JP6161003B2 - Program, reconstruction apparatus and tomography apparatus - Google Patents

Description

  The present invention relates to a non-CT type tomography apparatus such as a computed tomography apparatus (CT (Computed Tomography) apparatus) and a tomosynthesis apparatus (also referred to as a laminograph).

  In a conventional CT apparatus, a so-called FBP (Filtered Back Projection) method is often used as a method for reconstructing a cross-sectional image from projection data obtained by scanning.

  FIG. 4 shows a basic flowchart of FBP method reconfiguration of a conventional RR CT apparatus. In the RR-type CT apparatus (a method in which scanning is performed by rotating an object or an integrated rotation of a radiation source and a radiation detector), one view of projection data is obtained for each rotation position. This projection data is subjected to so-called | ω | filtering in a direction orthogonal to the direction of the projected axis of rotation for each view (S21), and then a reconstruction area set on a desired section of the subject ( (Multiple grid points) are back-projected and added (S22). By performing this operation on the 360 ° view, a cross-sectional image is reconstructed in the reconstruction area.

  In the FBP method, enlargement reconstruction processing (referred to as zooming) is often performed. In the magnified reconstruction, a good cross-sectional image can be obtained by setting a reconstruction area in a part of the subject and reconstructing only this area.

  On the other hand, as a reconstruction method, a successive approximation method may be used for the purpose of improving image quality.

  FIG. 5 is a basic flowchart of a conventional general successive approximation method reconstruction. In general, a CT apparatus can obtain projection data of a subject along various paths by scanning.

  First, an initial reconstruction image is set for a reconstruction area including the entire subject set on a desired cross section of the subject (S31), and this reconstruction image is reprojected for each path (S32). The difference between the reprojection data and the measured projection data is obtained (S33), and the obtained difference data is backprojected (added) onto the reconstructed image and updated (S34). Then, by repeating reprojection or backprojection while changing the path, the difference gradually approaches 0, and a reconstructed image is obtained (reconstructed) as a cross-sectional image of the subject.

  This successive approximation method can be variously modified. For example, the successive approximation method may be repeated for each view instead of being repeated for each path.

  The conventional FBP method and the successive approximation method are described in, for example, Patent Document 1 and Patent Document 2, respectively.

  As described in Patent Document 1, the FBP method can be applied even when the subject protrudes from the scan region (tomographic field of view), and a good cross-sectional image can be obtained.

JP 2005-308633 A JP 2013-141608 A

  The successive approximation method can improve the image quality. However, the successive approximation method cannot perform a good reconstruction when the subject protrudes from the reconstruction area on the cross section.

  This is because the reprojection outside the reconstruction area is not reprojected, but the measurement projection data is also projected outside the reconstruction area, so that if the reconstructed image is correct, the reprojection data and the measurement projection data This is because the premise of the successive approximation method that the difference should be 0 is broken. In other words, when the subject protrudes from the reconstruction area, the difference is not zero but is natural.

  For this reason, first, in the successive approximation method, it is not possible to perform good reconstruction with respect to the enlargement reconstruction process (called zooming) performed by setting a reconstruction area in a part of the subject often performed in the FBP method. .

  Secondly, in the successive approximation method, for the reconstruction of the ROI portion of the ROI scan (also referred to as a protrusion scan) that fits the region of interest (ROI) of the subject in the scan region, the subject similarly protrudes from the reconstruction region. Therefore, good reconstruction cannot be performed.

  The present embodiment has been made in view of the above circumstances, and an object thereof is to provide a successive approximation method capable of good reconstruction even when the subject protrudes from the reconstruction region.

  In order to achieve the object, the invention according to claim 1 is a radiation source that emits radiation toward a subject, a radiation detector that detects radiation transmitted through the subject and outputs the radiation data as transmission data, and A cross-sectional image of the subject is reconstructed by processing the transmission data of a tomography apparatus having scanning means for relatively scanning the subject and the radiation so as to change the direction of transmission of the subject. A reconstruction apparatus program, wherein the transmission data is converted into measurement projection data for each view, which is a set of a plurality of radiation paths, and a reconstruction area set on a cross section of the subject. An initial reconstructed image is set, and the reconstructed image is reprojected in the view unit to create reprojection data, and a difference data is obtained by taking a difference between the reprojected data and the measured projection data. Creating, applying | ω | filtering to the difference data, back projecting the difference data subjected to the | ω | filtering onto the reconstructed image, and changing the view from the reprojection to the backprojection. By repeating the above, the gist is that the reconstructed image is updated to be a cross-sectional image of the subject.

  With this configuration, difference data between the reprojection data and the measurement projection data is obtained, and | ω | filtering and back projection are performed on the difference data, so that the same effect as the conventional FBP method is brought about and the subject is reconstructed Even if it protrudes from the region, it is possible to reconstruct a good successive approximation method.

  In order to achieve the above object, according to a second aspect of the present invention, in the program according to the first aspect, the scanning of the tomography apparatus may be performed by rotating the subject around the subject or the subject. Rotation of the radiation source and the radiation detector as a center, the view is a set of a plurality of radiation paths at one position of the rotation, and the | ω | filtering is projected onto the difference data The gist is to perform in a direction orthogonal to the direction of the axis of rotation.

  With this configuration, this program can be applied to an RR CT apparatus.

  In order to achieve the above object, according to a third aspect of the present invention, in the program according to the first aspect, the scanning of the tomography apparatus may be performed by rotating the subject around the subject or rotating the subject. Rotation of the radiation source and the radiation detector as a center, the view is a set of a plurality of radiation paths parallel to each other when viewed from the direction of the axis of rotation, and the | ω | filtering is performed on the difference data The gist is to perform in a direction orthogonal to the direction of the axis of rotation projected onto the screen.

  With this configuration, this program can be applied to an RR CT apparatus. Furthermore, since the radiation paths are parallel when viewed from the direction of the axis of rotation, reprojection and backprojection calculations are facilitated.

  In order to achieve the above object, according to a fourth aspect of the present invention, in the program according to the first aspect, the scanning of the tomography apparatus is a stepwise rotation of the subject around the subject or the step. Stepwise rotation of the radiation source and the radiation detector centered on the subject, and the radiation source and the radiation for the subject in a direction orthogonal to the axis of rotation performed in a state where the rotation is stopped Relative linear movement of the detector, the view is a set of radiation paths parallel to each other in the direction of the axis of rotation, and the | ω | filtering is the projected onto the difference data The gist is to perform in a direction orthogonal to the direction of the axis of rotation.

  With this configuration, this program can be applied to a TR CT apparatus.

  In order to achieve the object, the invention according to claim 5 is a reconfiguring apparatus having the program according to any one of claims 1 to 4.

  With this configuration, it is possible to provide a reconstruction device that can reconstruct a good successive approximation method even if the subject protrudes from the reconstruction region.

  In order to achieve the above object, a sixth aspect of the present invention is a tomographic apparatus having the reconstruction device according to the fifth aspect.

  With this configuration, it is possible to provide a tomographic apparatus that can reconstruct a good successive approximation method even if the subject protrudes from the reconstruction area.

  According to the present embodiment, it is possible to provide a successive approximation method that allows good reconstruction even when the subject protrudes from the reconstruction region.

It is a schematic diagram which shows CT apparatus of 1st embodiment of this invention. It is a flowchart of the reconstruction process in 1st embodiment of this invention. It is an example of a | ω | filter in the first embodiment of the present invention. It is a basic flowchart of FBP method reconfiguration of a conventional RR CT apparatus. It is a basic flowchart of the conventional general successive approximation method reconstruction.

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

(Configuration of the first embodiment of the present invention)
FIG. 1 is a schematic diagram showing a CT apparatus according to the first embodiment.

  The CT apparatus 1 relatively connects the radiation source 2, the radiation detector 5 that detects the radiation 4 transmitted through the subject 3 and outputs it as transmission data, and the subject 3 and the radiation 4 to change the direction of transmission. It comprises scanning means 6 for scanning, and a reconstruction device 7 for taking transmission data and reconstructing a cross-sectional image of the subject.

  Specifically, the scanning unit 6 includes a mechanism and a control unit that rotate the subject 3 with respect to the rotation axis RA around the subject 3 or rotate the radiation source 2 and the radiation detector 5.

  The radiation detector 5 detects the cone-shaped radiation 4 two-dimensionally on the detection surface 5a.

  The reconstruction device 7 is a normal computer and has a CPU, a storage unit, a display unit, and an input / output unit.

  In FIG. 1, a scan area A is an area covered with the radiation 4 during scanning, and can be reconstructed without difficulty.

  The reconstruction area R is an area in which reconstruction calculation is performed, and is an area set on a cross section orthogonal to the rotation axis RA. The reconstruction area R can be freely set as one or a plurality of parallel planes (even as a solid). be able to. The reconstruction area R may be set so as to cover the entire subject 3, or may be set so as to include a part of the subject 3 as shown in FIG. 1. In addition, the reconstruction area R can be set so as to protrude from the scan area A, but the image of the protruding area gradually deteriorates as the distance from the scan area increases.

(Operation of the first embodiment)
With reference to FIG.2 and FIG.3, the effect | action of 1st embodiment is demonstrated.

  The CT apparatus 1 of the first embodiment performs scanning in the same manner as a normal RR CT apparatus.

  Scanning is performed by capturing transmission data at predetermined angular intervals while relatively rotating the subject 3 once.

  Here, transmission data obtained at one position of rotation is transmission data of one view.

  The view means a set of radiation paths that do not intersect each other and cover the scan area A continuously.

  When the scan is completed and the transmission data of the view for one rotation is stored, the stored transmission data is used to perform reconstruction as follows.

  FIG. 2 is a flowchart of the reconstruction process in the first embodiment.

  First, in step S1, all stored transmission data is converted into measurement projection data and stored.

The measurement projection data P is an expression of transmission data I for each radiation path,
P = LOG (I0 / I) (1)
Obtained by logarithmic transformation. Here, I0 is air data (transmission data when there is no subject). The measurement projection data P is obtained for each view.

  The measurement projection data P obtained in this way is a value obtained by projecting the linear absorption coefficient of radiation along the radiation path (more precisely, line integration).

  Next, in step S2, an initial reconstruction image is set in the reconstruction region R set on the cross section of the subject 3.

  Here, the reconstruction area R is one or a plurality of parallel planes (solids), the reconstruction area R is represented by a plurality of lattice points, and the reconstruction image is a numerical value on the plurality of lattice points. Is represented as

  Next, the view is changed in step S3, and steps S4 to S8 are repeated for each view. For the view change, after one view for one scan is completed, one scan is repeated many times.

  In step S4, the reconstructed image is reprojected along the radiation path (more precisely, line integration) to obtain reprojection data.

  In step S5, the difference between the reprojection data and the measured projection data P is obtained. The difference is obtained between the same radiation paths and used as difference data.

  In step S6, | ω | (omega absolute value) filter is applied to the difference data. The | ω | filter is a filter in the frequency space and is applied in a direction orthogonal to the direction of the rotation axis RA projected on the difference data. The | ω | filter is the same as the filter used in the FBP method.

FIG. 3 shows an example of the | ω | filter. The | ω | filter F (ω) is basically a filter that multiplies a coefficient proportional to | ω | for each angular frequency ω, as shown in FIG. In FIG. 3, ω _N is the maximum angular frequency determined by the data pitch.

  The | ω | filter is not limited to FIG. 3A, but may be the one shown in FIG. 3B in which the high frequency region is lowered, or FIG. 3C in which the high frequency region is increased. This is used for performing high frequency cut and high frequency emphasis processing simultaneously with | ω | filtering.

  That is, the | ω | filter can be said to be a | ω | filter if it is proportional to | ω | in the low frequency region (around ω = 0 in FIG. 3).

  Returning to FIG. 2, in step S7, the difference data after the | ω | filter is back-projected (added) to the reconstructed image, and the reconstructed image is updated.

  Back projection is the same as the FBP method, and is a process of adding a difference value to a lattice point along the radiation path.

  In step S8, it is determined whether an end condition is satisfied. As an end condition, it may be determined whether the difference is equal to or less than a certain value, or may be determined based on whether the number of times of repeating the view for one scan has reached a predetermined value.

  If it is determined that the end condition is not satisfied (NO), the process proceeds to the next view and steps S4 to S8 are repeated.

  If it is determined in step S8 that the end condition is satisfied, the current reconstructed image is ended as a finally obtained cross-sectional image.

  The obtained cross-sectional image is a distribution image of the linear absorption coefficient.

(Effects of the first embodiment)
According to the first embodiment, difference data between reprojection data and measurement projection data is obtained for each view, and | ω | filtering and back projection are performed on the difference data, so that the same as in the conventional FBP method is performed. Even if the subject 3 protrudes from the reconstruction region R, it is possible to provide an iterative approximation method that can be reconstructed satisfactorily.

  As a result, even if the subject 3 protrudes from the reconstruction area R, a cross-sectional image with good image quality, which is a feature of the successive approximation method, can be obtained.

  Further, according to the first embodiment, it is possible to perform a good reconstruction with respect to the enlargement reconstruction process performed by setting the reconstruction region R in a part of the subject 3.

  In addition, according to the first embodiment, similarly to the reconfiguration of the ROI portion of the ROI scan in which the region of interest (ROI) of the subject 3 is stored in the scan region A, the same good reconfiguration can be performed.

(Modification of the first embodiment)
In addition, the present invention is not limited to the above-described embodiment, and various modifications can be made without departing from the scope of the invention.

(Modification 1)
In the first embodiment, the view is a view at one position of rotation and is a collection of cone-shaped radiation paths from the radiation focus F, but this view is converted to a different view and Repeated calculations can also be performed.

  As described above, the view means a set of radiation paths that do not intersect each other and cover the scan area continuously.

  For example, a set of paths that are parallel when viewed from the direction of the rotation axis RA and are fan-shaped when viewed from a direction orthogonal to the rotation axis RA can be used as the view. In this case, so-called fan-para conversion is performed to obtain projection data for each new view. Also, | ω | filtering is performed in a direction orthogonal to the direction of the rotation axis RA projected along the radiation path on the difference data in the new view.

  By adopting this view, since the radiation path is parallel when viewed from the direction of the rotation axis RA, reprojection and backprojection calculations are facilitated.

(Modification 2)
The CT apparatus 1 according to the first embodiment is a normal RR CT apparatus, but may be a so-called TR (translate rotation) CT apparatus.

  The TR type CT apparatus differs in the scanning of the scanning means 6. Scanning of the TR-type CT apparatus stops stepwise rotation of the subject 3 around the subject, or stepwise rotation of the radiation source 2 and the radiation detector 5 around the subject 3, and stops rotation. The relative linear movement (translation) of the radiation source 2 and the radiation detector 5 with respect to the subject 3 in the direction perpendicular to the rotation axis RA and intersecting with the radiation 4 is performed. Here, “relative” means that the subject 3 may move, the radiation source 2 and the radiation detector 5 may move, and in addition to the movement of the subject 3, the radiation source 2 and the radiation detection. It means that the vessel 5 may also move.

  In this case, the view is a set of a plurality of fan-shaped radiation paths that are parallel to the rotation axis RA and orthogonal to the rotation axis RA (translation direction). Projection data for each view can be obtained by rearranging projection data for each view.

  Also, | ω | filtering is performed in a direction orthogonal to the direction of the rotation axis RA projected along the radiation path on the difference data in this view.

(Modification 3)
The reconstruction apparatus 7 of the first embodiment can be applied to RR type helical scans, as well as any other type of CT apparatus (eg, TR type, SR type, fifth generation type, etc.). Applicable.

  Furthermore, the reconstruction apparatus 7 of the first embodiment can be applied to a non-CT type tomography apparatus. For example, the present invention can be applied to a so-called tilted CT apparatus in which the rotation axis is tilted from FIG. 1 and a tomography apparatus called a so-called tomosynthesis apparatus (also referred to as a laminograph).

  Note that the direction in which the | ω | filter is applied in the difference data of one view differs depending on the scanning method of the tomography apparatus, but is generally applied in the “direction in which the transmission image is shaken by scanning”.

  For example, in the RR CT apparatus, the direction orthogonal to the projected rotation axis is the direction in which the transmitted image is shaken, and in the TR CT apparatus, the direction in which the translation direction is projected (the direction orthogonal to the projected rotation axis). Is the direction in which the transmitted image shakes. Further, in the helical scan of the RR CT apparatus, the direction in which the tangent of the helical trajectory is projected is the direction in which the transmission image is shaken, and this can be approximated in the direction perpendicular to the projected rotation axis. In addition, in the case of a tomosynthesis apparatus that performs only linear movement, the direction in which the direction of linear movement is projected is the direction in which the transmission image is shaken.

(Modification 4)
In the first embodiment, only the basic part is described in the reconstruction process of the reconstruction device 7, but various techniques used in normal successive approximation can be added.

  For example, by changing the back projection weight of the difference data for each path according to the noise of the projection data, it is possible to perform reconstruction with reduced noise.

  Further, in the loop for changing the view, the convergence of the successive approximation can be accelerated by devising the change order different from the view collection order.

DESCRIPTION OF SYMBOLS 1 ... CT apparatus, 2 ... Radiation source, 3 ... Subject, 4 ... Radiation, 5 ... Radiation detector, 5a ... Detection surface, 6 ... Scanning means, 7 ... Reconstruction device,
F: Radiation focus, RA: Rotation axis, A: Scan area, R: Reconstruction area

Claims (6)

A radiation source that emits radiation toward the subject; a radiation detector that detects the radiation transmitted through the subject and outputs it as transmission data; and the subject and the subject to change the direction of the transmission with respect to the subject A program for a reconstruction device that processes the transmission data of a tomography apparatus having a scanning unit that relatively scans radiation, and reconstructs a cross-sectional image of the subject,
The program converts the transmission data into measurement projection data for each view, which is a set of a plurality of radiation paths, and further sets an initial reconstruction image in a reconstruction area set on the cross section of the subject. , Re-projecting the reconstructed image in the view unit to create re-projection data, taking the difference between the re-projection data and the measurement projection data to create difference data, and | ω | Applying filtering, backprojecting the | ω | filtered difference data onto the reconstructed image, and updating the reconstructed image by repeating the reprojection or backprojection while changing the view. And a cross-sectional image of the subject. The program according to claim 1, wherein the scanning of the tomography apparatus is rotation of the subject around the subject or rotation of the radiation source and the radiation detector around the subject. ,
The view is a collection of multiple radiation paths at one position of the rotation;
The | ω | filtering is performed in a direction orthogonal to the direction of the axis of rotation projected on the difference data. The program according to claim 1, wherein the scanning of the tomography apparatus is rotation of the subject around the subject or rotation of the radiation source and the radiation detector around the subject. ,
The view is a set of radiation paths parallel to each other when viewed from the direction of the axis of rotation;
The | ω | filtering is performed in a direction orthogonal to the direction of the axis of rotation projected on the difference data. 2. The program according to claim 1, wherein the scanning of the tomography apparatus includes stepwise rotation of the subject centered on the subject or the radiation source and the radiation detector centered on the subject. Stepwise rotation, and relative linear movement of the radiation source and the radiation detector relative to the subject in a direction orthogonal to the axis of rotation performed in a state where the rotation is stopped,
The view is a set of radiation paths parallel to each other when viewed from the direction of the axis of rotation;
The | ω | filtering is performed in a direction orthogonal to the direction of the axis of rotation projected on the difference data.   A reconfiguration apparatus having the program according to claim 1.   A tomography apparatus having the reconstruction device according to claim 5.

Images