JP2017000664A - Image processing system, tomographic image generation system, and program - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve representation performance of a peripheral region of a high absorber in a reconstructed image resulting from projection images obtained by tomosynthesis imaging or CT imaging.SOLUTION: A controller 91 of a console 90 performs pattern matching between a reconstructed image resulting from reconstruction of only high absorber region projection images provisionally extracted from projection images and shape models stored in a high absorber shape DB 953 to select a shape model of a high absorber implanted in a subject, specifies the shapes and positions of the high absorber regions in the projection images and the reconstructed image resulting from reconstruction of the projection images on the basis of the selected shape model of the high absorber and the reconstructed image resulting from reconstruction of the high absorber region images provisionally extracted, interpolates pixels in the high absorber regions specified in the projection images, generates a reconstructed image resulting from reconstruction of the interpolated projection images, and synthesizes the resulting reconstructed image with only the reconstructed high absorber region image, thereby generating a reconstructed image of the subject.SELECTED DRAWING: Figure 4

Description

  The present invention relates to an image processing apparatus, a tomographic image generation system, and a program.

  Conventionally, in the medical field, a technique is known in which a subject is radiographed by tomosynthesis imaging or CT (Computed Tomography) imaging, and the obtained projection image is reconstructed to generate a reconstructed image (tomographic image) of the subject. Yes.

  However, if the subject includes a high-absorber (hereinafter referred to as a high-absorber) having a very high absorption coefficient (mass absorption coefficient) of radiation such as metal, streak streak artifacts on the reconstructed image, Due to the beam hardening, an artifact is generated in which the periphery of the high-absorber is crushed in black.

Therefore, for example, in Patent Document 1, in a CT apparatus, a sinogram generated by imaging is binarized, and the binarized sinogram is simply backprojected into a two-dimensional space by calculation for each rotation angle. Is obtained, and a metal region is specified based on a sinogram obtained by projecting the back projection image by calculation. Then, the metal region in the sinogram is interpolated, and the interpolated sinogram is reconstructed by calculation to generate a reconstructed image, and the generated reconstructed image and the backprojected image are combined to form a high absorber. It is described to reduce the resulting artifacts.
Similarly, a projection image obtained by tomosynthesis imaging is subjected to threshold processing to extract a high-absorber region of radiation such as a metal region, an image obtained by reconstructing the extracted projection image of the high-absorber region, and a projection image Artifacts resulting from the high absorber are reduced by synthesizing an image obtained by reconstructing an image obtained by interpolating the high absorber region portion with surrounding pixels.

JP 2010-99114 A

However, since the thin portion of the edge portion of the superabsorber transmits X-rays to some extent, it is difficult to accurately extract the superabsorber region by binarization processing (threshold processing).
FIG. 8A shows the edge of the superabsorbent region in the projection image. FIG. 8B shows a profile before interpolation of the superabsorbent region in the direction indicated by the arrow X in FIG. FIG. 8C shows a profile after interpolation of the superabsorbent region in the direction indicated by the arrow X in FIG. As shown in FIG. 8B, the signal values of the superabsorber region and the region outside the superabsorber gradually change in the projection image. Therefore, if the superabsorber region is interpolated by dividing the superabsorber region and the superabsorber region by a predetermined threshold, in the projection image after interpolation of the superabsorber region, as shown in FIG. The edge of the superabsorbent that has passed through the radiation remains without being interpolated. As a result, in the projection image obtained by interpolating the high-absorber region, an edge is generated at the edge of the high-absorber as shown by arrows in FIG. Similarly, as shown by an arrow in FIG. 10 (b) on an image obtained by synthesizing a reconstructed image obtained by reconstructing a projection image obtained by interpolating the superabsorber region and a reconstructed image including only the superabsorber region. (FIG. 10 (b) is an enlarged view of the partial region β of FIG. 10 (a)), an edge is generated at the edge of the high absorber region, resulting in an artifact, and the edge of the high absorber cannot be drawn correctly.

  An object of the present invention is to improve the rendering performance of a high-absorber edge in a reconstructed image obtained by reconstructing a projection image obtained by tomosynthesis imaging or CT imaging.

In order to solve the above-mentioned problem, the invention described in claim 1
A reconstructed image of the subject is generated using a plurality of projection images obtained by arranging a subject between the radiation source and the radiation detector and changing a positional relationship between the radiation source and the radiation detector. An image processing apparatus for performing
Temporary extraction means for temporarily extracting a high-absorber region of radiation from the projection image;
A selecting means for selecting a shape model of the superabsorber embedded in the subject from a database storing a shape model of the superabsorber of a plurality of types of radiation;
High absorption in the reconstructed image obtained by reconstructing the projection image and the projection image based on the reconstructed image obtained by reconstructing the shape model of the selected superabsorber and the temporarily extracted superabsorber region image Identifying means for identifying the shape and position of the body region;
Interpolation means for interpolating pixels of the superabsorbent region specified by the specifying means in the projection image;
First reconstructed image generation means for reconstructing a projection image after interpolation by the interpolation means to generate a reconstructed image;
Second reconstructed image generating means for generating a reconstructed image of only the superabsorbent region based on the shape and position of the superabsorbent region in the reconstructed image specified by the specifying means;
By synthesizing the reconstructed image generated by the first reconstructed image generating means and the reconstructed image generated by the second reconstructed image generating means, artifacts due to the superabsorber are reduced. Third reconstructed image generating means for generating a reconstructed image of the subject;
Is provided.

The invention according to claim 2 is the invention according to claim 1,
The selecting means performs pattern matching between each of the shape models stored in the database and a reconstructed image obtained by reconstructing the projection image of only the temporarily extracted superabsorbent region, and the obtained similarity is obtained. Based on the database, a shape model of the superabsorber embedded in the subject is selected from the database.

The invention according to claim 3 is the invention according to claim 1,
The selection means performs pattern matching between each of the projection images obtained by orthographic projection of each shape model stored in the database and the projection image of only the temporarily extracted superabsorbent region, and the obtained similarity is obtained. Based on the database, a shape model of the superabsorber embedded in the subject is selected from the database.

The invention according to claim 4 is the invention according to claim 1,
The selecting means performs pattern matching between each shape model stored in the database and a reconstructed image obtained by reconstructing a projection image of only the temporarily extracted superabsorbent region. The sum of the obtained first similarity and the second similarity obtained by performing pattern matching between the projection image obtained by orthographic projection of the shape model and the projection image of only the temporarily extracted superabsorbent region Is calculated as the similarity, and based on the similarity calculated for each shape model, the shape model of the superabsorbent body embedded in the subject is selected from the database.

The invention according to claim 5 is the invention according to any one of claims 2 to 4,
It is determined whether or not the similarity with the shape model selected by the selection means is equal to or higher than a predetermined reference value. If the similarity is equal to or higher than a predetermined reference value, the interpolation means and the first reconstructed image generation Control means for generating a reconstructed image of the subject in which artifacts due to the high-absorbent body are reduced by performing processing of the means, the second reconstructed image generating means, and the third reconstructed image generating means Is provided.

The invention according to claim 6 is the invention according to any one of claims 1 to 5,
The database stores each shape model of the superabsorbent body in association with the model number of the superabsorbent body,
Designation means for the user to designate the model number or shape of the superabsorbent body embedded in the subject;
When selecting the shape model of the superabsorbent embedded in the subject from the database by the selecting means, whether to select by pattern matching or based on the model number or shape specified by the specifying means A selection means for the user to select,
Is provided.

The invention according to claim 7 is the invention according to claim 1,
The database stores each shape model of the superabsorbent body in association with the model number of the superabsorbent body,
Comprising designation means for the user to designate the model number or shape of the superabsorbent body embedded in the subject;
The selection means selects a shape model of the superabsorbent body embedded in the subject from the database based on the model number or shape designated by the designation means.

The invention according to claim 8 is the invention according to any one of claims 1 to 7,
Second selection means is provided for allowing a user to select whether or not to generate a reconstructed image of the subject in which artifacts due to the high absorber are reduced.

The invention according to claim 9 is the invention according to any one of claims 1 to 8,
Storage means for storing the patient information of the patient who is the subject and the identification information in the database of the selected shape model in association with each other;

The invention according to claim 10 is the invention according to any one of claims 1 to 9,
Display means for displaying the model number and shape of the shape model selected by the selecting means and / or the position of the superabsorbent region specified by the specifying means.

The invention according to claim 11 is the invention according to any one of claims 1 to 10,
The database is associated with the shape models of the plurality of types of superabsorbers, the material of the superabsorbers, the direction, size and / or volume embedded in the subject, the imaging region where the shape model is used, and / or At least one of the shooting directions is stored.

The invention according to claim 12
A radiation source that irradiates a subject with radiation, and a radiation detector that two-dimensionally arranges radiation detection elements that detect radiation and generate electrical signals, and obtain a projection image corresponding to the irradiated radiation. Imaging means for acquiring the projection image of the subject arranged between the radiation source and the radiation detector a predetermined number of times while changing the positional relationship between the radiation source and the radiation detector;
Image processing means for generating a tomographic image that is a reconstructed image of the subject using the projection image acquired by the imaging means;
A tomographic image generation system comprising:
The image processing means includes
Temporary extraction means for temporarily extracting a high-absorber region of radiation from the projection image;
A selecting means for selecting a shape model of the superabsorber embedded in the subject from a database storing a shape model of the superabsorber of a plurality of types of radiation;
High absorption in the reconstructed image obtained by reconstructing the projection image and the projection image based on the reconstructed image obtained by reconstructing the shape model of the selected superabsorber and the temporarily extracted superabsorber region image Identifying means for identifying the shape and position of the body region;
Interpolation means for interpolating pixels of the superabsorbent region specified by the specifying means in the projection image;
First reconstructed image generation means for reconstructing a projection image after interpolation by the interpolation means to generate a reconstructed image;
Second reconstructed image generating means for generating a reconstructed image of only the superabsorbent region based on the shape and position of the superabsorbent region in the reconstructed image specified by the specifying means;
By synthesizing the reconstructed image generated by the first reconstructed image generating means and the reconstructed image generated by the second reconstructed image generating means, artifacts due to the superabsorber are reduced. Third reconstructed image generating means for generating a reconstructed image of the subject;
Is provided.

The program of the invention described in claim 13 is:
A reconstructed image of the subject is generated using a plurality of projection images obtained by arranging a subject between the radiation source and the radiation detector and changing a positional relationship between the radiation source and the radiation detector. A computer used in the image processing apparatus
Temporary extraction means for temporarily extracting a high-absorber region of radiation from the projected image;
A selecting means for selecting a shape model of the superabsorber embedded in the subject from a database storing a shape model of the superabsorber of a plurality of types of radiation;
High absorption in the reconstructed image obtained by reconstructing the projection image and the projection image based on the reconstructed image obtained by reconstructing the shape model of the selected superabsorber and the temporarily extracted superabsorber region image A specifying means for specifying the shape and position of the body region;
Interpolation means for interpolating pixels of the superabsorbent region specified by the specifying means in the projection image;
First reconstructed image generation means for reconstructing a projection image after interpolation by the interpolation means to generate a reconstructed image;
Second reconstructed image generating means for generating a reconstructed image of only the superabsorbent region based on the shape and position of the superabsorbent region in the reconstructed image specified by the specifying means;
By synthesizing the reconstructed image generated by the first reconstructed image generating means and the reconstructed image generated by the second reconstructed image generating means, artifacts due to the superabsorber are reduced. A third reconstructed image generating means for generating a reconstructed image of the subject;
To function as.

  ADVANTAGE OF THE INVENTION According to this invention, the rendering performance of the super absorbent body edge part in the reconstructed image which reconfigure | reconstructed the projection image obtained by tomosynthesis imaging | photography or CT imaging | photography can be improved.

It is a figure showing the whole tomographic image generation system composition concerning this embodiment. It is a figure for demonstrating tomosynthesis imaging | photography. It is a block diagram which shows the functional structure of the console of FIG. It is a flowchart which shows the to-be-reconstructed image production | generation process A performed by the control part of FIG. It is a figure which shows an example of the projection image by which the high absorber area | region produced | generated by this embodiment was interpolated. (A) is a figure which shows an example of the reconstruction image produced | generated by this embodiment, (b) is an enlarged view of the partial area | region (alpha) of (a). It is a flowchart which shows the to-be-reconstructed image generation process B performed by the control part of FIG. (A) is a figure which shows the edge part of the superabsorbent body area | region in a projection image. (B) is a profile before interpolation of the superabsorbent region in the direction indicated by the arrow X in (a). (C) is the profile after interpolation of the superabsorbent region in the direction indicated by the arrow X in FIG. It is a figure which shows an example of the projection image by which the high absorber area | region produced | generated by the conventional method was interpolated. The figure which shows an example of the reconstruction image produced | generated by the conventional method, (b) is an enlarged view of the partial area | region (beta) of (a).

  Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. The present invention is not limited to the illustrated example.

<First Embodiment>
[Configuration of the tomographic image generation system 100]
First, a schematic configuration of the tomographic image generation system 100 according to the present invention will be described. The tomographic image generation system 100 is a system that generates a reconstructed image (tomographic image) of the subject H using a projection image obtained by tomosynthesis imaging of the subject H (part of the human body). FIG. 1 shows a schematic configuration of a tomographic image generation system 100 according to the present embodiment. As shown in FIG. 1, the tomographic image generation system 100 mainly includes a radiation imaging apparatus 1 and a console 90.
In the following description, the longitudinal direction of the subject table 54 (the body axis direction of the subject H arranged on the subject table 54) is orthogonal to the y-axis direction, and the imaging surface (surface irradiated with radiation) is orthogonal to the y-axis direction. The description will be made assuming that the direction is the x-axis direction and the radiation irradiation direction (thickness direction of the subject H) is the z-axis direction.

  The tomographic image generation system 100 is provided inside and outside an imaging room 101a and a front room (also referred to as an operation room) 101b. In the imaging room 101a, an imaging table 50 of the radiation imaging apparatus 1, a radiation source 61, and the like are provided. In the radiographing room 101a, an access point AP for relaying wireless communication between the radiation detector F and a console 90 described later is also provided.

  The front chamber 101b is provided with a console 62 of the radiation irradiation device 60, an exposure switch 63, and the like. Further, FIG. 1 shows a case where the control BOX 80, the console 90, and the like are provided outside the front chamber 101b, but they can also be provided inside the front chamber 101b.

  As shown in FIG. 1, the radiation imaging apparatus 1 as an imaging unit includes a radiation detector F, an imaging table 50 that holds the radiation detector F and the subject H, and a radiation irradiation apparatus 60. Yes. In addition, in FIG. 1, the figure which looked at the radiography apparatus 1 which image | photographs the to-be-photographed object H in the supine position as an example is shown.

The radiation detector F is configured by a semiconductor image sensor such as an FPD (Flat Panel Detector). The FPD has, for example, a glass substrate, and detects radiation (X-rays) emitted from the radiation source 61 and transmitted through at least the subject H at a predetermined position on the substrate according to its intensity. A plurality of detection elements (pixels) that convert radiation into electrical signals and store them are arranged in a matrix. Each pixel is configured to include a switching unit such as a TFT (Thin Film Transistor), for example. The electrical signal stored in each pixel is switched by the switching unit and stored in the radiation detector F. The projection image of the subject H is obtained by reading the electrical signal. The FPD includes an indirect conversion type in which radiation is converted into an electric signal by a photoelectric conversion element via a scintillator, and a direct conversion type in which radiation is directly converted into an electric signal, and either may be used.
The radiation detector F has a function of communicating with the console 90 via the network N1 and the control BOX 80, and a wireless communication function of communicating with the console 90 via the access point AP.

The imaging table 50 includes a detector loading unit 51, a loading unit support unit 52, a transport device 53, a subject table 54, and the like.
The detector loading unit 51 holds the radiation detector F.
The loading unit support unit 52 is movable in the longitudinal direction of the subject table 54 (the body axis direction of the subject H, the y-axis direction) on the side of the subject table 54 opposite to the surface on which the subject H is placed. It is provided and supports the detector loading unit 51.
Although not shown, the transport device 53 includes, for example, a drive motor, and transmits the rotational force of the drive motor to the loading unit support unit 52 with a rack and pinion, so that the loading unit support unit 52 is moved in the longitudinal direction of the subject table 54. Move in (y-axis direction). The transport device 53 can adopt any configuration, mechanism, or the like as long as it can move the loading unit support unit 52 in the longitudinal direction of the subject table 54. It is not limited to the structure using a pinion. For example, a linear movement of an actuator or the like can be transmitted to the loading unit support unit 52 to move the loading unit support unit 52.
The subject table 54 is a table that supports the subject H provided in the radiation irradiation direction of the radiation source 61, and is a resin plate such as an acrylic plate, a plate made of an inorganic material such as a carbon plate, a metal plate, or the like. It is configured. The subject table 54 is provided with a guide (not shown) for moving the loading unit support 52 along the longitudinal direction (y-axis direction) of the subject table 54.

The radiation irradiation device 60 includes a radiation source 61 that irradiates the radiation detector F through the subject H, and an operator console that allows a radiographer or other photographer to set imaging conditions such as tube current, tube voltage, and irradiation time. 62, an exposure switch 63 that is operated by the photographer to instruct irradiation of radiation from the radiation source 61, and the radiation source 61 is moved along the body axis direction of the subject H on the subject table 54 (in the y-axis direction). And a radiation source moving mechanism 64 that tilts the irradiation angle of the radiation source 61 according to the position so that the radiation detector F is irradiated with the radiation irradiated from the radiation source 61 at the moved position. . When the imaging condition is set from the console 90 via the control BOX 80 or the console 62 and the exposure switch 63 is pressed, the radiation irradiation device 60 transmits a pressing signal of the exposure switch 63 to the console 90. Based on the control signal from the console 90, the radiation source 61 is irradiated with radiation while the radiation source moving mechanism 64 moves the radiation source 61 under the set imaging conditions.
Further, a collimator 75 is provided in the radiation direction of the radiation source 61 to limit an irradiation area of the radiation emitted from the radiation source 61.

  In the present embodiment, as the radiation source 61 of the radiation irradiation apparatus 60, a radiation source that irradiates radiation toward the subject H or the radiation detector F in a cone shape, that is, a radiation source that irradiates a so-called cone beam is used. The radiation source 61 may be configured to use a radiation source that irradiates radiation that spreads in a substantially planar shape (that is, a so-called fan beam) like a fan. In the case of using a radiation source that irradiates a fan beam, radiation is emitted from the radiation source so that the fan beam expands in the certain direction.

  The radiation source moving mechanism 64 and the transport device 53 synchronize with each other in accordance with a control signal transmitted from the console 90 via a control box 80 described later, and rotate the radiation source 61 and the loading unit support unit 52 around the rotation center O ( 2), the radiation source 61 and the radiation detector F are moved in opposite directions as shown in FIG. 2 by moving in the opposite directions along the subject table 54 (that is, in the y-axis direction). Move to.

  The radiation imaging apparatus 1 configured as described above is configured tomosynthesis imaging a predetermined number of times (a plurality of times) while the radiation source 61 and the radiation detector F move in the opposite direction from the predetermined imaging start position to the end position in synchronization. And the projection image is acquired by the radiation detector F for every photographing. At this time, the optical axis of the radiation source 61 is configured to irradiate the center of the radiation detector F.

  At that time, for example, it is possible to continuously irradiate the radiation from the radiation source 61 without interruption, and the radiation detector F may perform a predetermined number of times of projection image acquisition processing during that time. Alternatively, the radiation source 61 may be irradiated with a predetermined number of times (pulse irradiation), and the projection image may be acquired by the radiation detector F each time the radiation is irradiated.

  The radiation detector F may be configured to transmit the acquired projection image to the console 90 as the image processing apparatus via the control BOX 80 every time the projection image is acquired. It is also possible to store the image once in a storage unit (not shown) and transmit the projection images to the console 90 together when a predetermined number of times the projection image acquisition process is completed.

  A control BOX (also referred to as a repeater) 80 is connected to each unit of the radiation imaging apparatus 1, the radiation detector F loaded in the detector loading unit 51, the console 90, and the like via a network N1. In the control BOX 80, a LAN (Local Area Network) communication signal transmitted from the console 90 or the like to the radiation irradiation device 60 is converted into a signal for the radiation irradiation device 60, or vice versa. Does not have a built-in converter.

  As shown in FIG. 3, the console 90 includes a control unit 91, an operation unit 92, a display unit 93, a communication unit 94, and a storage unit 95, and is a computer device configured by connecting each unit via a bus 96. is there.

  The control unit 91 is configured by a CPU, a RAM, and the like. The CPU of the control unit 91 reads out various programs such as a system program and a processing program stored in the storage unit 95 and expands them in the RAM, and starts subject reconstructed image generation processing A described later according to the expanded programs. Perform various processes. The control unit 91 cooperates with the program stored in the storage unit 95 to temporarily extract means, selection means, identification means, interpolation means, first reconstructed image generation means, and second reconstructed image generation means. , Function as third reconstructed image generating means, control means, storage control means.

  The operation unit 92 includes a keyboard having character input keys, numeric input keys, various function keys, and the like, and a pointing device such as a mouse, and a key pressing signal pressed by the keyboard and an operation signal by the mouse. Are output to the control unit 91 as an input signal.

  The display unit 93 includes, for example, a monitor such as a CRT (Cathode Ray Tube) or an LCD (Liquid Crystal Display), and displays various screens according to instructions of a display signal input from the control unit 91.

  The communication unit 94 is configured by a LAN card or the like, and transmits and receives data to and from external devices connected to the networks N1 and N2 via a switching hub.

  The storage unit 95 includes, for example, an HDD (Hard Disk Drive), a semiconductor nonvolatile memory, or the like. The storage unit 95 stores the system program and various processing programs as described above.

The storage unit 95 is provided with a projection image storage unit 951 that stores the projection image received from the radiation detector F, a reconstructed image storage unit 952 that stores the generated reconstructed image, and the like.
Further, the storage unit 95 stores a high-absorber shape DB (Data Base) 953, received patient imaging order information (patient information, imaging region, imaging direction, etc.), and the like. The high-absorber shape DB 953 is a three-dimensional shape model (hereinafter referred to as a shape) of a high-absorber body whose shape is known in advance, for example, a metal plate, a bolt, a pacemaker, an artificial joint, or a cochlear implant that is embedded in the human body. Model) data with identification information stored.

  The console 90 transmits a wake-up signal to the radiation detector F, for example, via the access point AP or the control BOX 80 by the communication unit 94 to change the radiation detector F from a sleep state to a wake-up state. Then, the radiation detector F is controlled, the tube current or the like set by the radiographer or the like through the operation unit 92 is transmitted to the radiation irradiating apparatus 60 via the control BOX 80, or set via the control BOX 80. Thus, the transport device 53 and the radiation source moving mechanism 64 can be controlled.

  In the present embodiment, the console 90 also functions as an image processing device. When the projection image acquired by the radiation detector F is transmitted from the radiation imaging device 1, the received projection image is converted into the received projection image. Based on this, a reconstructed image of the subject H (a two-dimensional tomographic image of a cross section indicated by a one-dot chain line in FIG. 1) is generated. Note that the image processing apparatus can be configured as a separate apparatus from the console 90.

  Further, as shown in FIG. 1, an access point AP is connected to the console 90 via a network N2. In addition, the console 90 is connected via a network N2 to a non-illustrated HIS (Hospital Information System), RIS (Radiology Information System), PACS (Picture Archiving and Communication System). Etc. are connected. The console 90 is configured to perform various processes such as acquisition of imaging order information from HIS, RIS, etc., and transmission of the generated reconstructed image to the PACS.

  Note that it is not necessary to configure the network connecting the devices by a plurality of networks N1 and N2 as in the present embodiment, and the tomographic image generation system 100 is configured by connecting each device to one network. Is also possible. Further, when a plurality of networks are used as a network connecting the devices as in the present embodiment, which device is connected to which network can be changed as appropriate.

[Operation of the tomographic image generation system 100]
Next, the operation of the tomographic image generation system 100 in the present embodiment will be described.
In the tomographic image generation system 100, the control unit 91 of the console 90 executes subject reconstructed image generation processing A described below, thereby controlling each unit of the radiation imaging apparatus 1 to control the radiation source 61 and the radiation detector F. Is taken a predetermined number of times while moving the, and a reconstructed image of the subject H is generated based on the obtained series of projection images.

  FIG. 4 shows a flowchart of the subject reconstructed image generation process A executed by the control unit 91 of the console 90. The subject reconstructed image generation process A is executed in cooperation with the control unit 91 and a program stored in the storage unit 95. 4 and 7, the rounded rectangle indicates an image generated by the upper processing step.

  First, the controller 91 performs tomosynthesis imaging, and obtains a plurality (n) of projection images P1 of the subject H (step S1). Specifically, when the imaging order information is selected by the operation unit 92 and the exposure switch 63 is pressed, the control unit 91 controls each apparatus of the radiation imaging apparatus 1 via the control BOX 80 to control the radiation source 61. Then, the radiation detector F is moved in the opposite direction along the body axis direction of the subject H around the rotation center O, and a predetermined number of times of imaging are performed. A series of projection images obtained by imaging is transmitted to the console 90 by the radiation detector F. In the console 90, each of a series of projection images received by the communication unit 94 is used to display patient information and imaging conditions of the projection images (imaging site, imaging direction, tube voltage, radiation source 61 when each projection image is acquired, And stored in the projection image storage unit 951 in association with the position of the radiation detector F).

  Here, the signal value (pixel value) of each pixel of the projection image acquired by the radiation detector F is a value obtained by converting the radiation intensity reaching the radiation detector F into an electric signal, that is, reaching the radiation detector F. The pixel value increases as the reached radiation intensity increases. On the other hand, the console 90 handles each pixel value of the projection image as representing an absorbed dose. In other words, the control unit 91 converts each pixel value of a series of projection images received by the communication unit 94 into a value representing an absorbed dose, and stores the value in the projection image storage unit 951. Each pixel value of the projected image increases as the absorbed dose increases, and is rendered white (with a low density) on the projected image.

Next, the control unit 91 temporarily extracts a high-absorber region from each acquired projection image P1, and acquires a projection image P2 of only the high-absorber region (step S2).
For example, if a metal plate, bolt, pacemaker, artificial joint, cochlear implant or the like is embedded in the body of the subject H, these are high-absorbers with a very high radiation absorption coefficient, and thus the projected image P1. Above, the pixel value in that region is high. In step S2, the superabsorber region having a high pixel value is temporarily extracted. As a method for extracting the superabsorbent region, for example, threshold processing (binarization processing), graph cut processing that is advanced region extraction processing, or the like can be performed. In order to increase the recognition accuracy, various correction processes such as a scattered radiation correction process may be performed on the projection image P1 in advance.

  Next, the control unit 91 reconstructs the projection image P2 of only the superabsorber region, and generates a reconstructed image C1 that is a reconstructed image of only the superabsorber region (step S3). Here, “reconstructing a projection image” means generating a reconstruction image (tomographic image) by back projecting the projection image. For example, using a known method such as FBP (Filtered Back Projection) method, successive approximation image reconstruction method, Feldkamp method, shift addition method, etc., the projection image P2 of only the high absorber region is reconstructed to obtain a high absorber. It is possible to generate the reconstructed image C1 of only the region.

  Next, the controller 91 sets 1 to the variable i and sets the number of shape models stored in the high absorber shape DB 953 to imax (step S4), performs pattern matching, and sets the i-th in the high absorber shape DB 953. The similarity between the shape model and the reconstructed image C1 is calculated (step S5).

In step S5, for example, a three-dimensional shape model of the superabsorber is generated based on the image of each slice plane of the reconstructed image C1, and the three-dimensional space is changed while changing the position and angle of the i-th shape model. The process of calculating the similarity (for example, normalized cross-correlation coefficient) between the i-th shape model and the generated three-dimensional shape model is repeated, and the highest similarity among the calculated similarities is determined as the i-th shape model. The similarity of the reconstructed image C1 of only the superabsorbent region is assumed.
Alternatively, the i-th shape model is converted into a two-dimensional reconstructed image while changing the position and angle of the i-th shape model (converted into a tomographic image at the same slice interval as that of the reconstructed image C1). ) In the two-dimensional space, the process of calculating the similarity between the i-th shape model and the reconstructed image C1 is repeated, and the highest similarity among the calculated similarities is determined as the i-th shape model and the reconstructed image C1. It is good also as similarity of. Here, since the reconstructed image includes reconstructed images of a plurality of slice planes, similarity is calculated between reconstructed images having the same slice position, and representative values of the calculated plurality of similarities (for example, , The integrated value or the average value) as the similarity at the position and angle.
The calculated highest similarity is stored in the RAM in association with the shape model identification information and the position and angle of the shape model when the similarity is calculated.

  When the calculation of the similarity between the i-th shape model and the reconstructed image C1 is completed, the control unit 91 increments i (step S6) and determines whether i> imax is satisfied (step S7). When it is determined that i> imax is not satisfied (step S7; NO), the controller 91 returns to step S4.

  When it is determined that i> imax is satisfied (step S7; YES), the control unit 91 embeds a shape model having the highest similarity with the reconstructed image C1 in the subject H among the shape models of the high absorber shape DB 953. It is selected as a shape model of the selected superabsorbent (step S8), and it is determined whether or not the similarity between the selected shape model and the reconstructed image C1 is equal to or greater than a predetermined reference value (step S9). In step S <b> 9, the predetermined reference value is a value for determining whether or not the selected shape model is a shape model of the superabsorber that is truly embedded in the subject H.

  When the degree of similarity between the selected shape model and the reconstructed image C1 is equal to or greater than a predetermined reference value (step S9; YES), the control unit 91 determines a high level in the reconstructed image C1 based on the selected shape model. By identifying and replacing the shape and position of the absorber region, a reconstructed image C2 of only the high absorber region is generated (step S10). For example, by arranging a shape model based on the position and angle stored in the RAM in pattern matching and generating a reconstructed image with the same slice interval as the slice interval of the reconstructed image C1, high absorption in the reconstructed image C1 The shape and position of the body region can be specified.

  Next, the control unit 91 ortho-projects the reconstructed image C2 on the computer to generate a projection image P3 of only the superabsorbent region (step S11). The orthographic projection is based on the voxel value of the reconstructed image, and simulates the X-ray absorption information absorbed by the subject (here, a high-absorber) when the radiation source 61 irradiates the projected image. Is to generate. By steps S10 and S11, the shape and position of the superabsorbent region in the projection image can be specified.

  Next, the control unit 91 removes an area corresponding to the projection image P3 of only the superabsorbent area in the projection image P1 (step S12). As a removal method, for example, the pixel value of the region corresponding to the projection image P3 is replaced with a value indicating that there is no data, such as 0 or a negative value.

  Next, the control unit 91 generates a projection image P4 in which the superabsorbent region is interpolated by interpolating the region removed in step 12 with surrounding pixels in the projection image P1 (step S13). As an interpolation method, for example, a known interpolation method such as linear interpolation or polynomial interpolation can be used.

Next, the control unit 91 reconstructs the projection image P4 in which the superabsorber region is interpolated to generate a reconstructed image C3 in which the superabsorber region is interpolated (step S14).
Then, the control unit 91 generates a reconstructed image C5 by synthesizing the reconstructed image C2 including only the superabsorber region and the reconstructed image C3 obtained by interpolating the superabsorber region (step S15), and the subject reconstructed image. The generation process A ends. For example, the corresponding pixels of the slice planes corresponding to the reconstructed image C2 and the reconstructed image C3 are added to generate the reconstructed image C5.

  On the other hand, when it is determined that the degree of similarity between the selected shape model and the reconstructed image C1 is not equal to or greater than a predetermined reference value (step S9; NO), the control unit 91 includes only the superabsorber region of the projection image P1. An area corresponding to the projected image P2 is removed (step S16). For example, the pixel value in the region corresponding to the projection image P2 is replaced with a value indicating that there is no data, such as 0 or a negative value.

  Next, the control unit 91 generates a projection image P5 in which the superabsorbent region is interpolated by interpolating the region removed in step S16 with surrounding pixels in the projection image P1 (step S17). As an interpolation method, for example, a known interpolation method such as linear interpolation or polynomial interpolation can be used.

Next, the control unit 91 reconstructs the projection image P5 in which the superabsorber region is interpolated to generate a reconstructed image C4 in which the superabsorber region is interpolated (step S18).
Then, the control unit 91 generates a reconstructed image C6 by synthesizing the reconstructed image C1 including only the superabsorber region and the reconstructed image C4 obtained by interpolating the superabsorber region (step S19), and the subject reconstructed image. The generation process A ends. For example, the corresponding pixels of the slice planes corresponding to the reconstructed image C1 and the reconstructed image C4 are added to generate the reconstructed image C6.

  FIG. 5 shows an example of the projection image P4. FIG. 6A shows an example of the reconstructed image C5. FIG. 6B shows an enlarged view of the partial region α in FIG.

  As shown in FIG. 5, in the projection image P4, the edge artifact (see FIG. 9) of the superabsorber that has occurred in the past is suppressed. As described above, the projection image P4 specifies and interpolates the superabsorber region in the projection image P1 based on the superabsorber shape model embedded in the subject selected from the superabsorber shape model DB 951. It is the image produced | generated by this. Therefore, even if there is a thin portion at the edge of the superabsorber, interpolation is performed as a superabsorber region including that portion, so that as shown in FIG. In addition, the image of the edge of the superabsorber is suppressed.

  Further, as shown in FIGS. 6A and 6B, the artifacts of the edge of the high-absorbent body (see FIGS. 10A and 10B) that have been generated in the past are also suppressed in the reconstructed image C5. ing. As described above, the reconstructed image C5 is obtained by projecting the reconstructed image C2 of the superabsorber region generated based on the superabsorber shape model embedded in the subject selected from the superabsorber shape model DB 951. This is an image synthesized with the reconstructed image C3 obtained by reconstructing the image P4. Accordingly, since the reconstructed image of the superabsorber region is synthesized with the reconstructed image of the projection image P4 in which the occurrence of the edges of the superabsorber is suppressed, as shown in FIGS. 6 (a) and 6 (b). Thus, an image in which the artifacts of the edge of the superabsorbent material which has been generated conventionally is suppressed is obtained. In this way, by suppressing artifacts due to the high-absorber in the reconstructed image, it is possible to improve the rendering performance of the peripheral region of the high-absorber, for example, a high joint state between the artificial joint and the bone, etc. The diagnostic performance of the absorber edge can be improved.

<Second Embodiment>
Next, a second embodiment of the present invention will be described.
In the first embodiment, the case where the superabsorber shape model embedded in the subject is selected from the superabsorber shape model DB 951 based on the reconstructed image C1 of only the superabsorber region has been described as an example. In the second embodiment, a case where a superabsorber shape model embedded in a subject is selected from the superabsorber shape model DB 951 based on the projection image P2 of only the superabsorber will be described as an example.

  Since the configuration in the second embodiment is the same as that described in the first embodiment, the description will be referred to and the operation in the second embodiment will be described below.

  FIG. 7 shows a flowchart of the subject reconstructed image generation process B executed by the control unit 91 of the console 90 in the second embodiment. The subject reconstructed image generation process B is executed in cooperation with the control unit 91 and a program stored in the storage unit 95.

  First, the control part 91 performs the process of step S21-S24. Since the process of S21-S24 is the same as the process of step S1-S4 of FIG. 4, description is used.

Next, the control unit 91 performs pattern matching to calculate the similarity between the projection image Pm and the projection image P2 generated from the i-th shape model of the superabsorbent body shape DB 953 (step S25). For example, while changing the position and angle of the i-th shape model, the i-th shape model is orthographically generated to generate a projection image Pm of the shape model, and the similarity between the generated projection image Pm and the projection image P2 is calculated. The process is repeated, and the highest similarity among the calculated similarities is set as the similarity between the projection image Pm of the i-th shape model and the projection image P2. Here, since there are a plurality of projection images Pm and projection images P2, the similarity is calculated between the projection image Pm and the projection image P2 having the same positions of the radiation source 61 and the radiation detector F, and the calculated plurality of projection images Pm and projection images P2 are the same. A representative value (for example, an integrated value or an average value) of the similarity is set as the similarity between the projection image Pm and the projection image P2 at the position and angle.
The highest similarity is stored in the RAM in association with the identification information of the shape model and the position and angle of the shape model when the similarity is calculated.

  When the calculation of the similarity between the projection image Pm generated based on the i-th shape model and the projection image P2 is completed, the control unit 91 increments i (step S26), and determines whether i> imax. Judgment is made (step S27). If it is determined that i> imax is not satisfied (step S27; NO), the controller 91 returns to step S24.

  When it is determined that i> imax is satisfied (step S27; YES), the control unit 91 generates a projection image Pm having the highest similarity to the projection image P2 among the shape models of the high absorber shape DB 953. A shape model is selected as a shape model of the superabsorbent body embedded in the subject (step S28), and whether or not the similarity between the projection image Pm and the projection image P2 of the selected shape model is greater than or equal to a predetermined reference value. Judgment is made (step S29). In step S29, the predetermined reference value is a value for determining whether or not the selected shape model is a shape model of the superabsorbent body that is truly embedded in the subject H.

  When the similarity between the projection image Pm of the selected shape model and the projection image P2 is greater than or equal to a predetermined reference value (step S29; YES), the control unit 91 reconstructs the reconstructed image based on the selected shape model. By identifying and replacing the shape and position of the superabsorber region in C1, a reconstructed image C2 of only the superabsorber region is generated (step S30). For example, by arranging a shape model based on the position and angle stored in the RAM in pattern matching and generating a reconstructed image with the same slice interval as the slice interval of the reconstructed image C1, high absorption in the reconstructed image C1 The shape and position of the body region can be specified.

  About the process after step S31, since it is the same as the process after step S11 of FIG. 4, description is used.

  Also by the subject reconstructed image generation processing B, the projection image P4 shown in FIG. 5 and the reconstructed image C5 shown in FIGS. 6A and 6B can be obtained, and the performance of rendering the superabsorber edge is improved. Can be made.

  As described above, according to the control unit 91 of the console 90, the high-absorber region of radiation is temporarily extracted from the projection image captured by the radiation imaging apparatus 1, and, for example, only the temporarily extracted high-absorber region is projected. By performing pattern matching between the reconstructed image obtained by reconstructing the image and the shape model stored in the high absorber shape DB 953, the shape model of the high absorber embedded in the subject is selected from the high absorber shape DB 953. . Next, the control unit 91 reconstructs the projection image and the projection image based on the reconstructed image obtained by reconstructing the selected superabsorber shape model and the temporarily extracted superabsorber region only image. The shape and position of the superabsorber region in the image are specified, the pixels of the superabsorber region specified in the projection image are interpolated, and the reconstructed projection image is reconstructed to generate a reconstructed image. By combining with a reconstructed image of only the region, a reconstructed image of the subject with reduced artifacts due to the high absorber is generated.

  Therefore, even if there is a thin portion at the edge of the superabsorber, the interpolation is performed by specifying the superabsorber region including that part, so artifacts due to the superabsorber are not generated. It is possible to obtain a reconstructed image that is reduced and the performance of rendering the edge portion of the high absorber is improved.

  Further, the control unit 91 determines whether or not the similarity between the selected shape model and the superabsorbent region is equal to or greater than a predetermined reference value. A reconstructed image of a subject with reduced artifacts due to the body is generated. Therefore, when the shape model of the high absorber embedded in the subject H is not stored in the high absorber shape DB 953, it is possible to prevent erroneous processing using the shape model.

  In addition, the description content in the said embodiment is a suitable example of the tomographic image generation system which concerns on this invention, and is not limited to this.

For example, as a pattern matching technique for selecting the closest model to the high absorber embedded in the subject from the shape models stored in the high absorber shape DB 953 in the control unit 91, the first embodiment is used. The method of the second embodiment may be combined with the method of the second embodiment. For example, for each of the shape models stored in the high-absorber shape DB 953, the shape model and the reconstructed image C1 of only the high-absorber region obtained by reconstructing the projection image of only the temporarily extracted high-absorber region. The first similarity obtained by performing pattern matching and the second similarity obtained by performing pattern matching between the projection image Pm of the shape model and the projection image P2 of only the temporarily extracted superabsorbent region The shape of the superabsorber in which the shape model of the projection image Pm when the highest similarity is calculated among the similarities calculated for each shape model is embedded in the subject. A model may be selected.
Thus, by combining the method of the first embodiment and the method of the second embodiment, it is possible to give stability to the result of pattern matching.

  In the high-absorber shape DB 953, the model data of the high-absorber corresponding to the model is stored in association with the data of each shape model, and corresponds to the shape model of the high-absorber selected by pattern matching. The model number, position (position when the degree of similarity is maximum), shape, etc. are displayed on the display unit 93, and if the model number, shape, or position is different from the actual, it may be corrected by the operation unit 92. . Thereby, the user can confirm the model number, shape, and position of the selected superabsorbent body, and if it is incorrect, can correct it. When the correction is input, for example, the control unit 91 corrects the reconstructed image C2 including only the high absorber based on the correction content input by the operation unit 92, and performs the subsequent processing again.

  In the above-described embodiment, a subject is selected from the shape models stored in the high-absorber shape DB 953 by pattern matching between the shape model stored in the high-absorber shape DB 953 and the reconstructed image or the projection image. Although the shape model of the embedded superabsorbent is selected, the selection method is not limited to pattern matching. For example, in the high-absorber shape DB 953, the model number or shape of the high-absorber embedded in the subject is stored in association with the data of each shape model in association with the model number of the high-absorber corresponding to the model. The control unit 91 selects the shape model corresponding to the specified model number and shape as the shape model of the superabsorbent body embedded in the subject. It is good to do. Thereby, the processing time of pattern matching can be shortened.

  Whether the processing for selecting the shape model of the high-absorber embedded in the subject from the shape models stored in the high-absorber shape DB 953 is performed by pattern matching or by specifying the model number or shape by the user Can be selected by the operation unit 92 (that is, provided with a selection unit), and the control unit 91 may select a shape model of the superabsorbent body embedded in the subject by the selected method. Thereby, the shape model of the superabsorbent body embedded in the subject can be selected by a method desired by the user.

Further, the model number and identification information of the selected shape model may be stored in the storage unit 95 in association with the patient information of the subject H. Then, when the subject reconstructed image generation process A or B is performed next time, the control unit 91 searches the storage unit 95 and associates the model number and identification information of the shape model with the patient information of the patient who is the subject H. If stored, the shape model corresponding to the model number or identification information may be selected as the shape model of the superabsorbent body embedded in the subject.
Alternatively, the model number and / or shape of the superabsorbent stored in the storage unit 95 in association with the patient information and a message such as “Is this superabsorbent embedded?” Are displayed on the display unit 93. Then, the operation unit 92 may input YES or NO. If YES is input, the superabsorbent region is specified by the previous shape model. If NO is input, the superabsorbent shape model is selected by pattern matching, and the selected shape is selected. The superabsorbent region may be specified based on the model.
Alternatively, the shape model of the superabsorbent is selected by pattern matching as in the above embodiment, and the model number and identification information of the selected shape model are stored in the storage unit 95 in association with the patient information. A warning may be output when the identification information does not match. The warning may be output by displaying a warning on the display unit 93 or outputting a warning message or a warning sound by voice.

In addition, the configuration may be such that the user can select whether or not to generate a reconstructed image with reduced artifacts due to the superabsorbent by the operation unit 92 (that is, a second selection unit may be provided). ). When generation of a reconstructed image with reduced artifacts due to the high-absorber is selected, the control unit 91 reconstructs by subject reconstructed image generation processing of the first embodiment or the second embodiment. When generating an image and generating a reconstructed image with reduced artifacts due to the high-absorber is not selected, the reconstructed image may be generated by reconstructing the projection image P1 obtained by imaging. .
As a result, when the user desires to reduce artifacts caused by the high-absorber, an image with reduced artifacts caused by the high-absorber can be provided. For example, the user may want to shorten the processing time. However, when it is not desired to reduce artifacts caused by the high-absorber, the processing time for generating the reconstructed image can be shortened.

  Further, in the high-absorber shape DB 953, the high-absorber material is associated with the shape model of the high-absorber, direction information (information indicating in which direction the object is embedded with respect to the front of the subject), dimensions, and / or Alternatively, at least one of the volume, the imaging region where the shape model is used, and the imaging direction may be stored in association with each other.

  For example, since the X-ray transmittance varies depending on the material, it can be used to change the threshold value when the high-absorber region is temporarily extracted.

  Further, in the high-absorber shape DB 953, direction information (information indicating in which direction the image is embedded with respect to the front of the subject) is stored in association with the shape model of the high-absorber, so that a projection image can be captured. Since the direction of the shape model in the projection image or the reconstructed image can be specified according to the shooting direction of the image, the range for changing the angle of the shape model at the time of pattern matching can be limited, and the processing time for pattern matching Can be shortened.

  Further, in the superabsorber shape DB 953, by storing the size and / or volume in association with the superabsorber shape model, for example, the superabsorber actually embedded based on the reconstructed image is stored. The size and volume can be estimated, the pattern matching target can be narrowed down to a shape model close to the estimated size and volume, and the processing time for pattern matching can be shortened.

  Also, in the high-absorber shape DB 953, by storing the imaging part and the imaging direction in which the shape model is used in association with the shape model of the high-absorber, depending on the imaging part and the imaging direction at the time of projection image shooting. The target of pattern matching can be narrowed down, and the processing time concerning pattern matching can be shortened.

  Moreover, in the said embodiment, the process which specifies a high absorber area | region based on the selected shape model, and produces | generates a reconstructed image by whether a similarity is more than a predetermined reference value (FIG. 4). Steps S10 to S15, Steps S30 to S35 in FIG. 7, and processing for generating a reconstructed image based on the superabsorber region temporarily extracted by the threshold processing (Steps S16 to S19 in FIG. 4, Step S36 in FIG. 7). -S39), both processes are performed, both reconstructed images are displayed on the display unit 93, and the user can select which reconstructed image is to be adopted by the operation unit 92. May be.

  In the above embodiment, the temporary extraction of the superabsorbent region is performed on the projection image. However, it may be performed on the reconstructed image obtained by reconstructing the projection image.

  In the above embodiment, the superabsorber shape DB 953 is described as being stored in the storage unit 95, but may be stored in an external server or the like connected via the network N1 or N2. Good.

  Moreover, in the said embodiment, the radiation detector F is what is called a portable type | mold (it is also called a cassette type | mold etc.), and it is the detector loading part 51 (the figure mentioned later) of the imaging stand 50 which comprises the radiation imaging device 1. 1), the radiation detector F is used for radiography, but the radiation detector F is not portable, but is a so-called dedicated type radiation detector formed integrally with the imaging table 50. In addition, the present invention can be applied.

  Moreover, in the said embodiment, although the radiography apparatus 1 demonstrated as an apparatus which image | photographs in a supine position, it is good also as an apparatus which performs imaging | photography of a standing position.

In the above embodiment, the radiation imaging apparatus 1 is described as performing tomosynthesis imaging by moving the radiation source 61 and the radiation detector F in the opposite directions along the body axis direction of the fixed subject H. The radiation source 61 may be moved while the radiation detector F and the subject H are fixed. Alternatively, the radiation detector F may be fixed and the subject H and the radiation source 61 may be moved. Alternatively, the radiation source F may be fixed and the radiation detector F and the subject H may be moved.
Further, the present invention can be applied not only when generating a tomographic image from a projection image obtained by tomosynthesis imaging but also when generating a tomographic image from a projection image obtained by CT imaging.

  In the above description, an example in which a hard disk, a semiconductor nonvolatile memory, or the like is used as a computer-readable medium for the program according to the present invention is disclosed, but the present invention is not limited to this example. As another computer-readable medium, a portable recording medium such as a CD-ROM can be applied. A carrier wave is also applied as a medium for providing program data according to the present invention via a communication line.

  In addition, the detailed configuration and detailed operation of each apparatus constituting the tomographic image generation system can be changed as appropriate without departing from the spirit of the invention.

DESCRIPTION OF SYMBOLS 100 Tomographic image generation system 1 Radiation imaging apparatus 50 Imaging stand 51 Detector loading part 52 Loading part support part 53 Conveyance apparatus 54 Subject stand 60 Radiation irradiation apparatus 61 Radiation source 62 Operation desk 63 Exposure switch 64 Radiation source moving mechanism 80 Control BOX
90 Console 91 Control unit 92 Operation unit 93 Display unit 94 Communication unit 95 Storage unit 96 Bus F Radiation detector H Subject

Claims (13)

A reconstructed image of the subject is generated using a plurality of projection images obtained by arranging a subject between the radiation source and the radiation detector and changing a positional relationship between the radiation source and the radiation detector. An image processing apparatus for performing
Temporary extraction means for temporarily extracting a high-absorber region of radiation from the projection image;
A selecting means for selecting a shape model of the superabsorber embedded in the subject from a database storing a shape model of the superabsorber of a plurality of types of radiation;
Based on the shape model of the selected superabsorber and the reconstructed image obtained by reconstructing only the temporarily extracted superabsorber region, the shape of the superabsorber region in the projection image and the reconstructed image, and A specifying means for specifying the position;
Interpolation means for interpolating pixels of the superabsorbent region specified by the specifying means in the projection image;
First reconstructed image generation means for reconstructing a projection image after interpolation by the interpolation means to generate a reconstructed image;
Second reconstructed image generating means for generating a reconstructed image of only the superabsorbent region based on the shape and position of the superabsorbent region in the reconstructed image specified by the specifying means;
By synthesizing the reconstructed image generated by the first reconstructed image generating means and the reconstructed image generated by the second reconstructed image generating means, artifacts due to the superabsorber are reduced. Third reconstructed image generating means for generating a reconstructed image of the subject;
An image processing apparatus comprising:   The selecting means performs pattern matching between each of the shape models stored in the database and a reconstructed image obtained by reconstructing the projection image of only the temporarily extracted superabsorbent region, and the obtained similarity is obtained. The image processing apparatus according to claim 1, wherein a shape model of a superabsorber embedded in the subject is selected from the database.   The selection means performs pattern matching between each of the projection images obtained by orthographic projection of each shape model stored in the database and the projection image of only the temporarily extracted superabsorbent region, and the obtained similarity is obtained. The image processing apparatus according to claim 1, wherein a shape model of a superabsorber embedded in the subject is selected from the database.   The selecting means performs pattern matching between each shape model stored in the database and a reconstructed image obtained by reconstructing a projection image of only the temporarily extracted superabsorbent region. The sum of the obtained first similarity and the second similarity obtained by performing pattern matching between the projection image obtained by orthographic projection of the shape model and the projection image of only the temporarily extracted superabsorbent region The image processing apparatus according to claim 1, wherein the shape model of the superabsorbent body embedded in the subject is selected from the database based on the similarity degree calculated for each shape model.   It is determined whether or not the similarity with the shape model selected by the selection means is equal to or higher than a predetermined reference value. If the similarity is equal to or higher than a predetermined reference value, the interpolation means and the first reconstructed image generation Control means for generating a reconstructed image of the subject in which artifacts due to the high-absorbent body are reduced by performing processing of the means, the second reconstructed image generating means, and the third reconstructed image generating means An image processing apparatus according to any one of claims 2 to 4. The database stores each shape model of the superabsorbent body in association with the model number of the superabsorbent body,
Designation means for the user to designate the model number or shape of the superabsorbent body embedded in the subject;
When selecting the shape model of the superabsorbent embedded in the subject from the database by the selecting means, whether to select by pattern matching or based on the model number or shape specified by the specifying means A selection means for the user to select,
An image processing apparatus according to any one of claims 1 to 5. The database stores each shape model of the superabsorbent body in association with the model number of the superabsorbent body,
Comprising designation means for the user to designate the model number or shape of the superabsorbent body embedded in the subject;
The image processing apparatus according to claim 1, wherein the selection unit selects a shape model of a superabsorbent body embedded in the subject from the database based on a model number or a shape designated by the designation unit.   8. The apparatus according to claim 1, further comprising second selection means for a user to select whether or not to generate a reconstructed image of the subject in which artifacts due to the high absorber are reduced. Image processing apparatus.   9. The storage control unit according to claim 1, further comprising: a storage control unit that stores the patient information of the patient who is the subject and the identification information in the database of the selected shape model in association with each other. Image processing device.   10. The display device according to claim 1, further comprising a display unit configured to display a model number of the shape model selected by the selection unit, a shape, and / or a position of the superabsorbent region specified by the specification unit. Image processing device.   The database is associated with the shape models of the plurality of types of superabsorbers, the material of the superabsorbers, the direction, size and / or volume embedded in the subject, the imaging region where the shape model is used, and / or The image processing apparatus according to claim 1, wherein at least one of the shooting directions is stored. A radiation source that irradiates a subject with radiation, and a radiation detector that two-dimensionally arranges radiation detection elements that detect radiation and generate electrical signals, and obtain a projection image corresponding to the irradiated radiation. Imaging means for acquiring the projection image of the subject arranged between the radiation source and the radiation detector a predetermined number of times while changing the positional relationship between the radiation source and the radiation detector;
Image processing means for generating a tomographic image that is a reconstructed image of the subject using the projection image acquired by the imaging means;
A tomographic image generation system comprising:
The image processing means includes
Temporary extraction means for temporarily extracting a high-absorber region of radiation from the projection image;
A selecting means for selecting a shape model of the superabsorber embedded in the subject from a database storing a shape model of the superabsorber of a plurality of types of radiation;
Based on the shape model of the selected superabsorber and the reconstructed image obtained by reconstructing only the temporarily extracted superabsorber region, the shape of the superabsorber region in the projection image and the reconstructed image, and A specifying means for specifying the position;
Interpolation means for interpolating pixels of the superabsorbent region specified by the specifying means in the projection image;
First reconstructed image generation means for reconstructing a projection image after interpolation by the interpolation means to generate a reconstructed image;
Second reconstructed image generating means for generating a reconstructed image of only the superabsorbent region based on the shape and position of the superabsorbent region in the reconstructed image specified by the specifying means;
By synthesizing the reconstructed image generated by the first reconstructed image generating means and the reconstructed image generated by the second reconstructed image generating means, artifacts due to the superabsorber are reduced. Third reconstructed image generating means for generating a reconstructed image of the subject;
A tomographic image generation system comprising: A reconstructed image of the subject is generated using a plurality of projection images obtained by arranging a subject between the radiation source and the radiation detector and changing a positional relationship between the radiation source and the radiation detector. A computer used in the image processing apparatus
Temporary extraction means for temporarily extracting a high-absorber region of radiation from the projected image;
A selecting means for selecting a shape model of the superabsorber embedded in the subject from a database storing a shape model of the superabsorber of a plurality of types of radiation;
Based on the shape model of the selected superabsorber and the reconstructed image obtained by reconstructing only the temporarily extracted superabsorber region, the shape of the superabsorber region in the projection image and the reconstructed image, and Identification means for identifying the position,
Interpolation means for interpolating pixels of the superabsorbent region specified by the specifying means in the projection image;
First reconstructed image generation means for reconstructing a projection image after interpolation by the interpolation means to generate a reconstructed image;
Second reconstructed image generating means for generating a reconstructed image of only the superabsorbent region based on the shape and position of the superabsorbent region in the reconstructed image specified by the specifying means;
By synthesizing the reconstructed image generated by the first reconstructed image generating means and the reconstructed image generated by the second reconstructed image generating means, artifacts due to the superabsorber are reduced. A third reconstructed image generating means for generating a reconstructed image of the subject;
Program to function as.