Device and method to reconstruct three dimensional shape of blood vessel

Abstract

To provide a device and a method to reconstruct a three dimensional shape of a blood vessel.SOLUTION: An electronic device to reconstruct a three dimensional shape of a blood vessel, comprises: an image acquisition unit configured to acquire medical image groups captured at a plurality of positions on the basis of a C arm imaging device; and a processor configured to generate tree structure data of three-dimensional candidate point groups corresponding to point groups along center lines of blood vessel regions of a reference image and a sub-image of the medical image groups, determine a three-dimensional path on the basis of the tree structure data, determine a three-dimensional diameter for each of the points on the three-dimensional path, and reconstruct the three-dimensional shape of the blood vessels corresponding to the three-dimensional path on the basis of the determined three-dimensional diameter.SELECTED DRAWING: Figure 4

Description

[Technical field]

[0001] The present disclosure relates to three-dimensional shape reconstruction of blood vessels. [Background technology]

[0002] Angiography is a diagnostic procedure that uses X-rays to visualize blood vessels and their condition, and is a useful tool for examining vascular diseases. In angiography, which is useful for diagnosing vascular diseases, accurate understanding of the length and shape of blood vessels is important for determining the severity of the disease and selecting an appropriate treatment method. In particular, in relation to selecting a treatment method, it is required to accurately calculate an index related to blood flow, and for this purpose, it is necessary to quickly and accurately restore the three-dimensional shape of blood vessels.

[0003] The above-mentioned background art is owned or acquired by the inventors in the process of deriving the contents of the disclosure of this application, and is not necessarily publicly known art that was disclosed to the general public prior to the filing of this application. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Korean Patent Publication No. 10-2024-0045980 Summary of the Invention [Problem to be solved by the invention]

[0005] In one embodiment, the 3D shape reconstruction of a blood vessel can determine the 3D path of the blood vessel using tree-structured data of nodes indicating 3D candidate points corresponding to points along the center lines of two blood vessel images.

[0006] In one embodiment, the 3D shape reconstruction of a blood vessel can prevent reconstruction errors in sections with severe curvature, which occur in active contours due to the influence of force vectors that extend in both directions rather than a force vector that points inward.

[0007] In one embodiment, the 3D shape reconstruction of a blood vessel can prevent reconstruction errors that may be induced by multiple overlapping matching in sections where the epipolar line and the center line of the vessel are parallel in a path finding algorithm (e.g., dynamic programming).

[0008] However, the technical issues are not limited to those mentioned above, and other technical issues may exist. [Means for solving the problem]

[0009] A method for restoring a three-dimensional shape of a blood vessel, which is embodied by a processor according to one embodiment, includes the steps of acquiring a group of medical images taken from multiple positions, generating tree-structured data of a group of three-dimensional candidate points corresponding to a group of points along center lines of a group of blood vessel regions in a reference image and a sub-image from among the group of medical images, determining a three-dimensional route based on the tree-structured data, determining a three-dimensional diameter for each point on the three-dimensional route, and restoring a three-dimensional blood vessel shape corresponding to the three-dimensional route based on the determined three-dimensional diameter.

[0010] An electronic device according to one embodiment may include an image acquisition unit that acquires a group of medical images taken from multiple positions, and a processor that generates tree-structured data of a group of three-dimensional candidate points corresponding to a group of points along center lines of a group of vascular regions in a reference image and a sub-image among the group of medical images, determines a three-dimensional route based on the tree-structured data, determines three-dimensional diameters for each point on the three-dimensional route, and reconstructs a three-dimensional vascular shape corresponding to the three-dimensional route based on the determined three-dimensional diameters. Effect of the Invention

[0011] The three-dimensional shape reconstruction of a blood vessel according to an embodiment can accurately reconstruct the three-dimensional shape of the blood vessel even at a position where the blood vessel is severely curved.

[0012] The 3D shape reconstruction of a blood vessel according to an embodiment can more quickly reconstruct the 3D shape of a blood vessel.

[0013] The 3D shape reconstruction of a blood vessel according to an embodiment can accurately reconstruct the 3D shape of the blood vessel even in a section in which the epipolar line and the center line of the blood vessel are parallel. [Brief description of the drawings]

[0014] [Figure 1] 1 is a diagram illustrating a structure of a medical electronic device according to an embodiment. [Diagram 2] 11A and 11B are diagrams illustrating a change in the projection angle of radiation due to rotation of a C-arm included in an electronic device according to an embodiment. [Diagram 3] 1 is a diagram illustrating a process in which an electronic device according to an embodiment calculates a straight-line distance from a radiation source to a target blood vessel. [Figure 4] 1 is a flowchart showing a method for restoring a three-dimensional shape of a blood vessel according to an embodiment. [Diagram 5] FIG. 11 illustrates the generation of three-dimensional location candidates according to one embodiment. [Figure 6] FIG. 11 illustrates the generation of three-dimensional location candidates according to one embodiment. [Figure 7] FIG. 11 illustrates the generation of three-dimensional location candidates according to one embodiment. [Figure 8] FIG. 13 illustrates filtering of three-dimensional candidate points according to one embodiment. [Figure 9] FIG. 11 is a diagram illustrating generation of tree structure data according to an embodiment. [Figure 10] FIG. 11 is a diagram illustrating generation of tree structure data according to an embodiment. [Figure 11] FIG. 11 is a diagram illustrating generation of tree structure data according to an embodiment. [Figure 12]FIG. 13 is a diagram illustrating edge weights in tree structure data according to an embodiment. [Figure 13] 11A and 11B are diagrams illustrating an example of a three-dimensional path selected using edge weight values ​​of tree structure data according to an embodiment. [Figure 14] FIG. 13 is a diagram illustrating matching between a 3D point cloud and a 2D point cloud by pair matching according to one embodiment. [Figure 15] FIG. 13 is a diagram illustrating matching between a 3D point cloud and a 2D point cloud by pair matching according to one embodiment. [Figure 16] FIG. 1 is a diagram illustrating the reconstruction of a three-dimensional shape according to an embodiment. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0015] <Summary of the Invention> According to one embodiment, the step of generating tree structure data may include a step of determining a medical image selected from a plurality of medical images based on at least one of centerline length or lesion information as a reference image.

[0016] According to one embodiment, the step of generating tree structure data may include a step of determining, for each of a group of reference points along a center line of a vascular region segmented from the reference image, a group of three-dimensional candidate points corresponding to a group of sub-points along the center line of the vascular region segmented from the sub-image.

[0017] According to one embodiment, the step of determining the group of three-dimensional candidate points may include the steps of generating a group of nodes corresponding to the group of three-dimensional candidate points based on the depth according to the arrangement order of the group of reference points along the center line of the vascular region segmented from the reference image, and connecting the generated group of nodes via edges.

[0018] According to one embodiment, the step of connecting the generated node groups via edges may include a step of connecting the node groups of adjacent depths according to whether the distance between the three-dimensional candidate points corresponding to the node groups of adjacent depths among the generated node groups is within a threshold distance.

[0019] According to one embodiment, the step of connecting the generated nodes via edges may include determining an edge weight value for the edge based on the distance between the 3D candidate points corresponding to both nodes connected to the edge.

[0020] According to one embodiment, the step of connecting the generated node group via edges may include a step of determining an edge weight value for the edge based on the distance between an epipolar line for a reference point corresponding to one of the 3D candidate points of both nodes connected to the edge and an epipolar line for a sub-point.

[0021] According to one embodiment, the step of determining a three-dimensional route may include a step of determining the three-dimensional route by selecting a group of intermediate points from a group of three-dimensional candidate points from a start point to an end point of the reference center line.

[0022] According to one embodiment, the step of determining the three-dimensional route may include a step of determining the three-dimensional route based on edge weight values ​​of nodes corresponding to the three-dimensional candidate points from the tree structure data.

[0023] According to one embodiment, the step of determining a three-dimensional route may include a step of determining a three-dimensional route from a group of potential routes based on a group of three-dimensional candidate points based on a cost calculated using edge weight values ​​of the tree structure data.

[0024] According to one embodiment, determining the three-dimensional diameter for each point on the three-dimensional path can include identifying two-dimensional points from the reference image and the sub-image that correspond to points on the three-dimensional path, and calculating the three-dimensional diameter based on at least one of the widths of the identified two-dimensional points.

[0025] According to one embodiment, calculating the three-dimensional diameter may include calculating an average width of the identified two-dimensional points as the three-dimensional diameter.

[0026] According to one embodiment, the step of restoring the 3D vascular shape may include the steps of generating a group of vertices along a circumference corresponding to a 3D diameter for each point on the 3D path, and restoring the 3D vascular shape using a mesh based on the generated group of vertices.

[0027] A computer readable computer program may be provided for carrying out the method according to an embodiment on a computer.

[0028] <Detailed Description of the Invention> Specific structural or functional descriptions in the embodiments are merely examples and may be modified and embodied in various forms. Therefore, the actual embodied forms are not limited to the specific embodiments disclosed, and the scope of this specification includes modifications, equivalents, or alternatives included in the technical ideas described in the embodiments.

[0029] Although the terms "first," "second," etc. are used to describe various components, such terms are only used to distinguish one component from another component. For example, a first component can be referred to as a second component, and similarly, a second component can be referred to as a first component.

[0030] When an element is described as being "coupled" to another element, it should be understood that the elements may be directly coupled or connected to each other, but there may be other elements in between.

[0031] The singular includes the plural unless the context specifically dictates otherwise. As used herein, terms such as "comprise" and "have" specify the presence of stated features, numbers, steps, operations, components, parts, or combinations thereof, but do not exclude the presence or addition of one or more other features, numbers, steps, operations, components, parts, or combinations thereof.

[0032] Unless otherwise defined, all terms used herein, including technical or scientific terms, generally have the same meaning as would be understood by a person of ordinary skill in the art. Terms commonly used and predefined should be interpreted to be consistent with the meaning they have in the context of the relevant art, and should not be interpreted as having an ideal or overly formal meaning unless otherwise specified in this specification.

[0033] Hereinafter, embodiments will be described in detail with reference to the accompanying drawings. In the description with reference to the accompanying drawings, the same components are given the same reference numerals, and duplicated descriptions thereof will be omitted.

[0034] FIG. 1 is a diagram illustrating a structure of a medical electronic device according to an embodiment.

[0035] A medical electronic device 100 (hereinafter referred to as "electronic device") according to an embodiment can restore the three-dimensional shape of a target from a medical image. A medical image of a subject's blood vessels can be called a blood vessel image. The electronic device 100 can restore the three-dimensional shape of the blood vessel, and can also be called a blood vessel three-dimensional shape restoration device. The three-dimensional shape of the blood vessel restored by the electronic device 100 is used to calculate an index related to blood flow (e.g., quantitative flow ratio (QFR) and fractional flow reserve (FFR)).

[0036] The electronic device 100 may include an image acquisition unit and a processor 110. In this specification, the image acquisition unit is exemplified as an imaging device, but is not limited thereto, and may be a device that receives medical images (e.g., blood vessel images) from an external imaging device based on wired and / or wireless communication.

[0037] The image acquisition unit may include a device that irradiates the blood vessels of the subject with radiation to capture a blood vessel image. Examples of types of blood vessels include the left main coronary artery (LM), the left anterior descending artery (LAD), the left circumflex artery (LCX), and the right coronary artery (RCA). The image acquisition unit may capture images of blood vessels using coronary angiography. For example, the image acquisition unit may include a main body unit 11, a C-arm 12, a radiation irradiation unit 13, and a radiation detection unit 14. An imaging device having the C-arm 12 may also be called a C-arm imaging device.

[0038] The C-arm 12 may have a partially open C-shaped arc shape. For example, when the C-arm 12 is vertically installed with the bottom surface (floor) of the electronic device 100 as a reference, the C-arm 12 may have a shape that is parallel to the bottom surface and symmetrical with respect to a plane including an isocenter 111 as a reference. The isocenter 111 is a part or point that is the center of the linear velocity of radiation emitted from various positions regardless of the rotation of the C-arm, and can be defined as a point where a first rotation axis and a second rotation axis of the C-arm intersect. The isocenter 111 may refer to the center of the rotation orbit of the radiation irradiation unit 13 (or a radiation source (not shown)) generated by the rotation of the C-arm 12. The C-arm 12 may have a shape that is open toward the isocenter 111.

[0039] The main body 11 may be connected to the C-arm 12. The main body 11 may be mechanically coupled to the C-arm 12. The C-arm 12 may rotate with respect to the main body 11. The C-arm 12 may rotate within the YZ plane with respect to the main body 11. For example, a protrusion included in the main body 11 and a moving guide included in the C-arm 12 may be coupled to each other, and the C-arm 12 may rotate within the YZ plane along the moving guide with respect to a rotation axis that is parallel to the X-axis and passes through the isocenter 111. Also, the C-arm 12 may rotate within the XZ plane with respect to the main body 11. For example, with the point where the main body 11 and the C-arm 12 abut fixed, the C-arm 12 may rotate within the XZ plane with respect to a rotation axis that is parallel to the Y-axis and passes through the isocenter 111.

[0040] The radiation irradiation unit 13 and the radiation detection unit 14 can be disposed on the inner surface of the C-arm 12 so as to face each other across the isocenter 111. Each of the radiation irradiation unit 13 and the radiation detection unit 14 can be connected to the C-arm 12. The radiation irradiation unit 13 can include one or more radiation sources, and can emit radiation to the subject using the one or more radiation sources. The radiation detection unit 14 can include a radiation detection sensor, and can detect radiation emitted from the radiation irradiation unit 13 and transmitted through the blood vessels of the subject using the radiation detection sensor. Since the radiation irradiation unit 13 and the radiation detection unit 14 are connected via the C-arm 12 and rotated as an integral body, when the radiation detection unit 14 moves, the radiation irradiation unit 13 can be positioned on the opposite side of the radiation detection unit 14 with respect to the isocenter 111. As will be described later, the electronic device 100 can capture medical images (e.g., blood vessel images) from a plurality of imaging positions while changing the position (e.g., imaging position) of the radiation detection unit 14.

[0041] A subject can lie on the table 15. More specifically, the table 15 can include a table top 15-1 on which the subject can lie, and a table support 15-2 that supports the table top 15-1. The table support 15-2 can be fixed to a bottom surface. For example, the table top 15-1 can be movably coupled to the table support 15-2. The electronic device 100 can further include an actuator (not shown) that changes the position of the table top 15-1 with respect to the table support 15-2. The electronic device 100 can move the table top 15-1 with respect to the table support 15-2 by the actuator (not shown). The actuator (not shown) can include a motor (e.g., an electric motor) and a power transmission structure.

[0042] The table top 15-1 can move left and right and up and down with the table support 15-2 as a reference. For example, the electronic device 100 can move the table top 15-1 left and right on a plane (e.g., XY plane) parallel to the surface of the table. The surface of the table can refer to the upper surface of the table top 15-1. The electronic device 100 can also move the table top 15-1 up and down on an axis (e.g., Z axis) perpendicular to the plane (e.g., XY plane) parallel to the surface of the table. In the following description, moving the table 15 can refer to moving the table top 15-1 with the table support 15-2 as a reference. For example, the electronic device 100 can move the table 15 to accurately irradiate the target blood vessel of the subject with radiation.

[0043] The electronic device 100 can control the C-arm 12 so that the C-arm 12 rotates around the subject positioned on the table 15. After fixing the C-arm 12, the electronic device 100 can irradiate a target blood vessel of the subject on the table 15 with radiation using a radiation source of the radiation irradiation unit 13 connected to the C-arm 12. The radiation irradiated to the blood vessel of the subject can be X-rays.

[0044] In one embodiment, a subject 120 may be positioned on a table top 15-1 of the electronic device 100. The electronic device 100 may use radiation emitted from a radiation source 131 to generate a vascular image 130 of a target blood vessel 320 of the subject 120.

[0045] The processor 110 of the electronic device 100 can derive quantitative analysis results, such as the diameter of the target blood vessel 320 and the length of the lesion, from the generated blood vessel image 130. The processor 110 can convert the distances measured in the blood vessel image 130 into real-world physical distances, such as μm, mm, or cm, to derive quantitative analysis results related to the target blood vessel 320 from the blood vessel image 130. For example, the processor 110 can utilize a calibration factor to convert the distances measured in the blood vessel image 130 into physical distances.

[0046] Specifically, the physical distance can be calculated by multiplying the correction coefficient by the number of pixels corresponding to the distance measured in the blood vessel image 130. For example, the physical distance can be expressed by the following formula 1.

[0047]

number

[0048] Referring to the formula 1, the correction coefficient can indicate a physical distance corresponding to the length (e.g., the horizontal length or the vertical length) of one pixel constituting the blood vessel image 130. In order to accurately calculate the physical distance, it is necessary to accurately calculate the correction coefficient.

[0049] The correction factor may refer to the spatial relationship between an object in the vascular image and an object in reality. The correction factor may be calculated by dividing the distance from the radiation source to the target object (SOD: source to object distance) by the distance from the radiation source to the image receiver (SID: source to image receptor distance), and multiplying the result by the imager pixel spacing. The imager pixel spacing may refer to the length (e.g., the horizontal length or vertical length) of one pixel constituting the vascular image. That is, the correction factor may be expressed as the following formula 2.

[0050]

number

[0051] In the formula of Equation 2, SOD may represent a linear distance from a source (e.g., a radiation source) to an object to be imaged, and SID may represent a linear distance from the source to an image receiver. The source-to-image receiver distance (SID) and imager pixel spacing may be extracted from DICOM (Digital Imaging and Communications in Medicine) metadata. DICOM may refer to a standard for storing and transmitting data associated with images generated from medical electronic devices.

[0052] In one embodiment, the electronic device 100 can irradiate a target blood vessel 320 of the subject 120 located on the table top 15-1 using a radiation source 131 included in the radiation irradiation unit 13. The electronic device 100 can calculate a target distance from the radiation source 131 to the target blood vessel 320 based on a first vertical distance from the isocenter 111 to the table 15 and a second vertical distance from the target blood vessel 320 to the table 15. Here, the first vertical distance from the isocenter 111 to the table 15 can refer to the vertical distance from the isocenter 111 to the top surface of the table top 15-1. Similarly, the second vertical distance from the target blood vessel 320 to the table 15 can refer to the vertical distance from the target blood vessel 320 to the top surface of the table top 15-1.

[0053] More specifically, the electronic device 100 can adjust the position of the target blood vessel 320 by moving the table 15 before irradiating the target blood vessel 320 with radiation using the radiation source 131. The electronic device 100 can move the table 15 so that the radiation source 131, the target blood vessel 320, and the isocenter 111 are aligned in a straight line. In one embodiment, the electronic device 100 can position the radiation source 131, the target blood vessel 320, and the isocenter 111 in a straight line by moving the table top 15-1 only in the left-right direction on the XY plane. In another embodiment, the electronic device 100 can position the radiation source 131, the target blood vessel 320, and the isocenter 111 in a straight line by moving the table top 15-1 all in the left-right direction on the XY plane and in the up-down direction on the Z axis.

[0054] In one embodiment, if the position of the radiation source 131 changes due to rotation of the C-arm 12, the table top 15-1 can be moved along with it so that the radiation source 131, the target vessel 320 and the isocenter 111 are aligned.

[0055] In an embodiment, the electronic device 100 can position the isocenter 111 on the same line as the radiation source 131 and the target blood vessel 320 by moving the table 15 in the left-right direction on a plane including the surface of the table. The height of the table 15 can be fixed, but is not limited thereto, and can be variable. For example, the electronic device 100 can calculate the height of the target blood vessel 320 from the bottom surface. The electronic device 100 can calculate the height of the target blood vessel 320 from the bottom surface by adding the height of the target blood vessel 320 from the surface of the table and the height of the table 15 from the bottom surface. However, without being limited thereto, the electronic device 100 can obtain the height of the target blood vessel 320 from the table 15 (e.g., the surface of the table). The electronic device 100 can generate a plane that is parallel to the bottom surface (or the XY plane) and includes a point corresponding to the target blood vessel 320 based on the position (e.g., the height) of the target blood vessel 320. The electronic device 100 can calculate the point (the "intersection point") where the target blood vessel 320 intersects with a plane that is parallel to the bottom surface and includes the target blood vessel 320, and with a linear axis connecting the radiation source 131 and the isocenter 111. The electronic device can move the table top 15-1 left and right on the XY plane so that the target blood vessel 320 is located at the calculated intersection point.

[0056] In one embodiment, the electronic device 100 can calculate a target distance indicating a linear distance from the radiation source 131 to the target blood vessel 320 using a number of parameters after adjusting the position of the target blood vessel 320 by moving the table top 15-1. The number of parameters can include, for example, the position of the table 15, the position of the target blood vessel 320, and the irradiation angle of the radiation emitted from the radiation source 131 (for example, the projection angle 231 of the radiation in FIG. 2). More specifically, the electronic device 100 can accurately calculate the target distance from the radiation source 131 to the target blood vessel 320 based on a first vertical distance from the isocenter 111 to the table 15, a second vertical distance from the target blood vessel 320 to the table 15, and the irradiation angle of the radiation emitted from the radiation source 131.

[0057] FIG. 2 is a diagram illustrating a change in a projection angle of radiation due to rotation of a C-arm included in an electronic device according to an embodiment.

[0058] In one embodiment, the electronic device 100 can use a radiation source 131 included in the radiation application unit 13 to apply radiation to a target blood vessel of a subject positioned on the table 15 .

[0059] The electronic device 100 according to an embodiment can rotate the radiation source based on at least one of a first rotation axis (A-A') including the isocenter 111 and parallel to the table plane and a second rotation axis (B-B') different from the first rotation axis (A-A') based on the rotation of the C-arm. For example, the radiation source can rotate in one of a plane perpendicular to the first rotation axis (A-A') including the isocenter 111 and parallel to the table plane, or a plane perpendicular to the second rotation axis (B-B') including the isocenter 111 and different from the first rotation axis (A-A'). The second rotation axis (B-B') can intersect the first rotation axis (A-A') and can be orthogonal to each other at the isocenter 111, for example. For reference, for convenience of explanation, an example is described in which the C-arm and the radiation source rotate in one plane, but this is not limited to this, and they may also rotate simultaneously based on two rotation axes (A-A', B-B').

[0060] For example, the electronic device 100 can rotate the C-arm 12 relative to the main body 11 around a first rotation axis (A-A'). The first rotation axis (A-A') can refer to an axis connecting a point where the C-arm 12 and the main body 11 are coupled to the isocenter 111. The position of the radiation source 131 can be changed by rotating the C-arm 12. For example, the radiation source 131 can rotate in a rotational orbit on a plane 190 that includes the isocenter 111 and is perpendicular to the first rotation axis (A-A') by rotating the C-arm 12 around the first rotation axis (A-A'). In FIG. 2, the plane 190 perpendicular to the first rotation axis (A-A') is exemplified as an XZ plane.

[0061] By the rotation of the C-arm 12, the radiation source 131 can rotate in a rotational orbit shaped like a circle on the plane 190. When the C-arm 12 rotates, the rotational orbit along which the radiation source 131 moves is not geometrically a perfect circle, and a certain part of the circle may be flattened due to sagging or vibration of the C-arm 12.

[0062] The electronic device 100 can also rotate the C-arm 12 relative to the main body 11 around a second rotation axis (B-B'). The second rotation axis (B-B') can be an axis that passes through the isocenter 111 and intersects (e.g., perpendicular to) the first rotation axis (A-A'). For example, by rotating the C-arm 12 around the second rotation axis (B-B'), the radiation source 131 can rotate along a rotational orbit on a plane that includes the isocenter 111 and is perpendicular to the second rotation axis (B-B'). For reference, FIG. 2 shows an example in which the C-arm rotates around the first rotation axis (A-A') while also rotating around the second rotation axis (B-B'). However, when rotation of the C-arm 12 occurs around the second rotation axis (B-B') while the radiation source 131 and the radiation detection unit 14 are positioned on the Z axis, the second rotation axis (B-B') is parallel to the X axis, and the radiation source 131 can rotate in a rotational orbit on the YZ plane.

[0063] When the position of the radiation source 131 is changed by rotating the C-arm 12, the projection angle 231 (e.g., θ) of the radiation emitted from the radiation source 131 may be changed. In other words, the projection angle 231 of the radiation emitted from the radiation source may be changed by rotating the C-arm 12.

[0064] The radiation projection angle 231 may refer to the angle between an axis (e.g., the Z axis) perpendicular to a plane including the surface of the table 15 and an axis 211 corresponding to the irradiation direction of radiation emitted from the radiation source 131. The radiation irradiation direction may refer to a direction from the radiation irradiation unit 13 (or the radiation source 131) toward the radiation detection unit 14 (or the radiation detection sensor). The axis 211 corresponding to the radiation irradiation direction may refer to an axis connecting the radiation irradiation unit 13 and the radiation detection unit 14. The radiation projection angle 231 may be an angle between -180° and 180°.

[0065] The isocenter 111 may be defined in various ways based on the rotation of the C-arm 12. For example, the isocenter 111 may refer to a point at which radiation irradiated from various positions of the radiation source due to the rotation of the C-arm 12 is concentrated. As another example, the isocenter 111 may refer to the geometric center of the rotation orbit of the radiation source 131. As another example, the isocenter 111 may refer to the center of rotation of the C-arm 12.

[0066] FIG. 3 is a diagram illustrating a process in which an electronic device according to an embodiment calculates a straight-line distance from a radiation source to a target blood vessel.

[0067] 3 is a side view of an electronic device (e.g., electronic device 100 of FIG. 1) as viewed from the XZ plane. Before using the radiation source 131 to irradiate the target blood vessel 320 with radiation, the electronic device can move the table top 15-1 to position the radiation source 131, the target blood vessel 320, and the isocenter 111 in a straight line. As described above, the position of the radiation detection unit 14 (e.g., the imaging position) can be moved by rotating the C-arm 12. For example, the radiation detection unit 14 and the radiation irradiation unit 13 can be moved within a region range 310 corresponding to the surface of a hemisphere by rotating the C-arm 12.

[0068] JPEG2025071812000004.jpg89157

[0069]

number

[0070] JPEG2025071812000006.jpg43157

[0071] In one embodiment, the electronic device can calculate the target distance and then calculate a correction factor based on the calculated target distance. As described above, the electronic device can convert the distance in the blood vessel image 130 to a physical distance using the calculated correction factor. As described below, the electronic device in this specification can calculate the physical distance corresponding to the diameter of the target vessel by multiplying the correction factor by the number of pixels corresponding to the diameter of the target vessel shown in the captured blood vessel image.

[0072] FIG. 4 is a flowchart showing a method for restoring a three-dimensional shape of a blood vessel according to an embodiment.

[0073] In step 410, the electronic device (e.g., electronic device 100 of FIG. 1) can acquire a set of medical images taken from a plurality of positions. For example, the electronic device can acquire a set of medical images at a plurality of shooting positions based on a C-arm imaging device. For example, the electronic device can acquire medical images at each of the plurality of shooting positions by moving a radiation irradiation unit and a radiation detection unit of the C-arm imaging device by rotating the C-arm. For example, the electronic device can acquire a plurality of medical images using the C-arm imaging device.

[0074] The electronic device can determine imaging parameters of the captured medical images. The imaging parameters can be parameters that indicate the imaging position and / or orientation of the medical images, such as the position of the radiation emitter and / or radiation detector that captured the medical images and / or the radiation irradiation angle. The imaging parameters can include positions recorded in a DICOM file.

[0075] The electronic device can calibrate the imaging parameters of each medical image. The electronic device can perform an offset-correction algorithm. When the medical image is an image captured using a C-arm imaging device, noise may occur when a machine (e.g., a C-arm) rotates due to characteristics of the C-arm, and a difference may occur between an actual position and a position recorded in a DICOM (Digital Imaging and Communications in Medicine) file. For reference, imaging information (e.g., imaging parameters) of a DICOM file may include, for example, an angle, a source to image detector distance (SID), and a table position. In order to correct noise caused by characteristics of the C-arm, the electronic device can optimize imaging parameters through imaging information and a Common-Image point (CIP) pair of the acquired DICOM files. The Common-Image point (CIP) pair is a pair of points indicating a common position in a group of medical images captured from different imaging positions and / or irradiation angles, and may include, for example, a proximal point indicating a start point of a blood vessel region and a distal point indicating an end point of a blood vessel region. For example, a near point in the reference image and a near point in the sub-image may be a CIP pair. Additionally, a far point in the reference image and a far point in the sub-image may be a CIP pair. The electronic device may optimize the objective function (e.g., minimize the objective function value) by changing the imaging parameters of the second DICOM file and the SID of the first DICOM file through an optimizer such as LM or SGD. Exemplarily, the objective function may be a function that calculates a distance value between the 2D CIP pair and the 3D point when back projected after generating the 3D point through the 2D CIP pair and the imaging parameters.

[0076] In step 420, the electronic device can generate tree-structured data of 3D candidate points corresponding to points along centerlines of vascular regions of a reference image and a sub-image of the medical images. For example, the electronic device can set one of two medical images as a reference image and determine the other as a sub-image. The electronic device can segment vascular regions from each of the reference image and the sub-image. The electronic device can determine centerlines for the segmented vascular regions from each image.

[0077] In this specification, the centerline of a vascular region may be a line passing through the center of the blood vessel and connecting the center points of the blood vessel inner diameters longitudinally. A point along the centerline may refer to a point at a position that is a certain length away from a starting position in the blood vessel region of a medical image (e.g., a blood vessel image) along the centerline. A diameter at a position along the centerline may refer to a length (or interval) between the inner walls of the blood vessel along a line perpendicular to the centerline from that position, and may also be referred to as a width. The centerline of the blood vessel region in the reference image may be referred to as a reference centerline, and the centerline of the blood vessel region in the sub-image may be referred to as a sub-centerline.

[0078] The electronic device can determine 3D candidate points by points (e.g., sub-points) of sub-centerlines of the sub-images based on points (e.g., reference points) of the reference centerline of the reference image. Figure 5 shows pairing of multiple sub-points based on an arbitrary reference point, Figure 6 shows determining 3D candidate points by pair, Figure 7 shows back-projecting the 3D candidate points onto the reference image, and Figure 8 shows filtering of the 3D candidate points. Figure 9 shows an example of the generated tree structure data.

[0079] In step 430, the electronic device can determine a three-dimensional path based on the tree-structured data. Figure 12 shows an example of edge weights of the tree-structured data, and Figure 13 illustrates an example of a three-dimensional path determined using the tree-structured data.

[0080] In step 440, the electronic device can determine the three-dimensional diameter for each point on the three-dimensional path. Figures 14 and 15 illustrate an example of determining the three-dimensional diameter using vessel width calculated from two-dimensional images.

[0081] Based on the determined three-dimensional diameter, the electronic device can reconstruct a three-dimensional vessel shape corresponding to the three-dimensional path in step 450. An example of reconstructing a three-dimensional vessel shape is illustrated in FIG.

[0082] 5 to 7 are diagrams illustrating the generation of 3D point candidates according to one embodiment.

[0083] The electronic device according to an embodiment can segment a vascular region from a set of captured medical images (e.g., a set of vascular images) as described above. The illumination angle at which each medical image was captured can be different from the illumination angle at which other medical images were captured. The electronic device can determine a centerline (e.g., a vascular centerline) for the segmented vascular region of each medical image. The electronic device can also acquire CIP points and camera information in addition to the centerline. The electronic device can generate a 3D centerline through the centerline acquired from the set of medical images (e.g., angiography images), the CIP points (e.g., proximal points and distal points) input by the user, and camera information at two capture positions (e.g., camera parameters optimized by the offset-calibration described above). The 3D centerline can include a set of waypoints selected from a set of 3D candidate points (e.g., a point cloud that is a collection of 3D candidate points).

[0084] According to one embodiment, the electronic device can determine a medical image selected from the plurality of medical images as a reference image based on at least one of centerline length or lesion information.

[0085] For example, the electronic device may select, as a reference image, a medical image having the shortest centerline length among a plurality of medical images. For example, the electronic device may select, as a reference image, an image having a short two-dimensional centerline among two medical images. The electronic device may select, as a sub-image, an image having a long two-dimensional centerline among two medical images. As another example, the electronic device may select, as a sub-image, a medical image based on lesion information among a plurality of medical images. The lesion information is information about a lesion shown in a medical image, and may include, for example, at least one of the severity, size, location, or shape of the lesion. The severity of the lesion is information about vascular stenosis caused by the lesion, and may be, for example, %DS (Percentage diameter stenosis). %DS=1-(diameter of lesion) / (normal diameter). The normal diameter is the diameter when it is assumed that the lesion is normalized by surgery, and may be determined, for example, as a statistical value (for example, an average value) of a group of diameter values ​​of normal regions surrounding the lesion. The electronic device may identify a group of lesions for each of a plurality of medical images, and calculate lesion information for each identified lesion. For example, the electronic device can calculate a %DS for each lesion from each medical image. The electronic device can calculate multiple %DS values ​​for each medical image if multiple lesions are present in each medical image. The electronic device can determine a maximum %DS value from the multiple %DS values ​​for each medical image. The electronic device can compare the maximum %DS values ​​of the multiple medical images with each other. The electronic device can determine the medical image with the largest %DS value (e.g., the maximum %DS value) from the multiple medical images as a reference image. As another example, the electronic device can select the reference image taking into account all of the centerline length and lesion information described above.

[0086] The electronic device can determine, for each of a set of reference points along a centerline of the vascular region segmented from the reference image, a set of three-dimensional candidate points corresponding to a set of sub-points along the centerline of the vascular region segmented from the sub-image. The electronic device can determine N points (e.g., reference points) along the centerline (e.g., reference centerline) from the reference image. The electronic device can determine M points (e.g., sub-points) along the centerline (e.g., sub-centerline) from the sub-image, where N and M can be integers equal to or greater than 1. The electronic device can determine M sub-point-specific three-dimensional points (e.g., three-dimensional candidate points) for each of the N reference points.

[0087] 5, the electronic device can map one reference point 511 and M sub-points 529. For example, the electronic device can determine M three-dimensional candidate points based on the assumption that they indicate the same physical point (e.g., a physical point in a three-dimensional space). The reference point 511 can be a pixel point in a reference image 510 that is a two-dimensional image, and the sub-points can be pixel points in a sub-image 520 that is also a two-dimensional image. Since the electronic device maps the M sub-points 529 to each of the N reference points, N×M mappings can be formed between the reference points and the sub-points 529.

[0088] 6, the electronic device may determine a 3D candidate point for each mapping described above in FIG. 5. For example, the electronic device may determine a 3D point corresponding to the mapping of a reference point (e.g., reference point 511 in FIG. 5) and a first sub-point (e.g., first sub-point 521 in FIG. 5) based on an assumption that the reference point (e.g., reference point 511 in FIG. 5) and a first sub-point (e.g., first sub-point 521 in FIG. 5) indicate the same physical point (e.g., 3D point). Similarly, the electronic device may determine a 3D point corresponding to the mapping of a reference point (e.g., reference point 511 in FIG. 5) and a second sub-point (e.g., second sub-point 522 in FIG. 5) based on an assumption that the reference point (e.g., reference point 511 in FIG. 5) and a second sub-point (e.g., second sub-point 522 in FIG. 5) indicate the same physical point. In FIG. 6, for ease of explanation, a simplified example is illustrated in which three sub-points are mapped to one reference point 611, but a three-dimensional candidate point group can be determined for mapping all M sub-points.

[0089] For reference, Fig. 6 is a simplified side view of an electronic device (for example, the electronic device 100 in Fig. 1) as viewed from the XZ plane, and the radiation irradiation units 603-1, 603-2 and radiation detection units 604-1, 604-2 shown here correspond to the radiation irradiation unit 13 and the radiation detection unit 14 in Fig. 3. In this example, the radiation irradiation units 603-1, 603-2 and the radiation detection units 604-1, 604-2 move along a circular orbit within the area range 310, and the radiation detection unit 604-1 can capture an image based on radiation irradiated from the radiation irradiation unit 603-1 at the reference position, and the radiation detection unit 604-2 can capture an image based on radiation irradiated from the radiation irradiation unit 603-2 at the sub-position.

[0090] According to an embodiment, the electronic device can determine the 3D candidate points corresponding to the reference point 611 and the sub-point 621 by triangulation. For example, the electronic device can determine, as the 3D candidate point, a midpoint 699 on a line 698 perpendicular to a line 691 passing through a point on the image plane corresponding to the reference point 611 of the reference image 610 from the radiation irradiation unit 603-1 at the reference position where the reference image 610 was captured, and a line 692 passing through a point on the image plane corresponding to the sub-point 621 of the sub-image 620 from the radiation irradiation unit 603-2 at the sub-position where the sub-image 620 was captured, or a point where both lines intersect. The image plane corresponding to the reference image 610 may be a plane corresponding to the radiation detection unit 604-1 at the reference position, and the image plane corresponding to the sub-image 620 may be a plane corresponding to the radiation detection unit 604-2 at the sub-position. The straight line passing through a point on the image plane of each image from the radiation irradiation unit may be referred to as an epipolar line. If the epipolar line corresponding to the reference point and the epipolar line corresponding to the sub-point intersect, the intersecting point is determined as the 3D candidate point, and if they do not intersect, a midpoint 699 on the line intersecting the group of points where the distance between both epipolar lines is the smallest may be determined as the 3D candidate point. The electronic device can generate 3D candidate points for multiple sub-points corresponding to any reference point among the multiple reference points by repeating the above-mentioned operation. M 3D candidate points can be generated for one reference point. The electronic device can generate N×M 3D candidate points. A collection of 3D candidate points can be called a point cloud.

[0091] 7, the electronic device can back-project multiple 3D candidate points corresponding to the same reference point 711 onto the reference image 710. For example, the electronic device can determine a point where a ray emitted from a radiation emitter 703-1 at the reference position passes through each of multiple 3D candidate points 799 and reaches a reference image plane (e.g., a plane corresponding to a radiation detector 704-1 at the reference position) as a projected point 719 of the 3D candidate point.

[0092] FIG. 8 is a diagram illustrating filtering of 3D candidate points according to one embodiment.

[0093] The electronic device can refine (or filter) the plurality of 3D candidate points based on their respective projected positions onto the reference image 810. For example, the electronic device can keep the 3D candidate points corresponding to points 815 projected around the reference point 811 and reject the remaining 3D candidate points 819. The electronic device can filter out the plurality of 3D candidate points having points projected within a first threshold distance from the reference point.

[0094] Therefore, when the generated 3D candidate points are projected onto a reference image, a group of points 815 projected in the vicinity of the reference points used in generation can be used as candidate points for the 3D centerline (e.g., 3D candidate points).

[0095] 9 to 11 are diagrams illustrating the generation of tree structure data according to an embodiment.

[0096] The electronic device according to an embodiment can generate a group of nodes corresponding to the group of three-dimensional candidate points according to the depth according to the arrangement order of the reference points along the center line of the vascular region segmented from the reference image 910. For example, the electronic device can generate a group of nodes corresponding to the group of three-dimensional candidate points from node 931 corresponding to a proximal point to node 939 corresponding to a distal point. The electronic device can connect the generated group of nodes via edges.

[0097] The electronic device can generate tree structure data 930 based on the filtered group of three-dimensional candidate locations. The tree structure data 930 includes a plurality of nodes, and each node can be connected via an edge. For example, each node of the tree structure data 930 can point to the corresponding three-dimensional candidate location. The electronic device can generate the tree structure data 930 by generating a group of nodes corresponding to the filtered group of three-dimensional candidate locations and connecting the generated group of nodes via edges.

[0098] The electronic device can generate tree structure data 930 including a group of nodes built hierarchically by setting a depth for each node. The depth of each node can be set based on the order in which a plurality of reference points are arranged along the centerline from a start point (e.g., a proximal point) to an end point (e.g., a distal point). Each of the reference points arranged along the reference centerline from the reference image 910 can have an index (e.g., a centerline index) according to the arranged order. The electronic device can determine the depth of a node indicating the reference point based on the centerline index of the reference point. For example, a node indicating a three-dimensional candidate point generated based on the first point (e.g., index=0) from the reference centerline can have a depth of 0. A node indicating a three-dimensional candidate point generated based on the second point (e.g., index=1) from the reference centerline can have a depth of 1. 7 and 8, multiple nodes 932 at the same depth (e.g., depth 1) may be generated because multiple 3D candidate points are generated for the same reference point 912. Similarly, multiple nodes 933 corresponding to multiple candidate points may be generated for the reference point 913 at the next depth.

[0099] The electronic device can generate edges that form connections between the nodes based on the depths set for the nodes. The electronic device can generate edges between the nodes corresponding to adjacent depths. Each edge can form a connection between a node at a previous depth (e.g., depth 0) and a node at a next depth (e.g., depth 1). The edge can be directional and can have a node at the previous depth as a start node and a node at the next depth as an end node.

[0100] The electronic device may also connect adjacent depth nodes among the generated nodes according to whether a distance between 3D candidate points corresponding to adjacent depth nodes is within a threshold distance. The electronic device may generate an edge between a node of the first depth and a node of the second depth if a distance between the 3D candidate point indicated by the node of the first depth and the 3D candidate point indicated by the node of the second depth is within a second threshold distance. The first depth and the second depth may be adjacent depths (e.g., adjacent depths with no additional depth therebetween). The electronic device may exclude generation of an edge between a node of the first depth and a node of the second depth if a distance between the 3D candidate point indicated by the node of the first depth and the 3D candidate point indicated by the node of the second depth exceeds a second threshold distance. For example, if the distance between the third point in the group of 3D candidate points corresponding to the first reference point and the fourth point in the group of 3D candidate points corresponding to the second reference point exceeds the second threshold distance, the edge 990 between the two points may not be connected. FIG. 10 shows an example of a group of 3D candidate points 1010 classified according to the depth of the tree structure data being back-projected onto an image plane (e.g., a reference image plane). FIG. 11 shows a point cloud 1110 consisting of the group of 3D candidate points of the tree structure data visualized from an arbitrary view. In FIGS. 10 and 11, the group of points corresponding to the 0th index (e.g., depth 0) to the 800th index (e.g., depth 800) are displayed in different hues (e.g., gradation), but this is merely an example and is not limited to the index (e.g., a group of tepra) from the start point to the end point being 800, and the interval between the groups of points by depth is not limited to the one shown in the figure. The closer to the start point (eg, index 0), the darker the index is, and the closer to the end point (eg, index 800), the brighter the index is; however, this is for ease of understanding.

[0101] FIG. 12 is a diagram illustrating edge weights in tree structure data according to an embodiment.

[0102] According to an embodiment, the electronic device may determine an edge weight 1240 for an edge between nodes while generating the above-mentioned tree structure data 1230. The edge weight 1240 of an edge connecting both nodes may be determined based on a distance 1242 (e.g., point distance) between 3D candidate points indicated by the two nodes and a distance 1241 (e.g., line distance) between an epipolar line with respect to a reference point and an epipolar line with respect to a sub-point corresponding to the 3D candidate point. The electronic device may determine the sum of the point distance and the line distance as the edge weight 1240. For example, the electronic device may determine the edge weight 1240 for an edge based on a distance 1242 between 3D candidate points corresponding to both nodes connected to the edge. The electronic device may also determine the edge weight 1240 for an edge based on a distance 1241 between an epipolar line with respect to a reference point and an epipolar line with respect to a sub-point corresponding to one of the 3D candidate points of both nodes connected to the edge.

[0103] The distance 1242 between the 3D candidate points indicated by both nodes may be the distance in 3D space between the 3D candidate point 1291 indicated by the parent node (e.g., node 1231 at the first depth) and the 3D candidate point 1292 indicated by the child node (e.g., node 1232 at the second depth). The distance 1242 between the 3D candidate points may act as a weighting value that mitigates the phenomenon in which the restored 3D shape becomes severely curved due to noise.

[0104] The distance 1241 between the epipolar lines may be a distance between epipolar lines for 2D points (e.g., reference points and sub-points) used to generate the 3D candidate point 1291 indicated by the parent node. The epipolar lines may refer to lines connecting the radiation emitters 1203-1 and 1203-2 and 3D points generated when transforming points (e.g., 2D points) on the image planes 1210 and 1220 corresponding to the radiation detectors into 3D space. For reference, the distance 1241 between the epipolar lines may act as a weighting value for closely matching the 2D centerline in each 2D image plane when the 3D centerline is projected onto each 2D image plane.

[0105] For reference, an example has been described in which the sum of the distance 1242 between the 3D candidate points and the distance 1241 between the epipolar lines is determined as the edge weight 1240, but the present invention is not limited to this. Each edge may have various edge weights. For example, each edge may have a first edge weight indicating the distance between the 3D candidate points and a second edge weight indicating the distance between the epipolar lines.

[0106] FIG. 13 is a diagram illustrating an example of a three-dimensional path selected using edge weight values ​​of tree structure data according to an embodiment.

[0107] According to one embodiment, the electronic device can determine the three-dimensional route 1392 by selecting a group of waypoints from a group of three-dimensional candidate points from a start point to an end point of the reference center line. The electronic device can determine the three-dimensional route 1392 based on a cost calculated using edge weights of the tree structure data 1330 from among the potential routes based on the group of three-dimensional candidate points.

[0108] For example, the electronic device can generate a three-dimensional path 1392 by selecting a node corresponding to a waypoint from among the nodes of the tree structure data 1330 based on the edge weight value described above. The electronic device can determine the three-dimensional path 1392 from among potential paths formed by selecting a waypoint node group 1391 from a node corresponding to a proximal point (e.g., a start node) to a node corresponding to a distal point (e.g., an end node) using a cost based on the edge weight value group of the tree structure data 1330. For example, the electronic device can calculate a path cost for each potential path group by adding up edge weight values ​​of edges connecting nodes forming the potential path. The electronic device can determine a path with the smallest calculated path cost (e.g., a path with a minimized cost) from among the potential paths as the three-dimensional path 1392. For example, the electronic device can search for a shortest path with the smallest sum of edge weight values ​​from a proximal point to a distal point through the Dijkstra algorithm. The determined three-dimensional path 1392 may refer to the centerline of the blood vessel in three-dimensional space, and may also be referred to as the three-dimensional centerline.

[0109] 14 and 15 are diagrams illustrating matching between 3D point clouds and 2D point clouds by pair matching according to one embodiment.

[0110] In one embodiment, the electronic device can pair the way-points of the 3D centerline to the reference points of the reference image and the sub-points of the sub-images. The electronic device can match the determined 3D centerline to the 2D centerline of the reference image and the 2D centerline of the sub-image, respectively.

[0111] 14, for example, the electronic device can project a three-dimensional centerline 1430 onto a reference image. The electronic device can match a group of points 1431 where the three-dimensional centerline 1430 is projected onto the reference image with a group of indexes of points (e.g., reference points 1411) of a two-dimensional centerline 1410 of the reference image. The electronic device can determine an index of the reference point 1411 that is matched to each of the group of points 1431 of each three-dimensional centerline 1430 based on a DTW algorithm. The DTW algorithm can be an algorithm that calculates a similarity between both curves to define a relationship between the points.

[0112] The electronics can similarly project the three-dimensional centerline onto the sub-image, and can match indices of points where the three-dimensional centerline is projected onto the sub-image with points (e.g., sub-points) of the two-dimensional centerline of the sub-image.

[0113] 15, the electronic device can determine a diameter of the three-dimensional points that define a three-dimensional centerline. For example, the electronic device can identify two-dimensional points from the reference image and the sub-image that correspond to points on the three-dimensional path. The electronic device can calculate the three-dimensional diameter based on a width of the identified two-dimensional points.

[0114] For example, the electronic device can identify reference points and sub-points that are matched to a group of three-dimensional points that define a three-dimensional centerline based on the pair matching described above. Each three-dimensional point can be matched with one or more reference points and one or more sub-points. In some examples, a plurality of two-dimensional points can be matched to one three-dimensional point. The electronic device can obtain diameter information of the group of two-dimensional points that are matched to the three-dimensional point. The diameter information is a diameter value of a blood vessel at the two-dimensional point, and can also be referred to as a blood vessel width. The electronic device can determine the diameter information based on the width of the two-dimensional point. For example, the electronic device can determine the width of the two-dimensional point as a length along a direction perpendicular to a two-dimensional centerline (e.g., a reference centerline or a sub-centerline) of the two-dimensional point in the divided blood vessel region.

[0115] The electronic device can calculate a width at each of the two-dimensional points when multiple two-dimensional points are matched to the same three-dimensional point. For example, the electronic device can determine a width at a reference point and a width at a sub-point identified for the three-dimensional point. The electronic device can determine a three-dimensional diameter value based on the determined widths.

[0116] 15, the electronic device can back-project the three-dimensional centerline 1530 onto the reference image 1510 and the sub-image 1520, respectively. The electronic device can map each point of the three-dimensional centerline 1530 to a point 1515 along the reference centerline of the reference image 1510 and a point 1525 along the sub-centerline of the sub-image 1520. The electronic device can calculate the distance (e.g., vessel width) 1591 between the top contour line 1511 and the bottom contour line 1512 from the reference image 1510. Similarly, the electronic device can calculate the distance (e.g., vessel width) 1592 between the top contour line 1521 and the bottom contour line 1522 from the sub-image 1520. The electronic device can calculate the average value of the widths of the identified two-dimensional points as the three-dimensional diameter. For example, the electronic device can determine the average value of the determined widths as the three-dimensional diameter value. The electronic device can determine the three-dimensional diameter value for each of the three-dimensional points.

[0117] FIG. 16 is a diagram illustrating the reconstruction of a three-dimensional shape according to an embodiment.

[0118] In one embodiment, the electronic device can generate a group of vertices 1620 along a circumference 1610 corresponding to a three-dimensional diameter for each point on the three-dimensional path. The electronic device can reconstruct the three-dimensional blood vessel shape using a mesh 1630 based on the generated group of vertices.

[0119] The electronic device can use the determined 3D diameter values ​​for each 3D point to generate circles corresponding to the vessel surface at the corresponding point. For example, the electronic device can generate circles defining a vessel cross section perpendicular to a 3D centerline from each 3D point. The electronic device can generate vertices 1620 corresponding to the circle defining the vessel cross section. For example, the electronic device can arrange vertices 1620 along a circumference 1610 corresponding to the vessel cross section. The electronic device can determine a surface normal by calculating a tangent of the 3D centerline from the 3D point. The electronic device can generate a circle having a 3D diameter value on the surface based on the determined normal.

[0120] For reference, the diameter information may be obtained from the two-dimensional image illustratively in units of pixels, and the electronic device may convert the diameter values ​​in pixels to diameter values ​​in longitudinal units (e.g., in millimeters (mm)) using the calibration factor described above. The electronic device may use the longitudinally converted diameter values ​​to generate a circle that defines a vessel cross-section at the three-dimensional point. The vessel cross-section may be perpendicular to a three-dimensional centerline (e.g., a vessel centerline).

[0121] The electronic device can convert the vertices 1620 that define the blood vessel cross-sections generated for a plurality of 3D points into a mesh 1630. The electronic device can convert the vertices 1620 corresponding to the blood vessel cross-sections into a mesh 1630 through a Poission surface reconstruction algorithm. The electronic device can generate faces by grouping the vertices 1620 based on the Poisson surface reconstruction. For example, the faces can be triangle primitives and / or quadrilateral primitives. A triangle primitive can be defined and / or formed by three vertices. Similarly, a quadrilateral primitive can be defined and / or formed by four vertices. The electronic device can reconstruct the 3D shape of the blood vessel by forming primitives corresponding to the faces from the vertices 1620 as described above.

[0122] In addition, the electronic device may apply texture information to a primitive corresponding to a surface. For example, the electronic device may determine a texture value for a surface as an FFR (Fractional Flow Reserve) value corresponding to a blood vessel position of the surface. The FFR value is a blood flow pressure value for each position in a cardiac blood vessel, and may be calculated from three-dimensional blood vessel information. Therefore, the electronic device may provide a user with an FFR value for each position in a blood vessel more intuitively by applying the FFR value as a texture value to a primitive for each position in the blood vessel.

[0123] The above-described embodiments may be embodied using hardware components, software components, and / or a combination of hardware and software components. For example, the devices, methods, and components described in the embodiments may be embodied using a general-purpose computer or a special-purpose computer, such as a processor, a controller, an arithmetic logic unit (ALU), a digital signal processor, a microcomputer, a field programmable gate array (FPGA), a programmable logic unit (PLU), a microprocessor, or a different device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and software applications executed on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to the execution of software. For ease of understanding, the processing device may be described using one, but one skilled in the art will appreciate that the processing device may include multiple processing elements and / or multiple types of processing elements. For example, the processing device may include multiple processors or one processor and one controller. Other processing configurations are also possible, such as parallel processors.

[0124] Software may include computer programs, code, instructions, or a combination of one or more of these, and may configure or, independently or collectively, instruct a processing device to operate as desired. The software and / or data may be permanently or temporarily embodied in some type of machine, component, physical device, virtual device, computer storage medium or device, or transmitted signal wave to be interpreted by or provide instructions or data to a processing device. The software may be distributed over computer systems coupled to a network, stored or executed in a distributed manner. The software and data may be stored on a computer-readable recording medium.

[0125] The method according to the present embodiment may be embodied in the form of program instructions to be executed by various computer means and may be recorded on a computer-readable recording medium. The computer-readable medium may include program instructions, data files, data structures, and the like, either alone or in combination, and the program instructions recorded on the medium may be specially designed and constructed for the embodiment or may be known and available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine language code, such as that generated by a compiler, but also high-level language code executed by a computer using an interpreter, etc.

[0126] The hardware devices described above may be configured to operate as one or more software modules to perform the operations of an embodiment, and vice versa.

[0127] As described above, the embodiments and the like have been described based on the accompanying drawings, but a person having ordinary skill in the art can apply various technical modifications and variations based on the above description. For example, the above techniques may be carried out in a different procedure than the above methods, or the components of the above systems, structures, devices, circuits, etc. may be combined or combined in a different form than the above methods, or may be replaced with other components or equivalents, and still achieve appropriate results.

[0128] Accordingly, other implementations, other embodiments, and equivalents of the claims are within the scope of the following claims.

Claims

1. A method for reconstructing a three-dimensional shape of a blood vessel, the method being implemented by a processor, comprising: acquiring a set of medical images taken from a plurality of positions; generating tree structure data of three-dimensional candidate points corresponding to points along centerlines of vascular regions of a reference image and sub-images of the medical images; determining a three-dimensional path based on the tree structure data; determining a three-dimensional diameter for each point on the three-dimensional path; and restoring a three-dimensional blood vessel shape corresponding to the three-dimensional path based on the determined three-dimensional diameter.

2. The method for restoring the three-dimensional shape of a blood vessel as described in claim 1, wherein the step of generating the tree structure data includes a step of determining, as the reference image, a medical image selected from the plurality of medical images based on at least one of the length of the center line or lesion information.

3. 2. The method for restoring the three-dimensional shape of a blood vessel as described in claim 1, wherein the step of generating the tree structure data includes a step of determining, for each of a group of reference points along a center line of a blood vessel region divided from the reference image, a group of three-dimensional candidate points corresponding to a group of sub-points along a center line of a blood vessel region divided from the sub-image.

4. The step of determining a group of three-dimensional candidate locations includes: generating a group of nodes corresponding to the group of three-dimensional candidate points by depth according to an arrangement order of the group of reference points along a center line of the blood vessel region segmented from the reference image; The method for reconstructing a three-dimensional shape of a blood vessel according to claim 3 , further comprising the step of: connecting the generated group of nodes via edges.

5. 5. The method for restoring the three-dimensional shape of a blood vessel as described in claim 4, wherein the step of connecting the generated node groups via edges includes a step of connecting the node groups of adjacent depths depending on whether a distance between three-dimensional candidate point groups corresponding to node groups of adjacent depths among the generated node groups is within a threshold distance.

6. 5. The method for reconstructing the three-dimensional shape of a blood vessel according to claim 4, wherein the step of connecting the generated group of nodes via edges includes a step of determining an edge weight value for the edge based on the distance between groups of three-dimensional candidate points corresponding to both nodes connected to the edge.

7. 5. The method for reconstructing the three-dimensional shape of a blood vessel according to claim 4, wherein the step of connecting the generated node group via edges includes a step of determining an edge weight value for the edge based on a distance between an epipolar line for a reference point corresponding to one of the three-dimensional candidate points of both nodes connected to the edge and an epipolar line for a sub-point.

8. The method for restoring the three-dimensional shape of a blood vessel as described in claim 1, wherein the step of determining the three-dimensional route includes a step of determining the three-dimensional route by selecting a group of intermediate points from the group of three-dimensional candidate points from a start point to an end point of a reference center line of the reference image.

9. The method for reconstructing the three-dimensional shape of a blood vessel according to claim 1, wherein the step of determining the three-dimensional route includes a step of determining the three-dimensional route based on a group of edge weight values ​​of a group of nodes corresponding to the group of three-dimensional candidate points from the tree structure data.

10. The method for reconstructing the three-dimensional shape of a blood vessel as described in claim 1, wherein the step of determining the three-dimensional route includes a step of determining the three-dimensional route from a group of potential routes based on the group of three-dimensional candidate points, based on a cost calculated using edge weight values ​​of the tree-structured data.

11. The step of determining a three-dimensional diameter for each point on the three-dimensional path includes: identifying a set of two-dimensional points from the reference image and the sub-image that correspond to a set of points on the three-dimensional route; The method of claim 1 , further comprising: calculating the three-dimensional diameter based on at least one of the widths of the identified two-dimensional points.

12. The method for reconstructing the three-dimensional shape of a blood vessel according to claim 11 , wherein the step of calculating the three-dimensional diameter includes a step of calculating an average value of widths of the identified two-dimensional points as the three-dimensional diameter.

13. The step of reconstructing the three-dimensional blood vessel shape includes: generating a group of vertices along a circumference corresponding to the three-dimensional diameter for each point on the three-dimensional path; The method for reconstructing a three-dimensional shape of a blood vessel according to claim 1 , further comprising: reconstructing the three-dimensional blood vessel shape using a mesh based on the generated vertex group.

14. 14. A computer readable recording medium having stored thereon one or more computer programs comprising instructions for performing the method of any one of claims 1 to 13.

15. A computer readable program for carrying out the method according to any one of claims 1 to 13 on a computer.

16. In an electronic device, an image acquisition unit that acquires a group of medical images taken from a plurality of positions; a processor that generates tree-structured data of three-dimensional candidate points corresponding to points along center lines of vascular regions of a reference image and a sub-image among the medical images, determines a three-dimensional route based on the tree-structured data, determines a three-dimensional diameter for each point on the three-dimensional route, and reconstructs a three-dimensional vascular shape corresponding to the three-dimensional route based on the determined three-dimensional diameter.

Images