JPH10243941A - Processor for re-constituting image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To realize the correct re-constitution of an image photographed through the use of a cone beam by converting projecting data obtained by the cone beam into a projecting data of an extended variation parallel beam and correcting a distortion in a Z-axis direction so as to execute reverse projection to a voxcel. SOLUTION: A testee is irradiated with the conical X-ray cone beam which is irradiated from an X-ray source 3 and an X-ray transmitted through it is detected by a second-dimensional X-ray detector 5 which is obtained by arraying first-dimensional X-ray detectors in plural lines. Then, the tomographic image of an object is re-constituted by an image re-constitution processing part 12 based on detection data obtained by rotating the X-ray source 3 and the second- dimensional X-ray detector 5 around the testee. In the image re-constitution, the rotary axis direction component of the beam is neglected, detection data of the extended variation parallel beam where transmissive angles are parallel with the object is collected at every transmissive angle so as to be processed and the tomographic image of the object is re-constituted based on detection data of the extended variation parallel beam where the Z-axis direction distortion is corrected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of irradiating an object with a cone-shaped X-ray cone beam emitted from an X-ray source, and detecting X-rays transmitted through the object by a plurality of X-ray detecting elements. A one-dimensional X-ray detector arranged in a row is detected by a two-dimensional X-ray detector arranged in a plurality of rows, and a detection is obtained by rotating an X-ray source and a two-dimensional X-ray detector around an object. The present invention relates to an image reconstruction processing device that reconstructs a tomographic image of an object based on data.

[0002]

2. Description of the Related Art Conventional X-ray CT (computed tomography)
As shown in FIG. 25, an apparatus using a fan beam in which an X-ray beam is emitted from an X-ray source 101 in a fan shape (fan shape) is known. Such an X-ray CT apparatus irradiates an X-ray beam emitted from the X-ray source 101 onto a subject, and transmits X-rays (transmission X-ray intensity) transmitted through the subject.
Is detected by an X-ray detector 102 in which a plurality of X-ray detection elements are arranged in a row in a fan-like manner with about 1000 channels, and data is collected. The X-ray source 101 and the X-ray detector 102 are rotated around the subject. Data is collected about 1000 times during one rotation while making one rotation. (One data collection is called one view.)
), And reconstructs an X-ray transmission image (tomographic image) of the subject based on the collected data. FOV103
Indicates an effective field of view. The image reconstruction equation when using this fan beam is calculated by (Equation 1).

[0003]

(Equation 1)

As can be seen from the equation (1), in the reconstruction of the fan beam, it is necessary to multiply the data obtained from the X-ray detector by a weight depending on the position of the pixel to be reconstructed and backproject. Complex processing. That is, as shown in FIG. 26, pixels constituting an image to be reconstructed for the effective field of view FOV 103 are set, and data obtained in each channel of the X-ray detector 102 is weighted with a focus (X The radiation point of the X-ray beam of the source 101) is multiplied by the reciprocal of the square of the pixel-to-pixel distance FpixelD (X) and backprojected to the corresponding pixel. In addition, Fpix
elD is Focus-Pixel-Distance. There is also a method of direct back projection, but in this case, polar coordinate conversion is required, and the calculation becomes complicated.

Therefore, an X-ray CT using a fan beam
For the device, at present two types of image reconstruction methods have been devised. One method is called the fan-para conversion method, which is shown in FIGS. 27 (a) and 27 (b).
As shown in (2), the projection data obtained from the X-ray detector by the fan beam is rearranged and interpolated to create parallel beam projection data (including conversion), and the obtained data is converted into a conventional parallel beam. X-ray CT using
This is a method of back projection as performed in the device. While data conversion calculations and interpolation processing are required, at the time of back projection, different weighting processing is not required for each reconstructed pixel in the case of a fan beam, and one data (parallel beam projection data) is transferred to the beam path (parallel beam projection data). The processing can be simplified since it is only necessary to back-project all pixels on a straight line connecting the radiation point and the X-ray detector channel when a parallel beam is formed.

That is, the processing of the fan beam is performed as shown in FIG.
As shown in FIG. 29A and FIG. 29, an arbitrary coordinate system xy is defined as projection data [Rf] (ψ, φ). This [Rf] (ψ, φ) converts the subject f (x, y) to (ψ, φ).
) Indicates the line integral (assumed). First, this projection data [Rf] (ψ, φ) is multiplied by cosψ. Next, the result and a filter [1 / (sinψ squared)
)] And convolution. Next, the result is weighted by [d / (R square)] and back-projected according to the X-ray path. Note that R has a numerical value that fluctuates depending on the position of back projection. The image reconstruction equation (
Equation 2) is

(Equation 2)

[0007]

In the parallel beam processing, as shown in FIG. 28B, an arbitrary coordinate system xy is defined as projection data [Rf] (s, θ). This [Rf] (s, θ
) Indicates that the subject (x, y) is linearly integrated in the (s, θ) direction (assumed).

First, the projection data [Rf] (s, θ)
And the filter [1 / (square of s)] are repeatedly performed over 180 ° or 360 °. Next, the result is back-projected according to the X-ray path. The image reconstruction equation obtained above is

(Equation 3)

## EQU1 ##

Another method is a fan beam reconstruction method using a centering axis.
No. 5-99240. On the other hand, FIG.
As shown in the figure, a detector element is arranged on a cylindrical surface such that a cone beam in which an X-ray beam is radiated from an X-ray source 201 in a conical shape and N rows of detector rows for fan beams are stacked in the Z-axis direction. X-ray C for capturing an X-ray fluoroscopic image using the two-dimensional X-ray detector 202 in which (M channels × N columns) are arranged
T devices have been devised.

X-ray C using such a cone beam
Typical cone beam reconstruction in T-device (Feldkamp
Reconstruction) is disclosed in the following document. "Practical cone-beam algorithm" LAFeldkamp, LCDavis, and JWKress J. Opt. Soc. Am. A / Vol.1, No.6, pp.612-619 / June 19
84 This is a mathematically exact reconstruction of the fan beam (2
Reconstruction algorithm [Filtered-Backproje
ction (filtered backprojection method)] in the Z-axis direction is an approximate three-dimensional reconstruction algorithm.

This cone beam reconstruction method targets a conventional scan using a cone beam, and comprises the following steps. In this cone beam, voxels as three-dimensional pixels are used as shown in FIG. 31 instead of pixels as two-dimensional pixels. 1. Weighting of projection data The projection data is multiplied by a term dependent on the Z coordinate and a cos term. 2. Convolution operation The convolution operation of the data obtained by the process 1 and the same reconstruction function as that of the fan beam is performed. 3. BackProjection (Backprojection) The data obtained by the process 2 is backprojected onto the path (from the focal point to the channel of the detector) through which the X-ray has passed. That is, the point at which the straight line passing through the voxel backprojected from the focal point intersects the detector plane is calculated, and backprojected data is created by interpolation or the like from the data of the two processes around that point, and it is expressed as FvoxelD (X ) Is weighted by the reciprocal of the square of (2) and backprojected.
This back projection is performed over 360 ° (one rotation).

When expressed by a formula similar to the fan beam reconstruction formula, the following is obtained. Note that FvoxelD (X) = Focus-Voxel-Di
stance represents the line segment between the focal point and the voxel as the cone beam Mi.
The length of the line segment projected on the dplane plane (plane parallel to).

(Equation 4)

The three-dimensional reconstruction equation (cone beam reconstruction equation) (Equation 4) is very similar to the fan beam reconstruction on the equation, but explains that the back projection method of Data-Back is greatly different. I do.

In the two-dimensional fan beam reconstruction, as shown in FIG. 32, it is necessary to back-project all pixels (pixels) in the reconstruction plane from data of detectors arranged one-dimensionally. On the other hand, in the reconstruction of a cone beam (Feldkamp), as shown in FIG. 33, a point where a straight line connecting a focal point and a voxel to be reconstructed intersects with a two-dimensional X-ray detector plane is determined, and the intersection is determined. Is back-projected to all voxels located on that line. Therefore, when a certain surface is reconstructed by cone beam reconstruction, such as fan beam reconstruction, data of a specific detector row and channel is back-projected to only some voxels of the reconstruction surface. Since it is necessary to select the data (detector rows and detector channels) to be back-projected for each voxel, the three-dimensional positional relationship between the reconstructed voxel, the focused straight line, and the X-ray detector surface must be determined. Becomes important. Moreover, considering a detector row having the same Z coordinate and a straight line focusing on the detector element, a voxel passing through the straight line on a certain plane (reconstructed plane) is detected around the focus. Similar figure of the surface (in the case of cylindrical detector,
The calculation of this positional relationship is very complicated because they are arranged on concentric circles).

[0017]

As described above, in a conventional X-ray CT apparatus, when a cone beam and a two-dimensional X-ray detector are used, the calculation is complicated and enormous in image reconstruction. However, there is a problem that the processing time is too long for a computer or the like which is generally used, and it cannot be realized. Therefore, the present invention provides an image reconstruction processing apparatus that can establish the application of the fan-para conversion method in a cone beam and realize an accurate reconstruction of an image photographed using the cone beam. Aim.

[0018]

The invention corresponding to claim 1 is:
An object is irradiated with a cone-shaped X-ray cone beam emitted from an X-ray source, and X-rays transmitted through the object are arranged in a plurality of one-dimensional X-ray detectors in which X-ray detection elements are arranged in one line. A tomographic image of the object is reconstructed based on detection data obtained by rotating the X-ray source and the two-dimensional X-ray detector around the object, detected by the arranged two-dimensional X-ray detector. In the image reconstruction processing device, based on detection data obtained by rotating the X-ray source and the two-dimensional X-ray detector by a predetermined angle, a beam from the X-ray source to the two-dimensional X-ray detector is converted into an X-ray source and a two-dimensional X-ray detector. Irregular parallel data conversion means for collecting and processing, for each transmission angle, detection data of an irregular parallel beam whose transmission angle to the object is parallel ignoring the rotation axis direction component of the two-dimensional X-ray detector; Anomalous parallel data collected and processed by the anomalous parallel data conversion means It is provided with a the parallel reconstruction means for reconstructing a tomographic image of the object based on the detection data over arm.

According to a second aspect of the present invention, an object is irradiated with a conical X-ray cone beam emitted from an X-ray source, and the X-rays transmitted through the object are arranged in a row of X-ray detecting elements. The detected one-dimensional X-ray detector is detected by a two-dimensional X-ray detector arranged in a plurality of rows, and based on detection data obtained by rotating the X-ray source and the two-dimensional X-ray detector around the object. In an image reconstruction processing apparatus for reconstructing a tomographic image of an object, a two-dimensional X-ray source from the X-ray source is obtained from detection data obtained by rotating the X-ray source and the two-dimensional X-ray detector by a predetermined angle. The beam to the X-ray detector collects the detection data of the anomalous parallel beam whose transmission angle to the target is parallel ignoring the rotation axis direction components of the X-ray source and the two-dimensional X-ray detector for each transmission angle. Parallel data conversion means for processing by means of An irregular parallel data correction means for correcting distortion in the rotation axis direction with respect to the detection data of the irregular parallel beam collected and processed by the irregular parallel beam, and an object based on the irregular parallel beam detection data corrected by the irregular parallel data correction means. And a parallel reconstruction means for reconstructing a tomographic image.

According to a third aspect of the present invention, in the second aspect of the present invention, the anomalous parallel data correction means determines a detection position of the detection data collected and processed by the anomalous parallel data conversion means according to a transmission angle of the parallel beam. Re-sampling processing for rearranging the position in the rotation axis direction. According to a fourth aspect of the present invention, in the second aspect of the present invention, the irregular parallel data correcting means sets a row of voxels as spatial pixels to be back-projected in parallel to a predetermined fixed axis. Assuming the trajectory of the focal point of the anomalous parallel beam parallel to this voxel row, detection data of the anomalous parallel beam collected and processed by the anomalous parallel data conversion means on a plane parallel to this trajectory is set in advance. A similar centering process of performing projection as a pre-process for correcting distortion in the direction of the rotation axis onto the centering plane is performed.

According to a fifth aspect of the present invention, there is provided the invention according to any one of the first to fourth aspects, wherein the X-ray source and the second
The detection data is obtained by a conventional scan in which the dimensional X-ray detector is rotated without moving around the object in the direction of the rotation axis. The invention according to claim 6 is the invention according to any one of claims 1 to 4, wherein
The detection data is obtained by helical scanning in which the radiation source and the two-dimensional X-ray detector are rotated while moving around the object in the direction of the rotation axis. The invention according to claim 7 is the invention according to any one of claims 1 to 6, wherein the correction of the artifact and the quality hardening for the tomographic image reconstructed by the parallel reconstruction means uses the central section theorem. It is to be corrected.

According to an eighth aspect of the present invention, in the first aspect of the present invention, the correction of the artifact and the hardening of the quality of the tomographic image reconstructed by the parallel reconstruction means is performed in two dimensions. The correction is performed as a two-dimensional image using the central section theorem.

[0023]

FIG. 1 to FIG. 1 show a first embodiment of the present invention (conventional scan).
This will be described with reference to FIG. The cone beam reconstruction formula used in the embodiment of the present invention is:

(Equation 5)

Equation (5) is an approximation obtained by expanding the parallel beam reconstruction equation of Equation (3) with respect to the apparent cone angle. (Equation 3) gives the cone beam reconstruction equation
The same correction term as in (Equation 6) is added, and the backprojection processing is replaced with a three-dimensional irregular parallel beam backprojection processing.

FIG. 1 is a block diagram showing a schematic configuration of an X-ray CT apparatus equipped with an image reconstruction processing apparatus to which the present invention is applied. A gantry 1 as a projection data measuring system is an X-ray generating a fan-shaped fan beam X-ray flux.
A radiation source 3 and a two-dimensional X-ray detector 5 of a two-dimensional array type are accommodated. The X-ray source 3 and the two-dimensional X-ray detector 5 are
The subject is mounted on the rotating ring 2 so as to face the subject placed on the slide top of the bed 6.

The two-dimensional X-ray detector 5 includes a plurality of
(1000 channels) detecting elements arranged one-dimensionally in one row are stacked in six rows (six segments) and mounted on the rotating ring 2. Where 1
One detecting element is defined as corresponding to one channel. X-rays from the X-ray source 3 are emitted to a subject via an X-ray filter 4. X-rays that have passed through the subject are detected by the two-dimensional X-ray detector 5 as electric signals.

The X-ray controller 8 supplies a trigger signal to the high voltage generator 7. This high voltage generator 7 applies a high voltage to the X-ray source 3 at the timing of receiving the trigger signal. Thus, X-rays are emitted from the X-ray source 3. Gantry bed controller 9
Controls the rotation of the rotating ring 2 of the gantry 1 and the sliding of the slide top plate of the bed 6 in synchronization. The system controller 10 as a control unit main body of the entire system controls the X-ray so that the X-ray source 3 can execute a so-called continuous rotation (for example, helical scan) in which the X-ray source 3 moves in a spiral trajectory as viewed from the subject. It controls the line controller 8 and the gantry bed controller 9.

More specifically, the rotary ring 2 rotates continuously at a constant angular velocity, the slide top of the bed 6 moves at a constant speed, and the X-ray source 3 continuously or intermittently at a constant angle. X-rays are regularly emitted. The two-dimensional X-ray detector 5
Is amplified by the data collection unit 11 for each channel and converted into a digital signal. The projection data output from the data collection unit 11 is transmitted to the reconstruction processing unit 1
2

The reconstruction processing unit 12 obtains back projection data reflecting the X-ray absorption rate for each voxel based on the projection data. In a continuous rotation type X-ray CT apparatus using a fan beam, the effective field of view (FOV, imaging area)
It becomes a cylindrical shape around the rotation center axis of the continuous rotation, and the reconstruction processing unit 12 includes a plurality of voxels (
Pixels arranged three-dimensionally) are defined, and backprojection data of each voxel is obtained from projection data from the two-dimensional X-ray detector 5. The three-dimensional image data or tomographic image data created based on the back projection data is sent to the display device 13 and
It is displayed visually as a two-dimensional image or tomographic image.

As shown in FIG. 2, the geometry of this X-ray CT apparatus is such that the number of detector rows M rows (X-ray detection element rows) The number of channels N channels The height of each row in the Z-axis direction Dmm, (Focus-center-Distance) FCD (Focus-Detector-Distance) Effective field diameter FOV (Field of View) Fan angle α Cone angle β ing.

The processing performed by the reconstruction processing unit 12 will be described below. In the first embodiment, a case of a conventional scan will be described, and in a second embodiment described later, a case of a helical scan will be described. (1) An irregular cone beam / parallel beam conversion is performed, distortion is corrected, and an irregular parallel beam back projection is performed. (2) Make the resampling plane parallel to the voxel row. (3) The voxel row is made parallel to a certain axis (centering similarity method described in a modified example).

(1) will be described. When a method similar to the fan-para conversion method is applied as the first processing step, as shown in FIG.
When observing the OV, the cone beam and the fan beam look the same, so that the conversion into the parallel beam in the channel direction can be performed by the fan-para conversion method. If the fan angle is α, the channel angle of the focal position Fs ＝ s =
(α / 2), channel angle of focal position Fe {e = −ψ}
s = − (α / 2). At this time, the focal angle φs of the focal position Fs = θ + (α / 2), and the focal angle φe of the focal position Fe = θ− (α / 2). Focus position Fs, that is,
As shown in FIG. 3B, the focal angle φs of the focal position Fs
= Θ + (α / 2) Channel angle ψs = (α / 2)
From the beam, the focal angle φi of the focal position Fi = θ + ψi
At the channel angle ψi at the focal position F
e, that is, the focal angle φe of the focal position Fe = θ− (α /
The beams up to the channel angle チ ャ ン ネ ル e = − (α / 2) in 2) are collected to form a parallel beam having a phase θ.

However, distortion occurs in the Z-axis direction due to a change in apparent cone angle. The plane P perpendicular to the parallel beam having the phase θ (FIG. 3B) is represented by a straight line.
Considering this perpendicular plane as the center section of the phase θ), the center section (section along the straight line) of the phase θ passing through the focal position Fs is as shown in FIG. 4 (a). 4 (b) at the focal position Fi, FIG. 4 (c) at the focal position Fc, and FIG. 4 (d) at the focal position Fe. Although FIG. 4 shows only a portion above the midplane of the X-ray, the same applies to a portion below the midplane.

Here, in FIGS. 4A to 4D, the distance between the focal point and the detector is constant at FDD, and the cone angle of the irradiated X-ray beam is also constant at β. The distance Lx between the focal position Fx and the central cross section of the phase θ is a function cos 、 x of the channel angle ψx according to the focal position, Ls, Li, L
It changes to c and Le. That is, from the focal position Fx to FO
Assuming that the distance to the center of V is FCD, referring to FIG. 3, the distance Ls = FCD · cosψs = FCD · cos (α /
2) = Distance Le. Distance Li = FCD · cosψi,
Distance Lc = FCD · cosψc. Here, ψc = 0
Therefore, the distance Lc = FCD.

Therefore, the apparent cone angle β for the same height H is different. That is, for the apparent cone angle βs at the focal position Fs,
e, the apparent cone angle βe, tanβs = (H / Ls) = (H / [FCD · cos (α /
2)]) = tanβe holds. Further, for the apparent cone angle βi at the focal position Fi, tanβi = (H / Li) = [H / (FCD · cosψi
)], Tan βc = (H / Lc) = (H / FCD) holds for the apparent cone angle βc at the focal position Fc.

The height of the two-dimensional X-ray detector at this time is the detection surface height Zs = FDD · tanβs = (FDD / FCD) · [H /
cos (α / 2)] = Ze, and the detection surface height Zi at the focal position Fi is Zi = FDD · tanβi = (FDD / FCD) · (H /
cosψi), detection surface height Zc at focal position Fc = FDD · tanβc =
(FDD / FCD) · H. As a result, assuming that a grid-like object having a height H exists in each central cross section,
In the cylindrical two-dimensional X-ray detector, the projection data having a spread in the Z-axis direction at the phase θ causes nonlinear distortion as shown in FIG.

When a planar two-dimensional X-ray detector is used, as shown in FIG. 6, even larger distortion is generated. That is, the minimum value of the distance between the focal position and the planar two-dimensional X-ray detector is defined as FDDo. Distance FD from focus position to detector where X-rays at angle ψi arrive
D (ψi) is FDD (ψi) = (FDDo / cosψi), and the detection surface height at the focal position Z (ψi) = FDD (ψi) · tanβi = (FDDo / cos ［i) · [H / (FCD) Cos @ i)] = (FDDo / FCD). [H / (cos @ i.cos @ i)].

There are the following three methods for equalizing the sampling pitch in the channel direction in the projection data of the parallel beam using the cone beam. 1. The pitch of the projection data acquisition (view pitch and pitch in the channel direction) is changed irregularly so that the sampling pitch becomes uniform. 2. When generating parallel beam projection data from cone beam projection data, the
(2 views × 1 channel, or 1 view × 2 channels, or 2 views × 2 channels) to calculate projection data of uniform pitch by interpolation. 3. The parallel data of the extracted phase θ is interpolated in the channel direction to calculate projection data at an equal pitch.

As a second processing step, a distortion correction processing (resampling processing) is performed on the uniform parallel beam obtained in the first processing step. Since the distortion of the parallel beam occurs only in the slice direction (Z-axis direction), resampling is performed so that the projection data becomes data at equal intervals in the slice direction. FIGS. 7, 8, 9 and 10 show examples of resampling the data area necessary for reconstructing the reconstructed image Z0 into six columns. Although the resampling pitch and the resampling number are optional, the smaller the resampling pitch and the larger the resampling number, the more effectively nonlinear distortion can be corrected. As a result, projection data using a triangular beam (an irregular parallel beam with distortion correction) obtained when the focal point moves on a straight line n (see FIG. 11) perpendicular to the phase θ is obtained.

In the third processing step, the resampling data obtained in the second processing step is multiplied by a correction term. 4th
As a processing step, a convolution process in the channel direction is performed. This is the same processing as the conventional two-dimensional reconstruction. As a final step, an irregular parallel beam back projection process is performed. As shown in FIG. 11, in the irregular parallel beam backprojection processing, assuming that the distance from the X-ray focal point Fc to the voxel array to be backprojected is Li, Zi = (FDD / Li) · Z0
The resampling data to be back-projected is determined for each voxel column, and back-projected as it is. Thereby, three-dimensional back projection can be performed.

FIG. 12 is a block diagram showing a configuration of a main part of the reconstruction processing section 12. As shown in FIG. The reconstruction processing unit 12 includes an image reconstruction control unit 21, an irregular parallel beam backprojection processing unit 2,
2 and a memory 23. The image reconstruction control unit 21 reads out data (collected data) from the memory 23 to create an irregular parallel beam, performs distortion correction by resampling processing, and performs predetermined processing such as multiplication with a correction term and convolution. Then, the processed data is output to the irregular parallel beam backprojection processing unit 22.
Further, the image reconstruction control unit 21 determines a resampling sequence to be backprojected to each voxel sequence, and outputs the processed data to the irregular parallel beam backprojection processing unit.

This irregular parallel beam backprojection processing section 22
Performs backprojection processing on a predetermined area according to the data input from the image reconstruction control unit 21 and the resampling sequence for backprojection, and stores the result in the data storage unit 23-1 of the memory 23. . At the time of back projection, data of a plurality of resampling sequences may be weighted and added. In the first embodiment, the method has been described in which the fan-parallel conversion in the channel direction and the distortion correction in the slice direction are performed separately. However, the present invention is not limited to this. 1 by view × 2 channel 4-point interpolation
In this case, the irregular parallel beam projection data with distortion correction may be obtained from the cone beam projection data. This makes it possible to increase the processing speed of the image reconstruction processing.

As a modification of the first embodiment, a centering similarity method will be described. In the backprojection processing of the irregular parallel beam as shown in FIG. 11, since the resampling data sequence and the voxel sequence are set in parallel, the resampling data to be backprojected to the voxel sequence is fixed resampling for each voxel sequence. I can pull it out of the line,
Although the selection of the resampling sequence is easy, the voxel sequence rotates according to the phase θ, so that it becomes difficult to determine the voxel to be back-projected.

Therefore, as shown in FIG. 13, even if the phase changes, the voxel row is parallel to a certain axis (fixed axis) such as the X-axis or the Y-axis (the voxel contains the FOV,
It is tilted according to the phase change, and it becomes 90 °
Trajectory of a straight line n '(virtual parallel beam X-ray source (focus)) parallel to the voxel row
Assuming a resampling plane parallel to), the cone beam projection data is subjected to an irregular cone beam / parallel beam conversion processing to obtain second irregular parallel beam projection data on the resampling plane (similar centering processing). This is projection data by the second irregular parallel beam obtained when the focal point F moves on the straight line n '. This is back-projected as shown in FIG. According to this method,
Both the selection of voxels for backprojection and the selection of resampling data are facilitated, and the backprojection processing speed can be further increased.

For example, as shown in FIG. 14, the assumed focal point (φ) of the parallel beam at the phase φ with respect to the reconstructed image A in the inclination s direction moves on the trajectory B. The projection data of the focal side VE (φ) and the two-dimensional X-ray detector side VS (φ) of the reconstructed image A are Vce (φ, s) and Vce (φ, s) on the detector surface C of the two-dimensional X-ray detector 5. It is detected as Vcs (φ, s). This means that if similarity centering processing (processing for assuming a resampling plane parallel to a straight line n 'parallel to the voxel row) is not performed, the projection data Vce (φ, s) and Vcs as shown in FIG. (φ, s) is not parallel to the detector row, and the projection data of the reconstructed image A is detected as a distorted quadrangle, which complicates the backprojection processing and causes a reduction in the backprojection processing speed. . on the other hand,
When the similar centering process is performed, FIG. 15 (b)
As shown in the figure, the projection data Vce (φ, s) and Vcs (
φ, s) are both parallel to the centering row, and the projection data of the reconstructed image on the centering plane D is exactly a parallelogram, so that the backprojection processing speed can be increased. A method described in Japanese Patent Application No. 8-1015, "Image Reconstruction Processing Apparatus" may be applied as a method for making the voxel row and the resampling data parallel to a certain axis.

In the first embodiment, non-linear distortion that occurs in projection data when a cone beam is converted into a parallel beam is corrected as a correction process by taking additional time. If the reconstruction control unit 21 determines data to be back-projected for each channel in consideration of the distortion and gives the result to the irregular parallel beam back-projection processing unit 22, the distortion correction processing can be performed simultaneously in the back-projection processing. Can be. According to this, the time required for the backprojection processing is estimated to be long, but the time for performing only the distortion correction processing can be omitted. Therefore, if the distortion correction processing can be simultaneously performed efficiently in the backprojection processing, In addition, the processing speed of the entire image reconstruction process can be increased. In addition,
In the first embodiment, the cone beam Midplane (Z = 0) shown in FIGS. 4 to 6, 8 to 11, and 13 is used.
Although the upper half image reconstruction process has been described, the image reconstruction process of the lower half of the cone beam Midplane can be similarly reconstructed by performing the contrast process.

A second embodiment of the present invention will be described with reference to FIGS. In the first embodiment described above, the case of the conventional scan has been described. In the second embodiment, the case of the helical scan will be described. The case of the helical scan of the second embodiment is basically the same as the case of the conventional scan of the first embodiment, except for the irregular cone beam / parallel beam conversion processing and distortion. Only the correction processing is performed, and the subsequent multiplication processing of the correction term, the convolution processing in the channel direction, and the irregular parallel beam backprojection processing are the same.

In the helical scan, the state viewed from the Z-axis direction is the same as the state shown in FIGS. 3A and 3B. Therefore, the cone beam / parallel beam conversion in the channel direction is performed in the first embodiment. This is the same processing as the two-dimensional fan-para conversion processing same as the form. However, while the focal point F moves on the same circumferential trajectory in the case of the conventional scan, it moves on a helical trajectory in the case of the helical scan, so that the focal point F required to obtain an irregular parallel beam of phase θ is obtained. At the angles Fs to (Fi) to Fc to Fe, the focal point F becomes (I) a reconstructed image plane (Z axis coordinate =
ZI) below. (II) Reconstructed image plane (Z axis coordinate =
Case above ZI). (III) When transitioning from the lower side to the upper side of the reconstructed image plane (Z axis coordinate = ZI). (IV) When transitioning from the upper side to the lower side of the reconstructed image plane (Z-axis coordinate = ZI). Are classified into four types. Incidentally, the Z coordinate Z (φi) of the focal position Fi at the focal angle φi is given by Z (φi) = Z0 + [(φi / 360) · HP]. Note that HP is a helical pitch.

First, the case (I) will be described. In this case, a reconstructed image is detected (projection data is collected) on the upper half surface side of the two-dimensional X-ray detector 5. Assuming that the Z-axis coordinate of the reconstructed image is Z = ZI (constant), as shown in FIG. 16 and FIG. 17, at the focal position Fs where the beam becomes the first beam, it is viewed from the Z-axis coordinate Zs = Z (φs). The relative Z-axis coordinate of the reconstructed image is dZ = ZI-Z
(φs) and the focal position Fe (Z
In the axis coordinates Ze = Z (φe)), dZ = ZI−Z (φe
). Further, dZ = ZI−Z (φi) at the focus position FI (Z-axis coordinate Zi = Z (φi)) of the intermediate beam. In the case of this helical scan, compared to the case of the conventional scan of the first embodiment described above,
It can be seen that the non-linear distortion of the projection data is larger, and the area of the position and range where the necessary projection data exists varies depending on the focal angle φ. Since the projection data near As is not actually obtained, the projection data is obtained from the calculation by extrapolation.

Here, as shown in FIG. 18, resampling processing as described in the first embodiment is performed on the projection data having a large nonlinear distortion shown in FIG. Is performed. The subsequent multiplication processing of the correction term, the convolution processing in the channel direction, and the irregular parallel beam backprojection processing are the same as those in the first embodiment, and thus description thereof is omitted here.

Next, the case (II) will be described. In this case, a reconstructed image is detected on the lower half surface side of the two-dimensional X-ray detector 5 (projection data is collected).
In the case of (II), the projection data is an area that is mirror-inverted in the Z-axis direction in the case of (I) described above. That is, A (As, Ai, Ac, Ae) and B (Bs, B
i, Bc, Be) are the projection data of the area where it is inverted. Therefore, the distortion correction processing is performed as in the case of (I).

Next, the case (III) will be described. In this case, the detection area moves from the upper half surface to the lower half surface of the two-dimensional X-ray detector. As shown in FIGS. 19, 20, and 21, the irregular cone-to-parallel beam conversion calculates the focal angle at which the focal point F crosses the reconstructed image as φi,
Fi), from the focal position Fs to the focal position Fi, As to Ai corresponding to voxels on the focal side of the image are on the upper side, and Bs to Bi corresponding to voxels on the two-dimensional X-ray detector side of the image are on the lower side. At the focal position Fi, the voxels of all the images become one projection data, and Ai and Bi become the same. Further, from the focal position Fi to the focal position Fe, Ai to Ae are inverted with Bi to Be being upper and Bi to Be being upper. Therefore, in the distortion correction processing in the case of (III), the resampling processing and the upside-down inversion processing are performed to obtain irregular parallel beam projection data with distortion correction as shown in FIG. The subsequent multiplication processing of the correction term, the convolution processing in the channel direction, and the irregular parallel beam backprojection processing are the same as those in the first embodiment, and thus description thereof is omitted here.

In the case of (IV), (
As in the case of (II), the projection data is an area mirror-inverted in the Z-axis direction in the case of (III). That is, the detection area moves from the lower half surface to the upper half surface of the two-dimensional X-ray detector. From the focal position Fs to the focal position Fi, As to Ai are on the lower side, and Bs to Bi are on the upper side. At the focal position Fi, Ai and Bi are the same. Further, from the focal position Fi to the focal position Fe, A
i to Ae are on the upper side and Bi to Be are on the lower side. Therefore, the processing method is the same as in (III).

By the combination of the above (I) to (IV), a three-dimensional image can be reconstructed by performing an irregular parallel beam back projection by an irregular cone beam / parallel beam conversion even in a helical scan. The reconstruction control device determines a corresponding case of (I) to (IV) according to the phase θ, and performs an irregular cone beam / parallel beam conversion according to the corresponding case.
A three-dimensional image is reconstructed by performing a resampling process (when necessary, a vertical inversion process is used together), a distortion correction process, a multiplication of correction terms, a convolution process in a channel direction, and an irregular parallel beam backprojection process.

A third embodiment of the present invention will be described with reference to FIGS. After irregular cone beam / parallel beam conversion to irregular parallel beam back projection processing,
Compensate for artifacts (including hardening) using the median section theorem. The central section theorem states that "the primary Fourier transform of the projection data of the parallel beam to the phase θ of a certain image is equal to the distribution of the central section obtained by cutting the two-dimensional Fourier transform image of the original image at the corresponding angle θ." That is. That is, as shown in FIG. 23, consider a parallel beam in the phase θ direction of the image 31 in the real space,
As for the projection data 32 obtained by the parallel beam, the one-dimensional Fourier-transformed data is, as shown in FIG. 24, the center cross section of the two-dimensional Fourier transform image 33 in the frequency space of the image 31 cut at the corresponding angle θ. Equal to distribution 34. Therefore, if the distribution 34 of the central section is subjected to one-dimensional inverse Fourier transform, the projection data 32 can be obtained.

Therefore, after the image is once reconstructed, the image is subjected to two-dimensional Fourier transform, the distribution of the center section cut at a certain angle θ is obtained, and the distribution of the center section is subjected to one-dimensional inverse Fourier transform to obtain the phase of the image. The projection data of the parallel beam in the θ direction is obtained. A method of calculating a correction coefficient such as hardening based on the projection data, performing the second image reconstruction by parallel beam back projection, and correcting an artifact due to hardening is known as a two-pass BHC or the like. Have been.

However, since the first image reconstruction is performed by the back projection of the fan beam and the second image reconstruction is performed by the back projection of the parallel beam, the artifact caused by the spread of the fan beam is diffused like a fan at the time of the back projection. Therefore, there is a case where it cannot be removed by the correction by the parallel beam back projection. Similarly, in the case of a cone beam, an artifact due to the spread of the cone beam may not be able to be removed.

In the first and second embodiments described above, the backprojection processing is performed when viewed from the Z-axis direction.
Since the projection is performed in the same manner as the back projection of the parallel beam in the second pass, the correction effect in the second pass is large. Therefore, first, three-dimensional image reconstruction is performed by the method described in the first embodiment or the second embodiment, and then, ignoring the spread of the cone beam in the Z-axis direction, Instead of a two-dimensional image as one cross section of the three-dimensional image, a two-dimensional correction is performed on the two-dimensional image by applying the two-dimensional central cross-section theorem and considering the conventional two-dimensional image. Generally 3 artifacts
Although the cone angle is smaller than the fan angle, a sufficient correction effect can be obtained by the above-described correction using the two-dimensional central section theorem.

As a modified example, it is also possible to obtain projection data with a completely parallel beam using the three-dimensional central section theorem, and to perform correction using the data. Since the projection data by the irregular parallel beam and the projection data by the complete parallel beam are relatively similar, a certain correction effect can be expected.

[0060]

As described in detail above, according to the present invention,
It is possible to provide an image reconstruction processing device that can establish the application of the fan-para conversion method in a cone beam and can realize accurate reconstruction of an image captured using the cone beam.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an X-ray CT apparatus according to a first embodiment of the present invention (in the case of conventional scanning).

FIG. 2 is a diagram showing a geometry of the X-ray CT apparatus according to the embodiment;

FIG. 3 is a diagram for explaining conversion of a cone beam into an irregular parallel beam as viewed from the rotation axis direction of the X-ray source and the two-dimensional X-ray detector of the X-ray CT apparatus according to the embodiment.

FIG. 4 is a diagram showing a projection state of a plane perpendicular to the irregular parallel beam when the cone beam of the X-ray CT apparatus according to the embodiment is converted into an irregular parallel beam.

FIG. 5 is a diagram showing a projection state on a detection surface of a cylindrical two-dimensional X-ray detector using an irregular parallel beam of the X-ray CT apparatus according to the embodiment.

FIG. 6 is a diagram showing a projection state on a detection surface of a planar two-dimensional X-ray detector using an irregular parallel beam of the X-ray CT apparatus of the embodiment.

FIG. 7 is a diagram showing a relationship between an irregular parallel beam and a reconstructed image plane in the X-ray CT apparatus of the embodiment.

FIG. 8 is a diagram showing a projection state of a reconstructed image plane on an irregular parallel beam of the X-ray CT apparatus according to the embodiment.

FIG. 9 is a diagram showing a projection state of a reconstructed image plane on a detection plane of a two-dimensional X-ray detector using an irregular parallel beam of the X-ray CT apparatus according to the embodiment.

FIG. 10 is a diagram showing data obtained by resampling the projection data by the irregular parallel beam of the X-ray CT apparatus according to the embodiment.

FIG. 11 is an exemplary view for explaining backprojection processing of resampled data onto a reconstructed image plane by the X-ray CT apparatus according to the embodiment;

FIG. 12 is a block diagram showing a main configuration of a reconstruction processing unit 12 of the X-ray CT apparatus according to the embodiment;

FIG. 13 is a view for explaining backprojection processing from a resampling plane using similar centering processing on voxel rows parallel to the fixed axis of the X-ray CT apparatus according to the embodiment.

FIG. 14 is a view showing projection data of a reconstructed image A at a phase φ of the X-ray CT apparatus according to the embodiment.

FIG. 15 is a reconstructed image A of the X-ray CT apparatus of the embodiment.
FIG. 9 is a diagram showing backprojection data when similar centering processing is not performed and backprojection data when similar centering processing is performed.

FIG. 16 is an irregular parallel beam in the case where the focal point of the X-ray CT apparatus according to the second embodiment of the present invention (in the case of helical scan) is below the reconstructed image plane over the entire area (in the case of (I)) FIG. 7 is a diagram showing a projection state of a reconstructed image plane with respect to FIG.

FIG. 17 is a diagram showing a projection state of a reconstructed image plane on a detection plane of a two-dimensional X-ray detector using an irregular parallel beam in the case of (I) of the X-ray CT apparatus of the embodiment.

FIG. 18 is a diagram showing data obtained by performing resampling processing on projection data by the irregular parallel beam in the case of (I) of the X-ray CT apparatus according to the embodiment.

FIG. 19 shows a case where the focal point of the X-ray CT apparatus according to the embodiment shifts from the lower side to the upper side of the reconstructed image plane (case (III)).
FIG. 6 is a diagram showing a projection state of a reconstructed image plane on the first half of the irregular parallel beam of FIG.

FIG. 20 is a diagram showing a projection state of a reconstructed image plane with respect to the latter half of the irregular parallel beam in the case of (III) of the X-ray CT apparatus of the embodiment.

FIG. 21 is a diagram showing a projection state of a reconstructed image plane on a detection plane of a two-dimensional X-ray detector using an irregular parallel beam in the case of (III) of the X-ray CT apparatus of the embodiment.

FIG. 22 is a diagram showing data obtained by performing resampling processing on projection data using an irregular parallel beam in the case (III) of the X-ray CT apparatus according to the embodiment.

FIG. 23 is a view showing projection data of a parallel beam onto a phase θ of an image having a central section theorem used in the X-ray CT apparatus of the embodiment.

FIG. 24 is a view showing the distribution of the central section obtained by cutting the two-dimensional Fourier transform image of the original image of the central section theorem used in the X-ray CT apparatus of the embodiment at a corresponding angle θ.

FIG. 25 is a view showing a fan beam in a conventional X-ray CT apparatus.

FIG. 26 is a diagram showing a relationship among an X-ray source, an X-ray detector, a FOV, and a pixel of a conventional X-ray CT apparatus.

FIG. 27 is a view for explaining the fan-para conversion method of the conventional X-ray CT apparatus.

FIG. 28 is a first diagram for explaining an image reconstruction formula of the fan-para conversion method of the conventional X-ray CT apparatus.

FIG. 29 is a second diagram for explaining an image reconstruction equation of the fan-para conversion method of the conventional X-ray CT apparatus.

FIG. 30 is a diagram showing a cone beam in a conventional X-ray CT apparatus.

FIG. 31 is a diagram showing a relationship among an X-ray source, a two-dimensional X-ray detector, a FOV, and a voxel of a conventional X-ray CT apparatus.

FIG. 32 is a view for explaining back projection of detector data onto pixels in an X-ray CT apparatus using a fan beam according to a conventional example.

FIG. 33 is a view for explaining back projection of detector data onto voxels in a conventional X-ray CT apparatus using a cone beam.

[Explanation of symbols]

 Reference numeral 3: X-ray source, 5: two-dimensional X-ray detector, 11: data acquisition unit, 12: reconstruction processing unit, 21: image reconstruction control unit, 22: irregular parallel beam back projection processing unit, 23: memory, 23-1 Data storage unit.

Claims (8)

[Claims] 1. A one-dimensional X-ray detector in which a target is irradiated with a conical X-ray cone beam emitted from an X-ray source, and X-rays transmitted through the target are arranged in a row of X-ray detection elements. Are detected by a two-dimensional X-ray detector in which the detectors are arranged in a plurality of rows.
An image reconstruction processing device for reconstructing a tomographic image of the object based on detection data obtained by rotating a source and the two-dimensional X-ray detector around the object, wherein the X-ray From the detection data obtained by rotating the source and the two-dimensional X-ray detector by a predetermined angle, a beam from the X-ray source to the two-dimensional X-ray detector is divided into the X-ray source and the two-dimensional X-ray. Irregular parallel data conversion means for collecting and processing, for each of the transmission angles, detection data of an irregular parallel beam whose transmission angles to the object are parallel ignoring the rotational axis direction component of the detector; An image reconstruction processing apparatus, comprising: parallel reconstruction means for reconstructing a tomographic image of the object based on detection data of an irregular parallel beam collected and processed by data conversion means. 2. One-dimensional X-ray detection in which a target is irradiated with a cone-shaped X-ray cone beam emitted from an X-ray source, and X-rays transmitted through the target are arranged in a row of X-ray detection elements. Are detected by a two-dimensional X-ray detector in which the detectors are arranged in a plurality of rows.
An image reconstruction processing device for reconstructing a tomographic image of the object based on detection data obtained by rotating a source and the two-dimensional X-ray detector around the object, wherein the X-ray From the detection data obtained by rotating the source and the two-dimensional X-ray detector by a predetermined angle, a beam from the X-ray source to the two-dimensional X-ray detector is divided into the X-ray source and the two-dimensional X-ray. Irregular parallel data conversion means for collecting and processing, for each of the transmission angles, detection data of an irregular parallel beam whose transmission angles to the object are parallel ignoring the rotational axis direction component of the detector; An irregular parallel data correcting unit for correcting the distortion in the rotation axis direction with respect to the detected data of the irregular parallel beam collected and processed by the data converting unit; and an irregular parallel data corrected by the irregular parallel data correcting unit. Image reconstruction processing apparatus characterized by comprising a parallel reconstruction means for reconstructing a tomographic image of the object based on detection data Rerubimu. 3. The anomalous parallel data correction means rearranges the detection position of the detection data collected and processed by the anomalous parallel data conversion means with respect to the position in the rotation axis direction in accordance with the transmission angle of the parallel beam. 3. The image reconstruction processing device according to claim 2, wherein a resampling process is performed. 4. The irregular parallel data correcting means sets a row of voxels as spatial pixels to be back-projected in parallel to a predetermined fixed axis, and the trajectory of the focal point of the irregular parallel beam parallel to the row of voxels. As preprocessing for correcting the distortion in the rotation axis direction to a preset centering plane, the detection data of the irregular parallel beam collected and processed by the irregular parallel data conversion means on a plane parallel to this trajectory is assumed. 4. The image reconstruction processing apparatus according to claim 2, wherein a similar centering process for performing projection is performed. 5. The detection data is obtained by a conventional scan in which the X-ray source and the two-dimensional X-ray detector are rotated around the object without moving in the direction of a rotation axis. The image reconstruction processing device according to claim 4. 6. The detection data is obtained by a helical scan in which the X-ray source and the two-dimensional X-ray detector are rotated while moving around the object in the direction of a rotation axis. Item 5. An image reconstruction processing device according to any one of Items 4. 7. The method according to claim 1, wherein the correction of the artifact and the stiffness of the tomographic image reconstructed by the parallel reconstructing means is performed using a central section theorem. An image reconstruction processing apparatus described on page 1. 8. The apparatus according to claim 1, wherein the correction of the artifact and the hardening of the radiation quality for the tomographic image reconstructed by the parallel reconstructing means is performed as a two-dimensional image using a two-dimensional central section theorem. The image reconstruction processing device according to claim 1.