JP2011212314A - Blood vessel display control equipment and method, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To more easily confirm the result of extraction of a plurality of kinds of blood vessels in a three-dimensional medical image showing a subject.SOLUTION: When the blood vessels to be extracted are pulmonary arteries and pulmonary veins for example, a pulmonary blood vessel extraction section 31 extracts a vascular structure showing the pulmonary arteries or the pulmonary veins from the three-dimensional medical image showing the chest of the subject, stores relatedly each of the extracted vascular structure and blood vessel identification information showing the kind of the blood vessel of the vascular structure in a storage means, contact and intersection detection sections 31a and 31b detect contacts and intersections where the extracted vascular structures come into contact with or intersect each other, a display control section 33 displays the vascular structures in the contacts and intersections and the vicinities of the contacts and intersections in a mode permitting the blood vessel identification information related to the vascular structure to be identified for at least some of the detected contacts and intersections and displays the contacts and intersections in a mode permitting the contacts and intersections to be identified.

Description

  The present invention relates to an apparatus and method for displaying a result of extracting a plurality of types of blood vessels from a three-dimensional medical image representing a subject, and a program.

  There is a technique for automatically extracting and classifying pulmonary blood vessels as an elemental technique for enabling more accurate and detailed lung image diagnosis and surgical simulation.

  For example, a method of extracting a lung blood vessel region using a region growing method (Region Growing) based on a seed point representing a lung blood vessel designated by a user using a chest X-ray CT image as an input is known. (For example, Non-Patent Document 1).

  In addition, by performing multi-resolution conversion on a three-dimensional image, performing eigenvalue analysis of the Hessian matrix on each resolution image, and integrating the analysis results in each resolution image, various sizes in the three-dimensional image are obtained. A method of extracting the line structure (blood vessel) of the above is also known (for example, Non-Patent Document 2).

  Furthermore, a technique for classifying the extracted pulmonary blood vessels into pulmonary arteries and pulmonary veins is also known. For example, in Non-Patent Document 1 above, the pulmonary artery and bronchus run in parallel near the center of each area of the lung, whereas the pulmonary veins are near the boundary of each area of the lung, that is, between the bronchi and bronchus. A method that focuses on driving is proposed. Specifically, this method decomposes the extracted blood vessel structure into a plurality of blood vessel branch groups based on the contact relationship between blood vessels, extracts bronchi using a known method, and further extracts the extracted bronchi The boundary area of the lung area is estimated by performing three-dimensional Voronoi division using the set as the mother point set, and the average distance from the bronchus and the average distance from the boundary area of the lung area are calculated for each vascular branch group. The vascular branch group that is close to the pulmonary area and far from the boundary surface of the lung area is classified as a pulmonary artery, and the vascular branch group that is far from the bronchus and close to the boundary surface of the lung area is classified as pulmonary vein.

Sho Nakamura and 4 others, "Automatic classification of pulmonary arteries and pulmonary veins from chest X-ray CT images by tree structure analysis", IEICE technical report. MI, Medical Image, Japan, IEICE, January 21, 2006, Vol.105, No.580, pp.105-108, [Searched on November 20, 2009], Internet <URL: http://www.murase.nuie.nagoya-u.ac .jp / ~ ide / res / paper / J05-kenkyukai-snaka-1.pdf> Y Sato, et al., `` Three-dimensional multi-scale line filter for segmentation and visualization of curvilinear structures in medical images. '', Medical Image Analysis, June 1998, Vol.2, No.2, p.p.143-168

  In the automatic classification method of the pulmonary artery and pulmonary vein proposed in Non-Patent Document 1 and the like, about 10 to 20% of misclassification has been reported. If the result of automatic classification is used as it is, lung image diagnosis and surgery are performed. It will also affect the accuracy of the simulation.

  The present invention has been made in view of the above circumstances, and provides an apparatus, a method, and a program that can more easily confirm the extraction results of a plurality of types of blood vessels in a three-dimensional medical image representing a subject. It is intended to provide.

  The blood vessel display control device of the present invention extracts each of blood vessel structures representing a plurality of types of blood vessels from a three-dimensional medical image representing a subject, and determines each of the extracted blood vessel structures and the types of blood vessels of the blood vessel structure. At least one of the detected intersection point, a blood vessel extraction unit that associates and stores the blood vessel identification information to be expressed in the storage unit, a contact point detection unit that detects a contact point where the extracted blood vessel structures contact or intersect each other, and Display for displaying the tangent point and the vascular structure in the vicinity of the tangent point in such a manner that the vascular identification information associated with the vascular structure can be identified and the tangent point can be identified. And a control means.

  According to the blood vessel display control method of the present invention, blood vessel structures representing a plurality of types of blood vessels are each extracted from a three-dimensional medical image representing a subject, and each of the extracted blood vessel structures and the types of blood vessels of the blood vessel structure are determined. Storing the associated blood vessel identification information in the storage means, detecting the intersection point where the extracted blood vessel structures are in contact with or intersecting, and at least a part of the detected intersection point. Displaying the intersection and the vascular structure in the vicinity of the tangent point in a manner capable of identifying the vascular identification information associated with the vascular structure, and displaying the tangent point in a identifiable manner. Features.

  The blood vessel display control program of the present invention is for causing a computer to execute the above method.

  In the present invention, specific examples of the “three-dimensional medical image” include those representing the chest (lung) of the subject, those representing the head (brain), and those representing the abdomen (liver).

  Further, specific examples of the types of blood vessels of the extracted blood vessel structure include arteries, veins, and portal veins.

  A known extraction method can be adopted as a method for extracting a blood vessel structure representing a plurality of types of blood vessels. Specifically, after extracting the vascular structures of multiple types of blood vessels without identifying the types, the extracted blood vessels are classified by type to extract each of the vascular structures representing the multiple types of blood vessels. Alternatively, blood vessel structures representing a plurality of types of blood vessels may be extracted separately.

  Examples of the data structure for storing the blood vessel structure include a tree structure and a network structure.

  In addition, the association between the extracted blood vessel structure and the blood vessel identification information may be made to associate the blood vessel identification information for each portion constituting the blood vessel structure, for example, a blood vessel branch obtained by dividing the blood vessel for each branch.

  Specific examples of the display mode of the blood vessel structure and the blood vessel identification information include displaying each blood vessel type with a different color, and displaying the core line of each blood vessel in the region of the blood vessel structure with a different line type (solid line and different line type). For example, the display may be superimposed with a broken line.

  As a specific example of the display mode of the intersection point, a marker such as a circle surrounding the intersection point, an arrow indicating the intersection point, or the like is attached, or a color different from the blood vessel structure is attached to the intersection point. To do. These markers and coloring may be displayed translucently.

  When a plurality of intersection points are detected, all of the local images representing each intersection point and the blood vessel structure in the vicinity of the intersection point may be displayed in a list, or the entire chest image and the local image may be displayed side by side or sequentially. May be. Or you may make it display in order from the said local image near the base of a blood vessel.

  In the present invention, an input of correction content for the displayed blood vessel structure and / or blood vessel identification information associated with the blood vessel structure is received, and the blood vessel structure stored in the storage means based on the input correction content and / or the blood vessel structure The blood vessel identification information associated with the blood vessel structure may be corrected. Here, as a specific example of the correction of the blood vessel structure, correction of the connection relation of the blood vessels is given. Also, at the time of correction, input of correction contents of the blood vessel structure and blood vessel identification information for a part of the blood vessel structure is accepted, and from a part of the blood vessel structure to be corrected toward the periphery or the base of the blood vessel structure at once. The blood vessel structure and blood vessel identification information may be corrected.

  Further, the display of the tangent point and the vascular structure in the vicinity of the tangent point may be updated based on the input corrected vascular structure and vascular identification information.

  In the present invention, it is also possible to accept input of the correct / incorrect determination result of the displayed blood vessel structure and / or blood vessel identification information associated with the blood vessel structure.

  According to the present invention, each of the blood vessel structures representing a plurality of types of blood vessels extracted from the three-dimensional medical image representing the subject and the blood vessel identification information representing the types of blood vessels of the blood vessel structures are associated with each other in the storage means. Storing intersection points where the extracted vascular structures contact or intersect with each other are detected, and for at least some of the detected tangent points, the tangent points and the vascular structures in the vicinity of the tangent points can be associated with the vascular structures. The blood vessel identification information can be displayed in an identifiable manner, and the intersection point can be displayed in an identifiable manner. Here, when extracting a blood vessel structure representing a plurality of types of blood vessels, an error is likely to occur in the extraction result of the blood vessel structure or the blood vessel identification information at a tangent point where the blood vessel structures contact or intersect with each other. By displaying the identification information in an identifiable manner and displaying this intersection point in an identifiable manner, it is possible to more easily and more efficiently check the key points of the extraction results of blood vessel structures representing multiple types of blood vessels. It becomes possible to do.

  If correction is made to the blood vessel structure and / or the blood vessel identification information associated with the blood vessel structure, not only the correction target part can be easily grasped by the display but also the extraction result of the blood vessel structure can be corrected. It becomes possible to carry out simply.

  Furthermore, if the display of the vascular structure is updated based on the content of the correction, it is possible to immediately confirm the correction result, and how the correction made is applied to the vascular structure around the correction location. Since it is possible to easily grasp the spread, the efficiency of the correction work is improved.

  In addition, if the correctness of the blood vessel structure and / or the blood vessel identification information associated with the blood vessel structure is determined, not only the correction target portion can be easily grasped by the above display, but also the determination result of the blood vessel structure is determined. Can be performed more easily.

1 is a schematic configuration diagram of a medical image diagnostic system in which a pulmonary blood vessel display control device according to an embodiment of the present invention is introduced. The block diagram which showed typically the structure and process flow which implement | achieve the pulmonary blood vessel display control function in embodiment of this invention The flowchart showing the flow of the pulmonary blood vessel display control processing using the medical image diagnostic system in the embodiment of the present invention Schematic representation of pulmonary blood vessels and core / branch points Diagram showing an example of tree structure data representing pulmonary blood vessels A diagram schematically showing the contact of pulmonary blood vessels Schematic representation of how the pulmonary artery was first extracted by the region expansion method near the contact area of the pulmonary blood vessel Schematic representation of the pulmonary veins extracted after extracting the pulmonary artery by the region expansion method near the contact area of the pulmonary blood vessel A diagram schematically showing the crossing of pulmonary blood vessels Schematic representation of how the pulmonary artery was first extracted by the region expansion method near the intersection of pulmonary blood vessels Schematic representation of the pulmonary veins extracted after extracting the pulmonary artery by the region expansion method near the intersection of pulmonary blood vessels The figure which represented typically the detection process example of the intersection point by the 2nd intersection point detection part The figure which expressed typically the 1st display example of the extraction result of a pulmonary artery and a pulmonary vein A diagram schematically showing a second display example of the pulmonary artery and pulmonary vein extraction results A diagram schematically showing a third display example of the pulmonary artery and pulmonary vein extraction results Diagram showing an example of pulmonary artery and vein before correction of extraction results The figure which expressed typically the 1st collective correction example of the extraction result of the pulmonary artery and the pulmonary vein The figure which expressed typically the 2nd collective correction example of the extraction result of the pulmonary artery and the pulmonary vein The figure which expressed typically the 3rd collective correction example of the extraction result of the pulmonary artery and the pulmonary vein The figure which represented typically the tree structure data showing each vascular structure before performing the 1st batch correction example of the extraction result of a pulmonary artery and a pulmonary vein The figure which represented typically the tree structure data showing each vascular structure after performing the 1st modification example of the extraction result of a pulmonary artery and a pulmonary vein

  Hereinafter, a case where the present invention is implemented as a pulmonary blood vessel display control device that displays vascular structures representing pulmonary arteries and pulmonary veins extracted from a three-dimensional medical image representing the chest of a subject and corrects the extraction results. As an example, a medical image diagnostic system in which a blood vessel display control device according to an embodiment of the present invention is introduced will be described.

  FIG. 1 is a hardware configuration diagram showing an outline of this medical image diagnostic system. As shown in the figure, in this system, a modality 1, an image storage server 2, and an image processing workstation 3 are connected via a network 9 in a communicable state.

  The modality 1 generates image data of a three-dimensional medical image representing the region by imaging the region to be examined of the subject, and the image data is defined by the DICOM (Digital Imaging and Communications in Medicine) standard. A device for adding incidental information and outputting it as image information is included. Specific examples include CT and MRI. In the present embodiment, a case will be described in which three-dimensional image data of the chest of a human body is generated by scanning in the body axis direction of the human body that is the subject by CT.

  The image storage server 2 is a computer that stores and manages medical image data acquired by the modality 1 and image data of medical images generated by image processing at the image processing workstation 3 in an image database. A device and database management software (for example, ORDB (Object Relational Database) management software) are provided.

  The image processing workstation 3 performs image processing (including image analysis) on the medical image data acquired from the modality 1 or the image storage server 2 in response to a request from the interpreter, and displays the generated image. A computer having a well-known hardware configuration such as a CPU, main storage device, auxiliary storage device, input / output interface, communication interface, input device (mouse, keyboard, etc.), display device (display monitor), data bus, etc. The operating system is installed. The pulmonary blood vessel display control processing according to the present invention is implemented in the image processing workstation 3, and this processing is realized by executing a program installed from a recording medium such as a CD-ROM. The program may be installed after being downloaded from a storage device of a server connected via a network such as the Internet.

  The storage format of image data and communication between devices via the network 9 are based on a protocol such as DICOM.

FIG. 2 is a block diagram showing a part related to the pulmonary blood vessel display control process according to the embodiment of the present invention, among the functions of the image processing workstation 3. As shown in the figure, the pulmonary blood vessel display control process according to the embodiment of the present invention includes a pulmonary blood vessel extraction unit 31 including a seed point setting unit 31a and a blood vessel structure extraction unit (also serving as a first intersection detection unit) 31b. The second intersection detection unit 32, the display control unit 33, and the extraction result correction unit 34 are realized. Also, the three-dimensional medical image V, the pulmonary artery seed point SPA, the pulmonary vein seed point SPV, the pulmonary artery blood vessel structure PA, the pulmonary vein blood vessel structure PV, the first tangent point T1 _n , the second tangent point T2 _m , and the extraction result The determination result, the correction instruction information MI, and the display image I are data read / written from / to a predetermined memory area of the image processing workstation 3 by each processing unit.

  FIG. 3 is a flowchart showing a flow of user operations, calculation processing, display processing, and the like under the execution of the pulmonary blood vessel display control processing software according to the embodiment of the present invention. The flow of the pulmonary blood vessel extraction result confirmation and correction process according to an embodiment of the invention will be described.

  First, image data of the 3D medical image V to be processed is acquired from the image storage server 2 by image search / acquisition processing in a known image search system or a known ordering system (# 1).

Next, in the image processing workstation 3, in the pulmonary blood vessel extraction unit 31, the seed point setting unit 31a uses the pulmonary artery seed point SPA and the pulmonary vein seed point SPV as a reference for extracting each of the pulmonary artery and pulmonary vein. Accepting the setting (# 2), the vascular structure extraction unit 31b extracts the pulmonary artery vascular structure PA based on the set pulmonary artery seed point SPA, and the pulmonary vein vascular structure PV based on the pulmonary vein seed point SPV. Extracted and stored in a predetermined memory area of the image processing workstation 3. Further, the blood vessel structure extraction unit 31b also serves as a first intersection point detection unit, and detects a point T1 _n where each blood vessel contacts or intersects (referred to as a intersection point here) when extracting the pulmonary artery and pulmonary vein. (# 3). Here, n is a subscript for identifying a plurality of detected intersection points.

The second intersection detection unit 32 detects the intersection T2 _m of the pulmonary artery / venous vein by a method different from that of the first intersection detection unit 31b (# 4). Note that m is a subscript identifying a plurality of detected intersection points.

As illustrated in FIG. 8A, the display control unit 33 displays the blood vessel structures PA and PV in the vicinity of the detected intersection points T1 _n and T2 _m as well as the pulmonary artery and the pulmonary vein for at least a part of the detected intersection points T1 _n and T2 _m. The display image I having a circular marker at the intersection is generated and displayed on the display of the image processing workstation 3 (# 5).

The extraction result correction unit 34 receives an input of a determination result MI as to whether or not the displayed blood vessel structures PA and PV are correct as an extraction result, and the input determination result MI is displayed as the intersection points T1 _n and T2 _m displayed. The data is stored in a predetermined memory area of the image processing workstation 3 in association with each other. In addition, it accepts input of correction contents MI for the displayed blood vessel structures PA and PV, and stores blood vessel structures PA and pulmonary veins of the pulmonary artery stored in the memory area of the image processing workstation 3 based on the input correction contents MI. Modify the structure PV information (# 6).

Finally, the display control unit 34 generates the information in step # 5 based on the information of the tangent points T1 _n and T2 _m after the determination, and the information on the vascular structure PA of the pulmonary artery and the vascular structure PV of the pulmonary vein after the correction. The generated display image I is generated again, and the display on the image processing workstation 3 is updated (# 7).

  Next, details of processing performed in each processing unit will be described.

  The seed point setting unit 31a generates a cross-sectional image with a given axial section based on the three-dimensional medical image V, displays the cross-sectional image on the display of the image processing workstation 3, and sets the seed point in the displayed cross-sectional image. Provide a user interface that accepts For example, the user operates the mouse of the image processing workstation 3 to specify a desired point in the pulmonary artery in the displayed cross-sectional image, and displays a processing menu for the specified point by a right click operation. Select “Set as pulmonary artery seed point” from the displayed processing menu. In response to this user operation, the seed point setting unit 31a stores the coordinate value of the specified point in the three-dimensional medical image V in a predetermined memory area of the image processing workstation 3, thereby setting the pulmonary artery seed point SPA. Is done. Similarly, for the pulmonary vein seed point SPV, the user right-clicks on a desired point in the pulmonary vein in the displayed cross-sectional image and selects “Set as pulmonary vein seed point” from the displayed processing menu. Then, the seed point setting unit 31a sets the pulmonary vein seed point SPV by storing the coordinate value of the designated point in the three-dimensional medical image V in a predetermined memory area. The seed point setting unit 31a may be configured to set a plurality of seed points for each of the pulmonary artery and pulmonary vein. Also, when specifying the seed point, it accepts a change operation of the generation conditions of the cross-sectional image such as the cross-sectional position and cross-section direction of the displayed cross-sectional image, and displays the cross-sectional image generated based on the generation condition after the change It may be.

The blood vessel structure extraction unit 31b extracts a set of pixels in the blood vessel region based on the input three-dimensional medical image data V and seed point information using a known region expansion method, and is illustrated in FIG. 4A. As described above, the thinning process is performed on the extracted blood vessel region, and the pixels on the thin line are classified into end points, edges (sides), and branch points based on the connection relation of the thin lines representing the obtained blood vessels. Then, tree structure data representing the blood vessel structure as illustrated in FIG. 4B is obtained. Here, for example, B ₉ from the edge portion B ₁ of the thin line in FIG. 4A, associated from Node B ₁ of the tree structure of Figure 4B as B _9, to each of the B ₉ from Node B ₁ of the tree structure, branch vessel regions and vascular identification information for identifying which one of the pulmonary artery and pulmonary vein corresponding the edge portion B ₁ in B _9, the position information of the vessel branch region, branch vessel orientation is retained. Also, feature amounts such as the diameter information of the blood vessel branch and the length of the blood vessel branch at each edge may be held. The direction of the blood vessel branch is the direction of the direction vector of the blood vessel branch, and the direction close to the direction from the hilar portion (base of the blood vessel) to the chest wall direction (periphery of the blood vessel) in the blood vessel branch. it can.

  In the vascular structure extraction unit 31b, the pulmonary artery vascular structure PA based on the pulmonary artery seed point SPA and the pulmonary vein vascular structure PV based on the pulmonary vein seed point SPV may be sequentially extracted. However, they may be performed in parallel.

Furthermore, the blood vessel structure extraction unit 31b also functions as a first blood vessel intersection detection unit. Specifically, in the region expansion method performed by the blood vessel structure extraction unit 31b, similar to the conventional method, the adjacent region that satisfies the region expansion condition using the pixel value, the area of the region, and the like starting from the seed point is selected as the blood vessel. The process is sequentially expanded as an area, and the process ends when there is no adjacent area that satisfies the area expansion condition. However, in this embodiment, the adjacent area that satisfies the area expansion condition is already a blood vessel. The process is also ended when the region has already been extracted. In the present embodiment, the contact or intersection (intersection point) T1 _n between blood vessels is detected using the latter processing end condition. Here, the tangent point T1 _n may be any one of a contact point or intersection point between pulmonary arteries, a contact point or intersection point between pulmonary veins, or a contact point or intersection point between pulmonary artery and pulmonary vein.

FIGS. 5A to 5C illustrate the case of detecting a contact between blood vessels. As shown in FIG. 5A, when the two blood vessels VL ₁ and VL ₂ are in contact with each other and the blood vessel VL ₁ is extracted as a pulmonary artery in order from the upper side of the figure by the region expansion process based on the pulmonary artery seed point SPA, cross-section of the blood vessels in the SL ₄ from section SL ₁ shown at the left side of FIG. 5B is as shown on the right side of FIG. 5B, the shaded region in FIG. 5B, extracted as a blood vessel structure PA ₁ of the pulmonary artery having a branching Is done. Then, if the blood vessel VL ₂ by the region expansion processing based on the pulmonary veins seed point SPV is extracted as the pulmonary veins in order from the left in figure, in the vicinity of section SL ₅ are extracted as a blood vessel structure PV ₁ in the pulmonary vein, In the cross section SL ₆ , since the cross section of the blood vessel VL ₂ has already been extracted as the pulmonary artery vascular structure PA ₁ , the latter processing end condition is satisfied, and the pulmonary vein vascular structure extraction processing ends. As a result, although the entire blood vessel VL ₂ should be extracted as the pulmonary vein PV ₁ originally, only the position of the section SL ₆ was extracted as the pulmonary vein PV ₁ as shown in FIG. 5C. Become. As described above, since the blood vessel structure is erroneously extracted at the point where the blood vessels contact each other, the first blood vessel intersection detection unit (blood vessel structure extraction unit) 31b uses the latter processing end condition to The contact T1 _{n of} is detected.

6A to 6C illustrate the case where the intersection of blood vessels is detected. As shown in FIG. 6A, when the two blood vessels VL ₃ and VL ₄ intersect, and the blood vessel VL ₃ is extracted as a pulmonary artery in order from the left in the figure by the region expansion process based on the pulmonary artery seed point SPA, The cross section of the blood vessel in the cross sections SL ₇ to SL ₁₀ shown on the left side of FIG. 6B is as shown on the right side of FIG. 6B, and the hatched area in FIG. 6B is extracted as the vascular structure PA ₂ of the pulmonary artery having a branch. Is done. Next, when the blood vessel VL ₂ is extracted as a pulmonary vein in order from the left direction of the figure by region expansion processing based on the pulmonary vein seed point SPV, it is extracted as a vascular structure PV ₂ of the pulmonary vein in the vicinity of the cross section SL ₁₁ , in cross-section SL _12, since the cross section of the vessel VL ₄ is already extracted as vasculature PA ₂ of the pulmonary artery, it corresponds to the latter processing end conditions, extraction of vasculature pulmonary vein ends. As a result, despite originally should entire vessel VL ₄ are extracted as the pulmonary vein PV _2, only up to the position of the cross-section SL ₁₂ as shown in FIG. 6C, that has not been extracted as a pulmonary vein PV ₂ Become. In this way, since the blood vessel structure is erroneously extracted even at the point where the blood vessels intersect, the first blood vessel intersection detection unit (blood vessel structure extraction unit) 31b uses the latter processing end condition to The intersection point T1 _{n of} is detected.

Information on the extracted intersection point T1 _n is stored in a predetermined memory area of the image processing workstation 3. The contact information intersection T1 _n, other coordinates of the contact point of intersection T1 _n, extraction result status representing the accuracy determination condition of tangent intersection near the extraction results ( "unprocessed", "confirmed", "fixed") It is included. Details of the extraction result status will be described later.

  In the above embodiment, the case where the pulmonary blood vessel extraction unit 31 extracts the pulmonary artery and the pulmonary vein separately has been described. For example, the region expansion method described in Non-Patent Document 1 or the eigenvalue of the Hessian matrix described in Non-Patent Document 2 After extracting the blood vessel structure without distinguishing between the pulmonary artery and the pulmonary vein by the method using the analysis, the extracted blood vessel structure is classified into the pulmonary artery and the pulmonary vein using the method described in Non-Patent Document 1, etc. It may be.

Based on the pulmonary artery vascular structure PA and pulmonary vein vascular structure PV, the second tangent point detection unit 32 detects the tangent point T2 _m between the blood vessels by a method different from that of the first tangent point detection unit 32. To do. Specifically, for branch points detected by the thinning process, adjacent branch points whose distance between the branch points is equal to or less than a predetermined threshold value d ₁ are integrated into one branch point, and each branch after integration is integrated. For a point, the number of blood vessel branches that are directly connected to the branch point of interest among the branch points that are connected to the branch point within the predetermined distance d ₂ from the branch point (the target branch point) is counted, and When the number of directly connected blood vessel branches is four or more, the branch point of interest is determined as a point where blood vessels intersect or contact each other, and is detected as a tangent point T2 _m . For example, in the case of vascular structure as shown in FIG. 7, the threshold value d ₁ distance for the branch point integration can be represented as the diameter of a circle Th _1, of the BP ₈ from the branch point BP _1, circle Th ₁ The branch points BP ₅ and BP ₆ that can be included in the block are integrated into one branch point (here, BP _M1 ). In addition, since the distance d ₂ that serves as a reference for extracting the branch point to be counted for the vascular branch can be expressed as the radius of the circle Th ₂ , when the integrated branch point BP _M1 is the target branch point, the branch point Branch points BP ₁ to BP ₄ , BP ₇ , and BP ₈ included in the circle Th ₂ centered on BP _M1 are the branch points to be counted for vascular branches, and among the vascular branches connected to these branch points Since the vascular branches directly connected to the target branch point BP _M1 are the vascular branches BP ₂ -BP ₅ , BP ₄ -BP ₆ , BP ₆ -BP ₇ , and BP ₅ -BP ₈ , The intersection point determination condition is satisfied, and the branch point BP _M1 is detected as the intersection point T2 _m . Alternatively, when the distance between the blood vessel branches constituting the extracted blood vessel structure is smaller than a predetermined distance (for example, about the blood vessel diameter), it may be determined that the blood vessel branches intersect or contact each other. The blood vessel branch may be divided for each branch of the blood vessel structure obtained by the region expansion method, or may be a line segment structure obtained by eigenvalue analysis of the Hessian matrix.

The tangent intersection T2 _m may be any contact or intersection between pulmonary arteries, contact or intersection between pulmonary veins, or contact or intersection between arteries and pulmonary veins.

The display control unit 33 identifies, for at least a part of the detected intersection points T1 _n and T2 _m , the vascular structure PA and PV in the vicinity of the tangent point and the vascular identification information associated with the vascular structure. In addition to being displayed in a possible manner, the intersection point is displayed on the display of the image processing workstation 3 in such a manner that the intersection point can be identified. Here, the display control unit 33 may set only the extraction result status of “unprocessed” among the intersection points T1 _n and T2 _m to be displayed, or the user can select the extraction result status of the display target. It may be.

8A to 8C are specific examples of the image I to be displayed. As shown in the drawings, the contact point of intersection are denoted by M ₃ from the circular markers M _1. Also, in FIG. 8A, a known volume rendering process is performed by assigning different colors to the pixels belonging to the vascular structures of the pulmonary artery and pulmonary vein in the three-dimensional medical image V, so that the vascular structure near the detected intersection point is obtained. Is color-coded by pulmonary artery and pulmonary vein and expressed in a pseudo three-dimensional manner. In FIG. 8B, the core line of each blood vessel (the thinning result of each blood vessel or a curve connecting the centers of the cross sections of each blood vessel) and the branch point are superimposed and displayed on the blood vessel structure color-coded as in FIG. 8A. On the other hand, in FIG. 8C, only the core line and the branch point of each blood vessel are expressed using different line types and colors in the pulmonary artery and the pulmonary vein without performing color-coded display of the blood vessel structure.

8A to 8C are examples of the displayed image I, and other images may be displayed. For example, the graphic used as a marker for the intersection point may be rectangular, the intersection point may be indicated by an arrow or the like, or the intersection point may be displayed in a color different from the blood vessel structure. In this case, the display may be semi-transparent and superimposed. Further, the intersection point T1 _n detected by the first intersection point detection unit 31b and the intersection point T2 _m detected by the second intersection point detection unit 32 are displayed in different display modes (marker shape, color, etc.). You may make it do. Or you may make it display in a display mode which changes with the value of extraction result status. On the other hand, for the image representing the blood vessel structure, an MIP image or an MPR image generated by a known method may be used, or various images may be switched and displayed. In addition, an operation for changing the viewpoint position and the line-of-sight direction (projection plane) may be received so that an image representing a blood vessel structure viewed from a different viewpoint or angle may be displayed.

Further, the display control unit 33 may generate the image I representing the periphery of each of the plurality of intersection points T1 _n and T2 _m and display the generated images side by side as a list. Switching display may be performed according to the operation. Moreover, you may make it display sequentially from the image of the intersection point close | similar to the base of a pulmonary artery or a pulmonary vein. Alternatively, the intersection points T1 _n and T2 _m are marked and displayed on the entire image representing the vascular structure of the entire lung, and specified according to an operation in which the user designates a desired intersection point in the entire image. The local image representing the periphery of the intersection may be displayed.

  The extraction result correction unit 34 provides a user interface that accepts input of correction instruction information MI indicating the correctness determination result of the displayed blood vessel structure PA and PV, or correction contents for the displayed blood vessel structure PA and PV.

FIG. 9A schematically shows a certain intersection and the blood vessel structures PA ₁₁ and PV ₁₁ near the intersection displayed by the display control unit 33. In the figure, the pulmonary artery PA ₁₁ is represented by a solid line, and the pulmonary vein PV ₁₁ is represented by a broken line. The user checks the displayed blood vessel structure, in particular, the connection relation of the blood vessel branches B ₁₁ , B ₁₂ , B ₁₇ , B ₂₁ connected to the intersection point indicated by the marker M ₁₁ and the type of blood vessel, and the extraction result is If it is determined that the correct, specify the marker M _11, the image processing right-click operation of the work station 3, to display the processing menu for the specified point, the "extraction result from the displayed processing menu Select “Make sure”. Extraction result modifying unit 34, in response to this user operation, it updates the extracted result status markers M tangent intersection ₁₁ is attached T1 _n, T2 _m to "Confirmed" in the memory area of image processing workstation 3 Remember.

On the other hand, when there is an error in the extraction result and it is desired to correct it, for example, in FIG. 9A, the vascular branch among the vascular branches B ₁₁ , B ₁₂ , B ₁₇ , B ₂₁ connected to the intersection point indicated by the marker M ₁₁ in FIG. If you want to modify the blood vessels of the type and connection relationship of B _12, the user specifies a branch vessel B _12, the image processing right-click operation of the work station 3, the processing for a given branch vessel B ₁₂ Display the menu and select “Modify to“ Pulmonary vein ”” from the displayed processing menu. As a result, correction instruction information MI for correcting the blood vessel type of the blood vessel branch B ₁₂ to “pulmonary vein” is input to the extraction result correction unit 34. Further, the user specifies a branch vessel B ₂₁ and B _12, by the right-click operation of the mouse to display a processing menu for branch vessel B ₂₁ and B _12, which is designated, the "branch vessel from the displayed processing menu Select Connect. As a result, correction instruction information MI indicating correction of the connection relationship of connecting the blood vessel branches B ₂₁ and B ₁₂ is additionally input to the extraction result correction unit 34. Alternatively, the user specifies a branch vessel B ₂₁ and B _12, by the right-click operation of the mouse to display a processing menu for branch vessel B ₂₁ and B _12, which is designated, the "branch vessel from the displayed processing menu By simply selecting “Connect”, the blood vessel branches B ₂₁ and B ₁₂ are connected, and the blood vessel type of the blood vessel branch B ₁₂ on the downstream (peripheral) side of both blood vessel branches is set to the blood vessel on the upstream (base) side. Correction instruction information MI for correction to the same “pulmonary vein” as in the branch B ₂₁ may be input to the extraction result correction unit 34.

When the user selects “collective correction” by menu selection on the display screen or the like, the extraction result correction unit 34 is stored in the memory area of the image processing workstation 3 based on the input correction instruction information MI. Corrected information on pulmonary artery vascular structure PA and pulmonary vein vascular structure PV. FIG. 9B schematically shows a display image I displayed by the display control unit 33 after the blood vessel structure and the blood vessel identification information are collectively corrected based on the content of the correction instruction. As shown in the figure, the extraction result correction unit 34 connects the vascular branches B ₂₁ and B ₁₂ to not only correct the vascular branch B ₁₂ to “pulmonary veins” but also to connect to the downstream of the vascular branch B 12. The vascular branches B ₁₃ to B ₁₆ are also collectively corrected to “pulmonary vein”, and the display control unit 33 updates the display image I based on the corrected pulmonary artery vascular structure PA and pulmonary vein vascular structure PV.

10A and 10B schematically show changes in the tree structure data of the pulmonary artery blood vessel structure PA and the pulmonary vein blood vessel structure PV stored in the memory area by this modification. FIG. 10B represents after correction. As shown in FIG. 10A, before the correction, the pulmonary vein PV ₁₁ had only the node B ₂₁ representing the vascular branch B ₂₁ , but after the correction, as shown in FIG. 10B, the pulmonary artery PA ₁₁ node B ₁₂ represents a branch vessel B ₁₂ which is connected to the node B ₁₁ that represents a branch vessel B _11, and, B ₁₆ from the node B ₁₃ that represents the B ₁₆ from the branch vessel B ₁₃ downstream of the node B ₁₂ is, while maintaining the tree structure, and is connected to the node B ₂₁ that represents a branch vessel B ₂₁ of the pulmonary vein PV _11.

Further, the modifications at the same time, the extraction result modifying unit 34 updates the extraction result status tangent intersections marker M ₁₁ is attached to the "fixed", the display control unit 34, the extraction result status is "unprocessed" If the marker is attached only to the intersection point of, as shown in FIG. 9B, the marker may not be attached to the intersection point whose extraction result status is “corrected” or “confirmed”. it can. In this case, the extraction result modifying unit 34 may be configured to change the extraction result status tangent intersection corresponding to the marker M ₁₂ in the "fixed" but, vasculature near the contact point of intersection is automatically Since the correction is performed, the extraction result status is left as “unprocessed”, and as shown in FIG. 9B, the marker M ₁₂ is displayed and the intersection point corresponding to the marker M ₁₂ is displayed. It is preferable to perform processing in accordance with an extraction result determination / correction instruction by the user.

FIG. 9C shows another operation example of batch correction. As shown in the figure, the extraction result correction unit 34 connects the vascular branches B ₂₁ and B ₁₂ to not only correct the vascular branch B ₁₂ to “pulmonary veins” but also connect it downstream of the vascular branch B _12. However, if there is a tangent point (marker M _{12 in the} figure) whose extraction result status is “unprocessed” in the course of correction, the vascular branches B ₁₃ , B up to the tangent point are also corrected. Up to ₁₄ are subject to batch correction, and subsequent correction is not performed for the subsequent blood vessel branches B ₁₅ and B ₁₆ , but processing is performed according to the user's extraction result judgment / correction instructions for the intersection point corresponding to the marker M ₁₂ May be performed.

FIG. 9D shows still another operation example of batch correction. As shown in the figure, the extraction result correction unit 34 connects the vascular branches B ₂₁ and B ₁₂ to not only correct the vascular branch B ₁₂ to “pulmonary veins” but also connect it downstream of the vascular branch B _12. would do the bulk modify branch vessel that is, if the corrected developing extraction result status where there is contact the intersection of the "untreated" (against the intersection corresponding to a marker M ₁₂ in the drawing) is connected to the contact point of intersection The connection relation of each blood vessel branch is estimated from the directions of the blood vessel branches B ₁₄ , B ₁₅ , B ₁₆ , B ₁₈ and the batch correction is continued. Specifically, in the case of FIG. 9D, the extraction result correction unit 34 forms an angle formed between the vascular branches B ₁₄ , B ₁₅ , B ₁₆ , B ₁₈ connected to the intersection point represented by the marker M _12. And the connection relation of each blood vessel branch is determined so that the bending of the blood vessel is minimized. As a result, the branch vessel B ₁₄ and B ₁₆ are connected, the branch vessel B ₁₈ and B ₁₅ are connected. Since towards branch vessel B ₁₄ in the branch vessel B ₁₄ and B ₁₆ is upstream branch vessel B ₁₅ located more downstream is modified to the same "pulmonary vein" and branch vessel B _14. Meanwhile, since the blood vessel branch B ₁₈ and B ₁₅ towards the branch vessel B ₁₈ is upstream, more branch vessel B ₁₅ is the same "pulmonary artery" and branch vessel B ₁₈ located downstream, and the item is performed fixes Absent. As in the case of FIG. 9B, the extraction result modifying unit 34, but may be changed extraction result status tangent intersection corresponding to the marker M ₁₂ to "fixed", the extraction result status to "unprocessed leave to leave the ", as shown in FIG. 9D, and displays designated by the marker M _12, to perform the process in accordance with the determination and correction instructions for the extraction result by the user to the contact point of intersection corresponding to the marker M ₁₂ It is preferable to make it.

  The three operation modes for batch correction may be selectable at the time of execution or in advance by the user, or automatically according to the length of the blood vessel structure downstream from the blood vessel branch for which the batch correction instruction has been issued. It may be determined automatically, or may be determined automatically based on some other criteria. In addition to the batch correction mode, an individual correction mode in which each blood vessel branch is manually corrected individually may be selected.

  Further, each node of the tree structure representing the blood vessel branch is provided with a flag for identifying whether or not the manual correction has been performed, and the extraction result correction unit 34 performs the manual correction on the blood vessel branch and its Batch correction may not be performed for a vascular branch downstream from the vascular branch.

Further, since the blood vessel structure is changed by the processing of the extraction result correction unit 34, the second intersection detection unit 32 detects the intersection T2 _m again after the processing of the extraction result correction unit 34. May be.

As described above, according to the embodiment of the present invention, the display control unit 33 is detected by the intersection point T1 _n detected by the first intersection point detection unit 31b and the second intersection point detection unit 32. For at least a part of the intersection T2 _m , the vascular structure PA and PV in the vicinity of the intersection and the intersection are colored in different colors between the pulmonary artery and the pulmonary vein, and a circular marker is attached to the intersection A display image I is generated and displayed on the display. In this manner, each blood vessel is displayed in such a manner that the blood vessel identification information can be identified, and at the same time, a tangent point where blood vessel structures and blood vessel structures that are likely to cause errors in blood vessel identification information extraction results can be identified. Since it is displayed, it becomes possible to more easily and more efficiently confirm the key points of the extraction results of the pulmonary artery and pulmonary vein.

Further, the extraction result correction unit 34 receives the input of the determination result MI for the displayed blood vessel structures PA and PV, stores them in the memory area in association with the displayed intersection points T1 _n and T2 _m , Since the input of the correction content MI for the displayed vascular structure PA and PV is received, and the information on the vascular structure PA of the pulmonary artery and the vascular structure PV of the pulmonary vein stored in the memory area is corrected based on the input correction content MI The display makes it easy not only to grasp the correction target part, but also makes it easier to determine and correct the blood vessel identification information.

  In addition, the extraction result correction unit 34 corrects the downstream portion of the vascular structure based on the correction content MI for the vascular structure PA and PV near the intersection, thus reducing the correction load by the user. It becomes possible to do. In addition, by appropriately setting the range of batch correction, it becomes possible to balance the correction accuracy and load reduction.

Further, the display control unit 33 displays the display image I based on the information on the tangent points T1 _n and T2 _m after the determination, and the information on the corrected pulmonary artery vascular structure PA and pulmonary vein vascular structure PV. Because it is updated, it is possible to immediately confirm the correction result, and it is also possible to easily understand how the correction made has spread to the blood vessel structure around the correction location.・ The efficiency of correction work is improved.

At that time, the display control unit 33 prevents the display of markers other than the intersection points whose extraction result status is “unprocessed” in accordance with the extraction result statuses associated with the intersection points T1 _n and T2 _m . , It is possible to easily grasp which tangent point is judged and corrected for the extraction result of blood vessel structure, and which tangent point is undecided / uncorrected, which further improves the efficiency of judgment / correction work. improves.

In addition, the display control unit 33 displays the tangent points T1 _n and T2 _m by marking the entire image representing the vascular structure of the entire lung and displays the desired tangent point in the entire image. Accordingly, when the above-mentioned local image representing the periphery of the designated intersection point is displayed, the extraction result near the local intersection point can be determined while observing the entire image of the vascular structure of the lung. Corrections can be made, which contributes to an improvement in the efficiency and accuracy of judgment / correction work.

  In addition, if the display control unit 33 displays in order from the local image of the tangent point close to the base of the pulmonary artery or pulmonary vein, based on the determination result at the base with few errors in the extraction results, As a result, it is possible to determine and correct the extraction result, and to reduce the rework of the determination and correction work, thereby contributing to the improvement of work efficiency.

  The above embodiments are merely examples, and all the above descriptions should not be used to limit the technical scope of the present invention.

  In addition, various modifications made to the system configuration, processing flow, module configuration, specific processing contents, and the like in the above embodiment without departing from the spirit of the present invention are also within the technical scope of the present invention. included.

  For example, in the above embodiment, the system configuration has been described in which each process shown in FIG. 2 is performed by one image processing workstation 3, but each process is distributed to a plurality of workstations. May be configured to perform cooperative processing.

  As for the processing flow and the module configuration, the intersection point is detected during the automatic extraction / classification process of the pulmonary artery / vein by the pulmonary blood vessel extraction unit 31, and the automatic extraction process is interrupted when the intersection point is detected, The display control unit 33 performs marking display of the detected intersection point, manually extracts and classifies the blood vessel structure around the displayed intersection point, and then the pulmonary blood vessel extraction unit 31 resumes the automatic extraction and classification process. It is good also as such a structure.

  Further, regarding processing targets, when extracting arteries and veins from a three-dimensional medical image representing the brain, extracting two or more types of arteries, veins, and portal veins from a three-dimensional medical image representing the liver, etc. Also, the same method can be applied with the same apparatus as described above.

DESCRIPTION OF SYMBOLS 1 Modality 2 Image storage server 3 Image processing workstation 9 Network 31 Pulmonary blood vessel extraction part 31a Seed point setting part 31b Blood vessel structure extraction part (1st intersection point detection part)
32 Second intersection detection unit 33 Display control unit 34 Extraction result correction unit

Claims (7)

Each of the blood vessel structures representing a plurality of types of blood vessels is extracted from the three-dimensional medical image representing the subject, and each of the extracted blood vessel structures is stored in association with blood vessel identification information representing the type of blood vessel of the blood vessel structure. Blood vessel extraction means to be stored in the means;
A tangent point detecting means for detecting a tangent point where the extracted vascular structures contact or intersect with each other;
Displaying at least a part of the detected intersection point, the intersection point and the blood vessel structure in the vicinity of the intersection point in a manner capable of identifying blood vessel identification information associated with the blood vessel structure, and the intersection point A blood vessel display control device comprising: a display control means for displaying the information in a distinguishable manner.   An input of correction content for the displayed blood vessel structure and / or blood vessel identification information associated with the blood vessel structure is received, and the blood vessel structure and / or the blood vessel stored in the storage unit based on the input correction content The blood vessel display control device according to claim 1, further comprising extraction result correction means for correcting blood vessel identification information associated with the structure.   3. The blood vessel display according to claim 2, wherein the display control means updates the display of the tangent point and the vascular structure in the vicinity of the tangent point based on the input correction content. Control device.   4. The method according to claim 1, further comprising an extraction result determination unit that receives an input of a correct / incorrect determination result of the displayed blood vessel structure and / or blood vessel identification information associated with the blood vessel structure. The blood vessel display control device according to Item. The three-dimensional medical image represents a chest of a subject;
The blood vessel display control device according to any one of claims 1 to 4, wherein the types of blood vessels of the blood vessel structure extracted by the blood vessel extraction means are pulmonary arteries and pulmonary veins. Each of the blood vessel structures representing a plurality of types of blood vessels is extracted from the three-dimensional medical image representing the subject, and each of the extracted blood vessel structures is stored in association with blood vessel identification information representing the type of blood vessel of the blood vessel structure. Memorizing the means;
Detecting a tangent point where the extracted vascular structures contact or intersect;
Displaying at least a part of the detected intersection point, the intersection point and the blood vessel structure in the vicinity of the intersection point in a manner capable of identifying blood vessel identification information associated with the blood vessel structure, and the intersection point A blood vessel display control method comprising the steps of: On the computer,
Each of the blood vessel structures representing a plurality of types of blood vessels is extracted from the three-dimensional medical image representing the subject, and each of the extracted blood vessel structures is stored in association with blood vessel identification information representing the type of blood vessel of the blood vessel structure. Memorizing the means;
Detecting a tangent point where the extracted vascular structures contact or intersect;
Displaying at least a part of the detected intersection point, the intersection point and the blood vessel structure in the vicinity of the intersection point in a manner capable of identifying blood vessel identification information associated with the blood vessel structure, and the intersection point A blood vessel display control program for executing the step of displaying in a distinguishable manner.

Images