JP6413059B2 - Image processing apparatus, image processing method, and X-ray imaging apparatus - Google Patents

Description

  The present invention relates to an image processing apparatus and an image processing method for specifying a position of a metal body included in an X-ray imaging region, and an X-ray imaging apparatus including the image processing apparatus.

  When a CT (Computed Tomography) object includes a metal that hardly transmits X-rays, a metal artifact is generated on the reconstructed volume data as shown in FIG. The reason why such metal artifacts occur is that the overlap between metals cannot be expressed on the measurement image and the projection image because the X-rays do not reach the detector.

  The CT apparatus proposed in Patent Document 1 reduces metal artifacts by reducing contradiction as a projection image due to the inability to express the overlap between metals. Specifically, the CT apparatus proposed in Patent Document 1 extracts a metal projection image by binarizing the projection image, and the binarized projection image (metal projection image is extracted). The metal data is extracted by further binarizing the volume data obtained by back projecting the projected image), and the metal position on the projection image is specified by forward projecting the metal. Interpolation of pixel values from the surroundings reduces the contradiction as a projection image due to the inability to express metal overlap.

Japanese Patent No. 5324883

  However, the CT apparatus proposed in Patent Document 1 does not assume a case where a projection image contains many scattered rays, and is vulnerable to scattered rays.

  When a subject contains a metal that hardly transmits X-rays, the X-ray is blocked by the metal, so that the number of photons in a region where a projected image of the metal in the detector appears should be almost zero. However, in actuality, as shown in FIG. 20, the scattered rays generated in the surrounding area enter the region where the projected image of the metal is reflected, and therefore the number of photons in the region where the projected image of the metal is reflected in the detector is a value close to zero. Don't be. Furthermore, since the amount of scattered radiation varies depending on the shape of the subject, the projection angle, and the position of the metal, the number of photons in the region where the projected image of the metal appears on the detector also varies.

  In particular, the detector of a dental CT apparatus is not equipped with a collimator provided in the detector of a medical CT apparatus and is defenseless against scattered radiation. When the projection image shown in FIG. 21 actually obtained by the dental CT apparatus is binarized by threshold processing and a metal projection image is extracted as a white region, the result is as shown in FIG. It can be seen that a large area is also extracted as a white area.

  As described above, since it is difficult to binarize a projection image obtained by a dental CT apparatus so that a metal projection image can be extracted with high accuracy, the CT apparatus proposed in Patent Document 1 is used as a dental CT apparatus as it is. Even if applied, metal extraction and metal artifact removal are not performed properly. In particular, when the metal becomes large, the difficulty of binarization that can accurately extract the projected image of the metal becomes more prominent.

  In view of the above situation, an object of the present invention is to provide an image processing apparatus, an image processing method, and an X-ray imaging apparatus that perform metal extraction robust against scattered radiation.

  In order to achieve the above object, an image processing apparatus according to the present invention pays attention to one of voxels constituting volume data, and at least a part of pixel values of a projection image or a measurement image back-projected on the attention voxel. An acquisition unit that acquires the projection angle, and obtains a feature amount from a plurality of pixel values acquired by the acquisition unit, and whether the target voxel is inside or outside the metal region based on the feature amount An identification unit for identifying the voxel of the voxel that constitutes the volume data, changing the target voxel, and obtaining processing by the acquisition unit for each changed voxel and the identification unit A configuration (first configuration) in which the identification process is performed is described.

  In the image processing apparatus having the first configuration, the feature amount is a pixel of a projection image that is back-projected on the target voxel at all projection angles when the target voxel is in a metal region. Tends to have a higher value compared to pixels in the projected image that are back-projected to voxels that are outside the metal region, and pixels in the measurement image that are back-projected to the target voxel are back to voxels that are outside the metal region. It is preferable that the configuration includes a first feature amount (second configuration) that reflects at least one of tendencies having a lower value than the pixels of the measurement image to be projected.

  In the image processing device having the first or second configuration, the feature amount is an image that is back-projected onto the target voxel at all projection angles when the target voxel is in a metal region. The fluctuation of the pixel value of the measurement image tends to be smaller than the fluctuation of the pixel value of the projection image projected back to the voxel outside the metal region, and the fluctuation of the pixel value of the measurement image back-projected to the target voxel is metal. It is preferable that the configuration includes a second feature amount (third configuration) that reflects at least one of the tendencies that are smaller than the variation in the pixel value of the measurement image that is back-projected onto the voxel that is outside the region. .

  Further, for example, in the image processing apparatus having any one of the first to third configurations, the forward projection processing unit that forward projects the metal region extracted based on the identification result of the identification unit, and the forward projection processing unit. Using the generated metal forward projection image, the metal projection image is removed from the projection image or the measurement image, and the interpolation processing unit that interpolates the pixel values of the removed region from the surroundings, and generated by the interpolation processing unit An addition processing unit that adds the forward projection image of the metal generated by the forward projection processing unit to the generated image, and reconstructing the image generated by the addition processing unit to obtain the reconstructed volume data from which metal artifacts have been removed. It is good also as a structure (4th structure) further provided with the reconstruction process part to produce | generate. Further, for example, in the image processing apparatus having any one of the first to third configurations, the forward projection processing unit that forward projects the metal region extracted based on the identification result of the identification unit, and the forward projection processing unit. Using the generated metal forward projection image, the metal projection image is removed from the projection image or the measurement image, and the interpolation processing unit that interpolates the pixel values of the removed region from the surroundings, and generated by the interpolation processing unit Reconstructing the generated image to generate reconstruction volume data from which the metal has been removed, and reconstruction volume data from which the metal has been removed by the reconstruction processing unit to which the metal has been added An addition unit that adds the volume data extracted based on the identification result of the identification unit to generate reconstructed volume data after removal of metal artifacts (fifth configuration) It may be formed).

  In order to achieve the above object, an image processing method according to the present invention includes a projection image or a measurement image corresponding to at least a partial projection angle that is back-projected onto a target voxel that is one of the voxels constituting volume data. An acquisition step of acquiring a pixel value, a feature amount is obtained from a plurality of pixel values acquired by the acquisition step, and whether the target voxel is inside or outside the metal region based on the feature amount An identification step for identifying the target voxel between at least a part of the voxels constituting the volume data, and the acquisition step and the identification step for each of the changed target voxels. A configuration (sixth configuration) in which the identification process is performed is described.

  In order to achieve the above object, an X-ray imaging apparatus according to the present invention includes an X-ray irradiation unit that irradiates a subject with X-rays, an X-ray detection unit that detects X-rays transmitted through the subject, and the X-ray detection unit. An image generation unit that generates a projection image or a measurement image from a detection result of the line detection unit, and an image processing device having any one of the first to fifth configurations that performs image processing on the projection image or the measurement image; , (7th configuration).

  According to the present invention, a metal region can be extracted on volume data without binarizing the projection image as in Patent Document 1. Therefore, when identifying whether or not the target voxel is a metal region, it is possible to identify based on a feature amount reflecting a robust tendency against scattered radiation. Thereby, metal extraction robust against scattered radiation can be performed.

The figure which shows the external appearance of the main-body part of the X-ray imaging apparatus which concerns on one Embodiment of this invention. The figure which shows the orbit of local CT imaging mode The figure which shows the center of an imaging | photography object site | part at the time of setting the center of an imaging | photography object site | part to the position of an anterior tooth, and an image reconstruction range in local CT imaging | photography mode The figure which shows the center of an imaging | photography object site | part at the time of setting the center of an imaging | photography object site | part to the position of a left jaw, and an image reconstruction range in local CT imaging | photography mode The figure which shows the center of an imaging | photography object site | part at the time of setting the center of an imaging | photography object site | part to the position of a right 2nd premolar, and an image reconstruction range in local CT imaging | photography mode The figure which shows the orbit of all teeth CT imaging mode Diagram showing trajectory in all jaw CT imaging mode The figure which shows the structure of an image processing apparatus The figure which shows each example of a measurement image and a projection image Flow chart showing metal extraction process and metal artifact removal process Flow chart showing metal extraction process and other metal artifact removal process Diagram showing reconstructed volume data with metal artifacts Diagram showing binarized reconstructed volume data Figure showing a projected image Diagram showing volume data from which metal has been extracted Diagram showing forward projection image of metal The figure which shows the projection image which removed the projection image of the metal Diagram showing corrected projection image Diagram showing reconstructed volume data with metal removed Diagram showing reconstructed volume data with metal artifacts removed Diagram showing reconstructed volume data with metal artifacts Image diagram showing that X-rays enter the area where the projected image of the metal is reflected by the scattered radiation Figure showing a projected image The figure which shows the image which binarized the projection image of FIG.

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

  First, the configuration of a main body 1 (hereinafter referred to as “main body 1 of an X-ray imaging apparatus”) of an X-ray imaging apparatus according to an embodiment of the present invention will be described with reference to FIG. FIG. 1 is an external view of a main body 1 of an X-ray imaging apparatus. FIG. 1 (a) is a top view, FIG. 1 (b) is a front view, and FIG. 1 (c) is a side view.

  A main body 1 of an X-ray imaging apparatus is a main body of an X-ray imaging apparatus for dental use or otolaryngology, and is erected in a vertical direction from a base 2 placed on a floor surface and the base 2. The lower pole 3, the upper pole 4 connected to the lower pole 3 so as to be slidable in the vertical direction, the fixed arm 5 fixed to the upper end of the upper pole 4, and the swivel connected to the fixed arm 5 so as to be rotatable An arm 6 and a head holding portion 7 that is fixed to the central portion of the upper pole 4 and holds the head of a human body including a subject (for example, a tooth) are provided. In the embodiment, the fixed arm 5 is fixed to the upper pole 4. For example, the fixed arm 5 is directly or directly on the wall or ceiling of the room where the main body 1 of the X-ray imaging apparatus is installed. The aspect attached through the adjustment mechanism which can adjust distance may be sufficient.

  The swivel arm 6 has an X-ray irradiation unit 8 that irradiates the subject with X-rays and an X-ray detection unit 9 that detects X-rays transmitted through the subject. In the present embodiment, a two-dimensional X-ray detector in which conversion elements that generate electrical signals according to the irradiated X-rays are two-dimensionally arranged is used as the X-ray detection unit 9.

  The imaging mode of the main body 1 of the X-ray imaging apparatus is not particularly limited, and examples include a panoramic imaging mode and a CT imaging mode. In the panoramic imaging mode, the swivel axis of the swivel arm 6 is perpendicular to the swivel axis (X direction, X direction, so that the X-ray irradiation unit 8 and the X-ray detection unit 9 draw a predetermined locus along the shape of the dental arch. The tomography is performed while moving the swivel arm 6 about the swivel axis. In the CT imaging mode, tomographic imaging of the target imaging region (image reconstruction range) is performed while rotating the swivel arm 6 around the target imaging region (image reconstruction range) of the head.

  Here, the CT imaging mode will be described in more detail with reference to FIGS.

  The local CT imaging mode is a CT imaging mode in which a specific region narrower than the entire upper and lower tooth regions in the tooth jaw region is imaged. The image reconstruction range in the local CT imaging mode is, for example, a cylindrical space region having a diameter of 51 mm and a height of 55 mm. FIG. 2 shows the trajectory in the local CT imaging mode. In the local CT imaging mode, as shown in FIG. 2, the swivel arm 6 is placed so that the center of the X-ray detection unit 9 is on the extension line of the line connecting the X-ray irradiation unit 8 and the swivel axis center 206 of the swivel arm 6. Shooting is performed at a plurality of shooting positions while turning. In the local CT imaging mode, the pivot axis center 206 of the pivot arm 6 is normally at a fixed position as shown in FIG. In FIG. 2, four shooting positions are illustrated, but this is just an example, and the shooting positions are not limited to the illustrated positions.

  In the local CT imaging mode, since the X-ray beam width W on the X-ray detection unit 9 is narrower than the all-tooth CT imaging mode and the full jaw CT imaging mode described later, even if the size of the X-ray detection unit 9 is small. It can be implemented.

  In the local CT imaging mode, the position of the turning axis center 206 of the turning arm 6 is changed according to where the center of the region to be imaged (region of interest) is set. Usually, as shown in FIG. Further, the position adjustment is performed so that the center of the imaging target region (region of interest) and the position of the turning axis center 206 of the turning arm 6 coincide with each other. The center of the region to be imaged (region of interest) in the local CT imaging mode can be arbitrarily set. In addition to the position setting shown in FIG. 2, for example, as shown in FIG. 3, the center 208 of the imaging target region (region of interest) can be set to the position of the front tooth on the virtual dental arch 201. As shown in FIG. 5, the center 208 of the imaging target region (region of interest) can also be set to the position of the left jaw on the virtual dental arch 201. As shown in FIG. The position of the right second premolar on the virtual dental arch 201 can also be set, and various other position settings are possible.

  The all-tooth CT imaging mode is a CT imaging mode in which the entire upper and lower tooth regions are to be imaged. The image reconstruction range in the all-tooth CT imaging mode is, for example, a cylindrical space region having a diameter of 97 mm and a height of 100 mm. FIG. 6 shows the trajectory in the full-tooth CT imaging mode. In the all-tooth CT imaging mode, as shown in FIG. 6, the swivel arm 6 is such that the center of the X-ray detection unit 9 is on the extension line of the line connecting the X-ray irradiation unit 8 and the swivel axis center 206 of the swivel arm 6. Shooting is performed at a plurality of shooting positions while turning. In the all-tooth CT imaging mode, the pivot axis center 206 of the pivot arm 6 is normally at a fixed position as shown in FIG. In FIG. 6, four shooting positions are illustrated, but this is merely an example, and the shooting positions are not limited to the illustrated positions.

  Compared with the above-described local CT imaging mode, the full-tooth CT imaging mode has a wider imaging target and the X-ray beam width W on the X-ray detection unit 9 is wider, and therefore corresponds to the wide X-ray beam width W. The size of the X-ray detection unit 9 is required.

  The all jaw CT imaging mode is a CT imaging mode in which the entire range of the tooth jaw region is an imaging target. The image reconstruction range in the all jaw CT imaging mode is, for example, a cylindrical space region having a diameter of 161 mm and a height of 100 mm. FIG. 7 shows the trajectory in the full jaw CT imaging mode. In the full jaw CT imaging mode, as shown in FIG. 7, the swivel arm 6 is such that the center of the X-ray detection unit 9 deviates from the extended line of the line connecting the X-ray irradiation unit 8 and the swivel axis center 206 of the swivel arm 6. Shooting is performed at a plurality of shooting positions while turning. In the all jaw CT imaging mode, the pivot axis center 206 of the pivot arm 6 is normally at a fixed position as shown in FIG. In FIG. 7, four shooting positions are illustrated, but this is only an example, and the shooting positions are not limited to the illustrated positions.

  In the all jaw CT imaging mode, imaging is performed by shifting the center of the X-ray detection unit 9 from the extended line of the line connecting the X-ray irradiation unit 8 and the turning axis center 206 of the turning arm 6. The image reconstruction range 207 can be expanded as compared with the CT imaging mode. Accordingly, the entire range of the tooth and jaw region can be taken as an imaging target while suppressing the size increase of the X-ray detection unit 9.

  In the all-chin CT imaging mode, the X-ray detection unit 9 is sized up so that the X-ray beam width W on the X-ray detection unit 9 is larger than that shown in FIG. By using a cylindrical space region having a diameter of 230 mm and a height of 164 mm, it is possible to capture not only the entire region of the tooth and jaw region but also the entire region of the head and neck region.

  In the above-described CT imaging mode, the patient's dental arch 203 is present at the position assumed at the time of imaging (the position illustrated in FIGS. 2, 6, and 7), and thus imaging as intended by the photographer is performed. be able to. As a method for easily realizing the alignment of the patient dental arch 203 to the assumed position, for example, a method using a light beam can be cited. Examples of the light beam include a midline light beam indicating the position of the midline of the head, a horizontal light beam indicating the position of the line connecting the lower edge of the orbit and the ear canal, and a tomogram indicating the position of the canine (reference position for tomography). There are reference line light beams and the like, and an output unit of these light beams may be provided in the X-ray imaging apparatus, and the patient may finely adjust the position of the head with reference to these light beams.

  Further, a cephalometric unit (not shown) installed at a position away from the swivel arm 6 may be used to perform photographing in the cephalometric photographing mode. The cephalometric unit includes a cephalometric X-ray detection unit for detecting X-rays transmitted through the subject and imaging the subject, and a head fixing unit for fixing the head. Cephalometric imaging is used for orthodontic diagnosis and the like, and imaging is performed using a head-specific X-ray imaging method (cephalometric imaging method). In cephalometric imaging, for example, the ear rods of the head fixing part are inserted and fixed in the left and right outer ear hole parts of the head, and X-rays are irradiated from the X-ray irradiation part 8 provided on the turning arm 6 to obtain the subject. X-rays that have passed through are detected by a Cefaro X-ray detector.

  An X-ray imaging apparatus according to an embodiment of the present invention includes an image processing apparatus 10 shown in FIG. 8 in addition to the main body 1 of the X-ray imaging apparatus.

  The image processing apparatus 10 includes a CPU 101 that controls the entire image processing apparatus 10 according to programs stored in the ROM 102 and the HDD 107, a ROM 102 that records fixed programs and data, a RAM 103 that provides a working memory, and an X-ray imaging. A communication interface unit 104 for communicating with a control unit (not shown) that is stored in the main unit 1 of the apparatus and controls each unit of the main unit 1 of the X-ray imaging apparatus, and temporarily stores image data. VRAM 105, a display unit 106 for displaying an image based on image data stored in the VRAM 105, an imaging control program for controlling the X-ray imaging operation in cooperation with the control unit and the CPU 101, and reconstructed volume data Image reconstruction processing program for generating metal, metal extraction processing program for extracting metal HDD 107 for storing various programs such as a program, various parameter setting values used when executing the various programs, and various data such as reconstructed volume data, and an input unit 108 such as a keyboard and a pointing device. Yes.

  In the image processing apparatus 10, the communication method between the image processing apparatus 10 and the control unit may be wired communication, wireless communication, or communication combining wired and wireless. An example of the image processing apparatus 10 is a personal computer. In addition to image processing, the image processing apparatus 10 also performs remote operation and image display of the main body 1 of the X-ray imaging apparatus. Each program stored in the HDD 107 may be preinstalled in the image processing apparatus 10, distributed in a form stored in a storage medium such as an optical disk, and installed in the image processing apparatus 10. And installed in the image processing apparatus 10.

  An image output from the X-ray detection unit 9 and received by the image processing apparatus 10 is called a measurement image. Each pixel value of the measurement image can be regarded as representing a value approximately proportional to the number of X-ray photons that have reached each detection element of the X-ray detection unit 9.

  The reconstruction algorithm used in the image reconstruction processing program reconstructs the cross section using the line integral of the X-ray line attenuation coefficient, not the number of X-ray photons. For this reason, in the image reconstruction process, it is necessary to convert a measurement image indicating the distribution of the number of X-ray photons into an image (projection image) indicating the distribution of the line integral of the line attenuation coefficient. Hereinafter, the conversion (logarithmic conversion) from the measurement image to the projection image will be described in detail.

The pixel label is i, the pixel value of the measurement image when there is no subject is I ₀ (i), the pixel value of the measurement image when the subject to be reconstructed this time is placed is I (i), and the X-ray When the path is s and the linear attenuation coefficient at the position s is μ (s), the following equation (1) is established. This equation (1) is an equation representing how X-rays are attenuated.

While the left side of equation (1) is the pixel value of the measurement image, the content of the exponent function on the right side is the pixel value of the projection image obtained by inverting the sign. Therefore, by solving the equation (1) regarding the contents of the exponential function, a conversion equation for conversion from the measurement image to the projection image (logarithmic conversion) can be obtained. Thus, the logarithmic conversion formula is as shown in the following formula (2).

  Here, an example of the measurement image is shown in FIG. 9A, and a projection image obtained by logarithmically converting the measurement image of FIG. 9A is shown in FIG.

  Next, a metal extraction process and a metal artifact removal process performed by the image processing apparatus 10 will be described with reference to the flowchart of FIG. 10A.

  First, the image processing apparatus 10 reconstructs a projection image using a conventional general reconstruction algorithm, and generates reconstructed volume data shown in FIG. 11 (step S1). Note that the processing in step S1 is performed according to an image reconstruction processing program. Since the subject contains metal that hardly transmits X-rays, metal artifacts are generated on the reconstructed volume data shown in FIG.

  In step S2 following step S1, the image processing apparatus 10 performs threshold processing on the reconstructed volume data, sets voxels having a value equal to or greater than the threshold TH1 as metal candidate regions, and sets a predetermined value ( By giving a constant value (second constant value) smaller than the first constant value to voxels that do not belong to the metal candidate region, that is, by performing binarization processing, The binarized reconstructed volume data shown in FIG. 12 is obtained. In the above threshold processing, the threshold TH1 is set such that the metal region is included in the metal candidate region without being insufficient. In the binarized reconstruction volume data shown in FIG. 12, the metal candidate area is a white area.

  In step S3 following step S2, the image processing apparatus 10 acquires pixel values of projection images corresponding to all projection angles that are back-projected on the target voxel in the metal candidate region. Here, FIGS. 13A to 13C show examples of three projection images among the projection images corresponding to all projection angles that are back-projected onto the target voxel. For example, the value of the pixel P1 of the projection image shown in FIG. 13A, the value of the pixel P2 of the projection image shown in FIG. 13B, and the value of the pixel P3 of the projection image shown in FIG. This is a pixel value that is backprojected to the target voxel.

  In step S4 following step S3, the image processing apparatus 10 calculates the absolute value of the differential coefficient related to the projection angle of the pixel value of the projection image acquired in step S3 and the average of the pixel values of the projection image acquired in step S3 is equal to or greater than the threshold TH2. When the maximum value is equal to or less than the threshold value TH3, the target voxel is considered to be inside the metal region, and otherwise, considered to be outside the metal region.

  In step S5 following step S4, the image processing apparatus 10 confirms whether or not the processing in steps S3 and S4 has been performed on all voxels in the metal candidate region.

  If there is a voxel in the metal candidate region that has not been subjected to the processes in steps S3 and S4 (NO in step S5), the target voxel is a voxel in the metal candidate region and the processes in steps S3 and S4 are not yet performed. It changes to the voxel which is not given (step S6), and returns to step S3.

  On the other hand, if the processing in steps S3 and S4 has been performed on all voxels in the metal candidate region (YES in step S5), the image processing apparatus 10 completes the metal extraction (step S7). Thereby, the volume data from which the metal shown in FIG. 14 is extracted is obtained.

  In step S8 following step S7, the image processing apparatus 10 forward-projects the volume data shown in FIG. 14 to generate a metal forward projection image shown in FIG.

  In step S9 subsequent to step S8, the image processing apparatus 10 uses the metal forward projection image obtained in step S8 to generate a metal projection image from the original observed projection image (projection image used in step S3). And the pixel value of the removed area is interpolated from the surroundings. As a result, a projection image obtained by removing the metal projection image shown in FIG. 16 is obtained.

  In step S10 subsequent to step S9, the image processing apparatus 10 applies the metal forward projection image shown in FIG. 15 obtained in step S8 to the projection image obtained by removing the metal projection image shown in FIG. 16 obtained in step S9. Is added. Thereby, the corrected projection image shown in FIG. 17A is obtained.

  In step S11 following step S10, the image processing apparatus 10 confirms whether the processes in steps S8 to S10 have been performed on all projection angles.

  If there remains a projection angle that has not been subjected to the processes of steps S8 to S10 (NO in step S11), the projection angle is changed to a projection angle that has not been subjected to the processes of steps S8 to S10 (step S12). Return to S8.

  On the other hand, if the processes in steps S8 to S10 have been performed for all projection angles (YES in step S11), the image processing apparatus 10 reconstructs the corrected projected image obtained in step S10. As a result, the reconstructed volume data with the metal artifact removed shown in FIG. 18 is obtained.

  By performing the processing of steps S1 to S7 described above, metal can be extracted on the volume data without binarizing the projection image as in Patent Document 1. Therefore, even if it is a case where many scattered rays are contained like dental CT, a metal can be extracted precisely like FIG.

  Furthermore, metal artifacts can be removed by performing the processes in steps S8 to S13 described above.

  The present invention is not limited to dental CT, and can be applied to general CT. In particular, even with CT imaging using a detector equipped with a collimator, the scattered radiation cannot be completely blocked. Therefore, by applying the present invention, metal extraction and metal artifact removal can be performed more accurately. Will be able to. Furthermore, by applying the present invention, if the detector does not need to be provided with a collimator, the cost can be reduced.

  The embodiment of the present invention has been described above, but the embodiment can be variously modified within the scope of the gist of the present invention. Hereinafter, some modified examples will be described.

<Modifications of Steps S1, S2, S5, and S6>
(I) The normal reconstruction performed in step S1 and the binarization performed in step S2 may not be performed, and steps S3 and S4 may be performed for all voxels constituting the volume data.
(II) A region in which the normal reconstruction performed in step S1 and the binarization performed in step S2 are not performed and the metal region is included without a shortage (for example, a region where a dentition and an implant may be placed) ) May be determined in advance, and steps S3 and S4 may be performed on voxels in a predetermined area.

<Modification of Step S3>
(III) A measurement image may be used instead of the projection image. Note that both a projection image and a measurement image may be used.
(IV) Image pixel values corresponding to some projection angles may be acquired instead of image pixel values corresponding to all projection angles.
(V) As a storage format of the acquired pixel value, for example, there is a format in which the target voxel has a stack or a queue and the acquired pixel value is stored in the stack or the queue.

<Modification of Step S4>
(VI) Instead of the pixel value of the projection image, a pixel value obtained by normalizing the average or maximum value of the projection image to 1 may be used.
(VII) The differential coefficient may be calculated from the normalized pixel value of the projection image.
(VIII) The differential coefficient may be calculated as a difference from the projection image one sheet ahead, or may be calculated as a difference from the projection image two sheets or more ahead.
(IX) When the average pixel value of the projection image acquired in step S3 is equal to or greater than the threshold value TH2, the target voxel may be regarded as being in the metal region.
(X) When the maximum value of the absolute value of the differential coefficient related to the projection angle of the pixel value of the projection image acquired in step S3 is equal to or less than the threshold value TH3, the target voxel may be regarded as being in the metal region.
(XI) Instead of the condition that the average of the pixel values of the projection image acquired in step S3 is not less than the threshold value TH2, the number of times that the pixel value of the projection image acquired in step S3 exceeds the threshold value TH4 is not less than the threshold value TH5. Conditions may be used.
(XII) The projection angle of the pixel value of the projection image acquired in step S3 instead of the condition that the maximum value of the absolute value of the differential coefficient related to the projection angle of the pixel value of the projection image acquired in step S3 is not more than the threshold value TH3. A condition that the number of times that the absolute value of the differential coefficient with respect to the threshold value TH6 exceeds the threshold value TH7 may be used.
(XIII) The standard deviation of the pixel value of the projection image acquired in step S3 instead of the condition that the maximum value of the absolute value of the differential coefficient related to the projection angle of the pixel value of the projection image acquired in step S3 is not more than the threshold value TH3. Alternatively, a condition that is less than or equal to the threshold value TH8 may be used.
(XIV) When the difference between the maximum value and the minimum value of the pixel values of the projection image acquired in step S3 is equal to or less than the threshold value TH9, the target voxel may be regarded as being in the metal region.
(XV) A measurement image may be used instead of the projection image. In this case, the “condition that the average pixel value of the projection image is equal to or greater than the threshold value TH2” is replaced with the “condition that the average pixel value of the measurement image is equal to or less than the threshold value TH10”. Similarly, “the condition that the number of times that the pixel value of the projection image exceeds the threshold value TH4 is equal to or greater than the threshold value TH5” is replaced with “the condition that the number of times that the pixel value of the measurement image is less than the threshold value TH11 is equal to or greater than the threshold value TH12”. Similarly, “the condition that the maximum value of the absolute value of the differential coefficient related to the projection angle of the pixel value of the projection image is equal to or less than the threshold value TH3” is “the maximum value of the absolute value of the differential coefficient related to the projection angle of the pixel value of the measurement image” Is replaced by the condition that is less than or equal to the threshold TH13. Similarly, “the condition that the number of times that the absolute value of the differential coefficient related to the projection angle of the pixel value of the projection image exceeds the threshold TH6 is equal to or less than the threshold TH7” is “the absolute value of the differential coefficient related to the projection angle of the pixel value of the measurement image” Is replaced by the condition that the number of times that exceeds the threshold value TH14 is equal to or less than the threshold value TH15. Similarly, the “condition that the standard deviation of the pixel value of the projection image is equal to or less than the threshold value TH8” is replaced with the “condition that the standard deviation of the pixel value of the measurement image is equal to or less than the threshold value TH16”. Similarly, “the condition that the difference between the maximum value and the minimum value of the pixel value of the projection image is equal to or less than the threshold value TH8” is “the condition that the difference between the maximum value and the minimum value of the pixel value of the measurement image is equal to or less than the threshold value TH17”. Is replaced. Note that both a projection image and a measurement image may be used.
(XVI) Other than the above, “feature amount obtained from pixel values of a plurality of projection images” or “feature amount obtained from pixel values of a plurality of measurement images” is used to determine whether the target voxel is a metal region or not. May be.

<Modification of metal artifact removal processing>
The image processing apparatus 10 may perform the metal extraction process and the metal artifact removal process shown in the flowchart of FIG. 10B instead of the metal extraction process and the metal artifact removal process shown in the flowchart of FIG. 10A.

  The metal extraction process (steps S1 to S7) in the flowchart of FIG. 10B is the same as the metal extraction process in the flowchart of FIG. 10A. The metal artifact removal process of the flowchart of FIG. 10B is obtained by replacing steps S10 to S13 of the metal extraction process of the flowchart of FIG. 10A with steps S10 'to S13'.

  Hereinafter, the metal artifact removal process of the flowchart of FIG. 10B will be described. In step S8, the image processing apparatus 10 forward-projects the volume data shown in FIG. 14 to generate a metal forward-projected image shown in FIG.

  In step S9 subsequent to step S8, the image processing apparatus 10 uses the metal forward projection image obtained in step S8 to generate a metal projection image from the original observed projection image (projection image used in step S3). And the pixel value of the removed area is interpolated from the surroundings. As a result, a projection image obtained by removing the metal projection image shown in FIG. 16 is obtained.

  In step S10 'following step S9, the image processing apparatus 10 confirms whether the processes in steps S8 to S9 have been performed on all projection angles.

  If there remains a projection angle that has not been subjected to the processes of steps S8 to S9 (NO in step S10 ′), the projection angle is changed to a projection angle that has not been subjected to the processes of steps S8 to S9 (step S11 ′). Return to step S8.

  On the other hand, if the processing of steps S8 to S9 has been performed for all projection angles (YES in step S10 ′), the image processing apparatus 10 displays the metal projection image shown in FIG. 16 obtained in step S9. The removed projection image is reconstructed as it is (step S12 ′). Thereby, the reconstructed volume data from which the metal shown in FIG. 17B is removed is obtained. In the reconstructed volume data with the metal removed as shown in FIG. 17B, the metal region becomes an image like a soft tissue.

  In step S13 'following step S12', the image processing apparatus 10 adds the volume data shown in FIG. 14 to the reconstructed volume data obtained by removing the metal shown in FIG. 17B obtained in step S12 '. As a result, the reconstructed volume data with the metal artifact removed shown in FIG. 18 is obtained.

  Note that the modification to the flowchart of FIG. 10A described above can also be applied to the flowchart of FIG. 10B.

DESCRIPTION OF SYMBOLS 1 Body part of arm type X-ray imaging apparatus according to one embodiment of the present invention 2 Base 3 Lower pole 4 Upper pole 5 Fixed arm 6 Turning arm 7 Head holding part 8 X-ray irradiation part 8A X-ray focus 8B X-ray aperture 9 X-ray detection unit 10 Image processing device 101 CPU
102 ROM
103 RAM
104 Communication interface 105 VRAM
106 Display 107 HDD
DESCRIPTION OF SYMBOLS 108 Input part 201 Virtual dental arch 203 Patient dental arch 206 The pivot axis center of a turning arm 207 Image reconstruction range W X-ray beam width on an X-ray detection part

Claims (5)

An acquisition unit that pays attention to one of the voxels constituting the volume data and acquires pixel values of a projection image or measurement image that is back-projected on the target voxel for at least some projection angles;
Said calculated feature amounts of a plurality of the pixel values obtained by the obtaining unit, the identification unit identifies whether the voxel of interest based on the feature amount is outside the area of the metal that it is within the region of the metal,
With
The target voxel is changed between at least some of the voxels constituting the volume data, and the acquisition process by the acquisition unit and the identification process by the identification unit are performed for each of the changed target voxels ,
The acquisition unit acquires the pixel values that are not binarized,
The identification unit, without performing the binarization of the plurality of pixel values obtained by the obtaining unit, an image processing apparatus according to claim Rukoto determined the feature quantity. The feature amount is that, when the target voxel is in a metal region, the pixels that are not binarized in the projection image back-projected to the target voxel are outside the metal region at all projection angles. tend to have higher values than the pixel which is not binarized projected images backprojected voxels and pixels which are not binarized backprojection as measured image on the voxel of interest is outside the region of the metal The image processing apparatus according to claim 1, further comprising: a first feature amount that reflects at least one of tendencies having a lower value than a non-binarized pixel of a measurement image back-projected to a voxel. . When the target voxel is in a metal region, the feature amount is a region in which a change in a pixel value that is not binarized in a projection image back-projected on the target voxel is a metal region at all projection angles. small trend compared to the variation in the pixel values that have not been binarized projected images backprojected voxels outside, and, of the pixel values that have not been binarized backprojection as measured image on the voxel of interest 2. The second feature amount that reflects at least one of a tendency that is smaller than a variation of a non-binarized pixel value of a measurement image back-projected on a voxel whose variation is outside the metal region. Alternatively, the image processing apparatus according to claim 2. An acquisition step of acquiring a pixel value of a projection image or a measurement image corresponding to at least a part of the projection angle, which is back-projected onto a target voxel that is one of the voxels constituting volume data;
Obtains a feature from a plurality of the pixel values obtained by the obtaining step, an identification step of identifying whether the voxel of interest is a region outside of the metal or is in the region of the metal based on the feature quantity,
With
The target voxel is changed between at least some of the voxels constituting the volume data, and the acquisition process by the acquisition step and the identification process by the identification step are performed for each of the changed target voxels ,
In the obtaining step, the pixel values that are not binarized are obtained,
In the identification step, without performing the binarization of the plurality of pixel values obtained by the obtaining step, image processing method comprising Rukoto determined the feature quantity. An X-ray irradiation unit that irradiates the subject with X-rays;
An X-ray detector that detects X-rays transmitted through the subject;
An image generation unit that generates a projection image or a measurement image from the detection result of the X-ray detection unit;
The image processing apparatus according to any one of claims 1 to 3, wherein image processing is performed on the projection image or the measurement image;
An X-ray imaging apparatus comprising: