实现。
处理器70可能是一种集成电路芯片，具有信号的处理能力。在实现过程中，上述方法的各步骤可以通过处理器70中的硬件的集成逻辑电路或者软件形式的指令完成。上述的处 理器70可以是通用处理器，包括中央处理器(Central Processing Unit，简称CPU)、网络处理器(Network Processor，简称NP)等；还可以是数字信号处理器(Digital Signal Processing，简称DSP)、专用集成电路(Application Specific Integrated Circuit，简称ASIC)、现成可编程门阵列(Field-Programmable Gate Array，简称FPGA)或者其他可编程逻辑器件、分立门或者晶体管逻辑器件、分立硬件组件。可以实现或者执行本申请实施例中的公开的各方法、步骤及逻辑框图。通用处理器可以是微处理器或者该处理器也可以是任何常规的处理器等。结合本申请实施例所公开的方法的步骤可以直接体现为硬件译码处理器执行完成，或者用译码处理器中的硬件及软件模块组合执行完成。软件模块可以位于随机存储器，闪存、只读存储器，可编程只读存储器或者电可擦写可编程存储器、寄存器等本领域成熟的存储介质中。该存储介质位于存储器71，处理器70读取存储器71中的信息，结合其硬件完成上述方法的步骤。
本申请实施例所提供的可读存储介质的计算机程序产品，包括存储了程序代码的计算机可读存储介质，所述程序代码包括的指令可用于执行前面方法实施例中所述的方法，具体实现可参见前述方法实施例，在此不再赘述。
所述功能如果以软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。基于这样的理解，本申请的技术方案本质上或者说对相关技术做出贡献的部分或者该技术方案的部分可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(ROM，Read-Only Memory)、随机存取存储器(RAM，Random Access Memory)、磁碟或者光盘等各种可以存储程序代码的介质。
最后应说明的是：以上所述实施例，仅为本申请的具体实施方式，用以说明本申请的技术方案，而非对其限制，本申请的保护范围并不局限于此，尽管参照前述实施例对本申请进行了详细的说明，本领域的普通技术人员应当理解：任何熟悉本技术领域的技术人员在本申请揭露的技术范围内，其依然可以对前述实施例所记载的技术方案进行修改或可轻易想到变化，或者对其中部分技术特征进行等同替换；而这些修改、变化或者替换，并不使相应技术方案的本质脱离本申请实施例技术方案的精神和范围，都应涵盖在本申请的保护范围之内。因此，本申请的保护范围应所述以权利要求的保护范围为准。
工业实用性
本申请提供了一种肺部血管模型的建立方法、装置及服务器，其中该方法用于获取目 标对象的肺部扫描数据，并根据肺部扫描数据中起始像素点和预设血管约束条件，从肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至轮廓点子集合满足预设边界收敛条件，根据每个轮廓点子集合建立肺部血管模型。该方法可以在建立肺部血管模型时，根据预设血管约束条件有效区分血管与血管周围组织，可以显著提升肺部血管模型建立的精确度，并降低与周围组织发生的粘连现象，从而减少肺部血管模型的模拟难度。
此外，可以理解的是，本申请的肺部血管模型的建立方法、装置及服务器是可以重现的，并且可以用在多种工业应用中。例如，本申请的肺部血管模型的建立方法、装置及服务器可以用于医疗技术领域。

Claims (19)

  1. 一种肺部血管模型的建立方法，其特征在于，包括：
    获取目标对象的肺部扫描数据；
    根据所述肺部扫描数据中起始像素点和预设血管约束条件，从所述肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至所述轮廓点子集合满足预设边界收敛条件；
    根据每个所述轮廓点子集合建立肺部血管模型。
  2. 根据权利要求1所述的方法，其特征在于，所述根据所述肺部扫描数据中起始像素点和预设血管约束条件，从所述肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至所述轮廓点子集合满足预设边界收敛条件的步骤，包括：
    对于首个轮廓点子集合，从所述肺部扫描数据包含的像素点中确定第一级起始像素点，并提取所述第一级起始像素点的第一特征值，基于所述第一级起始像素点的第一特征值和预设血管约束条件，从所述肺部扫描数据的像素点中确定首个轮廓点子集合；
    对于除所述首个轮廓点子集合之外的轮廓点子集合，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点，并提取该起始像素点的第一特征值，基于该起始像素点的第一特征值和所述预设血管约束条件，从所述肺部扫描数据的像素点中确定该起始像素点对应的轮廓点子集合。
  3. 根据权利要求2所述的方法，其特征在于，其中，所述第一特征值用于表征所述第一级起始像素点的法线方向，所述基于所述第一级起始像素点的第一特征值和预设血管约束条件，从所述肺部扫描数据的像素点中确定首个轮廓点子集合的步骤，包括：
    根据所述法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向；其中，所述预设血管约束条件包括支气管伴行条件；
    对于每个所述轮廓点搜寻方向，基于该轮廓点搜寻方向从所述肺部扫描数据的像素点中确定候选轮廓像素点，计算所述候选轮廓像素点与所述第一级起始像素点之间的第一距离值，如果所述第一距离值大于或等于预设边界阈值，确定所述候选轮廓像素点属于所述第一级起始像素点对应的轮廓点子集合。
  4. 根据权利要求3所述的方法，其特征在于，所述根据所述法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向的步骤，包括：
    从所述肺部扫描数据的像素点中确定支气管像素点，并根据所述支气管像素点和所述预设血管约束条件对所述法线方向进行调整，确定至少一个轮廓点搜寻方向。
  5. 根据权利要求2至4中任一项所述的方法，其特征在于，所述基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤，包括：
    提取前一级起始像素点对应的轮廓点子集合中每个轮廓点的第二特征值；其中，所述第二特征值用于表征所述轮廓点的切线方向；
    对于每个所述轮廓点，基于该轮廓点的所述切线方向从所述肺部扫描数据的像素点中确定候选起始像素点；
    计算所述候选起始像素点与该轮廓点之间的距离值；
    如果所述距离值等于预设距离阈值，确定该候选起始像素点为该级起始像素点。
  6. 根据权利要求5所述的方法，其特征在于，所述距离阈值为当前轮廓点集合对下一轮廓点集合的搜寻距离。
  7. 根据权利要求2至6中任一项所述的方法，其特征在于，所述基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤，还包括：
    确定前一级起始像素点对应的轮廓点子集合的轮廓半径；
    如果所述轮廓半径满足预设血管分叉条件，从所述前一级起始像素点对应的轮廓点子集合中确定该起始像素点。
  8. 根据权利要求7所述的方法，其特征在于，所述轮廓点子集合满足预设边界收敛条件的步骤，包括：
    当该级起始像素点对应的轮廓点子集合的轮廓半径达到最小半径阈值时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件；
    和/或，根据所述肺部扫描数据确定肺裂隙点集合，当该级起始像素点对应的轮廓点集合到达所述肺裂隙处的点集合时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件。
  9. 根据权利要求1至8中任一项所述的方法，其特征在于，所述肺部扫描数据为CT设备扫描的医学数字成像和通信图像数据；或者所述肺部扫描数据为三维数据，不同组织的肺部扫描数据具有不同的CT值。
  10. 一种肺部血管模型的建立方法，其特征在于，包括：
    获取目标对象的肺部扫描数据，并确定第一级起始像素点；
    根据所述第一级起始像素点，确定轮廓点子集合；
    根据所述轮廓点子集合的轮廓半径、轮廓点子集合中各点的特征值及预设距离阈值确定下一级轮廓点子集合；
    所述轮廓点子集合满足预设边界收敛条件时停止计算；
    根据每个轮廓点子集合建立肺部血管模型。
  11. 一种肺部血管模型的建立装置，其特征在于，包括：
    信息获取模块，被配置成获取目标对象的肺部扫描数据；
    数据计算模块，被配置成根据所述肺部扫描数据中起始像素点和预设血管约束条件，从所述肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至所述轮廓点子集合满足预设边界收敛条件；
    模型建立模块，被配置成根据每个所述轮廓点子集合建立肺部血管模型。
  12. 根据权利要求11所述的装置，其特征在于，在进行根据所述肺部扫描数据中起始像素点和预设血管约束条件，从所述肺部扫描数据的像素点中确定至少一个轮廓点子集合，直至所述轮廓点子集合满足预设边界收敛条的步骤时，所述数据计算模块被配置用于：
    对于首个轮廓点子集合，从所述肺部扫描数据包含的像素点中确定第一级起始像素点，并提取所述第一级起始像素点的第一特征值，基于所述第一级起始像素点的第一特征值和预设血管约束条件，从所述肺部扫描数据的像素点中确定首个轮廓点子集合；
    对于除所述首个轮廓点子集合之外的轮廓点子集合，基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点，并提取该起始像素点的第一特征值，基于该起始像素点的第一特征值和所述预设血管约束条件，从所述肺部扫描数据的像素点中确定该起始像素点对应的轮廓点子集合。
  13. 根据权利要求12所述的装置，其特征在于，所述第一特征值用于表征所述第一级起始像素点的法线方向，在进行基于所述第一级起始像素点的第一特征值和预设血管约束条件，从所述肺部扫描数据的像素点中确定首个轮廓点子集合的步骤时，所述数据计算模块被配置用于：
    根据所述法线方向和预设血管约束条件，确定至少一个轮廓点搜寻方向；其中，所述预设血管约束条件包括支气管伴行条件；
    对于每个所述轮廓点搜寻方向，基于该轮廓点搜寻方向从所述肺部扫描数据的像素点中确定候选轮廓像素点，计算所述候选轮廓像素点与所述第一级起始像素点之间的第一距离值，如果所述第一距离值大于或等于预设边界阈值，确定所述候选轮廓像素点属于所述第一级起始像素点对应的轮廓点子集合。
  14. 根据权利要求13所述的装置，其特征在于，在进行根据所述法线方向和所述预设血管约束条件，确定至少一个轮廓点搜寻方向的步骤时，所述数据计算模块被配置用于：
    从所述肺部扫描数据的像素点中确定支气管像素点，并根据所述支气管像素点和所述预设血管约束条件对所述法线方向进行调整，确定至少一个轮廓点搜寻方向。
  15. 根据权利要求12至14中任一项所述的装置，其特征在于，在进行基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤时，所述数据计算模块被配置用于：
    提取前一级起始像素点对应的轮廓点子集合中每个轮廓点的第二特征值；其中，所述 第二特征值用于表征所述轮廓点的切线方向；
    对于每个所述轮廓点，基于该轮廓点的所述切线方向从所述肺部扫描数据的像素点中确定候选起始像素点；
    计算所述候选起始像素点与该轮廓点之间的距离值；
    如果所述距离值等于预设距离阈值，确定该候选起始像素点为该级起始像素点。
  16. 根据权利要求12至15中任一项所述的装置，其特征在于，在进行基于前一级起始像素点对应的轮廓点子集合确定该级起始像素点的步骤时，所述数据计算模块被配置用于：
    确定前一级起始像素点对应的轮廓点子集合的轮廓半径；
    如果所述轮廓半径满足预设血管分叉条件，从所述前一级起始像素点对应的轮廓点子集合中确定该起始像素点。
  17. 根据权利要求16所述的装置，其特征在于，在进行所述轮廓点子集合满足预设边界收敛条件的步骤时，所述数据计算模块被配置用于：
    当该级起始像素点对应的轮廓点子集合的轮廓半径达到最小半径阈值时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件；
    和/或，根据所述肺部扫描数据确定肺裂隙点集合，当该级起始像素点对应的轮廓点集合到达所述肺裂隙处的点集合时，确定该级起始像素点对应的轮廓点子集合满足预设边界收敛条件。
  18. 一种服务器，其特征在于，包括处理器和存储器，所述存储器存储有能够被所述处理器执行的计算机可执行指令，所述处理器执行所述计算机可执行指令以实现权利要求1至10中任一项所述的方法。
  19. 一种计算机可读存储介质，其特征在于，所述计算机可读存储介质存储有计算机可执行指令，所述计算机可执行指令在被处理器调用和执行时，计算机可执行指令促使处理器实现权利要求1至10中任一项所述的方法。
PCT/CN2023/099704 2022-06-28 2023-06-12 肺部血管模型的建立方法、装置及服务器 WO2024001747A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210746710.3 2022-06-28
CN202210746710.3A CN115049807B (zh) 2022-06-28 2022-06-28 肺部血管模型的建立方法、装置及服务器

Publications (1)

Publication Number Publication Date
WO2024001747A1 true WO2024001747A1 (zh) 2024-01-04

Family

ID=83162770

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2023/099704 WO2024001747A1 (zh) 2022-06-28 2023-06-12 肺部血管模型的建立方法、装置及服务器

Country Status (2)

Country Link
CN (1) CN115049807B (zh)
WO (1) WO2024001747A1 (zh)

Families Citing this family (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115049807B (zh) * 2022-06-28 2023-05-09 上海市胸科医院 肺部血管模型的建立方法、装置及服务器
CN116205920B (zh) * 2023-05-05 2023-07-18 天津医科大学总医院 基于肺扫描数据的重点区域检测模型生成方法及系统

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6842638B1 (en) * 2001-11-13 2005-01-11 Koninklijke Philips Electronics N.V. Angiography method and apparatus
US20070230809A1 (en) * 2006-03-31 2007-10-04 Canon Kabushiki Kaisha Image processing method and image processing apparatus
US20080317314A1 (en) * 2007-06-20 2008-12-25 Schwartz Lawrence H Automated Determination of Lymph Nodes in Scanned Images
CN110060254A (zh) * 2019-05-29 2019-07-26 深圳华声医疗技术股份有限公司 交互式血管测量方法、装置、计算机可读存储介质及系统
CN110717888A (zh) * 2019-09-06 2020-01-21 天津大学 一种血管内光学相干层析成像血管壁内轮廓自动识别方法
CN112508880A (zh) * 2020-11-23 2021-03-16 西安科锐盛创新科技有限公司 颅内血管图像配准方法、电子设备和计算机可读存储介质
CN114299357A (zh) * 2021-12-18 2022-04-08 深圳先进技术研究院 一种用于血管壁影像分割的自定义卷积路径方法
CN115049807A (zh) * 2022-06-28 2022-09-13 上海市胸科医院 肺部血管模型的建立方法、装置及服务器

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106296664B (zh) * 2016-07-30 2019-10-08 上海联影医疗科技有限公司 血管提取方法
CN107045721B (zh) * 2016-10-24 2023-01-31 东北大学 一种从胸部ct图像中提取肺血管的方法及装置
CN107230204B (zh) * 2017-05-24 2019-11-22 东北大学 一种从胸部ct图像中提取肺叶的方法及装置
US20220092791A1 (en) * 2018-04-12 2022-03-24 Veran Medical Technologies, Inc. Methods for the Segmentation of Lungs, Lung Vasculature and Lung Lobes from CT Data and Clinical Applications
CN109146854B (zh) * 2018-08-01 2021-10-01 东北大学 一种肺结节与肺部血管伴生关系的分析方法
CN112711831B (zh) * 2020-12-07 2022-10-28 上海联影医疗科技股份有限公司 血管仿真分析方法、设备、装置、计算机设备和存储介质

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6842638B1 (en) * 2001-11-13 2005-01-11 Koninklijke Philips Electronics N.V. Angiography method and apparatus
US20070230809A1 (en) * 2006-03-31 2007-10-04 Canon Kabushiki Kaisha Image processing method and image processing apparatus
US20080317314A1 (en) * 2007-06-20 2008-12-25 Schwartz Lawrence H Automated Determination of Lymph Nodes in Scanned Images
CN110060254A (zh) * 2019-05-29 2019-07-26 深圳华声医疗技术股份有限公司 交互式血管测量方法、装置、计算机可读存储介质及系统
CN110717888A (zh) * 2019-09-06 2020-01-21 天津大学 一种血管内光学相干层析成像血管壁内轮廓自动识别方法
CN112508880A (zh) * 2020-11-23 2021-03-16 西安科锐盛创新科技有限公司 颅内血管图像配准方法、电子设备和计算机可读存储介质
CN114299357A (zh) * 2021-12-18 2022-04-08 深圳先进技术研究院 一种用于血管壁影像分割的自定义卷积路径方法
CN115049807A (zh) * 2022-06-28 2022-09-13 上海市胸科医院 肺部血管模型的建立方法、装置及服务器

Also Published As

Publication number Publication date
CN115049807B (zh) 2023-05-09
CN115049807A (zh) 2022-09-13

Similar Documents

Publication Publication Date Title
WO2024001747A1 (zh) 肺部血管模型的建立方法、装置及服务器
Esneault et al. Liver vessels segmentation using a hybrid geometrical moments/graph cuts method
US9471989B2 (en) Vascular anatomy modeling derived from 3-dimensional medical image processing
US8682074B2 (en) Method for checking the segmentation of a structure in image data
WO2021208739A1 (zh) 眼底彩照图像血管评估方法、装置、计算机设备和介质
US11783488B2 (en) Method and device of extracting label in medical image
JP2006081906A (ja) 器官内の塞栓症により冒されている領域を分析するための自動化された方法および装置ならびに器官内の塞栓症により冒されている領域を分析するためのコンピュータプログラムコードが記憶されているコンピュータ読み取り可能媒体
EP1761897A1 (en) Displaying a tracheobronchial tree
JP2007061607A (ja) 第１のオブジェクト及び１つ以上の他のオブジェクトを含む画像を処理するための方法及び画像データからの画像を処理するためのシステム
US20220092788A1 (en) Segmentation method for pneumonia sign, medium, and electronic device
CN114419077B (zh) 一种基于ct图像的肺支气管自动分割方法及系统
Ma et al. A coronary artery segmentation method based on region growing with variable sector search area
Ukil et al. Smoothing lung segmentation surfaces in 3D X-ray CT images using anatomic guidance
TWI824829B (zh) 造影影像判定方法及造影影像判定裝置
CN114119645B (zh) 一种图像分割质量的确定方法、系统、设备及介质
CN112258533B (zh) 一种超声图像中小脑蚓部的分割方法
CN113706541B (zh) 一种图像处理方法及装置
Duan et al. Two-pass region growing combined morphology algorithm for segmenting airway tree from CT chest scans
CN114170258A (zh) 图像分割方法、装置、电子设备及存储介质
CN107689043B (zh) 一种血管段终端节点与分支节点的获取方法
Tan et al. A novel method for automated segmentation of airway tree
CN113192099A (zh) 一种组织提取方法、装置、设备和介质
JP5403431B2 (ja) 断層画像処理方法及び装置
Walczak et al. Segmenting lungs from whole-body CT scans
Zhou et al. A new segment method for pulmonary artery and vein

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 23829930

Country of ref document: EP

Kind code of ref document: A1