JP2011083499A - Radiation imaging apparatus, and phantom device used therein - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To acquire the positional relationship between an X-ray tube and a detector, the positional relationship between the X-ray tube and a reference tomographic plane and a parameter such as an X-ray projection direction to easily grasp the positional relationship between them in an imaging space at every radiation imaging apparatus. <P>SOLUTION: In the radiation imaging apparatus, a phantom (171) for calibrating a parameter for defining the imaging space including the three-dimensional position of the reference tomographic plane is used. Radiation is discharged from a radiation source where the phantom is arranged in the imaging space to collect data and an image for calibration is formed on the basis of the data. Further, the parameter is calibrated on the basis of the data of known amount possessed by the phantom and the projection position data of the phantom obtained from the image for calibration. <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a radiation imaging apparatus for imaging an object using radiation and a phantom apparatus used in the apparatus, and more particularly to a radiation imaging apparatus and apparatus for acquiring an image such as a panoramic image of an object based on a tomosynthesis method. The present invention relates to a calibration phantom device used in the above.

  In recent years, tomography of a subject based on tomosynthesis has been actively performed. Although the principle of this tomosynthesis method has been known for a long time (see, for example, Patent Document 1), in recent years, a tomographic method has also been proposed in which it is desired to enjoy the simplicity of image reconstruction based on the tomosynthesis method. (For example, see Patent Document 2 and Patent Document 3). Many examples are seen in dental and mammography (see, for example, Patent Document 4, Patent Document 5, and Patent Document 6).

  Conventionally, there is a dental panoramic imaging device as one of medical imaging devices to which this tomosynthesis method is preferably applied. This panoramic imaging apparatus has a tomographic plane (referred to as a reference tomographic plane) according to a trajectory mechanically set in the imaging space due to restrictions on the movement of the detector so that the reference tomographic plane is in focus. It has become. For this reason, the reconstructed image is not blurred only when the dentition is located along the reference tomographic plane. However, if the dentition is displaced from the reference tomographic plane, the image is blurred. Therefore, if you want to see the unclear part with high accuracy, reposition the subject so that the blurred part looks clear and re-collect the data, or perform intraoral photography of the blurred part to obtain a clearer part. I got an image.

  On the other hand, in recent years, as in Patent Document 7, a detector capable of collecting images at high speed (for example, 300 FPS) is used to capture all of the detected data into a computer, and a tomosynthesis method is used to soften the tomographic plane. An X-ray panoramic imaging apparatus that can be freely changed has been developed. In the case of this apparatus, information (gain) of distances of a plurality of tomographic planes parallel to the detection surface (X-ray incident surface) of the detector is obtained in advance using a phantom or by theoretical calculation. At the time of imaging, data is collected while rotating a pair of an X-ray tube and a detector around the jaw of the subject. The position of the rotation center at this time approaches or leaves the dentition. The collected data is processed by the tomosynthesis method using the above-described distance information, thereby creating an image with less blur.

JP-A-57-203430 JP-A-6-88790 JP-A-10-295680 JP-A-4-144548 JP2008-11098A US Patent Publication US2006 / 0203959 A1 JP2007-136163A

  In the case of the panoramic imaging apparatus described in Patent Document 7 described above, a panoramic image is created on the assumption that a plurality of tomographic planes to be subjected to the tomosynthesis method are parallel to the detection plane of the detector. For this reason, when the tomographic plane is changed, such as when creating a panoramic image of another tomographic plane, the magnification of the image changes, so distortion caused by this change occurs in the vertical direction of the image (vertical direction of the dentition). Arise. After the image is digitized, at least when the dentition is accurately positioned on the reference tomographic plane, it is possible to create an image having no distortion in both the vertical and horizontal directions only at the center of the front tooth. However, if the positioning condition is not met, the image will always be distorted. In addition, when the dentition is not located along the reference tomographic plane, lateral blurring also occurs in the reconstructed panoramic image. Therefore, even if an image with little blur is generated in focus, the distance between the two points is not correctly depicted on the image because the image is distorted in the vertical direction. Of course, the distance between the two points cannot be measured accurately.

  This distortion is caused by the fact that the magnification in the vertical direction changes according to the rotation angle of the X-ray tube / detector pair with respect to the dentition during data collection. Of course, this distortion also occurs when the fault plane is changed. Therefore, in the case of the panoramic imaging device described in Patent Document 7, the created panoramic image is not suitable for quantitative measurement, and it is difficult to see time-series changes such as subtraction, and its clinical application is limited. It has been. This is one reason why the conventional panoramic imaging device cannot be used as a substitute for true intraoral imaging and does not reach dental CT.

  By the way, the panorama image pickup device has a difference in mechanical operation for each device even if it is manufactured by the same manufacturer as well as there is a difference between the manufacturers. In particular, in the mechanism that rotates the pair of the X-ray tube and the detector, the influence of such variation affects the reconstructed panoramic image as an error. For this reason, factors such as the positional relationship between the X-ray tube and the detector, the movement state of the rotation center of the X-ray tube and detector pair, the speed of the movement, and the X-ray projection direction are as designed. For example, it is necessary to check whether each device has the variation information. Obtaining this information means that the positional relationship in the space (imaging space) in which the pair of the X-ray tube and the detector is rotated is grasped three-dimensionally. This information should be grasped for each device, and the obtained information should be reflected in the panorama reconstruction. Conventionally, there is no such need, and there is no such mechanism and method at present. there were.

  The present invention has been made in view of the above-described conventional situation. For example, even when there are differences in the rotation trajectories of the X-ray tube and the detector due to different manufacturers, there is no difference between the X-ray tube and the detector. To acquire parameters such as the positional relationship between the X-ray tube and the reference tomographic plane, and the X-ray projection direction so that the positional relationship in the imaging space can be easily grasped for each apparatus. Is the purpose.

  In order to achieve the above-described object, according to one aspect of the present invention, a radiation source that emits radiation, a radiation source that is disposed opposite to the radiation source, and that corresponds to the radiation when the radiation is incident are provided. A detector that outputs the two-dimensional data of the digital electric quantity in units of frames, a pair of the radiation source and the detector, the detector, or the object, the radiation source, the detector, And a moving means that moves relative to the remaining elements of the object, and a pair of the radiation source and the detector, the detector, or the object is moved by the moving means. In addition, data collection means for collecting the data output from the detector in units of frames, the data collected by the data collection means, and an imaging sky existing between the radiation source and the detector And a first image creating means for creating a three-dimensional image of the imaging region of the object using a reference tomographic plane set to the reference tomographic plane. A phantom for calibrating the parameters defining the imaging space including the three-dimensional position of the image, and the data collection means collects radiation when the radiation source emits radiation with the phantom arranged in the imaging space. Based on second image creation means for creating an image for calibration based on the obtained data, information on a known amount of the phantom and projection position information of the phantom obtained from the calibration image And calibration means for calibrating the parameters. Further, the image forming apparatus may further include modeling means for modeling the reference tomographic plane using the parameters calibrated by the calibration means.

  In order to achieve the above-described object, the present invention is another embodiment in which a radiation source that emits a radiation source and a detector that detects the radiation as an electrical signal are opposed to each other with an object interposed therebetween, and the radiation While rotating a pair of the source and the detector around the object, the radiation emitted from the radiation source and transmitted through the object is detected as an electric signal by the detector, and the electric signal is based on the electric signal. Provided is a phantom device for calibrating a three-dimensional spatial positional relationship between the radiation source and the detector in a radiation imaging apparatus that creates an image of an imaging region of an object. This phantom device. A first phantom for obtaining parameters of a three-dimensional positional relationship between the radiation source and the detector; and a three-dimensional relationship with respect to the object during relative movement of the radiation source and the radiation detector. And a second phantom for acquiring parameters of various positional relationships.

  As described above, according to the present invention, for example, even when there are differences in the rotation trajectories of the X-ray tube and the detector due to different manufacturers, the phantom is used to connect the X-ray tube and the detector. It is possible to acquire parameters such as the positional relationship between them, the positional relationship between the X-ray tube and the reference tomographic plane, and the X-ray projection direction. Therefore, the positional relationship in the imaging space can be easily grasped for each device.

In the accompanying drawings,
FIG. 1 is a perspective view showing an outline of the overall configuration of a panoramic imaging apparatus using X-rays as a radiation imaging apparatus according to an embodiment of the present invention. FIG. 2 is a rotation center when a dentition of a subject targeted by the panoramic imaging apparatus according to the embodiment, a 3D reference tomographic plane set in the dentition, and a pair of an X-ray tube and a detector rotate. FIG. FIG. 3 is a perspective view for explaining the geometry of the X-ray tube, 3D reference tomographic plane, and detector in the panoramic imaging apparatus. FIG. 4 is a block diagram for explaining an outline of an electrical configuration of the panoramic imaging apparatus. FIG. 5 is a flowchart showing an outline of processing for imaging that is executed in cooperation by the controller of the panoramic imaging apparatus and the image processor. FIG. 6 is a diagram for explaining the positional relationship among an X-ray tube, a 3D reference tomographic plane, a rotation center, and a detector. FIG. 7 is a graph illustrating the relationship between frame data and the panorama image mapping position. FIG. 8 is a diagram schematically illustrating an example of a reference panoramic image. FIG. 9 is a diagram schematically illustrating an example of an image when an ROI is set for a reference panoramic image. FIG. 10 is a flowchart for explaining the outline of processing for identifying the actual position and shape of a tooth executed by the image processor. FIG. 11 is a diagram for explaining a difference in projection angle from the same position in the Z-axis direction on the 3D panoramic image to the X-ray tube according to a change in the rotation center of the pair of the X-ray tube and the detector. FIG. 12 is a diagram schematically illustrating an example of a 3D reference image. FIG. 13 is a perspective view illustrating a plurality of parallel tomographic planes added to the 3D reference tomographic plane. FIG. 14 shows differences in positions on a plurality of tomographic planes when projected onto the X-ray tube from the same position in the Z-axis direction on the 3D panoramic image as the rotation center of the pair of the X-ray tube and the detector changes. Illustration to explain. FIG. 15 (1) is a diagram illustrating the process of identifying the optimally focused tomographic plane for each position on the 3D reference image in cooperation with FIGS. 15 (1) and 15 (2). The figure which shows-(D). FIG. 15 (2) is a diagram illustrating the process of identifying the optimally focused tomographic plane for each position on the 3D reference image in cooperation with FIGS. 15 (1) and 15 (2). The figure which shows-(H). FIG. 16 is a graph illustrating the result of frequency analysis in the process of specifying the optimum focus position. FIG. 17 is a graph illustrating an example of the position of the tomographic plane of the optimum focus in the process of specifying the optimum focus position. FIG. 18 is a graph illustrating a frequency characteristic pattern that changes according to a tomographic plane position. FIG. 19 is a diagram illustrating a state where the actual tooth position is deviated from the 3D reference tomographic plane. FIG. 20 is a diagram illustrating a state in which the tooth is shifted from the position of the 3D reference tomographic plane to the actual position according to the magnification ratio. FIG. 21 is a diagram for explaining a state in which teeth are shifted from the position of the 3D reference tomographic plane to the actual position according to the magnitude of the enlargement ratio. FIG. 22 is a diagram illustrating a state in which teeth are shifted from the position of the 3D reference tomographic plane to the actual position according to the magnitude of the enlargement ratio. FIG. 23 is a perspective view illustrating processing for moving a processing point on the 3D reference image for the position identification position. FIG. 24 is a perspective view illustrating identification of a tomographic plane position of an optimum focus specified for each processing point and abnormal identification thereof. FIG. 25 is a diagram schematically showing a 3D autofocus image created by identifying the optimum focus tomographic plane position and smoothing. FIG. 26 is a diagram for explaining the concept of processing for projecting a 3D autofocus image onto a 3D reference tomographic plane. FIG. 27 is a schematic diagram schematically illustrating an image projected on a 3D reference tomographic plane and an ROI set there. The figure explaining the concept of the process which projects a 3D autofocus image on the two-dimensional surface of a reference | standard panorama image. FIG. 28 is a diagram schematically illustrating a 2D reference image and an ROI set there. FIG. 30 is a diagram for explaining the principle of panoramic image reconstruction based on the tomosynthesis method. FIG. 31 is a graph showing the relationship between the frame data output from the detector and the projection angle. FIG. 32 is a perspective view for explaining the outline of the first phantom. FIG. 33 is a diagram for explaining the positional relationship between the X-ray tube and the detector when the first phantom is used. FIG. 34 is a plan view for explaining the outline of the second, third, and fourth phantoms. FIG. 35 is a side view illustrating the outline of the second, third, and fourth phantoms. FIG. 36 is a diagram for explaining the positional relationship between the X-ray tube and the detector when the second phantom is used. FIG. 37 is a diagram for explaining a deviation in projection angle. FIG. 38 is a diagram for explaining a method for calculating an actual value of a projection angle. FIG. 39 is a flowchart for explaining an outline of calibration using a phantom. FIG. 40 is a diagram exemplifying a positional relationship between modeling of a 3D reference tomographic plane and an entity to be measured. FIG. 41 is a perspective view illustrating a modification of the phantom. FIG. 42 is a diagram for explaining the positional relationship between the X-ray tube and the detector when the phantom shown in FIG. 41 is used. FIG. 43 is a diagram for explaining another modification of the phantom. 44 is a side view schematically showing the phantom shown in FIG. 43. FIG.

  Embodiments of the present invention will be described below with reference to the accompanying drawings.

  With reference to FIGS. 1-40, one Embodiment of the three-dimensional position identification apparatus, radiation imaging device, and imaging method using a radiation which concerns on this invention is described. Since these apparatuses and methods are implemented as a dental panoramic imaging apparatus using X-rays in this embodiment, the panoramic imaging apparatus will be described in detail below.

  FIG. 1 shows an external appearance of such a panoramic imaging device 1. This panoramic imaging apparatus 1 scans the subject's jaw with X-rays, identifies the actual position (actual position) of the dentition of the three-dimensional structure in the jaw from the digital amount of X-ray transmission data, And the panoramic image which compensated the change (difference) of the enlargement factor mentioned later of the dentition is created. In addition to this basic performance, this panoramic imaging apparatus 1 can provide epoch-making performance, such as being able to perform various forms of display and measurement from such panoramic images. In addition, an X-ray exposure dose can be reduced for the subject, and an imaging device that is easy for the operator to use can be provided. To obtain the basic performance described above, the tomosynthesis method is used.

  An outline of the configuration of the panorama imaging apparatus 1 will be described. As shown in FIG. 1, the panoramic imaging apparatus 1 includes a housing 11 that collects data from a subject (patient) P, for example, in a standing posture of the subject P, and data collection performed by the housing 11. Consists of a computer that controls and captures the collected data to create a panoramic image and performs post-processing of the panoramic image interactively or automatically with an operator (doctor, technician, etc.) And a control / arithmetic unit 12.

  The housing 11 includes a stand unit 13 and a photographing unit 14 that can move up and down with respect to the stand unit 13. The imaging unit 14 is attached to the support column of the stand unit 13 so as to be movable up and down within a predetermined range.

  Here, for convenience of explanation, an XYZ orthogonal coordinate system having the Z axis as the longitudinal direction of the stand unit 13, that is, the vertical direction, is set for the panoramic imaging device. Note that, for a two-dimensional panoramic image to be described later, the horizontal axis direction is expressed as j axis, and the vertical axis direction is expressed as i axis (= Z axis).

  The imaging unit 14 includes a vertical movement unit 23 having a substantially U-shape when viewed from the side, and a rotation unit 24 supported by the vertical movement unit 23 so as to be rotatable (rotatable). The vertical movement unit 23 is arranged in the Z-axis direction (vertical axis direction) over a predetermined range in the height direction through a vertical drive mechanism (for example, a motor and a rack and pinion) (not shown) installed on the stand unit 13. It can be moved. A command for this movement is issued from the control / arithmetic unit 12 to the vertical movement drive mechanism.

  As described above, the vertical movement unit 23 is substantially U-shaped when viewed from one side surface thereof, and includes an upper arm 23A and a lower arm 23B on each of the upper and lower sides, and upper and lower arms 23A and 23B. The connecting vertical arm 23C is integrally formed. The vertical arm 23C is supported by the above-described stand portion 13 so as to be movable up and down. Among the arms 23A to 23C, the upper arm 23A and the vertical arm 23C cooperate to form a photographing space (real space). Inside the upper arm 23A, a rotational drive mechanism 30A (for example, an electric motor and a reduction gear) for rotational drive is installed. The rotational drive mechanism 30A receives a rotational drive command from the control / arithmetic unit 12. The output shaft of the rotation drive mechanism 30A, that is, the rotation shaft of the electric motor is disposed so as to protrude downward (downward in the Z-axis direction) from the upper arm 23A, and the rotation unit 24 can rotate on this rotation shaft. Is bound to. That is, the rotation unit 24 is suspended by the vertical movement unit 23 and is rotated by being energized by the drive of the rotation drive mechanism 30A.

  The rotation drive mechanism 30A is connected to the movement mechanism 30B. The moving mechanism 30B is composed of an electric motor, a gear, etc. (not shown). The moving mechanism 30B also operates in response to a rotation driving command from the control / arithmetic unit 12, and is configured to be able to move the rotation driving mechanism 30A, that is, the rotation unit 24 along the XY plane. Thereby, the locus of the rotation center of a pair of an X-ray tube and a detector, which will be described later, can be moved two-dimensionally along a constant trajectory within a predetermined range along the XY plane.

  On the other hand, the lower arm 23B is extended to have a predetermined length in the same direction as the upper arm 23A, and a chin rest 25 is formed at the tip thereof. A bite block 26 (or simply called a bite) is detachably attached to the chin rest 25. The subject P holds this byte block 26. For this reason, the chin rest 25 and the bite block 26 serve to fix the oral cavity of the subject P.

  The rotating unit 24 has an appearance that is formed in a substantially U shape when viewed from one side surface in the state of use, and is rotatable to the motor output shaft of the upper arm 23A so that its open end can be rotated downward. It is attached. Specifically, a horizontal arm 24A that rotates (rotates) in the horizontal direction, that is, substantially parallel in the XY plane, and left and right vertical arms (first axis) extending downward (Z-axis direction) from both ends of the horizontal arm 24A. Vertical arm, second vertical arm) 24B, 24C. The horizontal arm 24 and the left and right first and second arms 24B and 24C are located in a photographing space (actual space) and are driven and operated under the control of the control / arithmetic apparatus 12.

  An X-ray tube 31 as a radiation emission source is equipped at the lower end portion inside the first vertical arm 24B. The X-ray tube 31 is constituted by a rotating anode X-ray tube, for example, and radiates X-rays radially from the target (anode) toward the second vertical arm 24C. The focal point of the electron beam colliding with the target is as small as about 0.5 mm to 1 mm in diameter, and therefore the X-ray tube 31 has a point-like X-ray source. On the X-ray exit side of the X-ray tube 31, a slit-shaped collimator that narrows a relatively thin beam-shaped X-ray incident on the detector 32 into an actual collection window (for example, a window having a width of 5.0 mm). 33 is attached. In addition, you may include this collimator 33 in the element which comprises a radiation emission source.

  On the other hand, a digital X-ray detector 32 in which X-ray detection elements as radiation detection means are arranged two-dimensionally (for example, in a 64 × 1500 matrix) is provided at the lower end of the second vertical arm 24C. X-rays incident from this incident window are detected. As an example, the detector 32 has a vertically long detection surface (for example, 6.4 mm wide × 150 mm long) made of CdTe. Since this embodiment employs the tomosynthesis method, the detector 32 must have a plurality of X-ray detection elements also in the lateral (width) direction.

  The detector 32 is arranged in the vertical direction with its vertical direction coinciding with the Z-axis direction. The effective width in the lateral direction of the detector 32 is set to, for example, about 5.0 mm by the collimator 33 described above. The detector 32 can collect incident X-rays as image data of digital electric quantity corresponding to the amount of the X-rays at a frame rate of 300 fps (one frame is, for example, 64 × 1500 pixels), for example. Hereinafter, this collected data is referred to as “frame data”.

  At the time of imaging, the pair of the X-ray tube 31 and the detector 32 are positioned so as to face each other with the oral cavity portion of the subject P interposed therebetween, and each pair is driven to rotate around the oral cavity portion. . However, this rotation is not a rotation that draws a simple circle. That is, the pair of the X-ray tube 31 and the detector 32 has a mountain-shaped constant trajectory in which the rotation center RC of the pair is formed by connecting two circular arcs inside the substantially horseshoe-shaped dentition as shown in FIG. Is driven to rotate. This constant trajectory is pre-designed to focus on the X-ray focus and follow the 3D reference tomographic plane along the standard tooth shape and size of the oral cavity (hereinafter referred to as the 3D reference tomographic plane). Orbit. When the X-ray focal point is made to follow the 3D reference tomographic plane SS, the X-ray tube 31 and the detector 32 do not necessarily rotate at the same angular velocity when viewed from the 3D reference tomographic plane. That is, this rotation can be referred to as “movement along the dentition”, and is rotated while appropriately changing the angular velocity.

  As described above, the locus on the XY plane when the 3D reference tomographic plane is viewed from the Z-axis direction has a substantially horseshoe shape, and an example is shown in FIG. The locus of this 3D reference fault plane is also known, for example, from the document “R. Molteni,“ A universal test phantom for dental panoramic radiography ”Medica Mudi, vol. 36, no. 3, 1991”. The geometric positional relationship between the X-ray tube 31, the 3D reference tomographic plane SS, the detector 32, the rotation axis AXz, and the rotation center RC through which the rotation axis AXz passes is as shown in FIG. The 3D reference tomographic plane SS is parallel to the entrance of the detector 32 (X-ray detection plane Ldet: see FIG. 6), is a curved cross section along the Z-axis direction, and is elongated and has a rectangular shape when developed in two dimensions. It is set as a cross section.

  FIG. 4 shows an electrical block diagram for control and processing of this panoramic imaging apparatus. As shown in the figure, the X-ray tube 31 is connected to the control / arithmetic apparatus 12 via a high voltage generator 41 and a communication line 42, and the detector 32 is connected to the control / arithmetic apparatus 12 via a communication line 43. ing. The high voltage generator 41 is provided in the stand unit 13, the vertical movement unit 23, or the rotation unit 24. It is controlled according to the shooting conditions and the sequence of exposure timing.

  The control / arithmetic unit 12 is constituted by, for example, a personal computer capable of storing a large amount of image data in order to handle a large amount of image data, for example. That is, the control / arithmetic unit 12 has, as its main components, interfaces 51, 52, 62, a buffer memory 53, an image memory 54, a frame memory 55, which are connected to each other via an internal bus 50. An image processor 56, a controller (CPU) 57, and a D / A converter 59 are provided. An operation device 58 is communicably connected to the controller 57, and a D / A converter 59 is also connected to a monitor 60.

  Of these, the interfaces 51 and 52 are connected to the high voltage generator 41 and the detector 32, respectively, and communicate control information and collected data exchanged between the controller 57 and the high voltage generator 41 and the detector 32. Mediate. Another interface 62 connects the internal bus 50 and a communication line, and the controller 57 can communicate with an external device. As a result, the controller 57 can also capture an intraoral image captured by an external intraoral X-ray imaging apparatus, and convert a panoramic image captured by the present imaging apparatus to an external device according to, for example, the DICOM (Digital Imaging and Communications in Medicine) standard. It can be sent to the server.

  The buffer memory 53 temporarily stores digital frame data received from the detector 32 via the interface 52.

  Further, the image processor 56 is placed under the control of the controller 57 and interactively creates a panoramic image of a predetermined 3D reference tomographic plane provided by the apparatus side and processes for subsequent use of the panoramic image with the operator. It has a function to execute. A program for realizing this function is stored in the ROM 61 in advance. Therefore, the ROM 61 functions as a recording medium that stores the program according to the present invention. This program may be stored in the ROM 61 in advance. However, in some cases, the program may be installed in a recording medium such as a RAM (not shown) via a communication line or a portable memory. .

  In the present embodiment, the 3D reference tomographic plane described above is prepared in advance on the apparatus side, and can be calibrated before shipment from the factory, before imaging, or during maintenance and inspection. Note that the 3D reference tomographic plane may be selected before photographing from a plurality of tomographic planes prepared in advance on the apparatus side. In other words, although it is a fixed cross section as a 3D reference tomographic plane, the selection operation enables the position of the 3D reference tomographic plane to be changed within a certain range in the depth (front and back) direction of the dentition. Also good.

  Frame data and image data processed by the image processor 56 or being processed are stored in the image memory 54 so as to be readable and writable. For the image memory 54, a large-capacity recording medium (nonvolatile and readable / writable) such as a hard disk is used. The frame memory 55 is used to display reconstructed panoramic image data, post-processed panoramic image data, and the like. The image data stored in the frame memory 55 is called by the D / A converter 59 at a predetermined cycle, converted into an analog signal, and displayed on the screen of the monitor 60.

  The controller 57 controls the overall operation of the constituent elements of the apparatus in accordance with a program responsible for overall control and processing stored in the ROM 61 in advance. Such a program is set so as to interactively receive operation information for each control item from the operator. Therefore, the controller 57 is configured to be able to execute collection (scanning) of frame data and the like, as will be described later.

  For this reason, as shown in FIG. 1, the patient places the jaw at the position of the chin rest 25 in the standing or sitting posture, holds the bite block 26, and presses the forehead against the headrest 28. As a result, the position of the patient's head (jaw) is fixed at substantially the center of the rotation space of the rotation unit 24. In this state, under the control of the controller 57, the rotation unit 24 rotates around the patient's head along the XY plane and / or along an oblique plane on the XY plane (see arrows in FIG. 1). .

  During this rotation, under the control of the controller 57, the high voltage generator 41 supplies a high voltage for exposure (specified tube voltage and tube current) to the X-ray tube 31 in a pulse mode with a predetermined cycle. The X-ray tube 31 is driven in the pulse mode. As a result, pulsed X-rays are emitted from the X-ray tube 31 at a predetermined cycle. The X-rays pass through the patient's jaw (dental portion) located at the imaging position and enter the detector 32. As described above, the detector 32 detects incident X-rays at a very high frame rate (for example, 300 fps), and sequentially outputs corresponding two-dimensional digital data (for example, 64 × 1500 pixels) in units of frames. Output. The frame data is temporarily stored in the buffer memory 53 via the communication line 43 and the interface 52 of the control / arithmetic apparatus 12. The temporarily stored frame data is then transferred to the image memory 53 and stored.

  Therefore, the image processor 56 reconstructs (creates) a tomographic image focused on the 3D reference tomographic plane SS as a panoramic image (reference panoramic image) using the frame data stored in the image memory 53. That is, this reference panoramic image is defined as “a panoramic image when it is assumed that a dentition is present along the 3D reference tomographic plane SS”. The image processor 56 performs processing such as creating a three-dimensional (3D) image and a three-dimensional (3D) autofocus image using the reference panoramic image. An outline of this processing is shown in FIG. The 3D reference image is defined as “a three-dimensional image when it is assumed that a dentition is present along the 3D reference tomographic plane SS”. The 3D autofocus image is defined as “a surface image obtained by automatically optimizing the dentition using frame data or reference panoramic image data from a 3D reference image”. That is, this 3D autofocus image is an optimum focus image that is less blurred and accurately represents the actual position and actual size of the dentition.

  In particular, the 3D autofocus image is an image that takes into account the fact that it is mostly different for each subject. As a practical matter, the dentition of each subject is not along the 3D reference tomographic plane SS (see FIG. 6), and is partially or entirely offset from the 3D reference tomographic plane SS or from that plane. It is tilted. For this reason, the 3D autofocus image automatically and accurately identifies the actual three-dimensional spatial position and shape of the dentition of each subject, and automatically displays the actual dentition shape from the identification result. To be created.

X-rays irradiated from the X-ray tube 31 (dotted X-ray source) pass through the oral cavity of the subject P and are detected by a vertically long detector 32 having a certain length in the Z-axis direction. For this reason, the X-ray irradiation direction becomes oblique as shown in FIGS. Therefore, the ratio between the actual size of the tooth and the size of the projected image formed by the shadow of the tooth on the detection surface Ldet of the detector 32 (in this embodiment, this ratio is referred to as “magnification rate”) is the rotation center RC. It changes according to the position. That is, in the example of FIG. 6 (however, only the tooth height is described), the ratio between the actual tooth height P ₁ real and the height P ₁ det on the detection surface Ldet is the rotation center RC. It changes according to the position. The position of the rotation center RC is set in advance so as to change during one scan (data collection) as illustrated in FIG. The reason for this is as follows. As shown in FIG. 6, the distance D _all between the X-ray tube 31 and the detector 32 is kept constant, and the distances D1 and D2 from the rotation center RC to the X-ray tube 31 and the detector 32 are also constant. Retained. On the other hand, in order to perform a scan focused on the 3D reference tomographic plane SS, during one scan (data collection), the trajectory at the position of the rotation center RC changes to a dentition curved in a horseshoe shape. On the other hand, it is designed to change to a mountain shape (see FIG. 2) as described above as an example.

  Specifically, the distance D3 from the rotation center RC to the 3D reference tomographic plane SS and the distance D4 from the detector 32 to the 3D reference tomographic plane SS (D3 + D4 = D2) change as scanning progresses. In response to this, the rotation center RC approaches or moves away from the dentition, so that the X-ray tube 31 also approaches or moves away from the dentition. Since the X-ray source of the X-ray tube 31 is regarded as a point-like shape, the projection onto the detection surface Ldet becomes closer as the X-ray tube 31 is closer to the dentition even if the teeth have the same height. The image gets bigger. That is, the enlargement rate is large. In the example of FIG. 2, the center of rotation RC is closer to the dentition when scanning the front tooth part than when scanning the molar part (back tooth side), and the enlargement ratio is increased accordingly. For example, referring to FIG. 2, the distance d1 when scanning the anterior tooth portion, for example, when the X-ray irradiation direction is 0 °, is the distance d2 when scanning the molar portion, for example, when the X-ray irradiation direction is 60 °, 75 °, With respect to d3, d1 <d2, d1 <d3, and d2 <d3. The trajectory of the rotation center RC shown in FIG. 2 is merely an example, but in the case of a panoramic imaging apparatus that scans while focusing on the 3D reference tomographic plane SS, the rotation center RC approaches and moves away from the dentition. This is the case.

  As described above, since the enlargement ratio varies depending on which tooth portion of the dentition is scanned, it becomes an obstacle when trying to quantitatively analyze the structure of the oral cavity and time-series changes.

  In addition to this, the above-described problem of the enlargement ratio has been described on the assumption that the dentition is along the 3D reference tomographic plane SS. Since the actual dentition of the subject is almost entirely absent from the position of the 3D reference tomographic plane SS, whether in whole or in part, this must be taken into account.

  A conventional panoramic image is created without considering the above-described problem due to the enlargement ratio and the actual dentition shift. For this reason, quantitative structural analysis from the conventional panoramic image is very difficult, and even if the dentition is in various shapes and positions for each subject, teeth in the dentition of the same subject A panoramic imaging device that can capture images with high accuracy regardless of the position of the camera has been desired.

  Therefore, the panoramic imaging apparatus according to the present embodiment eliminates image distortion caused by different magnification rates even for the same dentition, and the three-dimensional space of the actual dentition of the subject. One of the features is that the position (including the shape) is automatically and accurately identified. As a result, it is possible to provide a three-dimensional panoramic image that has an extremely high position (shape) identification accuracy, which is not present in the past.

  In this embodiment, a tomosynthesis method (tomosynthesis) is used to obtain an image of a tomographic plane. That is, among frame data (pixel data) collected at a constant rate by scanning, a plurality of frame data determined for each position of the locus projected on the XY plane of the three-dimensional 3D reference tomographic plane is an amount corresponding to the position. Only a shift (add & add) is performed by shifting each other. For this reason, “optimal focus” in this embodiment means “the focus is best, and there is little out of focus”, and the portion of interest has better resolution than other portions, or Says that the overall resolution of the image is higher.

  When the reference panoramic image is created, the data is stored in the image memory 54 and displayed on the monitor 60 in an appropriate manner. Among these, the operator's intention given from the operation device 58 is reflected in the display mode and the like.

  (Calibration of parameters for defining imaging space) Before describing imaging, calibration of parameters for defining imaging space using a phantom will be described with reference to FIGS. Processing accompanying this calibration is executed by the controller 57 and the image processor 56 in cooperation. A processor dedicated to calibration may be provided.

  <Principle of Reconstruction> Here, the basic principle of reconstruction in the panorama imaging apparatus will be described mathematically.

30A and 30B respectively show the trajectory Ts of the X-ray tube 31 and the trajectory TD of the detector 32 when the rotation center RC of the X-ray tube 31 and the detector 32 is at the position O, and the teeth. Indicates the positional relationship of the columns. In FIGS. 4A and 4B, the projection directions are different from each other. Here, the focal position of the X-ray tube 31 is S ₁ , S ₂ , the center position of the detector 32 in the width direction is C ₁ , C ₂ , the right end position of the detector 32 in the width direction is P ₁ , and the detector 32. The position of the left end in the width direction is P ₂ , the position of the rotation center RS of the X-ray tube 31 and the detector 32 is O, the rotation radius of the X-ray tube is R _s , the rotation radius of the detector 32 is RD, and the position of the rotation center If the distance from O to the dentition reconstruction position is d, and the shift amount of the frame data (data detected by the detector 32) for optimizing the focal points of the reconstruction points A and B is X,
(R _S + dcosθ): dsinθ = (R _S + R _D ): X
X = {(R _S + R _D ) / (R _S + dcosθ)} dsinθ (1)
Is required.

From this equation (1),
ΔX = {(R _S + R _D ) / (R _S + d)} dΔθ (2)
In addition,
ΔX / Δθ = {(R _S + R _D ) / (R _S + d)} d (3A)
θ = f (Fi) (3B)
Is required.

Using the formulas (3A) and (3B), ΔX / ΔFi = (ΔX / Δθ) (Δθ / ΔFi)
If Fi is frame data,
ΔX / ΔFi = [{(R _S + R _D ) / (R _S + d)} d] (Δθ / ΔFi) (4)
It becomes.

The left side ΔX / ΔFi of the equation (4) means a superimposition amount of frame data in shift and add according to the tomosynthesis method called gain (speed value). In addition, R _S + _{RD in} the right side of Expression (4) represents the distance between the detector and the X-ray tube (detector / X-ray tube distance), and R _S + d represents the distance between the X-ray tube and the focal point. (Focus position / X-ray tube distance). As a result, the gain curve (speed curve) is the detector ΔX-ray tube distance R _S + R _D , the focal position / X-ray tube distance R _S + d, and the slope Δθ of the curve representing the relationship between the frame data and the rotation angle. / ΔFi (see FIG. 31). If this gain curve is integrated so that the center of the front tooth becomes the center position of the image, a panoramic image focused at each position of the rotation angle can be reconstructed.

  As disclosed in Japanese Patent Application Laid-Open No. 2007-136163, the magnitude of the above-described gain ΔX / ΔFi is different from that of a normal electric circuit or the like, and the larger the gain ΔX / ΔFi, the more the frame data is mutually exchanged. The amount of overlay of frame data when adding is small. On the other hand, the smaller the gain ΔX / ΔFi, the larger the overlay amount.

In this embodiment, in order to reconstruct a focused panoramic image, various quantities such as the detector / X-ray tube distance R _S + R _D and the focus position / X-ray tube distance R _S + d are always maintained with high accuracy. ing. For this reason, these various amounts are measured at an appropriate timing, and calibration is performed to update information up to that time. For this calibration, a phantom described below is adopted.

  <Phantom> The phantom includes a first phantom 161 for measuring the position of the X-ray tube 31, a second phantom 171 for measuring the panorama magnification, and a third phantom for measuring the projection angle. Phantom 171A. The phantom may be provided as one accessory of individual panoramic imaging devices or may be provided as a common part. The first phantom 161, the second phantom 171 and the third phantom 171A are combined to constitute a phantom device.

  <First Phantom> FIG. 32 shows an example of the first phantom 161. The first phantom 161 is installed on the detection surface Ldet of the detector 32. Among the three columns 163, four columns 163 are installed so as to be orthogonal to the surface of the substrate 162. Are provided with two vertical bars 164 for erection of the two tip portions, and a horizontal bar 165 for connecting the two bars 164 to each other at two predetermined positions. The four struts 163 are stainless steel bars having a diameter of 3.0 mm and a length of 141 mm, for example, and are screwed to the substrate 162. The two vertical bars 164 are prisms made of, for example, stainless steel having an appropriate thickness and length (about 200 mm). Each horizontal bar 165 is installed on both vertical bars 164 at a predetermined distance, for example, 50 mm, from the position of the column 163 at each end of the vertical bar 164. The horizontal bar 165 is made of, for example, a round bar made of an aluminum alloy having a diameter of 0.8 mm and a length of 80 mm.

  FIG. 33 shows the geometrical positional relationship with the X-ray tube 31 when the first phantom 161 is mounted on the detector 32.

33, distances (heights) D2 to D6 from a reference 0 mm (detector 0 coordinate) in the vertical direction along the detector 32, the detection surface of the detector 32, and two upper and lower horizontal bars. The distance DP between 165 is a known quantity (unit is mm). For this reason, when the distance (height) from the reference coordinate plane of the X-ray tube 31 and the distance HX from the position of the X-ray tube 31 to the detection surface Ldet of the detector 32,
HX: (D2-D1) = DP: (D3-D2)
HX: (D6-D1) = DP: (D6-D5)
Therefore, from these relational expressions
D1 = (D2 (D6-D5) -D6 (D3-D2)) / (D2 + D6-D3-D5) (5)
HX = DP (D6-D2) / (D2 + D6-D3-D5) (6)
Is required. Furthermore, the longitudinal distance L between the longitudinal center position D4 of the detector 32 and the X-ray tube 31 is:
L = D4- (D2 (D6-D5) -D6 (D3-D2)) / (D2 + D6-D3-D5) (7)
Is calculated as

<Second and third phantoms>
Subsequently, an example of the second phantom 171 and the third phantom 171A will be described with reference to FIGS. FIG. 35 schematically shows the second phantom 171 and the third phantom 171A as viewed from above, that is, along the Z-axis direction (longitudinal direction), and viewed from the A direction in FIG. A side view is shown.

  The difference between the second and third phantoms 171 and 171A is in the way of installing a round bar for position measurement described later. In the present embodiment, the second and third phantoms 171 and 171A are provided as a set of phantoms, and can be selectively used as the second and third phantoms by moving the round bars.

  The second phantom 171 (and the third phantom 171A) is a transparent and substantially rectangular plate-like substrate (thickness is, for example, 10 mm) 172, and is movable on and off the substrate 172. And a plurality of round bars 173 arranged upright. The plurality of round bars 173 are linear bars made of an aluminum alloy having a diameter of 0.8 mm and a length of 50 mm, for example. This rod has a strength capable of maintaining linearity for a long time.

  The number of the round bars 173 is prepared by the number of representative representative projection directions Drec from −97.5 ° to + 97.5 ° stamped on the substrate 172. The projection direction Drec is a direction in which X-rays are irradiated while rotating the X-ray tube 31 and the detector 32, and a projection angle based on the center in the width direction of the detector when the XY plane is viewed from the Z-axis direction. This corresponds to θ.

  In this example, the number of round bars 173 is “15”. In a predetermined range RA of each projection direction Drec on the upper surface of the substrate 72, a round hole 174 into which the round bar 73 can be inserted and removed is formed. As an example, 30 round holes 174 (for example, a depth of 5 mm) with a predetermined interval are formed in the predetermined range RA. Therefore, by moving the round bar 73 to another round hole 174, the position of the round bar 173 can be changed before and after in each projection direction Drec. The plurality of round holes 174 function as a scale when the round bar 173 is moved.

  In the case of the second phantom 171, each round bar 173 (indicated as 173 f) traces the 3D reference tomographic plane SS, that is, the locus and projection direction Drec (that is, the 3D reference tomographic plane SS projected onto the XY plane). It is inserted at a position where the projection angle θ (the direction of the rotation angle when rotating the pair of the X-ray tube 31 and the detector 32) intersects.

  In the case of the third phantom 171A, each round bar 73 (denoted as 173r) has a predetermined distance outward in the projection direction from the position of the round bar 173f when used as the second phantom 171 in the predetermined range RA of each projection direction Drec. It is inserted into the round hole 174 at a position away from Dpre (here 20 mm).

  Note that, as is the case with the first phantom 161 described above, the second and third phantoms 171 and 171A can be installed in an imaging space using a tripod of a camera.

The geometric positional relationship between the second phantom 171 and the X-ray tube 31 and detector 32 is shown in FIG. In the figure, the distance HX from the X-ray tube 31 to the detector 32, the length S of the round bar 173, the projection position D7 (θ) on the detector 32 at the lower end of the round bar 173 at each rotational position θ, and Since the projection position D8 (θ) onto the detector 32 at the upper end of the round bar 173 at the rotation position θ is a known amount,
HX: (Rs + d) = (D8 (θ) -D7 (θ)): S
Holds. For this reason, the focal position / X-ray tube distance R _S + d is
(Rs + d) = HX * S / (D8 (θ) -D7 (θ)) (8)
It is requested from. That is, using the second phantom 171 and the information of the upper and lower positions D7 (θ) and D8 (θ) of the round bar 173f before the projection direction reflected in the panoramic image, all projection angles θ, that is, The focal position and X-ray tube distance R _S + d in all projection directions Drec can be calculated.

  In the second or third phantom 171 (171A), two round bars 173 (173f, 173r) are vertically spaced apart by a predetermined distance Dpre for each projection direction Drec, and this is projected at the projection angle described later. A fourth phantom 171B dedicated to measurement may be used.

<Measurement of projection angle>
Both the second and third phantoms 171 and 171A are used for measuring the projection angle of X-rays. For example, data is first collected using, for example, the second phantom 171, and then when the third phantom 171 </ b> A is used, the round bar 173 may be reinserted and moved. Note that data can be collected at once using the fourth phantom 171B and used for measurement of the projection angle θ.

First, the significance of measuring the projection angle will be described. As described above, the gain (speed value) in the 3D reference tomographic plane SS is as follows.
ΔX / ΔFi = [{(R _S + R _D ) / (R _S + d)} d] (Δθ / ΔFi) = G _θ (d) (9)
As given. However, since the calculation in the middle includes an error, the projected image of the round bar 173f placed at the position of the 3D reference tomographic plane SS of the second phantom 171 is the most focused gain. It is given as a gain G _θ (d) at the projection angle θ. Further, the frame F _θ corresponding to the position of the imaging center when the image is focused at each projection angle θ is also used to accurately determine the projection angle θ. The gain G _θ (d) (1−d ₀ / (Rs + d)) when the distance from the rotation center position O to the dentition reconstruction position changes by d + d ₀ is
G _θ (d) (1−d ₀ / (Rs + d)) ≈ (R _S + R _D ) / (RS + d + d ₀ ) · Dθ / DFi
(10)
It is requested from.

A specific method for obtaining the projection angle will be described. As described above, the focal length of the two-dimensional panoramic projection can be corrected almost accurately using the first and second phantoms 161 and 171. However, when a three-dimensional projection is required as in the panoramic imaging apparatus of the present embodiment, that is not sufficient, and the relationship between the projection angle θ of the detector 32 (that is, the projection direction Drec) matches the frame data Fi. It is important that That is, there is a relationship that which frame data Fi is used at which projection angle θ. The deviation of the projection angle θ is schematically shown in FIG.
The major factors that cause this relationship to be shifted are differences in specifications of manufacturers who design panoramic imaging devices and variations in mechanical adjustment of the devices. If the relationship between the projection angle θ and the frame data Fi is linear and the angle formed by the straight line connecting the X-ray tube and the detector with the reference line coincides with each projection direction Drec of the second phantom 71, the three-dimensional reproduction is repeated. The configuration process is facilitated. However, the relationship between the projection angle θ and the frame data Fi is usually not linear, and there are also panoramic imaging devices that do not exhibit the projection direction Drec defined by the second phantom 71. Therefore, as described below, it is necessary to measure the projection angle of X-rays that pass through each position of the projection angle θ, that is, each position of the dentition.

This measurement procedure will be described with reference to FIG. FIG. 9A shows the second phantom 171 and the third phantom 171A (or a fourth phantom instead thereof) when the gain G _θ (d) of the 3D reference tomographic plane SS is obtained at a certain projection angle θ. 171B) is a part of the panoramic image. This image assumes the right side in FIG. 34 of the phantom 171 (171A, 171B). In FIG. _{6A, the} object I _73f visible on the left side is an image of the front round bar 73f placed at the position of the 3D reference tomographic plane SS, and this is in focus. Further, the object I _73r visible on the right side is an image of the rear round bar 173f placed on the back side by a predetermined distance of 20 mm from the 3D reference tomographic plane SS, and is slightly blurred (in the figure, the blur is widened). Shown). If the projection direction of the X-rays emitted from the X-ray tube 31 at the projection angle θ is exactly the same as that of the second phantom 171, the images of the front round bar 73f and the rear round bar 173f are mutually Overlaid. For this reason, as shown in FIG. 6A, the fact that the images of the front round bar 173f and the rear round bar 173f are separated from each other means that the actual X-ray projection direction is designated by the second phantom 171. It means that it is deviated from the direction. In addition, in the case of the example shown in the figure, the projection angle is estimated to be larger (deeper) than that of the second phantom 171.

Therefore, storing the frame data F _theta in FIG (A) definitive front rod 173f to depend image _{I 73f} (i.e. image showing the position of the reference 3D tomographic plane SS). As the next step, the gain G _θ (d) is increased around the frame data F _θ to focus on the image of the rear round bar 173f. As a result, as shown in FIG. 5B, the rear round bar 173f is depicted as I _73r as a narrower and sharper object. Therefore, actually measuring the distance L between both images _{I 73f} and _{I 73r} on the panoramic image of FIG. (B). As a result, as shown in FIG. _5C , the geometric relationship of the actual projection position I _73r of the front round bar 173f, the rear round bar 173r, and the rear round bar 173r can be understood.

Therefore, using the relationship shown in FIG. 38C, the actual projection direction of X-rays is larger than the direction passing through the front round bar 173f as designed by ArcTan (L / 20 mm) degrees. Therefore, for each of the left half and right half sides of the phantom 171 (171A, 171B) shown in FIG.
Projection angle θ ′ = ArcTan (L / 20 mm) + θ (11)
The actual projection angle θ ′ is calculated from the following equation.

  In addition, for the central part of the phantom 171 (171A, 171B), that is, the round bars 173f, 173r with a projection angle of 0 degrees, the arrangement direction of the phantom itself is initially set so as not to be separated on the panoramic image as described above. For this reason, the actual value of each angle in a predetermined discrete projection direction is calculated for the round bars in the region swung left and right from the round bars 73f and 73r with a projection angle of 0 degrees.

  In the equation (10), the predetermined distance Dpre = “20 mm” is an example. The distance Dpre is preferably as large as possible in order to increase the detection accuracy of the projection angle θ, and should be close to the detector 32 side (that is, the back side in the projection direction). May be set to a value.

(3D modeling)
The series of calculations (estimation) described above are executed along the flow shown in FIG. 39, and a three-dimensional model of the 3D reference tomographic plane SS based on the parameters obtained by each calculation is created. With this three-dimensional model, the positional relationship of the 3D reference tomographic plane SS in the imaging space with respect to the X-ray tube 31 and the detector 32 can be grasped.

  Specifically, first, the first phantom 161 described above is installed on the detector 32 with a jig (not shown), and imaging is performed at a certain range of projection angles or pinpoint projection angles. Since this imaging does not necessarily have to be performed at all projection angles, it is not necessary to rotate the X-ray tube 31 and detector 32 pair all around the subject or most of it. This shortens the data collection time, reduces the amount of data to be processed, and reduces the calculation load. Then, using the known amount given by the first phantom 161 and the information given by the panoramic image, the position information of the X-ray tube 31 is obtained based on the equations (5), (6) and (7) (step S131). ). Note that image reconstruction is performed by a known tomosynthesis method.

Next, the second phantom 171 is placed in an imaging space between the X-ray tube 31 and the detector 32, for example, with a tripod, and imaging is performed over the range of all projection angles, and the panorama of the second phantom 171 displays an image. get. Then, the focal position / X-ray tube distance R _S + d and the gain G _θ (d) are calculated from the position information obtained from the panoramic image and the known amount due to the front round bar 173f of the second phantom 171 using the equation (8). , (9) (step S132).

The calculation related to step S132 described above is performed only at the position of the discrete projection angle θ. For this reason, it is better to obtain the focal position / X-ray tube distance R _S + d and the gain G _θ (d) at positions other than the discrete scan positions of the dentition to be imaged. Therefore, using the values calculated in step S132, the focal position / X-ray tube distance R _S + d and gain G _θ (d) at the position where the discrete scan position is filled are estimated by an appropriate interpolation method (step) S133). The various quantities calculated in steps S131 to S132 are stored in the memory (step S134).

  Next, the third phantom 171A is placed on, for example, a tripod in the imaging space, and imaging is performed over the range of all projection angles, and the panorama of the third phantom 171A acquires an image. Therefore, from the position information of the rear round bar 173r of the third phantom 171A and the position information of the front round bar 173f of the second phantom 171 acquired in step S132 in this image, the above-described formula (11 ), The actual projection angle θ ′ is measured (calculated) for each projection direction passing through each round bar 173 (173f, 173r), and the measured value is stored (step S135). It should be noted that a panoramic image of the fourth phantom 171B in which the front round bar 173f and the rear round bar 173r are suspended at the same time is acquired, and the above-described actual projection angle θ ′ may be calculated from this panoramic image. .

  Further, for the projection directions other than the projection direction passing through the round bar 173 (173f, 173r), the projection angle θ ′ is obtained and stored using an appropriate interpolation method (step S136).

  Next, a three-dimensional model for specifying the positional relationship in the imaging space of the 3D reference tomographic plane SS is created using the information acquired in steps S131 to S133, S135, and S136 as parameters (step S137). An outline of the creation of this three-dimensional model is shown in FIG. That is, projection data of the 3D reference tomographic plane SS is created on the imaging space for each projection angle θ, and a 3D model of the 3D reference tomographic plane SS is created. The information of the three-dimensional model created in step S137 and the information acquired in steps S131 to S133, S135, and S136 are taken in as calibration data at the time of actual imaging, and the existing calibration data up to that point is updated ( Step S138).

  As illustrated in FIG. 40, projection is performed along the oblique irradiation direction of X-rays from the modeled 3D reference tomographic plane SS, and the three-dimensional position of an imaging target (entity) such as a dentition is identified. be able to. This position identification process will be described later.

The phantom that can be employed in the radiation imaging apparatus according to the present invention is not limited to the above-described phantom, and can be implemented in various configurations other than the above. Hereinafter, this modification will be described.

<First Modification>
A first modification is shown in FIGS. A phantom 181 shown in FIG. 41 can be used in place of the first phantom 161 described above. As shown in the figure, the first phantom 181 is made of an acrylic prism having a cross section of 5 mm × 5 mm and a length of 92 mm. Four known positions in the longitudinal direction on one side surface of this prismatic body, for example, four positions of 10 mm, 24 mm, 24 mm, and 24 mm from the bottom, each made of aluminum or brass, for example, a small diameter of 0.6 mm in diameter. A rod 182 is attached. In this attachment, a semicircular shaving having a diameter of, for example, 0.6 mm is put on the surface of the first phantom 181 so that the surface of the phantom and the center of the small diameter rod 182 are flush with each other. The first phantom 181 is disposed in the imaging space using the attachment portion 183 at the lower end.

FIG. 42 shows the positional relationship between the X-ray tube 31, the detector 32, and the small-diameter rod 182 in the imaging space when this arrangement is made. In the quantities shown in the figure, the distances a, b, c, and d are known quantities, and Hx, Hz, Hf, and Ef are unknown quantities. Because there are four unknowns
Hz-Hf: Ef = Hz: H + a
Hz-Hf: Ef + e = Hz: Hx + b
Hz-Hf: Ef + 2e = Hz: Hx + c
Hz-Hf: Ef + 3e = Hz: Hx + d
Holds. From these equations
Hf / (Hx + a) = (Ef + e) / (Hx + b)
Hf / (Hx + a) = (Ef + 2e) / (Hx + c)
Hf / (Hx + a) = (Ef + 3e) / (Hx + d)
From these equations, (Ef + e) / (Hx + b) = (Ef + 2e) / (Hx + c)
(Ef + e) / (Hx + b) = (Ef + 3e) / (Hx + d)
Is required. Of these, from the former
Ef = (eHx + 2be-ec) / (cb)
But from the latter
Ef = (2eHx + 3be-de) / (cb)
Is required. From these two formulas,
(EHx + 2be-ec) / (cb) =
(EHx + 2be-ec) / (cb)
(12)
Is obtained.

  Therefore, since Hx is obtained from the equation (12), Ef, Hz, and Hf can also be obtained sequentially from the previous equation.

  Thereby, there are more unknown variables than the first phantom 161 described above, and the distance Hz between the X-ray tube 31 and the detector 32 can be calculated with a simple structure.

(Second modification)
This second modification will be described with reference to FIGS. 43 and 44. FIG. This modification can be used in place of the first phantom 161 and the second phantom 171 described above. A phantom 191 shown in FIGS. 43 and 44, like the above-described second phantom 171, is a transparent and substantially rectangular plate-like substrate 192 made of acrylic, and movably and upright on the substrate 192. And a plurality of square bars 193 to be provided. The second phantom 171 is different from the second phantom 171 described above in that the square bar 193 is used and the hole 194 for embedding the square bar 193 in the substrate 192 is formed in a square shape. It is the same as that of the second phantom 171.

  The plurality of square bars 193 include a front side square bar 193f planted on the front side in each projection direction Drec and a rear side square bar 193f planted on the rear side. Each of the plurality of front side square bars 193f has four horizontally placed small diameter rods 195 similarly to the small diameter rod 182 of the first phantom 181 described above, and the same members and mounting methods as the horizontally placed small diameter rod 195 are used. It also has a vertically mounted small-diameter rod 195V that is assembled vertically. Each of the plurality of rear side square bars 193r has only the small diameter rod 195V. Both the front square bar 192f and the rear square bar 192r have the same size as the first phantom 181 described above, and the mounting portion 196 at the lower end thereof is positioned in the same manner as the second phantom 171. It is inserted into the 194 and planted on the substrate 192.

For this reason, the position of the X-ray tube 31 can be obtained by the four horizontal small-diameter rods 195 of the front side square bar 193f, and the two vertical bars of the front side square bar 193f and the rear side square bar 193r in each projection direction Drec. The focal position, the X-ray tube distance, and the projection angle can be measured for each projection angle θ by the small-diameter rod 195V.
In the configuration shown in the above-described modification, the small-diameter rod is an element having a certain length in which the X-ray transmittance is different from the support part of the phantom and the medium in the imaging space. As long as the rate condition is satisfied, a point-like position indicating element may be provided instead of the small-diameter rod.

(Image processing)
Returning to FIG. 5, processing executed by the controller 57 and the image processor 56 in cooperation will be described. As described above, this processing includes data collection by scanning, reconstruction of a reference panoramic image as preprocessing, creation of a three-dimensional autofocus image (surface image) as main processing, and three-dimensional autofocus thereof. Display and measurement according to various aspects using an image are included.

<Data collection and reference panorama image reconstruction>
First, when preparation for imaging such as positioning of the subject P is completed, the controller 57 responds to an instruction from the operator given through the operation device 58 and commands a scan for data collection (step S1). Thereby, the rotation driving mechanism 30A, the moving mechanism 30B, and the high voltage generator 41 are instructed to be driven in accordance with a preset control sequence. For this reason, while rotating the pair of the X-ray tube 31 and the detector 32 around the jaw of the subject P, a pulsed or continuous wave X-ray is applied to the X-ray tube 31 during a predetermined period. Or continuously. At this time, the pair of the X-ray tube 31 and the detector 32 is rotated based on a driving condition set in advance so as to focus the 3D reference tomographic plane SS (see FIG. 6) calibrated as described above. Driven. As a result, the X-rays exposed from the X-ray tube 31 pass through the subject P and are detected by the detector 32. Therefore, as described above, digital amount of frame data (pixel data) reflecting the amount of X-ray transmission is output from the detector 32 at a rate of 300 fps, for example. This frame data is temporarily stored in the buffer memory 53.

  When the scan command is completed, the processing instruction is passed to the image processor 56. The image processor 56 reconstructs the reference panoramic image PIst based on shift & add based on the tomosynthesis method corresponding to the spatial position of the 3D reference tomographic plane SS, and stores each pixel value of the reconstructed image (step S2). ). In this reconstruction process, a process of multiplying a coefficient so that the ratio of the vertical and horizontal enlargement ratios is the same at the center of the front tooth portion is executed as in the conventional case.

  This reconfiguration method is well known, but will be explained a little. The set of frame data used for this reconstruction is, for example, mapping characteristics indicating the relationship between the horizontal mapping position of the panoramic image shown in FIG. 7 and the set of frame data to be mutually added to create an image of the mapping position. It is requested from. The curve indicating the mapping characteristics includes both a curved portion having a steep slope according to the molar portion on both sides in the frame data direction (horizontal axis) and a curved portion having a gentler slope than that of the molar portion according to the front tooth portion. Consists of. On this projection characteristic, a desired mapping position in the horizontal direction of the panoramic image is designated as shown in the figure. In accordance with this, a set of frame data used for creating an image at the mapping position and its shift amount (the degree of superposition: that is, the inclination) are obtained. Therefore, these frame data (pixel values) are added to each other while being shifted by the designated shift amount, and vertical image data of the designated mapping position (range) is obtained. The reference panorama image PIst when focused on the 3D reference tomographic plane SS is reconstructed by specifying the mapping position and shifting & adding the entire panoramic image in the horizontal direction.

  Next, the image processor 56 displays the reference panoramic image PIst on the monitor 60 (step S3). An example of the reference panoramic image PIst is schematically shown in FIG.

  Since the reference panoramic image PIst is an image obtained by adding frame data to each other while shifting, the reference panoramic image PIst is a rectangular two-dimensional image. Speaking of the enlargement ratio, the processing is performed by multiplying the coefficients so that the ratio of the vertical and horizontal enlargement ratios is the same at the center of the front tooth part. Has been improved to some extent. However, the aspect ratio of the teeth collapses as it proceeds to the molar part. That is, the teeth of the molar portion are drawn with a size smaller than the actual size. In the past, in many cases, panoramic images with such distortion were put up.

<ROI setting on the standard panorama image>
Next, the image processor 56 determines whether or not an ROI (region of interest) is set in the reference panoramic image PIst by using the operation device 58 from the operator (step S4). The ROI set here is, for example, a rectangular partial region in which the image interpreter is particularly interested. Of course, the ROI is not necessarily rectangular. This ROI may be set for a panoramic image created by autofocus described later, and this processing will also be described later.

  If the determination in step S4 is YES, the image processor 56 sets the ROI to the reference panoramic image PIst based on the operation information of the operator (step S5). Next, a partial image of the partial area set by the ROI is cut out, and the partial image is enlarged and displayed, for example (step S6). For example, as shown in FIG. 9, this partial image is displayed so as to be superimposed on the original reference panoramic image PIst. Further, the one or more partial images may be displayed so as to fit in a so-called template in which blocks are arranged in a predetermined order so as to schematically represent the upper teeth and lower teeth.

  Next, the image processor 56 determines whether or not to end the process. This determination depends on whether or not there is predetermined operation information from the operator (step S7). If it is determined that the process is not yet finished (NO in step S7), the process returns to step S4 and the above-described process is repeated. On the other hand, if it is determined that the process has been completed, the process shown in FIG. 5 is terminated.

  On the other hand, if the determination in step S4 is NO, that is, if it is determined that the ROI is not set, the image processor 56 proceeds to the next determination. That is, it is determined from the operation information of the operator whether or not to create a 3D autofocus image as the main process (step S8). When it is determined that this creation is not performed (step S8, NO), the process returns to step S7 to determine whether or not the process is ended as described above.

<Identification of optimal focus cross-sectional position>
On the other hand, if it is determined to create a 3D autofocus image (YES in step S8), the process proceeds to a subroutine process in step S9. The processing executed in step S9 is one of the features of the present invention, and is an automatic dentition existing while correcting distortion of the dentition size caused by the oblique X-ray irradiation direction. This is a position / shape identification process.

FIG. 10 shows a subroutine process for identifying the actual position / shape.
First, the image processor 56 converts the reference panorama image PIst (rectangular) into a curved surface parallel to the 3D reference tomographic plane SS (curved surface) to create a 3D panoramic image once. Then, each of the pixels of the 3D panoramic image is projected onto the 3D reference tomographic plane SS along the X-ray irradiation direction DRx by calculating the tomographic plane, calculating the frame data, and projecting the coordinate data by converting the frame data. A projection image of the reference tomographic plane SS is created (step S51). The pixel value of this projection image is stored in the image memory 54.

  The projection performed here is performed along an oblique projection direction toward the position of the rotation center RC (RC1, RC2), that is, the position of the X-ray tube 31, as described in FIG. In the example of FIG. 11, even in the pixel at the same position Pn in the height direction (Z-axis direction) on the 3D panoramic image, the difference in the position of the X-ray tube 31 causes the image on the 3D reference tomographic plane SS to be displayed. Projected to different positions SS1, SS2.

  A projection image created by this projection processing will be referred to as a 3D reference image PIref. The 3D reference image PIref is created by oblique projection in consideration of the above-described enlargement ratio for each position of the reference panoramic image PIst. The enlargement rate of the teeth of the anterior teeth was corrected to the actual size by the above projection, while the enlargement rate of the teeth of the molar portion was small, Corrected to actual size rather than projection. Therefore, the 3D reference image PIref is an image displayed at the actual size of the tooth, and is an image from which distortion due to the magnitude of the enlargement ratio due to the movement of the rotation center RC during scanning is removed. However, the 3D reference image PIref is also an image when it is assumed that the dentition exists along the 3D reference tomographic plane SS. Since the actual tooth of the subject P is rarely along the 3D reference tomographic plane SS, further real position identification processing described later is required.

  The image processor 56 displays the 3D standard image PIref on the monitor 60 for reference to the operator (step S52). This is shown in FIG.

  Thereafter, the image processor 56 adds a plurality of curved tomographic planes parallel to the 3D reference tomographic plane SS (step S53). This is shown in FIG. In the drawing, a plurality of tomographic planes are added before and after the X-ray irradiation direction DRx (dentation depth direction) of the 3D reference tomographic plane SS. As an example, a plurality of tomographic planes SFm to SF1 are set at a distance D1 (for example, 0.5 mm) on the front side of the 3D reference tomographic plane SS, and a plurality of tomographic planes SR1 to SRn are spaced at a distance D2 (for example, 0.5 mm) on the rear side. Is set in. The intervals D1 and D2 may be the same or different from each other. Further, the tomographic plane to be added may be one sheet before and after the 3D reference tomographic plane SS (m, n = 1), or may be one sheet or a plurality of sheets before or after the 3D reference tomographic plane SS.

  Since the position data of the tomographic planes SFm to SF1 and SR1 to SRn to be virtually added is stored in the ROM 61 in advance together with the position data of the 3D reference tomographic plane SS, this is read out to the work area of the image processor 56. Thus, such addition is performed. The heights of the tomographic planes SFm to SF1, SS, SR1 to SRn are appropriately set in consideration of the maximum inclination in the X-ray irradiation direction DRx and the height of the dentition. In addition, each time the identification process is performed, the position (distance D1, D2) and the number of tomographic planes to be added may be interactively changed.

  Next, the image processor 56 considers the angle of the X-ray irradiation direction DRx and changes the tomographic plane to each of the added tomographic planes SFm to SF1 and SR1 to SRn in consideration of the angle of the X-ray irradiation direction DRx as in step S51. The frame data is obtained by the above calculation, and is projected by converting the coordinates (step S54). As a result, projection images of the additional tomographic planes SFm to SF1 and SR1 to SRn are created. The pixel values of these projected images are stored in the image memory 54.

The projection images created here are referred to as 3D additional images PIsfm..., PIsf1, PIsr1,. These 3D additional images PIsfm, ..., PIsf1, PIsr1, ...,
PIsrn is also created for each position of the reference panoramic image PIst by oblique projection in consideration of the magnification described above. In the example of FIG. 14, even in the pixel at the same position Pn in the height direction (Z-axis direction) on the 3D panoramic image, the 3D additional image PIsfm,
..., PIsf1, PIsr1, ..., PIsrn are projected onto different positions.

Therefore, these 3D additional images PIsfm,..., PIsf1, PIsr1,..., PIsrn are also images displayed at the actual size of the teeth, and distortion due to the magnitude of the enlargement ratio due to movement of the rotation center RC during scanning is removed. It is an image that was made. However, these 3D additional images PIsfm,
.., PIsf1, PIsr1,..., PIsrn are images when it is assumed that the dentitions exist along the respective additional tomographic planes SFm to SF1 and SR1 to SRn.

  The generated plurality of 3D additional images PIsfm,..., PIsf1, PIsr1,..., PIsrn are displayed on the monitor 60 as they are as a three-dimensional image or as a rectangular two-dimensional image after coordinate conversion. It may be.

Thereafter, the image processor 56 designates the 3D reference image PIref, that is, the initial position P (x, y, z) = P (0, 0, 0) in the 3D reference tomographic plane SS (step S55: FIG. 15A). x, see). When this is completed, a line segment Lc having a certain length centered on the designated position P (x, y, z) is designated in the 3D reference image PIref (see step S56: FIG. 15B). The line segment Lc has a length corresponding to 2 ⁿ (n = 1, 2, 3,...; 128, for example) pixels. The line segment Lc may be curved along a part of the curved 3D reference tomographic plane SS, or may be set within a range that can be regarded as a straight line.

  Next, the image processor 56 virtually adds a plurality of line segments Ladd having the same length above and below the image of the specified line segment Lc (x, y, z) (step S57: FIG. 15C )reference).

Further, the pixel values P _ij of the 2 ⁿ pixels constituting each of the line segment L and the plurality of line segments Ladd described above are read from the image memory 54 and assigned to each line segment (step S58). This pixel value P _ij is a value that has already been acquired and stored in steps S51 and S54 described above.

Next, the pixel values P _ij of the pixels corresponding to the plurality of line segments L and Ladd are added to each other, and 2 ⁿ pixel values P _ij ^* for frequency analysis constituting the line segment Lc (x, y, z) ^. (Step S59: see FIG. 15D). By this addition, even when statistical noise is mixed in the original pixel value of the line segment L (x, y, z), the statistical noise when performing frequency analysis to be described later on the change of the pixel value is reduced. Can be made.

  Next, the image processor 56 adds the line segment Lc (x, y, z) currently specified on the above-mentioned 3D reference image PIref in each of the added 3D additional images PIsf1, ..., PIsf1, PIsr1, ..., PIsrn. Identifies the positions of the opposing line segments Lfm to Lf1 and Lr1 to Lrn in the X-ray irradiation direction DRx passing through the currently designated position P (x, y, z) (step S60: see FIG. 15E). ). At this time, since the current center position P (x, y, z) and the length of the line segment Lc and the rotational position of the X-ray tube 31 during scanning are known, both ends of the line segment Lc and X An X-ray irradiation range RA that is formed in a fan shape when viewed from the Z-axis direction can be calculated. For this reason, if the position P (x, y, z) is designated, the positions of the line segments Lfm to Lf1 and Lr1 to Lrn located in the X-ray irradiation range RA can be specified.

  Note that the process of step S60 for designating the position P (x, y, z) on the 3D reference image PIref is repeated until all the positions are designated. For this reason, the X-rays radiated from the X-ray tube 31 whose position is near the virtual tomographic planes SFm to SF1, SS, SR1 to SRn are in the range H1 to H2 (range in the Z-axis direction). It is transmitted in a fan shape (FIG. 15F). For this reason, the tomographic planes SFm to SF1, SS, SR1 to SRn themselves may be set as substantially horseshoe-shaped cross sections whose height changes for each scanning direction and is parallel to each other.

When the line segments Lfm to Lf1 and Lr1 to Lrn are determined as described above, the image processor 56 reads out the pixel values P _ij ^* of these line segments from the image memory 54 (step S61).

As shown in FIG. 15E, since the X-ray tube 31 is a point source, the X-ray irradiation range RA has a fan shape (when viewed from the Z-axis direction). For this reason, the number of pixels of each of the line segments Lfm to Lf1 and Lr1 to Lrn is deviated from ²ⁿ . Therefore, the image processor 56, the additional line segments Lfm～Lf1, line Lc of the number of pixels of Lr1~Lrn becomes a reference (x, y, z) to be the same as the number of pixels the ^{2 n,} the line segment Each of Lfm to Lf1 and Lr1 to Lrn is multiplied by a coefficient corresponding to the distances D1 and D2 (step S62). Accordingly, as schematically shown in FIG. 15G, all the line segments Lfm to Lf1, Lc, and Lr1 to Lrn are composed of ²ⁿ pixels which are parallel to each other.

  Thereafter, the image processor 56 performs frequency analysis on changes in the pixel values of all the prepared line segments Lf1 to Lfm, Lc, and Lr1 to Lrn (step S63). As a result, for each of the line segments Lf1 to Lfm, L, and Lr1 to Lrn, as shown in FIG. 15H, an analysis result is obtained with the horizontal axis representing the frequency and the vertical axis representing the Fourier coefficient (amplitude value).

  The frequency analysis uses fast Fourier transform (FFT), but may use wavelet transform. Further, instead of such a frequency analysis method, an equivalent process may be performed using a Sobel filter that performs a first-order differential operation for rendering an edge. When this filter is used, the position of the tomographic plane having the maximum edge can be regarded as the optimum focus position.

  Next, noise is removed from the result of frequency analysis for all line segments Lf1 to Lfm, Lc, and Lr1 to Lrn (step S64). FIG. 16 illustrates frequency analysis characteristics for one line segment. The coefficient of the frequency component in the region of the certain range of the analyzed highest frequency is excluded, and the coefficient of the remaining high frequency component is adopted. This is because the frequency component in a certain range on the highest frequency side is a noise component.

  Further, the image processor 56 square-adds the coefficient of the frequency analysis characteristic for each line segment for each line segment, sets the square addition value as the vertical axis, and the initial position P (x, y, z) = P. A plurality of tomographic planes SFm to SF1, SS, and SR1 to SRn passing through (0, 0, 0) in the X-ray irradiation direction DRx are calculated as profiles with the horizontal axis (step S65). An example of this profile is shown in FIG. In the figure, the cross-sectional position is a position in the X-ray irradiation direction DRx (the depth direction of the dentition) of the plurality of tomographic planes SF1 to SFm, SS, FR1 to FRn.

  FIG. 18 illustrates typical patterns of a plurality of types of profiles PR1, PR2, PR3, and PR4 when the material is enamel, cancellous bone, air, and bite block. If an enamel substance, that is, a tooth exists at any position in the X-ray irradiation direction DRx passing through the currently designated position P (x, y, z), the profile PR1 has a sharp peak. Have Further, when cancellous bone exists in the X-ray irradiation direction DRx, the profile PR2 is a gentle convex curve. Similarly, when only air exists in the X-ray irradiation direction DRx, the profile PR3 is a curve showing a tendency not to have a specific peak. Furthermore, when a bite block exists in the X-ray irradiation direction DRx, the profile PR4 has two sharp peaks. Among these, the peak corresponding to the inner side (X-ray tube side) of the X-ray irradiation direction DRx indicates the peak for the enamel substance, and the peak corresponding to the outer side (detector side) indicates the peak for the bite block. Data indicating the patterns PR1 to PR4 shown in FIG. 18 is stored in advance as a reference profile, for example, in the ROM 61 as a reference table.

  Therefore, the image processor 56 specifies the position of the optimum focus with respect to the tooth in the X-ray irradiation direction DRx passing through the currently designated position P (x, y, z) using the reference table (step S66). .

  That is, it is determined by the pattern recognition method which of the reference profiles PR1 to PR4 corresponds to the profile obtained in the previous step S65. First, when the obtained profiles are the reference profiles PR2 and PR4, they are excluded from processing. On the other hand, when the obtained profile corresponds to the reference profile PR1 (enamel), the cross-sectional position exhibiting the peak, that is, any one of the plurality of tomographic planes SF1 to SFm, SS, FR1 to FRn is the optimum focus. Identify as being. Further, when the obtained profile corresponds to the reference profile PR4, the cross-sectional position (enamel position) having a peak on the inner side (X-ray tube side), that is, a plurality of tomographic planes SFm to SF1, SS, FR1 to Any position of FRn is specified as the optimum focus.

  With these position specifying processes, the position of the tooth portion at the position P (x, y, z) currently specified is actually determined in the depth direction. That is, the tooth portion depicted in the 3D reference image PIref along the 3D reference tomographic plane SS may actually be on the front side or the rear side of the tomographic plane SS. This actual position is accurately determined by the above-described specifying process. In other words, it can be said that the tooth portion of the 3D reference image PIref drawn on the assumption that it is on the 3D reference tomographic plane SS is shifted to the actual position by the above-described specific processing.

  As a result, as shown in FIGS. 19 to 22, the position P1 on the 3D reference tomographic plane SS (3D reference image PIref) is set to P1real (or P2 P2real). In particular, the positions of line segments Lfm to Lf1 and Lr1 to Lrn set on the plurality of additional tomographic planes SFm to SF1, FR1 to FRn are set in consideration of the oblique angle θ in the X-ray irradiation direction DRx. For this reason, the shifted position P1real is larger when the oblique angle θ is smaller (see FIGS. 20A and 21A) (see FIGS. 20B and 21B). Becomes lower. Therefore, the shift position P1real is compensated for the distortion due to the oblique X-ray irradiation angle θ, that is, the magnification ratio. As shown in FIG. 22, when the tooth actually exists along the 3D reference tomographic plane SS, P1 = P1real is established, and the 3D reference tomographic plane SS assumed to be located is determined as the actual position. . In this case, the shift with the shift amount = 0 is executed.

In step S65, the image processor 56 stores the data indicating the identified actual tooth position in the work area for each position P (x, y, z).
In this way, the position P (x, y, z) currently specified in the 3D reference image PIref (that is, the 3D reference tomographic plane SS), that is, in this case, the initial position P (0, 0) specified first. , 0) to identify (filtering) whether or not a tooth part (enamel) is present in the depth direction, and if such a tooth part is present, the optimum in that depth direction The focus position is identified.

  After this, the image processor 56 determines whether or not the above-described specific processing has been completed for all the determination positions P set in advance on the 3D reference image PIref as shown in FIG. 23 (step S67). This determination is made by determining whether or not the currently processed position P (x, y, z) is the final position P (p, q, r). If this determination is NO and the specific processing has not been completed for all the determination positions P, the image processor 56 shifts the determination position P (x, y, z) by one (step S68). The process returns to step S55 described above, and the above-described series of specific processes is repeated.

  As shown in FIG. 23, the plurality of determination positions P are preliminarily arranged two-dimensionally at a predetermined interval along the 3D reference image PIref (that is, the 3D reference tomographic plane SS). In the example shown in the figure, the 3D reference image PIref is arranged along the vertical axis direction i and the horizontal axis direction j with the same predetermined distance d in the vertical and horizontal directions. However, the predetermined distance d may be different from each other in the vertical axis direction i and the horizontal axis direction j. The shift direction in the process of step S68 may be any of vertical, horizontal, and diagonal directions along the 3D reference image PIref. As shown in FIG. 23, after shifting along the vertical axis direction i of the 3D reference image PIref, shifting in the horizontal axis direction j and shifting along the vertical axis direction i may be repeated regularly ( (See symbol SC in the figure). Conversely, shifting in the horizontal axis direction j and then shifting in the vertical axis direction i may be repeated. Further, it may be shifted in an oblique direction.

  On the other hand, when the series of determinations described above are completed at all of the plurality of determination positions P, the determination in step S67 described above becomes YES in the above-described repeated determination. That is, the process of detecting the cross-sectional position of the optimal focus (including the determination of the presence or absence of the optimal focus position) is completed for each determination position P in the depth direction of the 3D reference tomographic plane SS. In this case, the process shifts to the process of combining the cross-sectional positions of the optimum focus.

<Process for combining cross-sectional positions of optimum focus>
If the determination in step S67 described above is YES, the image processor 56 reads data representing the cross-sectional position of the optimum focus specified and stored in step S65 (step S69). The data of the cross-sectional position is a position in the X-ray irradiation direction DRx that passes through each judgment position P (x, y, z). This is schematically shown in FIG. In the figure, black circles indicate the determination position P (x, y, z) of the 3D reference image PIref (3D reference tomographic plane SS). Here, the vertical and horizontal directions of the curved 3D reference image PIref are represented by (i,
j). In FIG. 24, as indicated by white circles, for example, the optimum focal section position with respect to the determination position P (x ₀₀ , y ₀₀ , z ₀₀ ) of i, j = 0, 0 is shifted to the inside (X-ray tube side) by one. The position of the tomographic plane SR1, which is the position of the tomographic plane SR2 where the optimum focal section position with respect to the judgment position P (x ₀₁ , y ₀₁ , z ₀₁ ) of i, j = 0, 1 next to it is one further inside It is the position of the tomographic plane SR3 where the optimum focal section position with respect to the judgment position P (x ₀₂ , y ₀₂ , z ₀₂ ) of i, j = 0, 2 next to it is one more inside, etc. Become. Note that FIG. 24 shows step S68 at one position in the Z-axis direction (vertical direction) for easier viewing of the figure, but the processing of step S68 is also performed for each of the other positions in the Z-axis direction. The

Next, the image processor 56 removes noise (step S70). In the example of FIG. 24, for example, there are m optimal focus cross-section positions with respect to the determination position P (x ₀₃ , y ₀₃ , z ₀₃ ) at positions i, j = 0, 3 in the vertical and horizontal directions of the image on the outside (on the detector side). It is assumed that it is the position of the tomographic plane SFm that has approached. In such a case, the image processor 56 determines that the difference between the cross-sectional positions is noise and abnormal by, for example, threshold determination. In this case, for example, the data of the positions of adjacent cross sections are smoothed so as to be smoothly connected, and replaced with the smoothed new position data, or data close to the outside of the detector is given priority, etc. Perform the process. It should be noted that the abnormal data may be simply excluded from the processing target without performing such compensation by replacement. Naturally, it is also possible to add an abnormality of data in the Z-axis direction to the exclusion of the abnormal data.

Thereafter, the image processor 56 combines the denoised positions (that is, enamel positions), smooths the data of the combined positions three-dimensionally, and generates a surface image obtained by tracing the shape of the enamel portion. Create (step S71). Further, the image processor 56 displays the surface image on the monitor 60 at a predetermined view angle as a three-dimensional panoramic image, that is, a 3D autofocus image PIfocus in which all the parts are automatically subjected to the optimum focus process ( Step S72).
As a result, as shown in FIG. 25, it is possible to provide a 3D autofocus image PIfocus that can be formed along the contour where the structure of the dentition of the oral cavity of the subject P can be seen most clearly when viewed at a predetermined view angle. In the figure, a curved horseshoe-shaped range S is a range for displaying a 3D autofocus image PIfocus, and a solid line portion represents an actual position and shape of the dentition. As shown by the AA 'and BB' lines, the gums (alveolar bone) part, mandibular sinus, temporomandibular joint, carotid artery, etc. are tomographic distances set at a fixed distance from the ends of the teeth (mainly enamel) It is possible to create a tomographic plane and project a 3D tomographic plane. In this case, it cannot be guaranteed that these parts are in the optimum focus, but a 3D panoramic image can be reconstructed as an image that does not feel uncomfortable. Of course, it is needless to say that a method of calculating and using these portions as they are can be used depending on the purpose of diagnosis.

  In this way, the 3D autofocus image PIfocus is curved along the dentition, but its surface is bumpy, and this “bumpy” gives the actual position and shape (contour) of each tooth to the pixel It is expressed by shading. Other parts can also be expressed as images with no sense of incongruity.

  In this way, the 3D autofocus image PIfocus representing the actual position and shape of the dentition of each subject P is created.

<Various display processes>
Thereafter, the image processor 56 gives the operator an opportunity to observe the 3D autofocus image PIfocus in another manner. That is, the image processor 56 determines whether or not to interactively display the 3D autofocus image PIfocus in another manner based on operation information from the operator.

  As an example, the image processor 56 determines whether or not to observe a partial region of the 3D autofocus image (3D panoramic image) PIfocus (FIG. 5, step S10). If the determination in step S10 is YES, the observation of the partial area is further performed on the 3D reference tomographic plane SS or on the rectangular plane (two-dimensional) of the reference panoramic image, based on information from the operator. (Step S11). If it is determined in step S11 that the 3D reference tomographic plane SS is to be used, the image processor 56 re-projects the 3D autofocus image PIfocus onto the 3D reference tomographic plane SS along the X-ray irradiation direction DRx passing through each pixel. (Step S12). The state of this reprojection is shown in FIG. This re-projection is executed by a sub-pixel method in which one pixel of a 3D reference tomographic plane is separated from a corresponding three-dimensional pixel by a sub-pixel and re-projected.

The reprojected image on the 3D standard tomographic plane SS is displayed on the monitor 60 as a 3D reference image PI _proj-3D (step S13). An example of the 3D reference image PI _proj-3D is shown in FIG.

On the other hand, when it is determined in step S11 that the rectangular plane of the reference panoramic image PIst is used, the image processor 56 re-projects the 3D autofocus image PIfocus onto the rectangular plane, that is, the plane of the reference panoramic image (step S14). Needless to say, this re-projection is also performed by a so-called sub-pixel method in which one pixel of a standard panoramic image plane is re-projected by dividing a corresponding three-dimensional pixel by a sub-pixel. The concept of this reprojection is shown in FIG. This reprojection image is displayed on the monitor 60 as a 2D reference image PI _proj-2D (step S15). An example of the 2D reference image PI _proj-2D is shown in FIG.

Therefore, the operator sets a desired, for example, rectangular ROI (region of interest) in this 3D reference image PI _proj-3D or 2D reference image PI _proj-2D (step S16: see FIGS. 27 and 29). The image of the partial area specified by this ROI is enlarged, for example, and is superimposed and displayed on, for example, the currently displayed 3D reference image PI _proj-3D or 2D reference image PI _proj-2D (step S17). Of course, this display may be a single image separate from the panoramic image, may be a divided display with the panoramic image, or may be one of templates composed of a plurality of blocks simulating a dentition. The stored display may be used.

  Thereafter, the image processor 64 determines whether or not to end the series of processes from the operation information (step S18). If this determination is YES, the process returns to step S7 described above. On the other hand, if NO, the process returns to step S10 and the above-described process is repeated.

  On the other hand, when it is determined in step S10 described above that the partial image is not observed, the image processor 56 displays the currently displayed 3D autofocus image PIfocus by rotating, moving, and / or enlarging / reducing it. Whether or not interactively is determined (step S19). If this determination is YES, the 3D autofocus image PIfocus is rotated, moved, and / or enlarged / reduced according to the command information, and the image is displayed (steps S20 and S21). Thereafter, the process is passed to step S81, and the same process as described above is repeated.

  Of course, the types of display modes are not limited to those described above, and various other modes such as colorization can be adopted.

  When the operator has instructed to end the process, the image processor 64 ends the process through steps S18 and S7.

  In addition, after performing the setting process of step S16 mentioned above, you may make it transfer to the process of step S19, without performing the display process of step S17. In this case, the set ROI is displayed in step S21 together with the rotated, moved, enlarged / reduced image.

  As described above, according to the present embodiment, the projection direction can be expressed three-dimensionally by grasping the panoramic imaging space three-dimensionally. Therefore, as long as the panoramic image is in focus, the three-dimensionally expressed image is not distorted and an accurate panoramic image can be constructed. As a result, the panoramic image can be displayed more stably regardless of whether the positioning is good or not, and a clear image can be created with the entire panoramic image.

In addition, the position of the 3D reference tomographic plane in the imaging space can be easily grasped using the first and second phantoms according to the present embodiment. For this reason, since calibration can be easily performed for each device and after installation in a clinic, it is possible to perform calibration that complements individual differences and temporal changes for each device. Thereby, a highly accurate and stable image with high distance accuracy is always obtained.
In such a measurement using a phantom, not only variation of the apparatus can be suppressed, but also a panoramic apparatus in which the trajectory of the apparatus is not known can be grasped. Therefore, as long as there is a detector that can make use of this feature and software (device) to which the present invention is applied, the feature of the present invention can be utilized in any device.

  By the way, the radiation imaging apparatus according to the present invention is not limited to the one implemented in the dental panoramic imaging apparatus, but widely used for grasping the three-dimensional shape (position) inside the object using the tomosynthesis method. Can be implemented. Such applications include, for example, medical applications such as mammography using the tomosynthesis method and lung cancer scanners.

  In addition, this invention is not limited to the thing of embodiment mentioned above, It can change suitably further in the range which does not deviate from the summary of this invention, and it cannot be overemphasized that these are also contained in this invention.

  According to the present invention, it is possible to easily and accurately carry out calibration of information such as the position of a 3D reference tomographic plane that defines the positional relationship of the imaging space, and to accurately image an object for medical use using radiation. Diagnostic equipment can be provided.

1 Dental panoramic imaging device (radiation imaging device)
12 Computer 14 Imaging unit 31 X-ray tube (radiation source)
32 Detector 33 Collimator 41 High voltage generator 53 Buffer memory 54 Image memory 55 Frame memory 56 Image processor 57 Controller 58 Controller 60 Monitor

Claims (16)

A radiation source that emits radiation; and
A detector that is arranged opposite to the radiation source and outputs two-dimensional data of a digital quantity of electricity corresponding to the radiation when the radiation is incident;
Any one of the radiation source and detector pair, the detector, or the object is moved relative to the remaining elements of the radiation source, the detector, and the object. Transportation means;
Data collecting means for collecting the data output from the detector in units of frames while the pair of the radiation source and the detector, the detector, or the object is moved by the moving means;
Using the data collected by the data collection means and a reference tomographic plane set in an imaging space existing between the radiation source and the detector, a three-dimensional image of the imaging region of the object is obtained. In a radiation imaging apparatus, comprising: a first image creation means for creating,
A phantom for calibrating parameters defining the imaging space including the three-dimensional position of the reference tomographic plane;
Second image creation means for creating an image for calibration based on data collected by the data collection means when radiation is emitted from the radiation source in a state where the phantom is arranged in the imaging space;
Radiation comprising: calibration means for calibrating the parameter based on information on a known amount of the phantom and projection position information of the phantom obtained from the calibration image. Imaging device. The radiation source is an X-ray tube that emits X-rays as the radiation,
The detector is a detector for detecting the X-rays irradiated from the X-ray tube;
The first image creating means creates a three-dimensional image in which the data collected by the data collecting means is processed as the three-dimensional image based on a tomosynthesis method to optimize the focus of the imaging part of the object. Means to
The second image creation means is means for processing the data collected by the data collection means based on a tomosynthesis method to create the calibration image.
The radiation imaging apparatus according to claim 1. The phantom is
A first phantom for obtaining parameters of a three-dimensional positional relationship between the radiation source and the detector;
A second phantom for obtaining parameters of a three-dimensional positional relationship with respect to the object at the time of relative movement of the radiation source and the radiation detector;
The radiation imaging apparatus according to claim 1, further comprising: The phantom is
The radiation imaging apparatus according to claim 3, further comprising a third phantom for measuring a projection angle of radiation emitted from the radiation source toward the detector.   The radiation imaging apparatus according to claim 4, wherein the third phantom and the second phantom are created as a set of phantoms. The second phantom is disposed at a position corresponding to the reference tomographic plane,
The second image creation means is means for creating the calibration image by optimizing the focus of the position of the reference tomographic plane,
The parameters acquired by the first phantom include distance information from the radiation source to the detector,
The parameters acquired by the second phantom include distance information between the position of the reference tomographic plane and the radiation source.
The radiation imaging apparatus according to claim 3.   The first phantom is disposed at each of a plurality of known positions including information whose distance in the direction along the detection surface of the detector in the imaging space is unknown, and is transparent to the medium of the imaging space. The radiation imaging apparatus according to claim 3, comprising a plurality of phantom bodies having different rates.   The second phantom has a plurality of rods arranged at positions corresponding to the reference tomographic plane for each projection direction of the radiation in the imaging space when distance information from the radiation source to the detector is known The radiation imaging apparatus according to claim 6, comprising a point-like phantom body.   The third phantom has two rod-like or dot-like phantom bodies arranged for each projection direction that are spaced apart from each other in the projection direction of the radiation in the imaging space. The radiation imaging apparatus according to claim 7.   Of the two rod-shaped phantom bodies for each projection direction in the third phantom, one phantom body is also used as the rod-shaped phantom body for each projection direction of the second phantom body. The radiation imaging apparatus according to claim 8. The X-ray tube and the detector are arranged so as to face each other with a human jaw including a dentition as an imaging part of the object,
The moving means is means for rotating the pair of the X-ray tube and the detector around the human jaw,
The radiation imaging apparatus according to claim 2, wherein the first image creating unit is configured to create the three-dimensional image along a cross section of the dentition. A radiation source that emits a radiation source and a detector that detects the radiation as an electrical signal are opposed to each other with an object interposed therebetween, and the pair of the radiation source and the detector is rotated around the object. The radiation source in a radiation imaging apparatus that detects radiation emitted from the radiation source and transmitted through the object as an electrical signal by a detector, and creates an image of the imaging region of the object based on the electrical signal A phantom device for calibrating a three-dimensional spatial positional relationship between the detector and the detector,
A first phantom for obtaining parameters of a three-dimensional positional relationship between the radiation source and the detector;
A second phantom for obtaining parameters of a three-dimensional positional relationship with respect to the object at the time of relative movement of the radiation source and the radiation detector;
A phantom device characterized by comprising:   The phantom device according to claim 12, further comprising a third phantom for measuring a projection angle of radiation radiated from the radiation source toward the detector.   The phantom device according to claim 13, wherein the third phantom and the second phantom are created as a set of phantoms. A radiation source that emits radiation; and
A detector that is arranged opposite to the radiation source and outputs two-dimensional data of a digital quantity of electricity corresponding to the radiation when the radiation is incident;
Any one of the radiation source and detector pair, the detector, or the object is moved relative to the remaining elements of the radiation source, the detector, and the object. Transportation means;
Data collecting means for collecting the data output from the detector in units of frames while the pair of the radiation source and the detector, the detector, or the object is moved by the moving means;
Using the data collected by the data collection means and a reference tomographic plane set in an imaging space existing between the radiation source and the detector, a three-dimensional image of the imaging region of the object is obtained. In a radiation imaging apparatus, comprising: a first image creation means for creating,
A phantom for calibrating parameters defining the imaging space including the three-dimensional position of the reference tomographic plane;
Second image creation means for creating an image for calibration based on data collected by the data collection means when radiation is emitted from the radiation source in a state where the phantom is arranged in the imaging space;
Calibration means for calibrating the parameter based on information of a known amount of the phantom and projection position information of the phantom obtained from the calibration image;
A radiation imaging apparatus comprising: modeling means for modeling the reference tomographic plane using the parameters calibrated by the calibration means. The parameter is
It includes distance information from the radiation source to the detector, distance information between the position of the reference tomographic plane and the radiation source, and an actual value of the projection angle of the radiation onto the detector. The radiation imaging apparatus according to claim 15.

Images