WO2019111920A1 - Dispositif de calcul de vecteur normal, procédé et programme - Google Patents

Dispositif de calcul de vecteur normal, procédé et programme Download PDF

Info

Publication number
WO2019111920A1
WO2019111920A1 PCT/JP2018/044648 JP2018044648W WO2019111920A1 WO 2019111920 A1 WO2019111920 A1 WO 2019111920A1 JP 2018044648 W JP2018044648 W JP 2018044648W WO 2019111920 A1 WO2019111920 A1 WO 2019111920A1
Authority
WO
WIPO (PCT)
Prior art keywords
points
circle
normal
point
acquisition
Prior art date
Application number
PCT/JP2018/044648
Other languages
English (en)
Japanese (ja)
Inventor
増谷　佳孝
Original Assignee
国立研究開発法人科学技術振興機構
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 国立研究開発法人科学技術振興機構 filed Critical 国立研究開発法人科学技術振興機構
Publication of WO2019111920A1 publication Critical patent/WO2019111920A1/fr

Links

Images

Classifications

    • A—HUMAN NECESSITIES
    • A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
    • A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
    • A61B6/00—Apparatus for radiation diagnosis, e.g. combined with radiation therapy equipment
    • A61B6/02—Devices for diagnosis sequentially in different planes; Stereoscopic radiation diagnosis
    • A61B6/03—Computerised tomographs
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01B—MEASURING LENGTH, THICKNESS OR SIMILAR LINEAR DIMENSIONS; MEASURING ANGLES; MEASURING AREAS; MEASURING IRREGULARITIES OF SURFACES OR CONTOURS
    • G01B21/00—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant
    • G01B21/20—Measuring arrangements or details thereof, where the measuring technique is not covered by the other groups of this subclass, unspecified or not relevant for measuring contours or curvatures, e.g. determining profile

Definitions

  • the present invention relates to an apparatus, method and program for calculating a normal.
  • Disclosed is a technique for estimating the original three-dimensional surface shape using radial basis functions (RBF) based on the coordinates of a point group on the three-dimensional surface and the normal of the surface at each point.
  • RBF radial basis functions
  • the surface normal is calculated at each point. If this normal calculation is not made correctly, accurate surface shape estimation can not be performed. In order to estimate the surface shape correctly, it is an issue to improve the accuracy of the normal calculation.
  • the present invention has been made from this point of view, and an object thereof is to provide a technique for calculating a normal with high accuracy based on a point cloud on a surface.
  • the apparatus selects a plurality of points from among the point acquisition unit for acquiring a plurality of acquisition points on a surface and the acquisition points, and centers the selected points.
  • a plurality of external points are selected from among the acquisition points other than the center of the circle, and virtual attraction is established between each of the selected external points and the nearest point on the circle's circumference
  • a normal line calculation unit that calculates a normal line in a direction in which the circle is directed by attractive force as a normal line at the center of the surface circle.
  • the normal line calculation unit may select all acquisition points when selecting a plurality of points from among the acquisition points.
  • the normal line calculation unit may select all acquisition points other than the center of the circle as external points for each of the circles.
  • the normal calculation unit replaces the virtual attractive force between each of the selected external points and the nearest point on the circumference of the circle with the force acting on the apex of the rigid body right cone whose base is the circle. , Normals may be calculated.
  • the normal line calculation unit may calculate the normal line again with the calculated normal line as an initial state.
  • the apparatus may further include a positive / negative direction unifying unit that aligns the positive / negative direction of the calculated normal.
  • the apparatus may further include a surface shape estimation unit that estimates a surface shape based on the calculated normal.
  • the surface may be a membrane on a medical image.
  • the apparatus further includes an image reconstructing unit configured to create a reconstructed image obtained by reconstructing a medical image based on the surface shape estimated by the surface shape estimating unit, and an image output unit configured to output the reconstructed image.
  • the image reconstruction unit recuts the range of the film based on the surface shape estimated by the surface shape estimation unit and identification information indicating whether each point is inside or at the end of the film. Configuration images may be created.
  • the identification information may be a label indicating an anatomical type according to the site of the membrane.
  • the apparatus selects a plurality of acquisition points, and selects a plurality of points from among the acquisition points, and relates to each of the circles centered on the selected points.
  • an inclination calculation unit that calculates an inclination of a circle.
  • the method according to another aspect of the present invention relates to each of a point acquisition step of acquiring a plurality of acquisition points on a surface, and selecting each of a plurality of points from among the acquisition points, each of which is centered on the selected point.
  • a virtual attractive force acts between each of the selected external points and the nearest point on the circle's circumference, Calculating a normal to the direction in which the circle is directed as a normal at the center of the surface circle.
  • the program according to still another aspect of the present invention relates to each of a point acquiring step of acquiring a plurality of acquisition points on a surface, a plurality of points selected from among the acquisition points, and a circle centered on each of the selected points.
  • a point acquiring step of acquiring a plurality of acquisition points on a surface, a plurality of points selected from among the acquisition points, and a circle centered on each of the selected points.
  • any combination of the above-described constituent elements, one obtained by converting the expression of the present invention between an apparatus, a method, a system, a recording medium, a computer program and the like is also effective as an aspect of the present invention.
  • the surface shape can be accurately estimated based on a point group given on the surface.
  • FIG. 6 illustrates the attraction of acting on the circle from all external points. It is a figure which shows the circle
  • FIG. 6 shows a point cloud labeled indicating the anatomic type of the mesentery.
  • FIG. 1 shows a normal line calculation process according to the technique described in Patent Document 1.
  • the normal to the surface at the point of interest is calculated using three neighboring points near the point of interest. That is, the normal calculation is performed based on the local point distribution for each focus point.
  • the accuracy of the calculated normal is greatly influenced by how to give and select the three nearby points. For example, in a portion with a large curvature on the surface, the normal should be calculated from sufficiently close proximity points. That is, the point cloud needs to be provided at a higher density as the curvature is larger. If the density of the point cloud is not high enough for the curvature, the calculated normal direction will not be accurate.
  • the present inventors can realize normal calculation based on the whole point distribution on the surface by using the following virtual physical model, and calculate the normal from the limited point group with high accuracy I found out what I could do.
  • FIG. 2 shows the configuration of an apparatus 100 according to an embodiment.
  • the apparatus 100 includes a point acquisition unit 10 and a normal calculation unit 20.
  • the point acquisition unit 10 acquires three-dimensional position coordinates of a plurality of points on the surface (hereinafter, the acquired points will be referred to as “acquisition points”). Hereinafter, the point acquisition unit 10 acquires n acquisition points. In one embodiment, the point acquisition unit 10 may directly acquire a point cloud plotted outside the device 100 via an input interface (not shown). In another embodiment, a point cloud database (not shown) for accumulating point cloud data plotted externally is additionally provided in the apparatus 100, and the point acquiring unit 10 acquires acquisition points from the point cloud database. You may get it. The point acquisition unit 10 transmits the position coordinates of each acquisition point to the normal calculation unit 20.
  • the normal line calculation unit 20 selects m points from the n acquisition points received from the point acquisition unit 10 (m ⁇ n). The normal line calculation unit 20 sets m virtual circles centered on the selected m points in the three-dimensional space. The normal line calculation unit 20 calculates a normal vector of the surface at the center of each circle using a virtual physical model based on the center coordinates of each circle and the normal vector of each circle.
  • FIG. 3 shows an example of a circle set by the normal line calculation unit 20.
  • the normal line calculation unit 20 selects m points P 1 ,... From the n acquisition points received from the point acquisition unit 10. . . , P m (m ⁇ n).
  • m points P 1 ,. . . , P m are represented by four points P i , P h , P j and P k (the same applies to FIG. 4 and the like).
  • the normal calculation unit 20 selects the selected points P 1 ,. . . , M number of the virtual circle C 1 centered on P m respectively. . . , C m in a three-dimensional space.
  • the normal line calculation unit 20 calculates the normal vector of the surface, with the direction of these normal vectors, that is, the direction in which each circle set here is directed, as the initial state.
  • the calculation process of the normal line by the normal line calculation unit 20 will be specifically described with reference to FIGS. 4 and 5.
  • Normal calculation is performed when virtual attraction is applied between the circle set above and an acquisition point outside the center of the circle (hereinafter referred to as “external point”) among acquisition points. It is performed based on the direction in which the circle is turned due to the attractive force.
  • all acquisition points are fixed in space and do not move.
  • all circles are rigid and all circles can freely rotate around the center of the circle.
  • the radius of the circle may be set to any size.
  • the radius of the circle may be determined based on the distance between the external point and the center of the circle. In this case, the radius of the circle is preferably set such that the distance between all the external points and the center of the circle is larger than the radius of the circle.
  • point P 1 Focusing on the circle C i whose center P m points extracted randomly from P i, and the point P i.
  • the normal line calculation unit 20 selects p external points from among n ⁇ 1 acquisition points other than the attention point P i (p ⁇ n ⁇ 1). Next, it is considered that virtual attraction acts on the circle C i from each of the selected external points.
  • the attractive force may be a central force that is inversely proportional to the power of ⁇ between the two points ( ⁇ is a real number of 1 or more).
  • FIG. 4 shows a state in which an attractive force acts on the circle C i from a point P j which is one of external points.
  • this attraction acts only on the point R ij on the circle C i where the distance to the point P j is minimum.
  • the circle C i rotates around the fixed center P i to change its direction due to the moment N ij (x represents a vector product).
  • FIG. 5 shows a state where an attractive force is applied to each corresponding point on the circle C i from all the external points.
  • the circle C i turns around the fixed center P i by the sum of the moments about the point P i of the attractive forces from these external points.
  • the circle C i comes to a stand in such a direction that the sum of these moments is balanced, that is, the direction in which the sum of j of the moment N ij is zero.
  • FIG. 6 shows how the moments of attraction from all external points are balanced as a result of the rotation of the circle C i .
  • the unit normal vector of the circle C i in this state is represented by n i .
  • the normal line calculation unit 20 calculates this unit normal vector n i as a surface normal vector at the point P i .
  • the normal vector calculated using this virtual physical model is based on the global point distribution on the surface, and it has been confirmed that it matches the actual surface normal vector with high accuracy .
  • the normal line calculating unit 20 may select all n acquisition points when selecting a plurality of points from the acquisition points. That is, n may be m. In this case, a circle is set at all n acquisition points. That is, the normal is calculated at all n acquisition points.
  • the estimation accuracy improves. Therefore, according to the present embodiment, the surface shape can be estimated with the highest accuracy for a given acquisition point.
  • the normal line calculation unit 20 may perform this selection according to the distance from the point P i .
  • the normal calculation unit 20, starting from the point closest to the point of interest P i may be selected p number of up to a point close to the p-th as an external point. In the case of this example, all the selected p external points are closer to the attention point P i than the non-selected np ⁇ 1 acquisition points.
  • the virtual attraction of this embodiment decreases as the distance between the two points increases. That is, the attractive force acting on the circle C i is larger as it is exerted from a point closer to the point P i . Therefore, by selecting a point closer to the point P i as the external point, it is possible to use a stronger one of the attractive forces acting on the circle C i for the normal calculation. This improves the accuracy of the calculated normal. That is, according to the present embodiment, when a fixed number of external points are selected for a given target point, the normal can be calculated with higher accuracy.
  • the calculation of the normal is more accurate because the point distribution on the surface is comprehensively considered as the number of points considering the relationship with the point of interest, ie, the number of external points interacting with the circle of interest, increases. improves. Therefore, according to the present embodiment, the normal can be estimated with the highest accuracy for a given target point.
  • the normal line calculation unit 20 sets the virtual attractive force between each of the selected external points and the nearest point on the circumference of the circle to the vertex of the rigid body right cone whose bottom surface is the circle.
  • the normal force may be calculated by replacing the force acting on the FIG. 7 shows how the attraction is applied to the circle C i from the point P j which is one of the external points under the same setting as FIG. Figure 7 is a circle C i and bottom, rigid straight cone D i to vertex point A i is added.
  • the force g ij acting on the vertex A i of the rigid rigid cone D i is defined as a force that generates the same moment N ij around the point P i .
  • the normal line calculation unit 20 calculates the normal line by replacing the force f ij with the force g ij .
  • FIG. 8 shows how attraction is applied to all the corresponding points on the circle C i from all the external points under the same setting as FIG.
  • the force f ij acting on corresponding point R ij on the circle C i from the external point P j instead of the force f ij acting on corresponding point R ij on the circle C i from the external point P j, using the force g ij acting on vertex A i of the rigid right cone D i Perform the calculation.
  • the sum of force g ij with respect to j that is, the resultant force of force g ij is represented by g i . Due to this resultant force g i , the same moment around the center point P i as in the case of FIG. 5 acts on the vertex A i .
  • Rigid right circular cone D i is the moment of the resultant force g i, changing the direction by rotating around a fixed center P i. Consequently rigid straight cone D i is still facing orientation as the moment
  • FIG. 9 shows the situation when the moment of the resultant force g i becomes zero as a result of the rotation of the rigid body straight cone D i .
  • the direction facing the bottom surface of the rigid right circular cone D i is consistent with the direction facing the circle C i in FIG. Therefore, the same vector n i as in the case of FIG. 6 is obtained as the surface normal vector.
  • the calculation of the moment around each attractive point P i shown in FIG. 5 results in a vector sum g i of forces at the apex A i of the rigid body right cone D i . Therefore, the calculation can be simplified as compared with the case where the points of attraction of each attractive force are dispersed on the circle C i as shown in FIG.
  • the normal line calculation unit 20 may calculate the normal line again with the calculated normal line as an initial state.
  • the normal calculated by the normal calculation unit 20 is an approximate value, and may include a constant error although the accuracy is higher than that of the prior art. This error can generally be reduced by re-executing the normal calculation described above. Therefore, the accuracy of the calculated normal can be gradually improved by repeatedly performing the normal calculation with the normal calculated once as the initial state.
  • the apparatus may further include a positive / negative direction unifying unit that aligns the positive / negative direction of the calculated normal.
  • FIG. 10 shows the configuration of an apparatus 110 according to the present embodiment. The difference from the configuration shown in FIG. 1 is that a positive / negative direction unifying unit 30 is provided.
  • the configurations and operations of the point acquisition unit 10 and the normal line calculation unit 20 are basically the same as those described above, so the following description focuses on different parts, in particular the positive / negative direction unifying unit 30.
  • the normal vectors calculated by the normal calculation unit 20 are not unified in positive and negative directions. That is, the calculated normal vector is a mixture of one that faces the front side of the surface and one that faces the back side.
  • the positive / negative direction unifying unit 30 is for aligning the direction of this normal.
  • the normal line calculating unit 20 transmits the calculated unit normal vector set n i to the positive / negative direction unifying unit 30.
  • the positive / negative direction unifying unit 30 unifies the direction of the unit normal vector set n i received from the normal calculating unit 20.
  • a known algorithm may be used as a specific method of normal direction unification.
  • a minimum spanning tree approach may be applied.
  • a minimum spanning tree is first created from a set of acquisition points on the surface, and starting from the start point, the inner product of unit normal vectors between two points close in distance becomes positive, We will unify the positive and negative directions sequentially. Thereby, the positive and negative directions are unified in the entire normal vector.
  • FIG. 11 b shows a normal vector in which the positive and negative directions are unified.
  • the positive and negative directions of the normal can be unified.
  • this By applying this to estimation of the surface shape, it is possible to estimate the surface shape with higher accuracy.
  • the apparatus may further include a surface shape estimation unit that estimates a surface shape based on the calculated normal.
  • FIG. 12 shows the configuration of an apparatus 120 according to the present embodiment. The difference from the configuration shown in FIG. 1 is that the plane shape estimation unit 40 is provided.
  • the configurations and operations of the point acquisition unit 10 and the normal line calculation unit 20 are basically the same as those described above, so the following description focuses on different parts, in particular, the surface shape estimation unit 40.
  • the normal calculation unit 20 transmits the center coordinates of each circle and the calculated normal vectors to the surface shape estimation unit 40.
  • the surface shape estimation unit 40 estimates the surface shape based on the center coordinates of each circle received from the normal calculation unit 20 and the calculated normal vectors.
  • the technique using the RBF disclosed in Patent Document 1 may be used to estimate the surface shape, and thus the description thereof is omitted here.
  • the configuration of one embodiment of the present invention has been described above in several embodiments.
  • An example of the operation viewed from the user of the device 100 according to the present embodiment is as follows.
  • the user obtains a limited number of point clouds distributed on the surface based on some information or preliminary knowledge about the surface shape that can not be directly viewed.
  • the three-dimensional coordinates of the point group are input to the point acquisition unit 10.
  • the normal line calculation unit 20 selects an appropriate number of points from the point group input to the point acquisition unit 10, and sets a circle centered on each of the selected points.
  • the normal of the surface at the center of the circle is calculated by the physical model on which the attractive force works.
  • the user can obtain the surface shape that could not be recognized originally by executing the RBF calculation using the calculated normal.
  • the shape of the surface can be estimated with high accuracy based on a finite point group on the surface.
  • One useful application of the present invention is medical imaging.
  • techniques for grasping the shape and structure inside the human body such as X-ray CT and MRI have been developed, but some human tissues can not be directly visualized even using these techniques.
  • the mesentery is an example of such a tissue.
  • Fig. 13 schematically shows the human mesentery M.
  • the mesentery is a tissue with a two-layered structure that supports the intestinal tract in the abdominal cavity. Understanding the shape and structure of the mesentery is important for safe and rapid abdominal surgery.
  • the mesentery has a similar X-ray absorption rate to nearby tissue. Therefore, in CT images, it is very difficult to distinguish the surrounding tissue and to clearly grasp the whole image.
  • the blood vessels run inside the dual structure of the membrane, and the terminal end of the membrane contacts the large intestine and the main blood vessels, it is possible to confirm the rough membrane structure based on the blood vessel shape based on the contrast CT image It is. That is, a user with anatomical knowledge can estimate the positions of a plurality of points on the membrane with respect to the mesentery that is not directly reflected in the contrast CT image.
  • FIG. 14 shows a point cloud in which a user such as a specialist having anatomic knowledge as described above estimates and plots the position of mesentery on a contrast-enhanced CT image.
  • FIG. 15a shows the shape of the mesentery estimated using the prior art described in Patent Document 1 based on the point cloud of FIG. As can be read from FIG. 15a, an error has occurred, such as the appearance of a hole that does not exist in the right part of the image.
  • FIG. 15b shows the shape of the mesentery estimated using the device according to the present embodiment based on the point cloud of FIG. It can be seen from FIG. 15 b that the estimation accuracy is greatly improved, and an image that more faithfully reproduces the actual shape of the mesentery can be obtained.
  • the shape of the film can be estimated with high accuracy based on the point group on the film on the medical image plotted by the user.
  • the present apparatus re-creates a reconstructed image in which a medical image is reconstructed based on the surface shape estimated by the surface shape estimation unit 40 of the apparatus shown in FIG. And an image output unit that outputs the configuration image.
  • FIG. 16 shows the configuration of an apparatus 130 according to this embodiment. The difference from the configuration shown in FIG. 12 is that an image reconstruction unit 50 and an image output unit 60 are provided.
  • the image reconstruction unit 50 receives the estimated surface shape from the surface shape estimation unit 40, and based on this, creates a reconstructed image reconstructed as a medical image.
  • an appropriate existing technique may be used according to the purpose of use such as a clinic or a research.
  • polygon data may be created using rendering techniques such as the marching cube method to create a three-dimensional reconstructed image.
  • the image reconstruction unit 50 transmits the created reconstructed image to the image output unit 60.
  • the image output unit 60 outputs the reconstructed image received from the image reconstruction unit 50 to the outside.
  • the image output unit 60 may be, for example, an image output device such as a display.
  • a user such as a medical professional can monitor a medical image reconstructed from a point cloud.
  • the image reconstruction unit 50 may create a reconstructed image in which the range of the film is cut out based on the surface shape estimated by the surface shape estimation unit 40.
  • the surface shape estimated by the surface shape estimation unit 40 generally continues to infinity beyond this end.
  • the boundary is used as a cutting line by giving information indicating that a specific point is a point on the boundary, in addition to the position coordinates of each point, with respect to the point group on the film. The membrane can be cut out of the shape.
  • a user such as a specialist who has the above-described anatomical knowledge estimates the position of mesentery on a contrast-enhanced CT image and plots a point cloud.
  • information hereinafter referred to as “identification information” indicating whether each point is inside the film or on the boundary of the film is given.
  • the point acquisition unit 10 of the device 130 of FIG. 16 acquires position coordinates and identification information of an acquisition point.
  • the point acquisition unit 10 transmits the position coordinates of the acquisition point to the normal line calculation unit 20 and transmits identification information of the acquisition point to the image reconstruction unit 50.
  • the image reconstruction unit 50 cuts out the range of the target film to create a reconstructed image.
  • the process described in Patent Document 1 may be used.
  • practicability is improved by cutting out and reconstructing only the target film.
  • the identification information may be a label indicating an anatomical type according to the site of the membrane. That is, in the present embodiment, this label is used as identification information.
  • FIG. 17 shows an example of a point cloud labeled with an anatomical type of mesentery.
  • four types of labels from type 1 to type 4 are given on the CT image of the mesentery.
  • Type 1 points are points on the line tangent to the large intestine and on the membrane border.
  • the point of type 2 is a point on the iliac artery and on the border of the membrane.
  • the point of type 3 is a point on the inferior mesenteric vein to the inferior mesenteric artery, which is on the membrane boundary.
  • the point of type 4 is the point inside the membrane. In this case, since the points labeled with types 1, 2 and 3 are on the boundary of the membrane, these points can be used as cut lines to cut out the range of the target membrane.
  • the point acquisition unit 10 of the device 130 of FIG. 16 transmits the position coordinates of the acquisition point to the normal line calculation unit 20 and transmits the label of the acquisition point to the image reconstruction unit 50.
  • the image reconstruction unit 50 uses the points given the labels of types 1, 2 and 3 received from the point acquisition unit 10 from the surface shape received from the surface shape estimation unit 40 as a cutting line, and the range of the target film To create a reconstructed image.
  • efficient extraction processing can be realized because identification information in a simple format such as a label is used.
  • an apparatus includes a point acquisition unit and a slope calculation unit.
  • the point acquisition unit acquires a plurality of acquisition points from among the point cloud distributed in the three-dimensional space.
  • the inclination calculation unit selects a plurality of points from among the acquired points, and selects and selects a plurality of external points from among the acquired points other than the center of the circle for each of the circles centered on the selected points.
  • virtual attraction acts between each of the external points and the nearest point on the circumference of the circle, the inclination of the circle is calculated based on the attraction.
  • the inclination of the circle calculated in this manner is considered to globally represent the spatial relationship between the point at the center of the circle and the other point group.
  • Various applications are possible by combining the information on the overall relationship between the points with, for example, the information on the spatial density distribution of the point cloud.
  • the normal line calculating process may be repeated a predetermined number of times. Generally, by repeating the normal calculation, the calculated normal converges to a constant value. The normal calculation process may be repeated until the calculated normal converges. Alternatively, means may be provided to compare the calculated normal line with the previously calculated normal line, and the calculation process may be repeated until the difference between the two is smaller than a predetermined threshold value. According to this modification, a target accuracy is provided in advance, and it is possible to calculate a normal of accuracy that satisfies the target.
  • the image output unit 60 may output the reconstructed image as it is, but is not limited to this, and may output another image data superimposed on the reconstructed image.
  • the texture of the original CT image data including blood vessels may be superimposed and output. According to this modification, the user can simultaneously monitor blood vessels traveling along the mesentery in addition to the reconstructed mesentery.
  • the same polygon representation as the reconstructed mesentery image may be given to the CT image data of the blood vessel.
  • the user can monitor the blood vessel while dynamically deforming with the mesentery.
  • the image reconstruction unit 50 cuts out the reconstructed image separated by a predetermined distance or more from the portion where the point group ends. It may be removed. If the number of point clouds is not sufficient for the curvature of the boundary portion of the target film, the identification information alone may not be sufficiently extracted, and an unnecessary part of the image may be left. According to this modification, it is possible to completely realize the cutout of the range of the target film by removing the unnecessary image that can not be cut off only by the identification information.
  • a variety of means may be used when a user with specialized knowledge as described above estimates the position on the surface and plots a point cloud. For example, points may be input using a mouse or the like, and plotting may be performed for each point. Alternatively, as a modification, a line may be input using a touch panel or the like, and a point group may be sampled and plotted from this line. According to this modification, the accuracy of plotting can be improved.
  • the points plotted on the previous tomographic screen may be displayed dimly on the current tomographic screen.
  • points plotted on the two or more previous tomographic screens may be displayed while sequentially decreasing the density as the tomographic screens are traced back one by one.
  • the plotted points may be three-dimensionally displayed according to the depth of the tomographic screen using a technique such as holography.
  • points plotted on each tomographic image may be connected to create a schematic image of a film having a wire frame structure.
  • the surface shape that can not be directly viewed in addition to human tissue that does not appear in X-ray CT, articles and natural objects that appear in photographs with unclear contrast, samples to be subjected to nondestructive inspection, objects that are buried in the ground or in water There are ultrasound images and the like.
  • the point cloud on the surface can be estimated by supplementing the expert's knowledge and the like with respect to these surfaces not directly visualized, and the original surface shape can be estimated and reconstructed by using the technique of the present invention. .
  • 10 point acquisition unit 20 normal calculation unit, 30 positive / negative direction unification unit, 40 plane shape estimation unit, 50 image reconstruction unit, 60 image output unit, 100 devices, 110 devices, 120 devices
  • the present invention is applicable to the industrial field that requires estimating and reconstructing the surface shape that can not be directly viewed.

Abstract

L'invention concerne une unité d'acquisition de points (10) d'un dispositif (100) qui acquiert une pluralité de points d'acquisition sur une surface. L'invention concerne également une unité de calcul de vecteur normal (20) qui : sélectionne une pluralité de points parmi les points d'acquisition ; sélectionne une pluralité de points externes parmi les points d'acquisition situés à l'extérieur du centre d'un cercle par rapport à des cercles situés autour de chacun des points sélectionnés ; et, lorsqu'une attraction virtuelle agit entre chacun des points externes sélectionnés et des points qui sont les plus proches de la circonférence des cercles, calcule un vecteur normal dans la direction allant vers le cercle en raison de l'attraction en tant que vecteur normal au centre du cercle sur la surface.
PCT/JP2018/044648 2017-12-08 2018-12-05 Dispositif de calcul de vecteur normal, procédé et programme WO2019111920A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2017236423A JP2021028574A (ja) 2017-12-08 2017-12-08 法線を算出するための装置、方法およびプログラム
JP2017-236423 2017-12-08

Publications (1)

Publication Number Publication Date
WO2019111920A1 true WO2019111920A1 (fr) 2019-06-13

Family

ID=66750220

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2018/044648 WO2019111920A1 (fr) 2017-12-08 2018-12-05 Dispositif de calcul de vecteur normal, procédé et programme

Country Status (2)

Country Link
JP (1) JP2021028574A (fr)
WO (1) WO2019111920A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030067461A1 (en) * 2001-09-24 2003-04-10 Fletcher G. Yates Methods, apparatus and computer program products that reconstruct surfaces from data point sets
JP2006331177A (ja) * 2005-05-27 2006-12-07 Dainippon Printing Co Ltd 表面にパイプを有する物体の三次元形状データの作成装置および作成方法
JP2011209055A (ja) * 2010-03-29 2011-10-20 Saki Corp:Kk 検査装置
WO2017043503A1 (fr) * 2015-09-11 2017-03-16 国立研究開発法人科学技術振興機構 Dispositif d'estimation de structure, procédé d'estimation de structure, et programme

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20030067461A1 (en) * 2001-09-24 2003-04-10 Fletcher G. Yates Methods, apparatus and computer program products that reconstruct surfaces from data point sets
JP2006331177A (ja) * 2005-05-27 2006-12-07 Dainippon Printing Co Ltd 表面にパイプを有する物体の三次元形状データの作成装置および作成方法
JP2011209055A (ja) * 2010-03-29 2011-10-20 Saki Corp:Kk 検査装置
WO2017043503A1 (fr) * 2015-09-11 2017-03-16 国立研究開発法人科学技術振興機構 Dispositif d'estimation de structure, procédé d'estimation de structure, et programme

Also Published As

Publication number Publication date
JP2021028574A (ja) 2021-02-25

Similar Documents

Publication Publication Date Title
US11547499B2 (en) Dynamic and interactive navigation in a surgical environment
KR101700847B1 (ko) 영상수술 트레이닝 제공방법 및 그 기록매체
CN108369736A (zh) 用于根据2d/2.5d术中图像数据计算切除的组织体积的方法和系统
AU2013317201A1 (en) Device and method for displaying three-dimensional image, and program
US8149237B2 (en) Information processing apparatus and program
JP7038377B2 (ja) 生体モデル生成装置、生体モデル生成方法、および生体モデル生成プログラム
Saito et al. [POSTER] deformation estimation of elastic bodies using multiple silhouette images for endoscopic image augmentation
TW202105406A (zh) 用於推薦手術程序之參數的系統及方法
US9092666B2 (en) Method and apparatus for estimating organ deformation model and medical image system
CN109191475A (zh) 椎体终板分割方法、装置及计算机可读存储介质
Ard et al. NIVR: Neuro imaging in virtual reality
WO2022219631A1 (fr) Systèmes et procédés de reconstruction d'images 3d à partir d'images ultrasonores et de caméras
US20130195323A1 (en) System for Generating Object Contours in 3D Medical Image Data
WO2019111920A1 (fr) Dispositif de calcul de vecteur normal, procédé et programme
Kolagunda et al. A mixed reality guidance system for robot assisted laparoscopic radical prostatectomy
CN116993805A (zh) 面向手术规划辅助的术中剩余器官体积估计系统
EP3432310A1 (fr) Procédé et système pour simulation d'intervention chirurgicale pré-opératoire
JP2015136480A (ja) ３次元医用画像表示制御装置およびその作動方法並びに３次元医用画像表示制御プログラム
US20170018079A1 (en) Image processing device, method, and recording medium having stored therein program
US20200234494A1 (en) Structure estimating apparatus, structure estimating method, and computer program product
JP2019010382A (ja) 画像位置合わせ装置、方法およびプログラム
Liu et al. Optimization of reconstruction of 2D medical images based on computer 3D reconstruction technology
Dangi et al. Endocardial left ventricle feature tracking and reconstruction from tri-plane trans-esophageal echocardiography data
Bohak et al. Neck veins: an interactive 3D visualization of head veins
EP2725546B1 (fr) Procédé et appareil de traitement d'affichage

Legal Events

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

Ref document number: 18885924

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 18885924

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP