JP2015033552A - X-ray machine and image processing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide an X-ray machine which can perform position correction by suppressing generation of distortion of each tomosynthesis tomographic image after reconstruction calculation by using an FBP (Filtered Back Projection) method.SOLUTION: An X-ray machine includes rotation means, storage means which sequentially stores signals outputted by an X-ray detection part as frame data when the rotation means moves an X-ray irradiation part and the X-ray detection part around an analyte, and image processing means which reconstructs a tomosynthesis tomographic image from the frame data by using the FBP method. The image processing means performs position correction for matching or substantially matching a length of at least one of lateral direction and lengthwise direction of each of the tomosynthesis tomographic images after reconstruction calculation on the basis of luminance of a voxel on an arbitrary sectional plane used as a correction reference at the time when luminance of a voxel on one sectional plane other than the arbitrary sectional plane used as the correction reference is projected on the arbitrary sectional plane used as the correction reference.

Description

  The present invention relates to an X-ray imaging apparatus that performs tomosynthesis for reconstructing a tomographic image having an arbitrary cutting height by one tomography, and reconstructs a tomosynthesis tomographic image using an FBP (Filtered Back Projection) method. The present invention relates to an image processing method for reconstructing a tomosynthesis tomographic image using the FBP method.

  Conventionally, in a dental panoramic X-ray imaging apparatus, a panoramic image is created on the assumption that a plurality of tomographic planes are parallel to a detection plane of a detector. For this reason, when the subject's dentition is located on the reference tomographic plane, an image without distortion could be created for the center of the front tooth, but the subject's dentition from the reference tomographic plane was The image was always distorted when it came off. In addition, if the subject's dentition was not positioned along the reference tomographic plane, the image was also blurred horizontally.

  Therefore, in recent years, as a dental panoramic X-ray imaging apparatus capable of obtaining an image with little blur even when the subject's dentition is deviated from the reference tomographic plane, a tomography having an arbitrary cutting height can be obtained by one tomography. A dental panoramic X-ray imaging apparatus that performs tomosynthesis to reconstruct an image has been proposed.

  One of the methods used when reconstructing a tomosynthesis tomographic image is the FBP method. The FBP method has an advantage that a high contrast image can be created and diagnosis of the apex, root canal, periodontal ligament and the like is easy.

JP2013-85667A (FIG. 12)

  However, each tomosynthesis tomographic image after reconstruction calculation using the FBP method does not match the length in the horizontal direction because the number of voxels in the horizontal direction of the anterior teeth portion is different, and the maximum of each tomographic plane from the X-ray source. Since the elevation angles at which the points are viewed are different, the lengths in the vertical direction do not match, and it is inconvenient if the tomosynthesis tomographic images of a plurality of layers are compared with each other to compare the state of teeth according to the tomographic depth. Therefore, in Patent Document 1, tooth position correction is performed to match the lengths in the horizontal and vertical directions of each tomosynthesis tomographic image after reconstruction calculation.

  However, each tomosynthesis tomographic image after reconstruction calculation using the FBP method has a different number of voxels in the lateral direction of only the front tooth part, and an elevation angle when the highest point of each tomographic plane is viewed from the X-ray source. Therefore, if each tomosynthesis tomographic image after reconstruction calculation is uniformly enlarged or reduced in the horizontal and vertical directions at the time of dentition position correction, the tomographic image after dentition position correction will be distorted. There was a problem.

  In view of the above circumstances, the present invention provides an X-ray imaging apparatus and an image processing method capable of correcting the position of each tomosynthesis tomographic image after reconstruction calculation using the FBP method while suppressing the occurrence of distortion. It is the purpose.

  In order to achieve the above object, an X-ray imaging apparatus according to the present invention provides an X-ray irradiator for irradiating a subject with X-rays and a digital electric signal corresponding to the incident X-rays at a constant frame rate. An X-ray detection unit for outputting, a turning means for moving the X-ray irradiation unit and the X-ray detection unit pair around the subject in a state of facing each other across the subject, Storage means for sequentially storing signals output by the X-ray detection unit as frame data as the turning unit moves the X-ray irradiation unit and the X-ray detection unit around the subject; and FBP method Image processing means for reconstructing a tomosynthesis tomographic image from the frame data using the image data, and the image processing means corrects the luminance of a voxel on a certain tomographic plane other than an arbitrary tomographic plane used as a correction reference. Any notice used as Based on the luminance of the voxel on an arbitrary tomographic plane used as a correction criterion when projected onto the plane, the lengths of at least one of the horizontal direction and the vertical direction of each of the tomosynthesis tomographic images after reconstruction calculation are matched or A configuration (first configuration) for performing position correction for approximately matching is adopted.

In the X-ray imaging apparatus having the first configuration, the image processing unit may include coordinates i on an arbitrary tomographic plane used as a correction reference (where i is a target in a voxel group having the same tomographic plane and height direction). The coordinates ii (i, i, i, i, i, i) of the voxel of a certain tomographic plane j through which the n (i) -th X-ray of N (i) frames that pass through the voxel) j, n (i)) is calculated for each frame and each tomographic plane, and is used as a correction criterion. The coordinate k on any tomographic plane (where k is a variable for specifying the position of the target voxel in the height direction) ), The coordinates kk (k, j, h (k)) of voxels of a certain tomographic plane j through which the h (k) -th X-ray of the H (k) frames that pass through the voxels of The frame and each tomographic plane are calculated, and the luminance value of the voxel of a certain tomographic plane j is calculated by the loop of voxel coordinates i, coordinates k, n (i), h (k). When N (i) ≠ 0 and H (k) ≠ 0, the brightness value datawa (i, j, k) when projected on an arbitrary tomographic plane used as a correction criterion is N according to the following equation (1). When (i) ≠ 0 and H (k) = 0, N (i) according to the following equation (2), and when N (i) = 0 and H (k) ≠ 0, according to the following equation (3), N (i ) = 0 and H (k) = 0, the configuration is calculated from the luminance value cal (ii, j, kk) of the point (ii, j, kk) on the fault plane j according to the following equation (4) ( The second configuration is desirable.
However, i1 and i2 in the above formulas (3) and (4) are when there is no frame corresponding to the coordinate i of the voxel on an arbitrary tomographic plane used as a correction reference (n (i) = 0). , The coordinates ii on the tomographic plane j corresponding to i satisfying n (i)> 0 in the preceding and succeeding frames, and satisfying i1 ≦ i2. In the above equations (2) and (4), k1 and k2 are obtained when there is no frame corresponding to the coordinate k of a voxel on an arbitrary tomographic plane used as a correction reference (h (k) = 0). The coordinates kk on the tomographic plane j corresponding to k satisfying h (k)> 0 in the preceding and following frames, and satisfying k1 ≦ k2.

In the X-ray imaging apparatus having the second configuration, the image processing unit is configured such that i1 = ii (i−Δ1, j, N (i−Δ1)), i2 when i1 + 1 <i2 and / or k1 + 1 <k2. = Ii (i + Δ2, j, N (i + Δ1)), and the coordinate on the tomographic plane j corresponding to the coordinate i on any tomographic plane used as the correction reference is a real number i _j expressed by the following equation (5) The following formula (6) is used instead of the above formula (2), the following formula (7) is used instead of the above formula (3), and the following formula (8) is used instead of the above formula (4) ( The third configuration is desirable.
However, Δ1 + Δ2-1 is the number of continuous voxels through which X-rays of a frame on an arbitrary tomographic plane used as a correction reference are not transmitted, and flr (x) in the above equations (6) to (8) is a real number x It represents the largest integer among the following integers.

  In order to achieve the above object, an image processing method according to the present invention is an image processing method for correcting the position of each tomosynthesis tomographic image after reconstruction calculation using an FBP (Filtered Back Projection) method, and is used as a correction reference. Reconstruction based on voxel brightness on an arbitrary tomographic plane used as a correction criterion when projected on an arbitrary tomographic plane using the luminance of a voxel on a certain fault plane other than an arbitrary tomographic plane as a correction criterion The position correction is performed to match or substantially match at least one of the horizontal and vertical lengths of the tomosynthesis tomographic images after calculation.

  According to the present invention, each tomosynthesis tomographic image after reconstruction calculation is not enlarged or reduced uniformly in each of the horizontal direction and the vertical direction to perform position correction, but a certain fault other than an arbitrary tomographic plane used as a correction reference. The position correction is performed based on the luminance of the voxel on the arbitrary tomographic plane used as the correction reference when the luminance of the voxel on the surface is projected onto the arbitrary tomographic plane used as the correction reference. As a result, it is possible to correct the position of each tomosynthesis tomographic image after reconstruction calculation using the FBP method while suppressing the occurrence of distortion.

It is a block diagram explaining the basic composition of the dental X-ray imaging apparatus concerning one embodiment of the present invention. It is a figure which shows the orbit figure of panoramic imaging. It is a figure which shows the example of an external appearance of a standing type X-ray imaging apparatus main body. It is a functional block diagram for demonstrating the image processing means of the dental X-ray imaging apparatus which concerns on one Embodiment of this invention. It is a figure which shows the tomographic plane of five layers. 5 is a flowchart illustrating an image processing operation according to an embodiment of the present invention. It is a figure which shows the coordinate transformation of a system. It is a perspective view which shows an example of the positional relationship of the pixel group which is a part of detection surface of an X-ray detector, and four voxels. It is a top view which shows the positional relationship shown in FIG. 8A. It is a perspective view which shows an example of the positional relationship of the pixel group which is a part of detection surface of an X-ray detector, and one voxel. It is a top view which shows the positional relationship shown in FIG. 9A. It is a figure which shows the thickness of the voxel with respect to a X-ray incident direction. It is a figure which shows an example of the diagonal square pole which approximates a voxel. It is a perspective view which shows the rectangle formed in a voxel. It is a figure which shows the other example of the diagonal square pole which approximates a voxel. It is a figure which shows the other example of the diagonal square pole which approximates a voxel. It is a figure which shows the other example of the diagonal square pole which approximates a voxel. It is a figure which shows the other example of the diagonal square pole which approximates a voxel. It is a figure which shows an example in case the X-ray which permeate | transmits a voxel inclines in the Z-axis direction with respect to the detection surface of an X-ray detector. It is a figure which shows the outline | summary of dentition position correction | amendment. It is the figure which arranged in a straight line for every tomographic plane the voxel whose height direction on a some tomographic plane is a predetermined position for convenience. It is a flowchart which shows an example of dentition position correction | amendment.

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

  FIG. 1 is a block diagram illustrating a basic configuration of a dental X-ray imaging apparatus according to an embodiment of the present invention. The dental X-ray imaging apparatus according to this embodiment can perform dental panoramic X-ray imaging and head and neck X-ray CT imaging. The dental X-ray imaging apparatus according to this embodiment includes an X-ray imaging apparatus main body 1 and an X-ray image display apparatus 2 and is configured to transmit and receive data via a communication cable or the like.

  The X-ray imaging apparatus body 1 includes an X-ray irradiation unit 110 that irradiates a subject (patient) with X-rays, an X-ray detection unit 120 that detects X-rays transmitted through the subject O, and an X-ray irradiation unit. 110 and the swivel arm 3 which has the X-ray detection part 120 facing each other.

  In panoramic imaging, tomographic imaging is performed while the swivel arm 3 is horizontally moved and swiveled so that the X-ray irradiation unit 110 and the X-ray detection unit 120 draw a predetermined locus along the shape of the dental arch.

  The X-ray imaging apparatus main body 1 includes a subject holding means (head fixing part) 4 that holds the maxillofacial surface of the subject O, a drive unit 160 that drives the turning arm 3, and an imaging apparatus main body control unit 170. And further. An operation panel 71 a is added to the photographing apparatus main body control unit 170.

  The X-ray irradiation unit 110 includes an X-ray generator 112 including an X-ray tube or the like that irradiates X-rays, and a collimator 111 including a slit or the like that controls the spread of the X-ray beam. The collimator 111 is composed of, for example, a mask in which a plurality of slits having different shapes are formed. The collimator 111 selects one of the slits and spreads the X-ray beam irradiated from the X-ray generator 112 by the selected slit. It has become a mechanism to limit. For example, in panoramic imaging, a panoramic slit is selected, and an X-ray slit beam corresponding to the selected shape of the panoramic slit is irradiated toward the X-ray detector 121 of the X-ray detector 120.

  The X-ray detection unit 120 includes a cassette 122 provided with an X-ray detector 121 including a CCD sensor spread in two dimensions, an X-ray indirect conversion method (FPD: flat panel detector) sensor, and the like. Although the cassette 122 is detachable from the X-ray detection unit 120, the X-ray detector 121 may be fixedly provided on the X-ray detection unit 120 without using the cassette 122. Since the dental X-ray imaging apparatus according to the present embodiment employs the tomosynthesis method, the X-ray detector 121 has a plurality of X-ray detection elements also in the lateral (width) direction. For example, the X-ray detector 121 can collect incident X-rays as image data having a digital electrical quantity corresponding to the amount of X-rays at a frame rate of 300 fps (one frame is, for example, 64 × 1500 pixels). it can. Hereinafter, this collected data is referred to as “frame data”.

  The drive unit 160 includes a turning axis position setting unit 161 that moves the turning shaft 3 c along the dental arch during panoramic photography, and a turning shaft rotating unit 162 that rotates the turning arm 3. The turning axis position setting means 161 includes an XY table for moving the turning axis 3c in the XY direction (horizontal direction), an X axis motor, and a Y axis motor. The turning shaft rotating means 162 includes a turning motor provided on the XY table of the turning axis position setting means 161. The turning motor is a hollow motor that has a highly accurate positioning function by a direct drive system in which a load is directly connected to the motor without using a reduction gear. The pivot axis 3c is set to coincide with the midline of the subject O at the start of imaging.

  The X-ray imaging apparatus main body 1 further includes turning arm position setting means 31 for moving the turning arm 3 straight in the horizontal direction. The turning arm position setting means 31 includes a turning table that can turn and a slide table that can move directly on the turning table.

  At the time of panoramic imaging or CT imaging, the pair of the X-ray irradiation unit 110 and the X-ray detection unit 120 are positioned so as to face each other across the oral cavity of the subject O, and each pair is integrated. Driven to rotate around the oral cavity. However, this rotation is not a rotation that draws a simple circle. That is, the pair of the X-ray generator 112 and the X-ray detector 121 is rotationally driven so that the rotational center RC of the pair moves along the Y axis as shown in FIG. The Y axis is an axis of symmetry of a substantially horseshoe-shaped standard dentition SS. In FIG. 2, one trajectory of the swivel arm 3 is shown every predetermined number of frames (for example, every 30 frames) in the right half of the page, and is not shown in the left half of the page. In addition, the X-ray irradiation unit 110 and the X-ray detection unit 120 are illustrated only for one trajectory, and illustration is omitted for the other trajectories.

  The imaging apparatus main body control unit 170 includes a control device (for example, a CPU) 171 that executes various control programs including a control program that controls the drive unit unit 160, and an X-ray generation unit control unit 172 that controls the X-ray irradiation unit 110. And X-ray detection unit control means 173 for controlling the X-ray detection unit 120. The operation panel 71a is composed of a small liquid crystal panel and a plurality of operation buttons. In addition to the operation buttons, input means such as a keyboard, a mouse, and a touch panel can also be used. In addition, the operation panel 71a may include display means including a display such as a liquid crystal monitor.

  For example, information such as characters and images necessary for operating the X-ray imaging apparatus main body 1 may be displayed on the display means provided on the operation panel 71a. Further, the display means connected to the X-ray image display device 2 for displaying the X-ray panoramic image or the X-ray CT image, and the display content displayed on the display means 26 of the X-ray image display device 2 is provided on the operation panel 71a. In addition, various commands may be issued to the X-ray imaging apparatus main body 1 through a pointer operation with a mouse or the like on characters or images displayed on the display means 26.

  The X-ray imaging apparatus main body 1 performs panoramic imaging of a dental arch in accordance with a command from the operation panel 71a or the X-ray image display apparatus 2. In addition, various commands, coordinate data, and the like are received from the X-ray image display device 2, while photographed image data is sent to the X-ray image display device 2. Further, the base of the imaging region (imaging target region) rr is based on the lower edge line of the cone beam emitted from the X-ray irradiation unit 110.

  The X-ray image display device 2 connected to the X-ray imaging device main body 1 includes a display device main body 20. The display device main body 20 is provided with display means 26 composed of a display device such as a liquid crystal monitor, and operation means 25 composed of a keyboard, a mouse, and the like. An X-ray image is displayed on the display means 26. Various commands can be given through a pointer operation with a mouse on characters or images displayed on the display means 26. Since the display means 26 can be constituted by a touch panel, the display means 26 also serves as the operation means 25 in this case.

  The image display device main body 20 includes a control device (for example, CPU) 21 that executes various programs, a storage unit 22 that is configured by a hard disk or the like and stores various types of shooting data, images, and the like, and an image generation unit 23 that performs image reconstruction. It has. The control device 21, the storage unit 22, and the image generation unit 23 constitute an image processing unit. The storage unit 22 also stores dentition coordinates and the like which will be described later. The control device 21 controls various operations according to a program stored in the storage unit 22 and operates so as to fulfill the function of the image generation means 23 according to the program. That is, the control device 21 also serves as the image generation unit 23.

  The storage unit 22 can store frame data, tomosynthesis tomographic images, and the like.

  FIG. 3 is a diagram showing an example of the appearance of the standing type X-ray imaging apparatus main body 1. Although the standing type will be described here, the present invention can be applied to a dental X-ray imaging apparatus including a sitting type X-ray imaging apparatus main body. 3A is a top view, FIG. 3B is a front view, and FIG. 3C is a side view.

  The X-ray imaging apparatus main body 1 includes a base 5 placed on the floor, a lower pole 6 erected in the vertical direction from the base 5, and an upper pole 7 connected to the lower pole 6 so as to be slidable in the vertical direction. A fixed arm 8 fixed to the upper end of the upper pole 7, a revolving arm 3 rotatably connected to the fixed arm 8, and a subject (for example, a tooth) fixed to the center of the upper pole 7 And a head holding portion 9 that holds the head of the human body including the head. The swivel arm 3 has an X-ray irradiation unit 110 and an X-ray detection unit 120, and the head holding unit 9 has a subject holding means (head fixing unit) 4. In the embodiment, the fixed arm 8 is fixed to the upper pole 7. However, for example, the fixed arm 8 is directly or directly on the wall or ceiling of the room where the X-ray imaging apparatus main body 1 is installed. The aspect attached with the adjustment mechanism which can be adjusted may be sufficient.

  FIG. 4 shows a block diagram for control and processing of the X-ray imaging apparatus according to the present embodiment. As shown in the figure, data from the X-ray imaging apparatus main body 1 is given to the X-ray image display apparatus 2 via a communication line.

  The X-ray image display device 2 is configured by, for example, a personal computer or a workstation capable of storing a large amount of image data in order to handle a large amount of image data. That is, the X-ray image display apparatus main body 20 includes a control device 21, a storage unit 22, an image generation unit 23, and interfaces 24 and 27 as main components, and these are the internal bus 28. Are connected so as to be able to communicate with each other. The control device 21 includes a CPU 21a and a ROM 21b in which a control program such as an OS is stored. The storage unit 22 includes a buffer memory 22a, an image memory 22b, and a frame memory 22c. An operation unit 25 and a display unit 26 are connected to the control device 21 via an interface 24.

  The interface 27 is connected to the X-ray imaging apparatus main body 1 and mediates communication of control information and collected data exchanged between the control device 21 and the X-ray irradiation unit 110 and the X-ray detection unit 120. Accordingly, the control device 21 can transmit a panoramic image taken by the X-ray imaging apparatus main body 1 to an external server according to, for example, DICOM (Digital Imaging and Communications in Medicine) standard.

  The buffer memory 22 a temporarily stores the digital electrical quantity frame data received from the X-ray detection unit 120 via the interface 27.

  The image generation unit 23 is placed under the control of the control device 21 and has a function of interactively executing a process for creating a panoramic image with an operator. A program for realizing this function is stored in advance in the ROM 21b. This program may be stored in the ROM 21b in advance. However, in some cases, the program may be installed in the ROM 21b or another non-illustrated non-volatile recording medium via a communication line or a portable memory. May be.

  Frame data and image data processed by the image generating means 23 or being processed are stored in the image memory 22b so as to be readable and writable. For the image memory 22b, a large-capacity recording medium (nonvolatile and readable / writable) such as a hard disk is used. The frame memory 22c is used for displaying reconstructed panoramic image data and the like. The image data stored in the frame memory 22 c is given to the display means 26 via the interface 24 and displayed on the screen of the display means 26.

  The CPU 21a of the control device 21 controls the overall operation of the components of the device in accordance with a program responsible for overall control and processing stored in the ROM 21b. Such a program is set so as to interactively receive operation information for each control item from the operator. Therefore, the CPU 21a is configured to be able to execute collection (scanning) of frame data and the like, as will be described later.

  When the subject O places his / her chin on the head holding portion 9 in a standing posture, the position of the head (jaw portion) of the subject O is fixed at substantially the center of the rotation space of the swivel arm 3. . In this state, the turning arm 3 rotates around the head of the subject O along the XY plane.

  While the revolving arm 3 is rotating, X-rays are emitted from the X-ray irradiation unit 110. The X-rays pass through the jaw (dentation portion) of the subject O located at the imaging position and enter the X-ray detection unit 120. As described above, the X-ray detection unit 120 detects incident X-rays at a very high frame rate (for example, 300 fps), and outputs corresponding two-dimensional digital data (for example, 64 × 1500 pixels) of digital electricity. Output sequentially in frame units. As described above, the frame data is temporarily stored in the buffer memory 22a via the communication line and the interface 27 of the X-ray image display apparatus 2. The temporarily stored frame data is then transferred and stored in the image memory 22b.

  In this embodiment, the tomosynthesis method is used to obtain a panoramic image. In the tomosynthesis method, X-rays are irradiated from the X-ray irradiation unit 110 to the subject O at a plurality of different angles, and X-rays transmitted through the subject O are detected by the X-ray detection unit 120, A photographed image (frame data) is photographed and stored in the image memory 22b. The image generation unit 23 reconstructs a tomosynthesis tomographic image based on an arbitrary tomographic plane from a plurality of captured images (frame data) stored in the image memory 22b. The tomosynthesis tomographic image is stored in the storage unit 22 as a panoramic image or displayed on the display means 26. In the present embodiment, as shown in FIG. 5, the reconstruction area R1 is set so that the standard dentition SS is positioned approximately at the center in the thickness direction of the reconstruction area R1, and the positive direction of the Y axis in the reconstruction area R1 60 layers of faults are set at a pitch of 0.5 mm from the (inside direction) toward the negative direction (outside direction).

  In this embodiment, a 60-layer tomosynthesis tomographic image corresponding to the 60-layer tomographic image is reconstructed from the frame data using the FBP method.

  The 60-layer tomosynthesis tomographic image is stored in the image memory 22b as a 60-layer panoramic image. The panoramic image stored in the image memory 22b is displayed on the display means 26 in various ways. The display mode is set based on a user instruction given from the operation means 25, for example.

  Next, an example of processing for reconstructing a tomosynthesis tomographic image using the FBP method will be described with reference to the flowchart of FIG.

  When generating a tomosynthesis tomographic image in accordance with a command from the operation panel 71a or the X-ray image display device 2, the image generating means 23 first reads defect registration data and creates a defect table (step S1). Then, the control device 171 reads the trajectory data of the frame (Step S2).

  Next, the image generation means 23 creates a black and white inversion lookup table (step S3). Subsequently, the image generation means 23 reads the density correction image and creates density correction data (step S4).

  Then, the X-ray imaging apparatus main body 1 executes a scan for data collection (step S5). Thereby, the turning arm 3 and the X-ray irradiation unit 110 are driven in accordance with a preset control sequence. That is, a pair of the X-ray irradiation unit 110 and the X-ray detection unit 120 rotates around the jaw of the subject O, and the X-ray irradiation unit 110 emits X-rays during the rotation operation. X-rays exposed from the X-ray detection unit 110 pass through the subject O and are detected by the X-ray detection unit 120. Therefore, as described above, the digital electric quantity frame data (projection data) reflecting the X-ray transmission amount is output from the X-ray detector 121 at a rate of, for example, 300 fps. This frame data is stored in the image memory 22b.

  The image generation unit 23 reads the projection data and performs defect correction, density correction, and black / white reversal correction on the projection data (step S6).

  For each projection data read in step S6, it is only necessary to calculate voxels through which the X-rays exposed from the X-ray detection unit 110 pass. In each projection data, since the range of voxels through which X-rays pass is narrow, the calculation time can be shortened by performing calculation only for voxels through which X-rays pass.

  Therefore, the image generation means 23 creates a dentition direction calculation range table for designating in advance the range of voxels to be calculated in each tomographic plane for each frame data (step S7).

  Each tomosynthesis tomographic image after the reconstruction calculation has different horizontal lengths because the number of voxels in the lateral direction of the front tooth part is different, and the elevation angle when the highest point of each tomographic plane is viewed from the X-ray source is different. Therefore, the lengths in the vertical direction do not match, and it is inconvenient if the tomosynthesis tomographic images of a plurality of layers are compared as they are to compare the state of teeth according to the tomographic depth. Therefore, the image generation means 23 uses each trajectory (each frame) as a table necessary for correcting the dentition position for making the horizontal and vertical lengths of the tomosynthesis tomographic images after the reconstruction calculation coincide with each other. In step S8, a table for specifying the ratio of the height and the table for specifying the voxel through which the X-ray incident on the center of the X-ray detector 121 passes is created.

  Next, the image generation means 23 calculates the actual position coordinates of each voxel vertex that occupies the reconstruction area (step S9).

  Subsequently, the image generation means 23 reads the trajectory data (arm angle and suspension axis position) and the projection data obtained by the data collection processing by the scan in step S5 (step S10), and convolves the projection data with the filter function. Integrate (step S11).

  Thereafter, in order to simplify the calculation, the image generation unit 23 sets the center of the X-ray generator 112 (X-ray source) as the origin and the X-ray detector 121 as shown in FIG. System coordinate conversion is performed by rotational movement and parallel movement of the system including the X-ray generator 112, the reconstruction region R1, and the X-ray detector 121 so that the center position is in the positive direction on the Y-axis (Step S1). S12).

  Next, reconstruction calculation using the FBP method executed by the image generation unit 23 in step S13 will be described with reference to FIGS. 8A to 16.

  First, let us focus on a certain pixel on the detection surface of the X-ray detector 121, and consider the case where the X-ray from the X-ray generator 112 is incident on the target pixel.

  X-rays incident on the pixel of interest are attenuated when passing through the subject, and X-rays captured by the X-ray detector 121 are reduced. That is, when the reconstruction area R1 shown in FIG. 7 configured by a plurality of voxels is set, X-rays transmitted through the voxel are incident on the target pixel, and the X-rays are attenuated by the amount of the voxel transmitted through the X-rays. This is reflected in the luminance value of the pixel of interest.

  When the X-rays incident on the target pixel pass through a part of the plurality of voxels in the tomographic plane, the X-rays are attenuated according to the volume ratio of the part where the X-rays of each voxel are transmitted, and the degree of attenuation is the target pixel. It is reflected in the brightness value.

  That is, each voxel through which X-rays incident on the pixel of interest are transmitted contributes to the luminance value of the pixel of interest at a ratio of the volume ratio of the portion of each voxel through which X-rays are transmitted. Conversely, if the luminance value of the target pixel is divided by the volume ratio of the portion of the voxel through which the X-ray incident on the target pixel is transmitted, the divided value is transmitted by the X-ray of each voxel. Corresponds to the X-ray attenuation of the part.

  FIG. 8A is a perspective view showing an example of a positional relationship between a pixel group PXG that is a part of the detection surface of the X-ray detector 121 and the voxels VX1 to VX4. FIG. 8B is a top view showing the positional relationship shown in FIG. 8A. The pixel group PXG is composed of pixels PX1 to PX16. 8A and 8B, if the pixel of interest is the pixel PX11, the voxels VX1 to VX4 are voxels through which X-rays incident on the pixel of interest are transmitted.

  On the other hand, the X-ray attenuation reflected on the pixel of interest, that is, the luminance value of the pixel of interest is the luminance value of the pixel of interest and the total volume of the portion through which the X-rays of all the voxels transmitted through the X-ray incident on the pixel of interest are transmitted. The multiplication value of the volume ratio of the portion through which X-rays of one voxel in all voxels are transmitted is the sum of the individual voxels in the whole voxel. Therefore, when attention is paid to a certain voxel, X-rays transmitted through the target voxel are incident on a plurality of pixels. Therefore, the influence of X-ray attenuation in the target voxel is applied to each pixel according to the volume ratio corresponding to each pixel. Will be given. In other words, for each pixel, among the affected X-ray attenuation, the ratio of the influence of the target voxel on the pixel (volume ratio) is multiplied by the luminance value, and the multiplied value is integrated for the pixel. By doing so, the X-ray attenuation of the entire target voxel can be obtained. That is, back projection onto the target voxel is obtained.

  FIG. 9A is a perspective view showing an example of a positional relationship between a pixel group PXG and a voxel VX1 which are a part of the detection surface of the X-ray detector 121. FIG. FIG. 9B is a top view showing the positional relationship shown in FIG. 9A. 9A and 9B, if the target voxel VX1 is the target voxel, the X-ray attenuation of the entire target voxel VX1 is (1) the pixel of the target voxel VX1 with respect to the luminance value of the pixel PX5 and the volume of the target voxel VX1. The product of the ratio of the volume of the portion through which X-rays incident on PX5 are transmitted, (2) the portion through which X-rays incident on the pixel PX6 of the target voxel VX1 with respect to the luminance value of the pixel PX6 and the volume of the target voxel VX1 are transmitted. (3) the multiplication value of the luminance value of the pixel PX7 and the ratio of the volume of the portion through which X-rays incident on the pixel PX7 of the target voxel VX1 are transmitted relative to the volume of the target voxel VX1; 4) Enter the pixel PX9 of the target voxel VX1 with respect to the luminance value of the pixel PX9 and the volume of the target voxel VX1. (5) the luminance value of the pixel PX10 and the volume of the portion through which the X-ray incident on the pixel PX10 of the target voxel VX1 is transmitted with respect to the volume of the target voxel VX1. A sum of a multiplication value of the ratio and (6) a luminance value of the pixel PX11 and a ratio of the volume of the portion through which the X-ray incident on the pixel PX11 of the target voxel VX1 is transmitted to the volume of the target voxel VX1. Become.

  Specifically, since the X-rays that pass through the voxels go straight, the detection surface of the X-ray detector 121 is projected to the voxel position in the X-ray incident direction for each voxel. Conversely, for each voxel, the voxel may be projected to the position of the detection surface of the X-ray detector 121 in the X-ray incident direction.

  If none of the constituent surfaces of the cuboid voxel is perpendicular to the X-ray incident direction, for example, as shown in FIG. 10, the thickness of the voxel with respect to the X-ray incident direction is not uniform in the voxel VX1, so depending on the pixel In some cases, X-rays transmitted through a thin part of the voxel may be incident. If this is strictly calculated, the calculation time of back projection becomes enormous.

  Therefore, in the present embodiment, back projection is performed by approximating a rectangular parallelepiped voxel to a rectangular prism having the same volume as the rectangular parallelepiped voxel and a uniform thickness in the X-ray incident direction. The oblique square column has two bottom surfaces by translating a pair of opposing constituent surfaces in which there are overlapping regions when viewed from the X-ray incident direction on a plane including both opposing constituent surfaces, or only one of them, It has the same volume as a cuboid voxel and a uniform thickness in the X-ray incident direction. Even if such approximation is performed, the image quality of the tomosynthesis tomographic image is hardly affected.

  FIG. 11A is a diagram illustrating an example of an oblique quadrangular column that approximates the voxel VX1. The oblique square column OP1 shown in FIG. 11A has a shape in which the rectangles RT2 and RT3 are the bottom surfaces and the thickness in the X-ray incident direction is uniform. The rectangle RT1 is formed on each side of the target voxel VX1 other than the sides included in the pair of opposing configuration surfaces in which the overlapping regions as seen from the X-ray incidence direction exist among the configuration surfaces of the voxel VX1 shown in FIGS. 9A and 9B. A rectangle with the midpoint as the vertex. A rectangle RT1 formed in the target voxel VX1 is shown in a perspective view in FIG. 11B. The rectangle RT2 detects the X-ray detection of the side closer to the X-ray detector 121 of the pair of opposing constituent surfaces where there is an overlapping region when viewed from the X-ray incident direction of the constituent surfaces of the voxel VX1 shown in FIGS. 9A and 9B. This is translated on a plane including the opposing constituent surface closer to the vessel 121. The rectangle RT3 detects the one farther from the X-ray detector 121 on the pair of opposing constituent surfaces where there is an overlapping region when viewed from the X-ray incident direction of the constituent surfaces of the voxel VX1 shown in FIGS. 9A and 9B. The parallel movement is performed on a plane including the opposing facing surface farther from the vessel 121. The outer peripheries of the rectangles RT1 to RT3 coincide with each other when viewed from the X-ray incident direction.

  If the target voxel is the voxel VX1 shown in FIGS. 9A and 9B, the X-ray attenuation of the entire target voxel VX1 is (1) the X-ray incident on the pixel PX5 of the rectangle RT1 with respect to the luminance value of the pixel PX5 and the area of the rectangle RT1. (2) Multiplication value of the luminance value of the pixel PX6 and the ratio of the area of the portion through which the X-rays incident on the pixel PX6 of the rectangle RT1 are transmitted relative to the area of the rectangle RT1. (3) Multiplying the luminance value of the pixel PX7 by the ratio of the area of the portion through which X-rays incident on the pixel PX7 of the rectangle RT1 are transmitted relative to the area of the rectangle RT1, and (4) the luminance value of the pixel PX9 and the rectangle Multiplying the ratio of the area of the portion through which X-rays incident on the pixel PX9 of the rectangle RT1 with respect to the area of RT1 are transmitted, (5) Pixel PX The product of the luminance value of 0 and the ratio of the area of the portion through which the X-ray incident on the pixel PX10 of the rectangle RT1 is transmitted to the area of the rectangle RT1; (6) the rectangle of the luminance value of the pixel PX11 and the area of the rectangle RT1 This is the sum of the product of the ratio of the area of the portion through which the X-rays incident on the pixel PX11 of RT1 are transmitted. The coordinates of each vertex of the rectangle RT1 can be calculated from the coordinates of each vertex of the target voxel VX1. Further, each area described above can be calculated from the X coordinate and Z coordinate of the rectangle RT1 and the X coordinate and Z coordinate of each lattice point of the pixel formed on the detection surface of the X-ray detector 121.

  As described above, the rectangle RT1 is the midpoint of each side of the target voxel VX1 other than the side included in the pair of opposing configuration surfaces in which there is an overlapping region when viewed from the X-ray incident direction among the configuration surfaces of the target voxel VX1. Is a rectangle with vertices. Thus, the oblique quadratic prism OP1 can be prevented from being biased with respect to the target voxel VX1 in the direction parallel to the detection surface of the X-ray detector 121, and the influence on the image quality of the tomosynthesis tomographic image due to approximation is minimized. Can be However, the setting of the position of the oblique quadrangle that approximates the target voxel VX1 is not limited to the setting of the present embodiment. For example, in the case where the oblique quadratic prism OP1 is not significantly deviated with respect to the target voxel VX1 with respect to the direction parallel to the detection surface of the X-ray detector 121, an oblique quadrangular column that approximates the target voxel VX1 is The oblique prism OP2 shown in FIG. 12, the oblique prism OP3 shown in FIG. 13, the oblique prism OP4 shown in FIG. 14, the oblique square OP5 shown in FIG.

  When the approximation shown in FIG. 12 is performed, if the target voxel is the voxel VX1 shown in FIGS. 9A and 9B, the X-ray attenuation of the entire target voxel VX1 is (1) a rectangle with respect to the luminance value of the pixel PX5 and the area of the rectangle RT4. The product of the ratio of the area of the portion through which the X-ray incident on the RT4 pixel PX5 is transmitted, (2) the luminance value of the pixel PX6 and the X-ray incident on the pixel PX6 of the rectangle RT4 with respect to the area of the rectangle RT4 transmitted (3) the multiplication value of the luminance value of the pixel PX7 and the ratio of the area of the portion through which X-rays incident on the pixel PX7 of the rectangle RT4 are transmitted to the area of the rectangle RT4, (4 ) The luminance value of the pixel PX9 and the ratio of the area of the portion through which X-rays incident on the pixel PX9 of the rectangle RT4 are transmitted to the area of the rectangle RT4 A multiplication value, (5) a multiplication value of a luminance value of the pixel PX10, and a ratio of an area of a portion through which an X-ray incident on the pixel PX10 of the rectangle RT4 is transmitted to an area of the rectangle RT4, and (6) a luminance value of the pixel PX11. The sum of the area of the rectangle RT4 and the ratio of the area of the portion through which X-rays incident on the pixel PX11 of the rectangle RT4 are transmitted. Note that the rectangle RT4 is farther from the X-ray detector 121 on the pair of opposing constituent surfaces where there is an overlapping region when viewed from the X-ray incident direction in the constituent surfaces of the voxel VX1 shown in FIGS. 9A and 9B.

  When the approximation shown in FIG. 13 is performed, if the target voxel is the voxel VX1 shown in FIGS. 9A and 9B, the X-ray attenuation of the entire target voxel VX1 is (1) a rectangle with respect to the luminance value of the pixel PX5 and the area of the rectangle RT5. Multiplying the ratio of the area of the portion through which the X-ray incident on the pixel PX5 of RT5 is transmitted, (2) the luminance value of the pixel PX6, and the X-ray incident on the pixel PX6 of the rectangle RT5 with respect to the area of the rectangle RT5 transmitted (3) the multiplication value of the luminance value of the pixel PX7 and the ratio of the area of the portion through which the X-rays incident on the pixel PX7 of the rectangle RT5 are transmitted to the area of the rectangle RT5, (4 ) The luminance value of the pixel PX9 and the ratio of the area of the portion through which X-rays incident on the pixel PX9 of the rectangle RT5 are transmitted to the area of the rectangle RT5 A multiplication value, (5) a multiplication value of the luminance value of the pixel PX10, and a ratio of an area of a portion through which X-rays incident on the pixel PX10 of the rectangle RT5 are transmitted to an area of the rectangle RT5, and (6) a luminance value of the pixel PX11. , The sum of the ratio of the area of the portion through which X-rays incident on the pixel PX11 of the rectangle RT5 are transmitted to the area of the rectangle RT5. Note that the rectangle RT5 is closer to the X-ray detector 121 on the pair of opposing constituent surfaces where there is an overlapping region when viewed from the X-ray incident direction of the constituent surfaces of the voxel VX1 shown in FIGS. 9A and 9B.

  When the approximation shown in FIG. 14 is performed, if the target voxel is the voxel VX1 shown in FIGS. 9A and 9B, the X-ray attenuation of the entire target voxel VX1 is (1) a rectangle with respect to the luminance value of the pixel PX5 and the area of the rectangle RT6. The product of the ratio of the area of the portion through which the X-ray incident on the pixel PX5 of RT6 is transmitted, (2) the luminance value of the pixel PX6, and the X-ray incident on the pixel PX6 of the rectangle RT6 with respect to the area of the rectangle RT6 transmitted (3) the multiplication value of the luminance value of the pixel PX7 and the ratio of the area of the portion through which X-rays incident on the pixel PX7 of the rectangle RT6 are transmitted to the area of the rectangle RT6, (4 ) The luminance value of the pixel PX9 and the ratio of the area of the portion through which X-rays incident on the pixel PX9 of the rectangle RT6 are transmitted to the area of the rectangle RT6 A multiplication value, (5) a multiplication value of the luminance value of the pixel PX10, and a ratio of the area of the portion through which X-rays incident on the pixel PX10 of the rectangle RT6 are transmitted to the area of the rectangle RT6, and (6) the luminance value of the pixel PX11 , The sum of the area of the rectangle RT6 and the ratio of the area of the portion through which the X-rays incident on the pixel PX11 of the rectangle RT6 are transmitted. Note that the rectangle RT6 has sides of the target voxel VX1 other than the sides included in the pair of opposing constituent surfaces in which overlapping regions are present when viewed from the X-ray incident direction among the constituent surfaces of the voxel VX1 shown in FIGS. 9A and 9B. Is a rectangle whose vertex is a dividing point that is divided 1: 2 from the far side of the X-ray detector 121.

  When the approximation shown in FIG. 15 is performed, if the target voxel is the voxel VX1 shown in FIGS. 9A and 9B, the X-ray attenuation of the entire target voxel VX1 is (1) a rectangle with respect to the luminance value of the pixel PX5 and the area of the rectangle RT7. The product of the ratio of the area of the portion through which the X-ray incident on the pixel PX5 of RT7 is transmitted, (2) the luminance value of the pixel PX6, and the X-ray incident on the pixel PX6 of the rectangle RT7 with respect to the area of the rectangle RT7 transmitted (3) the multiplication value of the luminance value of the pixel PX7 and the ratio of the area of the portion through which X-rays incident on the pixel PX7 of the rectangle RT7 are transmitted to the area of the rectangle RT7, (4 ) The luminance value of the pixel PX9 and the ratio of the area of the portion through which the X-rays incident on the pixel PX9 of the rectangle RT7 are transmitted to the area of the rectangle RT7 A multiplication value, (5) a multiplication value of the luminance value of the pixel PX10, and a ratio of an area of a portion through which X-rays incident on the pixel PX10 of the rectangle RT7 are transmitted to an area of the rectangle RT7, and (6) a luminance value of the pixel PX11 The sum of the area of the rectangle RT7 and the ratio of the area of the portion through which the X-rays incident on the pixel PX11 of the rectangle RT7 are transmitted. In addition, the rectangle RT7 has each side of the target voxel VX1 other than the sides included in the pair of opposing configuration surfaces in which the overlapping regions as seen from the X-ray incident direction exist among the configuration surfaces of the voxel VX1 shown in FIGS. 9A and 9B. Is a rectangle whose vertex is a dividing point that divides 2 by 2: 1 from the far side of the X-ray detector 121.

  In step S5, since the vertically long X-ray slit beam is irradiated from the X-ray generator 112, the X-ray direction irradiated from the X-ray generator 112 expands in the X-axis direction (= the horizontal direction of the X-ray detector 121). Is small, and the spread of the X-rays irradiated from the X-ray generator 112 in the Z-axis direction (= the vertical direction of the X-ray detector 121) is large. Therefore, regarding the X-axis direction (= lateral direction of the X-ray detector 121), the X-ray incident direction is set to the X-ray detector 121 as shown in FIGS. 10 to 15 described above regardless of the position of the target voxel. However, with respect to the Z-axis direction (= longitudinal direction of the X-ray detector 121), the X-ray incident direction is uniformly perpendicular to the detection surface of the X-ray detector 121. As the position of the target voxel is farther from the origin in the Z-axis direction, X-rays that pass through the target voxel enter the detection surface of the X-ray detector 121 obliquely in the Z-axis direction. Here, back projection in the case where X-rays transmitted through the target voxel are obliquely incident on the detection surface of the X-ray detector 121 in the Z-axis direction will be described with reference to FIG. FIG. 16 is a diagram illustrating an example in which X-rays transmitted through the target voxel VX1 are incident on the detection surface of the X-ray detector 121 obliquely in the Z-axis direction. Also in FIG. 16, similarly to FIGS. 10 to 15, a rectangular parallelepiped voxel (for example, the target voxel VX <b> 1 in FIG. 16) has the same volume as the rectangular parallelepiped voxel and has a uniform thickness in the X-ray incident direction. Back projection is performed by approximating a prism (for example, the oblique prism OP6 in FIG. 16). The oblique square column has two bottom surfaces by translating a pair of opposing constituent surfaces in which there are overlapping regions when viewed from the X-ray incident direction on a plane including both opposing constituent surfaces, or only one of them, It has the same volume as a cuboid voxel and a uniform thickness in the X-ray incident direction.

  When the approximation shown in FIG. 16 is performed, if the target voxel is the voxel VX1 shown in FIGS. 9A and 9B, the X-ray attenuation of the entire target voxel VX1 is (1) a rectangle with respect to the luminance value of the pixel PX5 and the area of the rectangle RT1. Multiplying the ratio of the area of the portion through which the X-ray incident on the RT1 pixel PX5 is transmitted, (2) the luminance value of the pixel PX6, and the X-ray incident on the pixel PX6 of the rectangle RT1 with respect to the area of the rectangle RT1 transmitted (3) a multiplication value of the luminance value of the pixel PX7 and the ratio of the area of the portion through which X-rays incident on the pixel PX7 of the rectangle RT1 are transmitted to the area of the rectangle RT1; ) The luminance value of the pixel PX9 and the ratio of the area of the portion through which X-rays incident on the pixel PX9 of the rectangle RT1 are transmitted to the area of the rectangle RT1 A multiplication value, (5) a multiplication value of a luminance value of the pixel PX10, and a ratio of an area of a portion through which X-rays incident on the pixel PX10 of the rectangle RT1 are transmitted to an area of the rectangle RT1, and (6) a luminance value of the pixel PX11. The sum of the ratio of the area of the portion through which X-rays incident on the pixel PX11 of the rectangle RT1 are transmitted to the area of the rectangle RT1. As described above, the rectangle RT1 is formed on each side of the target voxel VX1 other than the side included in the pair of opposing configuration surfaces in which the overlapping area as viewed from the X-ray incident direction exists among the configuration surfaces of the target voxel VX1. A rectangle with the midpoint as the vertex.

  In the above description, the case where none of the constituent surfaces of the target voxel is parallel to the detection surface of the X-ray detector 121 and the incident direction of X-rays transmitted through the target voxel are perpendicular to the detection surface of the X-ray detector 121. It was shown that both can perform the same approximation. However, none of the constituent surfaces of the target voxel is parallel to the detection surface of the X-ray detector 121 and transmits X-rays that pass through the target voxel. A similar approximation can be performed when the incident direction is not perpendicular to the detection surface of the X-ray detector 121.

Although only the voxel VX1 is shown in FIGS. 9A and 9B, the same back projection is performed for all the voxels in the reconstruction area R1. In the calculation of reconstruction, an orthogonal coordinate system defined by the X axis and the Y axis shown in FIG. 7 and the Z axis orthogonal to them may be used, and the radius r, the first deflection angle θ, and the A polar coordinate system defined by two declination angles φ may be used. When the polar coordinate system is used, the distance from the center (X-ray source) of the X-ray generator 112 to one voxel is set as a radius r. Further, the vertical field angle theta ₁ necessary camera from the center of the X-ray generator 112 (X-ray source) from the end of one voxel to the end in is very small, approximating the sin [theta ₁ to theta ₁ Can do. Similarly, the lateral angle phi ₁ necessary camera from the center of the X-ray generator 112 (X-ray source) from the end of one voxel to the end than a minute, to approximate the sin [phi ₁ to phi ₁ be able to.

  Although the shape of the voxel is a rectangular parallelepiped, the rectangular parallelepiped includes a cube that is a special example in which the length, width, and height are all equal. Similarly, each shape of the voxel constituent surface and the cross-section of the voxel parallel to the constituent surface is a rectangle, but the rectangle includes a square which is a special example in which the vertical and horizontal lengths are equal. In addition, when there are two pairs of opposing constituent surfaces in which the overlapping area when viewed from the X-ray incident direction is present on the side among the constituent surfaces of the target voxel, only a pair of opposing constituent surfaces is included in the constituent surfaces of the target voxel It is good to select as a pair of opposing structure surface where the area | region which overlaps seeing from a X-ray incident direction exists. In addition, when there are three pairs of opposing constituent surfaces in which the overlapping regions as seen from the X-ray incident direction exist among the constituent surfaces of the target voxel, only the pair of opposing constituent surfaces are included in the constituent surfaces of the target voxel. It is good to select as a pair of opposing structure surface where the area | region which overlaps seeing from a X-ray incident direction exists.

  When the calculation of back projection is completed for all the voxels in the reconstruction area R1, the reconstruction calculation using the FBP method in step S13 is completed. Thereafter, the image generating means 23 determines whether or not the trajectory data is finished (step S14), and if not finished, returns to step S10 and repeats the above-described operation.

  On the other hand, since the number (n) of transmission of X-rays through each voxel differs depending on the trajectory, the image generation means 23 calculates the number in the calculation process (step S15) and divides the final result by n (step S15). S16).

  As described above, each tomosynthesis tomographic image after reconstruction calculation does not match the horizontal length because the number of voxels in the horizontal direction of the front tooth part is different, and the highest point of each tomographic plane from the X-ray source The vertical lengths do not match because the elevation angles seen are different, so the image generation means 23 uses the tooth position correction table created in advance in step S8 and performs the reconstruction calculation. Tooth position correction is performed to match the horizontal and vertical lengths of each tomosynthesis tomographic image (step S17).

  Specifically, as shown in FIG. 17, it is used as a correction reference when the luminance of a voxel on a certain tomographic plane other than an arbitrary tomographic plane used as a correction reference is projected onto an arbitrary tomographic plane used as a correction reference. Based on the luminance of the voxel on an arbitrary tomographic plane, position correction is performed to match the lengths in the horizontal and vertical directions of the tomosynthesis tomographic images after reconstruction calculation.

  Projection onto an arbitrary tomographic plane used as a correction standard for the brightness of a voxel on a certain tomographic plane other than an arbitrary tomographic plane to be used as a correction reference is normally performed by examining the X-ray transmission path for all pixels in each frame. Although it should be calculated, the width of the X-ray detector 121 (= the horizontal length of the X-ray detector 121) is very narrow (for example, 64 pixels), and the change between adjacent frames is also small. It is assumed that the X-ray path that passes through a specific voxel on the tomographic plane does not substantially change between adjacent frames, and only the X-ray transmission path incident on the center of the X-ray detector 121 is considered for each frame. doing.

  Hereinafter, a voxel having a predetermined position in the height direction (= vertical direction of the X-ray detector 121) on an arbitrary tomographic plane (hereinafter referred to as a reference tomographic plane lay) used as a correction reference and two certain tomographic planes Position correction will be described in detail with reference to FIG. 18 in which voxels whose height direction on j = j1 and j2 is a predetermined position are arranged in a straight line for each tomographic plane for convenience.

  When the brightness of voxel i = b on a certain fault plane j = a other than the reference fault plane lay is projected onto the reference fault plane lay, the projected brightness is on voxel i = c on the reference fault plane lay. In this case, basically, the projected luminance is treated as the luminance of the voxel i = c on a certain tomographic plane j = a after position correction. a, b, and c are arbitrary natural numbers. When the X-rays of one frame pass through a plurality of voxels having the same tomographic plane and height direction, the X-rays of one frame are transferred to the intermediate plane (= viewed from the X-ray incident direction). A rectangle whose vertex is the midpoint of each side of a voxel other than the sides included in the pair of opposing constituent surfaces where there is an overlapping region (see the dotted line in FIG. 18) is only one frame. The voxel through which X-rays are transmitted.

  That is, in a certain frame, one voxel i = b through which X-rays on a certain tomographic plane j = a other than the reference tomographic plane lay pass one-to-one with voxel i = c on the reference tomographic plane lay. When it corresponds (pattern I shown in FIG. 18), the luminance projected on the reference tomographic plane lay is treated as the luminance of the voxel i = c on a certain tomographic plane j = a after position correction. Also, even if the correspondence is not one-to-one, a certain frame is also obtained when there is one X-ray frame passing through one voxel on the reference tomographic plane lay (pattern I ′ shown in FIG. 18). If a single voxel i = b through which X-rays on a certain fault plane j = a other than the reference fault plane lay pass is at voxel i = c on the reference fault plane lay, the projected luminance Is treated as the brightness of voxel i = c on a certain tomographic plane j = a after position correction.

  However, when X-rays of a plurality of frames pass through one voxel on the reference tomographic plane lay (pattern II shown in FIG. 18), no X-rays of any frame pass through the voxels on the reference tomographic plane lay. (Pattern III shown in FIG. 18) requires adjustment of the luminance value.

  The same concept applies to the height direction. However, since the arrangement of voxels is regular in the height direction, the number of X-rays that pass through one voxel on the reference tomographic plane lay in the pattern II shown in FIG. 18 is limited, as shown in FIG. The maximum value of the number of continuously arranged voxels on the reference tomographic plane lay corresponding to the pattern III is limited. In addition, the X-ray spread angle in the height direction should be examined for each frame, but in this embodiment, data of a frame incident in the front tooth direction is adopted in all frames in order to simplify the calculation.

  An example of the dentition position correction in step S17 in accordance with the above-described concept will be described with reference to the flowchart shown in FIG.

  In the example of the dentition position correction shown in FIG. 19, first, the image generation means 23 reads the number of voxels of each tomographic plane in the reconstruction area R1 from the storage unit 22 (step S171).

  Next, the image generation unit 23 reads the data of the luminance value cal (i, j, k) of each voxel after the reconstruction calculation calculated in step S16 from the storage unit 22 (step S172). i is a variable for specifying the order of target voxels in the voxel group having the same fault plane and height direction, j is a variable for specifying the fault plane to which the target voxel belongs, and k is a high This is a variable for specifying the position of the target voxel with respect to the vertical direction.

  Next, the image generation means 23 reads the dentition coordinate of the voxel through which the X-rays incident on the center of the X-ray detector 121 are transmitted from the storage unit 22 for each frame and for each tomographic plane, and the highest point of each tomographic plane. Is read from the storage unit 22 in the height direction ratio of each line segment connecting the X-ray source and the X-ray detector 121 (step S173).

  Next, the image generation means 23 uses the reading result of step S173, and the n (i) th X-ray of N (i) frames that pass through the voxel of the dentition coordinate i on the reference tomographic plane lay. Is calculated for each frame and each tomographic plane (step S174). The dentition coordinate ii (i, j, n (i)) of a voxel of a certain tomographic plane j is transmitted (step S174).

  Next, the image generation means 23 uses the reading result of step S173, and the h (k) th X-ray of the H (k) frames that pass through the voxel of the dentition coordinate k on the reference tomographic plane lay. The dentition coordinate kk (k, j, h (k)) of a voxel of a certain tomographic plane j through which is transmitted is calculated for each frame and each tomographic plane (step S175).

  Next, the image generating means 23 projects the voxel luminance value of the tomographic plane j onto the reference tomographic lay in a loop of the voxel dentition coordinate i, dentition coordinates k, n (i), and h (k). The brightness value datawa (i, j, k) is calculated from the brightness value cal (ii, j, kk) of the point (ii, j, kk) on the tomographic plane j (step S176).

When N (i) ≠ 0 and H (k) ≠ 0, that is, when both i and k correspond to pattern I or pattern II, the luminance value datawa (i, j, k) is expressed by the following equation (1). Is done.

When N (i) ≠ 0 and H (k) = 0, that is, i corresponds to pattern I or pattern II, and k corresponds to pattern III, the luminance value datawa (i, j, k) is It is represented by the formula (2).

When N (i) = 0 and H (k) ≠ 0, that is, i corresponds to pattern III, and k corresponds to pattern I or pattern II, luminance value datawa (i, j, k) is It is represented by the formula (3).

When N (i) = 0 and H (k) = 0, that is, when both i and k correspond to the pattern III, the luminance value datawa (i, j, k) is expressed by the following equation (4).

  However, i1 and i2 in the above equations (3) and (4) are the same when there is no frame corresponding to the dentition coordinate i of the voxel on the reference tomographic plane lay (n (i) = 0). It is the dentition coordinate ii on the tomographic plane j corresponding to i satisfying n (i)> 0 of the preceding and following frames, and satisfies i1 ≦ i2. Similarly, k1 and k2 in the above equations (2) and (4) are considered in the height direction. In addition, when at least one of the coordinate positions indicated by i1, i2, k1, and k2 is out of the region forming the tomosynthesis tomographic image, the luminance value datawa (i, j, k) is set to zero.

  Considering the case of i1 + 1 <i2 here, in the case of pattern III, the X-rays of the frame are transmitted on the fault plane (for example, the fault plane j = 2 shown in FIG. 18) having more voxels than the reference fault lay. There will be no voxels (eg, the voxel shown in FIG. 18), but it is not desirable not to use the luminance value of this voxel. Therefore, in this case, it is desirable to perform the following processing.

i1 = ii (i−Δ1, j, N (i−Δ1)), i2 = ii (i + Δ2, j, N (i + Δ1)) Is a real number i _j represented by the following equation (5). Here, Δ1 + Δ2-1 is the number of consecutive voxels through which the X-rays of the frame on the reference tomographic plane lay are not transmitted.
The same applies to k1 and k2 in the height direction. When i1 + 1 <i2 and / or k1 + 1 <k2, the following formula (6) is used instead of the above formula (2), the following formula (7) is used instead of the above formula (3), and the above (4 The following formula (8) is used instead of the formula. However, flr (x) in the following formulas (6) to (8) represents the maximum integer among the integers equal to or less than the real number x.

  When the calculation of the luminance value datawa (i, j, k) described above is completed, the position correction process (an example of the process of step S17) shown in FIG. 19 is completed.

  The 60-layer tomosynthesis tomographic image obtained by completing the processing in step S17 is stored in the image memory 22b as a 60-layer panoramic image. Finally, the control device 21 displays the reconstructed panoramic image on the display unit 26 in accordance with, for example, a user instruction given from the operation unit 25 or a display setting stored in advance in the ROM 21b (step S18).

  As described above, in the present embodiment, in the reconstruction calculation using the FBP method, instead of performing the back projection onto the voxels of the rectangular parallelepiped that forms the reconstruction area, the configuration surface of the rectangular parallelepiped that constitutes the reconstruction area The back projection is performed on one of a pair of opposing constituent surfaces where there is an overlapping region when viewed from the X-ray incident direction, or on the cross section of the voxel parallel to the opposing constituent surfaces. It can be greatly shortened. Therefore, a tomosynthesis tomographic image can be reconstructed with a small amount of reconstruction calculation using the FBP method.

  Furthermore, in this embodiment, each tomosynthesis tomographic image after reconstruction calculation is not enlarged or reduced uniformly in the horizontal direction or the vertical direction, and position correction is performed. The position correction is performed based on the luminance of the voxel on an arbitrary tomographic plane used as a correction reference when projected on an arbitrary tomographic plane using the luminance of the voxel on the tomographic plane as a correction reference. As a result, it is possible to correct the position of each tomosynthesis tomographic image after reconstruction calculation using the FBP method while suppressing the occurrence of distortion.

  The X-ray imaging apparatus according to the present invention is not limited to a dental X-ray imaging apparatus but can be applied to all X-ray imaging apparatuses that reconstruct a tomosynthesis tomographic image from frame data using the FBP method. For example, when a rib is taken as an imaging target, a standard rib coordinate group corresponding to a standard rib shape is used instead of the standard dentition coordinate group.

  In the above-described embodiment, the horizontal and vertical lengths of each tomosynthesis tomographic image after reconstruction calculation are matched by position correction. However, the horizontal length of each tomosynthesis tomographic image after reconstruction calculation is used. If it is not a big problem, the length of each tomosynthesis tomographic image after reconstruction calculation may be matched by position correction, and the length of each tomosynthesis tomographic image after reconstruction calculation When the difference in the lengths of the images does not cause much problem, the lengths of only the horizontal directions of the respective tomosynthesis tomographic images after the reconstruction calculation may be matched by position correction.

  In the above-described embodiment, the horizontal and vertical lengths of each tomosynthesis tomographic image after reconstruction calculation are made to coincide with each other by position correction. However, the horizontal direction of each tomosynthesis tomographic image after reconstruction calculation by position correction is used. Further, the lengths in the vertical direction may be substantially matched. That is, for example, when comparing tomosynthesis tomographic images of multiple layers and comparing the appearance of teeth according to the tomographic depth, the horizontal direction of each tomosynthesis tomographic image after reconstruction calculation after position correction There may also be a difference in length in the longitudinal direction.

  Further, in the above-described embodiment, in the reconstruction calculation using the FBP method, instead of performing the back projection onto the voxels of the rectangular parallelepiped constituting the reconstruction area, out of the constituent surfaces of the rectangular parallelepiped voxels constituting the reconstruction area Back projection was performed on one of a pair of opposing constituent surfaces in which an overlapping region as seen from the X-ray incident direction exists, or on the cross section of the voxel parallel to the opposing constituent surfaces, but the present invention is not limited to this. In the reconstruction calculation using the FBP method, back projection onto a rectangular parallelepiped voxel constituting the reconstruction area may be performed.

DESCRIPTION OF SYMBOLS 1 X-ray imaging apparatus main body 2 X-ray image display apparatus 3 Turning arm 110 X-ray irradiation part 120 X-ray detection part 20 Image display apparatus main body 21 Control apparatus 22 Memory | storage part 25 Operation means 26 Display means

Claims (4)

An X-ray irradiation unit for irradiating the subject with X-rays;
An X-ray detector that outputs a digital amount of an electrical signal corresponding to incident X-rays at a constant frame rate;
Swiveling means for moving the pair of the X-ray irradiation unit and the X-ray detection unit around the subject in a state of facing each other across the subject;
Storage means for sequentially storing, as frame data, signals output by the X-ray detection unit as the turning unit moves the X-ray irradiation unit and the X-ray detection unit around the subject;
Image processing means for reconstructing a tomosynthesis tomographic image from the frame data using an FBP (Filtered Back Projection) method,
The image processing means includes
Based on the brightness of a voxel on an arbitrary tomographic plane used as a correction reference when the luminance of a voxel on a certain tomographic plane other than an arbitrary tomographic plane used as a correction reference is projected on an arbitrary tomographic plane used as a correction reference An X-ray imaging apparatus characterized by performing position correction so that at least one of the lengths in the horizontal direction and the vertical direction of each tomosynthesis tomographic image after the reconstruction calculation is matched or substantially matched. The image processing means includes
N (transparent through voxels of coordinates i on any tomographic plane used as a correction criterion (where i is a variable for specifying the order of target voxels in a group of voxels whose height direction is the same as the tomographic plane) i) Voxel coordinates ii (i, j, n (i)) of a certain tomographic plane j through which the n (i) th X-ray of the frames is transmitted are calculated for each frame and each tomographic plane. ,
H (k) out of H (k) frames that pass through voxels at coordinates k on the arbitrary tomographic plane used as a correction criterion (where k is a variable for specifying the position of the target voxel in the height direction) The coordinates kk (k, j, h (k)) of the voxel of a certain fault plane j through which the X-ray passes are calculated for each frame and each fault plane;
Luminance value datawa (i,) when projected on an arbitrary tomographic plane using the voxel luminance value of a certain tomographic plane j as a correction reference in a loop of voxel coordinates i, coordinates k, n (i), h (k) j, k) according to the following formula (1) when N (i) ≠ 0 and H (k) ≠ 0, and when N (i) ≠ 0 and H (k) = 0, the following (2 ), When N (i) = 0 and H (k) ≠ 0, the following equation (3) is satisfied. When N (i) = 0 and H (k) = 0, the following equation (4) is satisfied. The X-ray imaging apparatus according to claim 1, wherein the X-ray imaging apparatus calculates the brightness value cal (ii, j, kk) of the point (ii, j, kk) on the tomographic plane j according to
However, i1 and i2 in the above formulas (3) and (4) are when there is no frame corresponding to the coordinate i of the voxel on an arbitrary tomographic plane used as a correction reference (n (i) = 0). , The coordinates ii on the tomographic plane j corresponding to i satisfying n (i)> 0 in the preceding and succeeding frames, and satisfying i1 ≦ i2. In the above equations (2) and (4), k1 and k2 are obtained when there is no frame corresponding to the coordinate k of a voxel on an arbitrary tomographic plane used as a correction reference (h (k) = 0). The coordinates kk on the tomographic plane j corresponding to k satisfying h (k)> 0 in the preceding and following frames, and satisfying k1 ≦ k2. The image processing means includes
If i1 + 1 <i2 and / or k1 + 1 <k2,
i1 = ii (i−Δ1, j, N (i−Δ1)), i2 = ii (i + Δ2, j, N (i + Δ1)), and a tomographic plane corresponding to a coordinate i on an arbitrary tomographic plane used as a correction reference The coordinate on j is a real number i _j represented by the following equation (5), and
The following formula (6) is used instead of the above formula (2), the following formula (7) is used instead of the above formula (3), and the following formula (8) is used instead of the above formula (4). X-ray imaging apparatus described in 1.
However, Δ1 + Δ2-1 is the number of continuous voxels through which X-rays of a frame on an arbitrary tomographic plane used as a correction reference are not transmitted, and flr (x) in the above equations (6) to (8) is a real number x It represents the largest integer among the following integers. An image processing method for correcting the position of each tomosynthesis tomographic image after reconstruction calculation using an FBP (Filtered Back Projection) method,
Based on the brightness of a voxel on an arbitrary tomographic plane used as a correction reference when the luminance of a voxel on a certain tomographic plane other than an arbitrary tomographic plane used as a correction reference is projected on an arbitrary tomographic plane used as a correction reference An image processing method characterized by performing position correction for making at least one of the lengths in the horizontal direction and the vertical direction of the tomosynthesis tomographic images after reconstruction calculation match or substantially match.