JP3392437B2 - Computer tomography equipment - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a helical scan type computer tomography apparatus for scanning a subject by continuously rotating radiation while moving the subject.

[0002]

2. Description of the Related Art An X-ray computed tomography apparatus (hereinafter abbreviated as X-ray CT apparatus) is an example of such an apparatus.
There is. The X-ray CT apparatus is widely used not only for medical purposes but also for industrial purposes as an apparatus for taking a tomographic image of a subject. In particular, the X-ray CT apparatus for medical use occupies an important position as an image diagnostic apparatus. Compared with the initial X-ray CT apparatus, the scanning speed is very fast and the overall imaging time is short compared to the initial one, but there is a demand for more rapid imaging. This is because, for example, in the case of a method called a dynamic scan, a short period of time is required to image the movement of the contrast medium injected into the body, and so on. Therefore, in recent years, an X-ray CT apparatus capable of continuous scanning using a slip ring has appeared.

Here, when the X-ray is rotated around the subject, the subject can be moved in the slice direction (body axis direction) to perform continuous scanning, thereby enabling volume scanning. According to the helical scan method, a large number of tomographic images can be taken in a short time. Note that X
A conventional method in which the subject is fixed while the line is rotating and the slices are imaged one by one is called a fixed position scanning method. For convenience of explanation, the slice direction perpendicular to the tomographic image, that is, the moving direction of the subject is defined as the Z axis. A set of projection data belonging to the same fan is called a view.

In the helical scan method, the subject moves continuously in the Z-axis direction during the rotation of the X-ray, so the position of the 360 ° view is not constant. Therefore, in order to reconstruct a tomographic image at a certain position, 360
It is necessary to create a virtual view of °.

As a typical example of creating a virtual view, a virtual view is created by interpolation from a projection view 360 degrees before and after the position of the tomographic image, a total of 720 degrees. This method is also widely used clinically. But,
In this method, since data for two rotations of X-rays is used, the subject moves a considerable distance during that time, which causes a problem of large blurring in the Z axis direction. Therefore, the following method has been proposed.

The concept of the second method is shown in FIG. FIG. 15 is a diagram called a sinogram, in which the horizontal axis represents the detector number n (channel CH) and the vertical axis represents the view number (projection number) m, which is a diagram schematically showing projection data. A collection of projection data belonging to the same projection number, that is, a horizontal line of a sinogram constitutes a view. If the number of views is large enough, as shown in FIG. 16, the beam related to the projection data detected by one detector always has opposite beams with opposite directions. The opposite beam data can be regarded as the same data because the beams have the opposite directions and the trajectories in the subject are the same.

Two circles connected by an arrow in FIG. 15 indicate projection data and its opposite beam data. Therefore, one piece of projection data sandwiching the slice position and one piece of opposite beam data are interpolated to obtain one virtual view at the slice position.
You can create projection data for books. The second method is
As shown in FIG. 15, in the view with a fan opening angle of 360 ° + 2 times (hereinafter also referred to as a double fan angle: 4α), the slice position from the projection data of the region surrounded by the rhombus is determined. Create a 360 ° virtual view. In this case, all the projection data of the area surrounded by the diamond is used twice.

Therefore, in the second method, the projection data of the triangular areas P1 and P2 are used twice even though they are located farthest from the slice position. Also,
The projection data of the triangular areas P3 and P4 is the area P, respectively.
2, the projection angle is the same as the projection data of P1, and the area P
1, which is closer to the slice position than P2, but is not used at all. As a result, there are the following drawbacks. (1) Compared with the first method, the blurring in the Z-axis direction is small, but it is not yet sufficient. (2) Blurring in the Z-axis direction is non-uniform depending on the pixel location, and the influence of blurring differs depending on the subject.

[0009]

As described above, in the conventional helical scan type computer tomography apparatus, the first method using the 720 ° view before and after the slice position is the first method for obtaining the virtual view of the slice position of 360 °.
Method and the second method using a view with a fan angle of 360 ° + 2 times before and after the slice position. The first method has a drawback that blur in the slice direction is large, and the second method has a drawback that blur in the slice direction is still present although it is less than the first method. Further, the second method has a problem that the blurring in the slice axis direction is non-uniform and is affected by the subject.

The present invention has been made to cope with the above-mentioned circumstances, and its object is a helical scan type computer tomography apparatus capable of reconstructing an image with little blurring in the slice direction and less unevenness in the blurring. Is to provide.

[0011]

In order to achieve the above-mentioned object, the present invention is a computer tomography for performing a helical scan to collect projection data and reconstructing an image at a predetermined slice position based on the collected projection data. In the apparatus, the projection data closest to the predetermined slice position and 2
Means for determining whether or not the projection data closest to the second sandwiches the predetermined slice position, and the projection data closest to the predetermined slice position and the projection data closest to the second slice data if the projection data sandwiches the predetermined slice position. Obtaining virtual projection data at the predetermined slice position by performing interpolation using the projection data and the second closest projection data,
When the projection data closest to the predetermined slice position and the projection data closest to the second slice position do not sandwich the predetermined slice position, the predetermined slice position is obtained by extrapolating the closest projection data and the second closest projection data. Of the virtual projection data, and means for reconstructing the image at the predetermined slice position based on the virtual projection data acquired by the interpolation means. Further, in order to achieve the above object, the present invention is a computer tomography apparatus for performing helical scan to collect projection data, and reconstructing an image at a predetermined slice position based on the collected projection data. Whether the projection data closest to the predetermined slice position and the projection data closest to the second among the projection data thus obtained and the projection data of the opposite beam existing in a rotation phase opposed to the projection data sandwich the predetermined slice position. When the projection data closest to the predetermined slice position and the second closest projection data sandwich the predetermined slice position, the closest projection data and the second closest projection data are interpolated. To obtain virtual projection data at the predetermined slice position, and to obtain the virtual projection data at the predetermined slice position. When the closest projection data and the second closest projection data do not sandwich the predetermined slice position, the virtual projection data at the predetermined slice position is obtained by extrapolating the closest projection data and the second closest projection data. And interpolation means for reconstructing the image at the predetermined slice position based on the virtual projection data acquired by the interpolation means.

[0012]

According to the present invention, the use ratio of the data located far from the slice position is reduced, and instead the data located near the slice position at the same projection angle is used. It is possible to obtain an image with less blur and less unevenness of the blur.

[0013]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of a computer tomography apparatus according to the present invention will be described below with reference to the drawings. Here, a so-called third generation X-ray CT apparatus capable of continuous scanning at 360 ° or more using a fan beam will be described, but the present invention is also applicable to other generations, for example, a fourth generation X-ray CT apparatus. .

First, the principle of the present invention will be described. Hereinafter, the projection data is assumed to be the projection data collected by the helical scan method. Since the helical scan is widely used in many third generation X-ray CT apparatuses, detailed description thereof will be omitted.

The projection data is defined as follows. Here, for convenience of explanation, a case will be described in which the X-ray detector array for detecting the fan beam is arranged symmetrically with respect to a line connecting the X-ray tube focal point and the rotation center (hereinafter referred to as a center line). The same can be applied to the case of asymmetrical arrangement.

The divergence angle of the fan beam is 2α, and the number of channels in the detector array (the number of detectors) is N (even number). Therefore, the divergence angle Δα of the beam of one channel is Δα
= 2α / N. The rotation angle of the X-ray between the views is Δθ, and the number of views per 360 ° is M (even number). Therefore, M × Δθ = 360 °. The view number is m, the channel number is n, and the projection data is D (m,
n). The angle formed by the center line of the m-th view is called the projection angle. Then, the projection angle of the m-th view is m × Δθ. The distance that the bed moves between the views is d,
The position of the m-th view (the position of the center line) is m × d. The total number of views scanned is MM.

Next, the opposing (reflection) beam in the helical scan type projection data will be described. In the fixed position scanning method, if the number of views is sufficient, one projection data always has an opposite beam. The opposite beam exists on both the ascending direction side (hereinafter referred to as the plus side) of the view number and the descending direction side (hereinafter referred to as the minus side) of the view number. Projection data D (m, n)
The projection data of the plus side opposite beam (hereinafter, also simply referred to as opposite beam data) is indicated as RP (m, n), and the minus side opposite beam data is indicated as RM (m, n). These projection data D, RP, and RM do not exactly match because of the spread of the fan beam, but D (m, n) and RP
(M, n) and RM (m, n) can be regarded as the same beam data in practical terms. for that reason,

[0018]

[Equation 1] , The channel is symmetric with respect to the center line, so nn, mp, and mm are obtained as follows. nn = N + 1-n (3) mp = m + M / 2- (N + 1-2n) * aa (4) mm = m-M / 2- (N + 1-2n) * aa (5) where aa = [Delta] [alpha] / [Delta] [theta]. (6) Furthermore, the following definitions are made. ci: (N-1) x integer part of aa cp: (N-1) x aa after the decimal point

In the helical scan system, since the position of the bed moves, there is no strict opposite beam, but
The beam having the relationship of the equations (1) to (5) with respect to the projection data is referred to as a counter beam for convenience. As shown in FIG.
One view lies on a horizontal line on the sinogram, while its opposite beam lies on a diagonal line.

Here, in equations (4) and (5), m
Since p and mm generally do not become integers, interpolation is performed using the data of two adjacent views with the same channel number nn. Therefore, if the integer part of mpi: mp is below the decimal point of mpp: mp and the integer part of mmi: mm is below the decimal point of mpp: mm, the opposite beam data RP (m, n), RM
(M, n) is calculated as follows. RP (m, n) = D (mpi, nn) * (1-mpp) + D (mpi + 1, nn) * mpp (7) where m = 1 to (MM-M / 2 + ci) RM (m, n) = D (mmi, nn) × (1-mmp) + D (mmi + 1, nn) × mmp (8) where m = (M / 2 + ci + 2) to MM

[0021]

[Equation 2] The relationship is established. Therefore, for convenience of explanation, the following definitions will be made.

[0022]

[Equation 3] Therefore, the opposite beam data can be defined as follows.

[0023]

[Equation 4] Therefore, assuming that mmr = mr + ci + 1- (N + 1-2n) * aa (15), the decimal part of the integer part mmrp: mmr of mmri: mmr, R (mr, n) is calculated as follows. R (mr, n) = D (mmri, nn) × (1-mmrp) + D (mmri + 1, nn) × mmrp (16) where mr = 1 to (MM-2 × ci-2)

FIG. 3 shows the relationship between projection data D (m, n) and counter beam data R (mr, n), and FIG. 4 shows counter beam data R (mr, n) and counter beam data RP (m, n).
n) and RM (m, n) are shown. The numbers in FIG. 4 indicate view numbers. Next, creation of virtual projection data at slice positions will be described.

The projection data of the virtual view on the slice can be obtained by linearly interpolating the projection data and the opposite beam data. The primary interpolation is performed in accordance with the distance between the two beams and the slice position, and two beams that are as close to the slice position as possible are used.

FIG. 5 shows the positional relationship between the projection data and the counter beam data. The projection data D1 and D2 is 1
Rotated (360 °) apart. Opposing beam data R (D
1 is RP and D2 is RM)
Is located between the projection data D1 and D2. Here, the projection data D1 and the counter beam data R are used as the slice positions.
And a slice position S1 between the opposite beam data R and the projection data D2.
2, consider S3.

Conventionally, primary interpolation has been performed using the two nearest beams sandwiching the slice position. Therefore, the data at the slice position S1 is interpolated using the projection data D1 and the opposite beam data R, and the slice positions S2 and S3 are obtained.
Data was interpolated using the opposite beam data R and the projection data D2. That is, conventionally, only interpolation is used as the primary interpolation.

For example, the data at the slice position S2 is interpolated (interpolated) using the opposite beam data R and the projection data D2.
When it does, it becomes X2. However, the data of the slice position S2 is extrapolated from the projection data D1 and the opposite beam data R (1
It can also be obtained by the following interpolation). When extrapolation is performed, the data becomes X1. The accuracy is the same because both interpolation and extrapolation are linear interpolation. Since interpolation is a method of estimating nonexistent data, it cannot be determined which is more suitable, but in general, a method of estimating from near data is often suitable. Therefore, the accuracy of the virtual view data is improved by using not only interpolation but also extrapolation. Therefore, in the present invention, not only interpolation but also extrapolation is used to interpolate the virtual view data. Therefore, as will be described later, when the virtual view data is created, more data near the slice position is used, and the image becomes more uniform.

Here, the position of the projection data D (m, n) in the Z direction is DP (m, n). As described above, the opposite beam used for creating the virtual projection data has the plus side and the minus side opposite beams. However, since it is undecided which one is used here, the opposite beam data is set. RR (m,
n), and its position is RRP (m, n). Also, assuming that the slice position is SP, virtual projection data DD (m, n) on the slice is obtained as follows. DD (m, n) = [D (m, n) * {RRP (m, n) -SP} + RR (m, n) * {SP-DP (m, n)}] / {RRP (m, n) ) -DP (m, n)} (17)

Here, the image (slice) number is defined. The virtual view numbers are j to (j + M-1) 36.
The image reconstructed using the 0 ° virtual view is defined as the j-th image. Next, the selection of the opposite beam used for creating the virtual data will be described.

Now, consider the case of reconstructing the j-th image.
The view numbers used for the reconstruction of this image are j to (j
+ M-1). In the present invention, it is necessary to create 3 virtual data.
A view with a fan angle (4α) of 60 ° + 2 times is used.
Therefore, the views used to create this virtual view are (j-ci-1) to (j + M + ci). This relationship is shown in FIG.

In FIG. 7, a portion extracted from the projection view D (m, n) is a DD1 portion and a DR1 portion. The RP1 portion is obtained by extracting a part from the plus side opposite beam. Here, the RP1 portion is created from the DR1 portion.

In the present invention, virtual views j- (j + M / 2-
1) to create projection views j- (j + M / 2-
1), that is, the DD1 portion, and the plus side opposing beam views j to (j + M / 2-1), that is, the RP1 portion are used. Therefore, assuming that the position of the opposite beam data RP (m, n) is RPP (m, n), the virtual projection data DD
(M, n) is obtained by converting RR (m, n) in the equation (17) into RP (m,
n) and RRP (m, n) as RPP (m, n), the following is obtained. DD (m, n) = [D (m, n) * {RPP (m, n) -SP} + RP (m, n) * {SP-DP (m, n)}] / {RPP (m, n) ) -DP (m, n)} (18)

In FIG. 7, a portion extracted from the projection view D (m, n) is a DD2 portion and a DR2 portion. The RM2 part is a part extracted from the minus side opposite beam. Here, the RM2 part is DR2
It is created from parts.

Similarly, virtual views (j + M / 2) to (j
+ M−1) to create a projection view (j + M /
2) to (j + M-1), that is, the DD2 portion and the negative side opposing beam RM views (j + M / 2) to (j +).
M-1), that is, the RM2 part is used. Therefore, the position of the opposite beam data RM (m, n) is set to RMP (m, n).
Then, in the virtual projection data DD (m, n), RR (m, n) in the equation (17) is RM (m, n), and RRP (m, n)
By setting n) to RMP (m, n), the following is obtained. DD (m, n) = [D (m, n) * {RMP (m, n) -SP} + RM (m, n) * {SP-DP (m, n)}] / {RMP (m, n) ) -DP (m, n)} (19) This relationship is shown in FIG. FIG. 8 shows a conventional method.

Projection data D (m, n) of the area Z1
Is the opposite beam data RP with the same projection angle as
It is positionally interpolated with the data of the part of the area Z5 of (m, n). The data of the area Z5 is obtained from the data of the area Z6. Therefore, the projection position of the area Z5 becomes the area Z6. Regions Z1 and Z6 (Z
In 5), since the slice position j + (M-1) / 2 is not sandwiched, interpolation cannot be performed. Therefore, in the conventional method (18)
The following formula was used instead of the formula. DD (m, n) = [D (m + M, n) * {SP-RPP (m, n)} + RP (m, n) * {DP (m + M, n) -SP}] / {DP (m + M, n) ) -RPP (m, n)} (18a)

Similarly, in the conventional method, the projection data D (m, n) in the area Z3 and the opposite beam data RM (m, n) in the area Z8 (Z7) are slice positions j + (M-).
Since 1) / 2 is not sandwiched, interpolation cannot be performed. for that reason,
The following formula was used instead of formula (19). DD (m, n) = [D (m-M, n) * {RMP (m, n) -SP} + RM (m, n) * {SP-DP (m-M, n)}] / {RMP (M, n) -DP (m-M, n)} (19a)

Comparing FIG. 7 and FIG. 8, regions Z1 and Z2
The areas Z3 and Z4 have the same projection angle, but the areas Z1 and Z3 are closer to the slice position than the areas Z2 and Z4. Therefore, the method of the present invention can obtain an image with less blurring and less nonuniformity than the conventional method. Since the slice position is normally the center of the view used, it is calculated as follows. SP = {j + (M-1) / 2} × d (20)

Since the conventional method uses interpolation, the slice position is limited to the center of the view to be used, but since the present invention also uses extrapolation, the slice position is not necessarily limited to the view center. This is also an advantage of the present invention.

Here, the interpolation coefficients RP (m, n) and RPP (m, n which are necessary when obtaining the equations (18) and (19).
n), RM (m, n), and RMP (m, n) are specifically obtained. The following parameters are introduced for clarity. k = m-j + 1 (21) However, k = 1 to M Therefore, m is calculated as follows. m = k + j-1 (22) However, m = j to (j + M-1) Therefore, the following relationship is obtained. SP-DP (m, n) = {j + (M-1) / 2} * d-m * d = {j + (M-1) / 2-k-j + 1} * d = {(M + 1) / 2- k} * d (23) RPP (m, n) -SP = mp * d- {j + (M-1) / 2} * d = {m + M / 2- (N + 1-2n) * aa} * d- { j + (M-1) / 2} * d = {k-1 / 2- (N + 1-2n) * aa} * d (24) RPP (m, n) -DP (m, n) = mp * d- m * d = {m + M / 2- (N + 1-2n) * aa} * d-m * d = {M / 2- (N + 1-2n) * aa} * d (25) RMP (m, n) -SP = Mm * d- {j + (M-1) / 2} * d = {m-M / 2- (N + 1-2n) * aa} * d- {j + (M-1) / 2} * d = { k-1 / 2- (N + 1-2n) * aa-M} * d (26) RMP (m, n) -DP (m, n) = mm * d-m * d = {m-M / 2- (N + 1-2n) * aa} * d-m * d = {-M / 2- (N + 1-2n) * aa} × d (27) Therefore, the equation (18) can be modified as follows. DD (k + j-1, n) = {D (k + j-1, n) * H2 (k, n) + RP (k + j-1, n) * H1 (k)} / {M / 2- (N + 1-2n) × aa} (28) However, similarly, k = 1 to (M / 2), the equation (19) can be modified as follows. DD (k + j-1, n) = {D (k + j-1, n) * H3 (k, n) -RM (k + j-1, n) * H1 (k)} / {M / 2 + (N + 1-2n) × aa} (29) where k = M / 2 + 1 to M where H1 (k) = (M + 1) / 2−k (30) where k = 1 to M, H2 (k, n) = k− 1 / 2- (N + 1-2n) * aa (31) However, k = 1- (M / 2), H3 (k, n) = M- {k-1 / 2- (N + 1-2n) * aa}. (32) However, k = M / 2 + 1 to M

Next, the configuration of an embodiment of the computer tomography apparatus according to the present invention based on such a principle will be described. FIG. 9 is a block diagram showing a schematic configuration of the first embodiment of the present invention. The fan beam emitted from the X-ray tube 10 and transmitted through the subject 12 is detected by the detector array 14 and input to the data acquisition unit 16. The data is sent to the projection data image reconstruction unit 18 obtained by the data collection unit 16. The CT image reconstructed by the image reconstructing unit 18 is displayed by the display device 20 and stored in the image memory 22. The data collection unit 16, the image reconstruction unit 18, the display device 20, and the image memory 22 are connected to each other via a bus line 24. An operator console 26 through which an operator inputs various instructions is connected to the bus line 24. The entire operation is controlled by a computer (not shown).

FIG. 10 is a block diagram showing a schematic structure of the image reconstructing section 18 in FIG. The projection data is supplied to the preprocessing circuit 30, and the output of the preprocessing circuit 30 is stored in the raw data memory 32. A reflection view is created by the reflection circuit 36 from the output of the raw data memory 32, and is stored in the reflection view memory 38. Raw data memory 32, reflection circuit 3
6. The reflection view memory 38 is controlled by the CPU 34. Instructions from the console 26 are input to the CPU 34. The outputs of the raw data memory 32 and the reflection view memory 38 are supplied to the virtual view creation circuit 40, and a virtual view is created based on the above equations (28) and (29). A virtual view memory 41 is connected to the virtual view creation circuit 40. Virtual view creation circuit 4
The output of 0 is the convolution (CONV) operation circuit 4
2 is supplied. The output of the CONV circuit 42 is supplied to the intermediate image memory 46 via the back projection (BP) operation circuit 44. The output of the intermediate image memory 46 is supplied to the image synthesizing circuit 48, and the image is reconstructed. The intermediate image memory 46 and the image synthesizing circuit 48 are provided to carry out functions other than the present invention. Therefore, when the present invention is implemented, it is bypassed by the control of the CPU 56.

The reconstructed image is supplied to the reconstructed image memory 59. A virtual view creation circuit 40, a CONV circuit 42,
BP circuit 44, intermediate image memory 46, image composition circuit 48
Are controlled by the CPU 56. In addition, the BP circuit 44
A constant setting unit 58 is connected to. Reconstructed image memory 5
The output of 0 is CT through the contrast scale circuit 52.
It is supplied to the image memory 54. The operation of the first embodiment will be described with reference to the flow chart shown in FIG. 1) In step # 10, a helical scan is performed to collect projection data, and the projection data is preprocessed by the preprocessing circuit 30 and then stored in the raw data memory 32.

2) When the scan is completed or a certain amount of raw data is collected, the counter beam data (reflection view data) R (mr, n) is obtained in step # 15 based on the equation (16). , And stores it in the reflection view memory 38. 3) Specify the number j of the image to be reconstructed in step # 20.

4) Initialization is performed in step # 25. Details of the initialization are shown in FIG. That is, here, first,
In step # 55, set 1 to the parameters k and kk,
In step # 60, the reconstructed image memory 50 is cleared.
In step # 65, θθ is calculated according to the image number j as follows. θθ = (j−1) × Δθ− {integer part of (j × Δθ / 360)} × 360 (33)

5) In step # 30, the virtual view creation circuit 40 creates virtual projection data of the (k + j-1) th view of the first to Nth channels. At this time, k = 1 to (M
/ 2), the equation (28) is used, and k = (M / 2 + 1)
If ~ M, the equation (29) is used. The created virtual projection data is stored in the virtual view memory 41 at the address (kk, n).

6) The virtual view obtained in step # 30 is transferred to the convolution circuit 35 in step # 35 to perform convolution, and then the convolution result is backprojected by the backprojection circuit 44. The projection angle θ at this time is θ
= Θθ + k × Δθ. However, when θ ≧ 360, θ = θ-360. 8) In step # 40, the parameters k and kk are updated. 9) In step # 45, it is determined whether k ≦ M, and k ≦ M
If so, the process returns to step # 30 to repeat the virtual view data creation. If k> M, the reconstruction is completed. 10) After this, the contrast scale is added to the reconstructed image.
Then, the CT image is converted into a CT image, transferred to the display device 20, and displayed.

As described above, according to the first embodiment,
Since the virtual view of the slice position is obtained by linear interpolation using not only the interpolation but also the extrapolation, the virtual view is created using the projection data and the facing data close to the slice position.
Therefore, it is possible to reconstruct an image with less blurring in the Z-axis direction and less unevenness in the blurring.

Next, a second embodiment of the present invention will be described. The first embodiment relates to a so-called fan direct method for directly reconstructing projection data (fan data) obtained by a fan beam. Next, the fan data is converted into parallel data (hereinafter referred to as fan / parallel conversion). A second embodiment relating to the method of performing the reconfiguration will be described.

First, the fan / parallel conversion in fixed position scanning will be described. If the number of views of fan projection data is sufficient, parallel beams can be collected from among them. The projection angle of the p-th view is p × Δθ, and a set of beams parallel to the center line is a parallel view.
Called. Let N be the number of channels in one view. Projection data PA (p, q) of the q-th channel of the p-th parallel view
Is calculated as follows. PA (p, q) = D (ppi, q) * (1-ppp) + D (ppi + 1, q) * ppp (51) where pp = p + (N + 1-2q) * bb (52) ppi: pp Integer part ppp: Decimal point of pp bb = Δα / (2 · Δθ) (53)

The set of PA (p, q) when p is constant and q = 1 to N is the p-th view of the parallel beam. As shown in FIG. 12A, this p-th view is a collection of projection data on diagonal lines on the sinogram.

In the helical scan system, since the position of the bed moves during the rotation of the X-ray, there is no parallel beam, but the beams having the relations of equations (51) to (53) are parallel beam for convenience. -It is called Mu. The parallel beams also exist at positions separated by 180 ° as shown in FIG. 12 (b). This is P
Assuming that B (p, q), PB (p, q) is obtained as follows. PB (p, q) = PA (p + M / 2, N−q + 1) (54) Similarly, the parallel beam at the position advanced by 360 ° is PC (p, p, q).
q), PC (p, q) is calculated as follows. PC (p, q) = PA (p + M, n) (55)

Next, slice position {p + (m-1) /
Creation of virtual projection data in 2) will be described.
Similar to the first embodiment, the virtual view data of the slice is obtained by linearly interpolating (interpolating or extrapolating) two parallel beams that are as close to the slice as possible according to the distance between the beam and the slice position. ). Therefore, assuming that the positions of the parallel beams PA and PB are PAP and PBP, the virtual parallel projection data PD (p, q) on the slice can be obtained as follows. PD (p, q) = [PA (p, q) * {PBP (p, q) -SP} + PB (p, q) * {SP-PAP (P, Q)}] / {PBP (p, q) ) -PAP (p, q)} (56) where p = j to (j + M / 2) n = 1 to N

Similar to the first embodiment, an image to be reconstructed by using a 360 ° virtual view whose virtual view numbers are j to (j + M-1) is the jth image (j = (ei + 1) to. (M
MM-ei)). In addition, the following parameters are introduced. r = p-j + 1 (57) However, r = 1 to M / 2 Therefore, the following relationship is obtained. p = r + j-1 (58) Therefore, the equation (56) can be modified as follows. PD (r + j−1, q) = {PA (p, q) × H4 (r, q) + PB (p, q) × H5 (r, q)} / {M / 2− (N + 1-2q) × aa } (59) However, r = 1- (M / 2) Here, H4 (r, q) = r-1 / 2- (N + 1-2q) * bb (60) H5 (r, q) = (M + 1) ) / 2-r- (N + 1-2q) * bb (61)

The structure of the second embodiment is similar to that of the first embodiment shown in FIG. The reflection circuit 36 can create a parallel beam only by changing the constant. The reflection view memory 38 also stores parallel beam data. These controls are performed by the CPU 34. Figure 1
3 is a flow chart showing the operation of the reconfiguration of the second embodiment. 1) In step # 110, a helical scan is performed to collect projection data, and the projection data is preprocessed by the preprocessing circuit 30 and then stored in the raw data memory 32.

2) When the scan is completed or a certain amount of raw data is collected, in step # 115 (51)
The parallel beam data PA (p, q) is calculated based on the equation,
This is stored in the reflection view memory. 3) In step # 120, the number j of the image to be reconstructed is designated.

4) Initialization is performed in step # 125. Details of the initialization are shown in FIG. That is, here, first, the parameters r and rr are set to 1 in step # 155, and the reconstructed image memory 50 is cleared in step # 160. In step # 165, (3
3) θ is calculated based on the equation.

5) In step # 130, the virtual view creation circuit 40 creates virtual parallel projection data of the (r + j-1) th view of the first to Nth channels based on the equation (59). The created virtual view data is stored in the virtual view memory 4
It is stored in the first (rr, q) address.

6) The virtual view obtained at step # 130 is transferred to the convolution circuit 35 at step # 135 to perform convolution, and then the convolution result is back-projected by the back-projection circuit 44. The projection angle θ at this time is θ
= Θθ + k × Δθ. However, when θ ≧ 360, θ = θ-360. 8) In step # 140, the parameters r and rr are updated.

9) In step # 145, it is determined whether or not r ≦ M. If r ≦ M, the process returns to step # 130 to repeat parallel beam data generation. If r> M, the reconstruction is completed. 10) After this, the contrast scale is added to the reconstructed image.
Then, the CT image is converted into a CT image, transferred to the display device 20, and displayed.

As described above, according to the second embodiment,
Even in the method in which fan data is converted into parallel data and then reconstructed, the virtual view of the slice position is obtained by linear interpolation using not only interpolation but first interpolation.
The same effect as the embodiment can be obtained.

The present invention is not limited to the above-described embodiments, but can be implemented with various modifications. For example, in the above description,
Although the third-generation X-ray CT apparatus has been described, the same applies to other generation apparatuses such as the fourth generation. Further, the interpolation is not limited to primary interpolation, and higher-order interpolation of secondary or higher order may be used.

[0063]

As described above, according to the present invention, Z
A helical scan type computer tomography apparatus capable of reconstructing an image with little axial blur is provided.

[Brief description of drawings]

FIG. 1 is a first computer tomography apparatus according to the present invention.
The flowchart which shows operation | movement of an Example.

FIG. 2 is a sinogram showing the relationship between the opposite beam and projection data.

FIG. 3 Projection data D (m, n) and counter beam data R
The figure which shows the relationship of (mr, n).

FIG. 4 is an opposite beam data R (mr, n) and a plus side,
Negative-side counter beam data RP (m, n), RM
The figure which shows the relationship of (m, n).

FIG. 5 is a diagram showing a principle of interpolating slice position data with two beams.

FIG. 6 is a diagram showing view data necessary for creating virtual view data.

FIG. 7 is a diagram showing view data used in an interpolation operation according to the present invention.

FIG. 8 is a diagram showing view data used for an interpolation operation according to a conventional method.

FIG. 9 shows a first computer tomography apparatus according to the present invention
The block diagram which shows the structure of an Example.

FIG. 10 is a diagram showing details of the image reconstruction unit shown in FIG. 9;

FIG. 11 is a flowchart showing details of the initialization shown in FIG. 1.

FIG. 12 is a diagram showing parallel view data used in the second embodiment of the present invention.

FIG. 13 is a flowchart showing the operation of the second embodiment.

FIG. 14 is a flowchart showing details of initialization shown in FIG.

FIG. 15 is a sinogram showing the reconstruction principle of a conventional helical tomographic computer tomography apparatus.

FIG. 16 is a diagram showing a relationship between projection data and counter beam data.

[Explanation of symbols]

10 ... X-ray tube, 12 ... Subject, 14 ... Detector array, 1
6 ... Data collection unit, 18 ... Image reconstruction unit, 20 ... Display device, 22 ... Image memory, 26 ... Operation console.

Claims (4)

(57) [Claims] 1. A computer tomography apparatus for performing a helical scan to collect projection data and reconstructing an image at a predetermined slice position based on the collected projection data, wherein the projection data closest to the predetermined slice position and the second Means for determining whether or not projection data close to the predetermined slice position sandwiches the predetermined slice position, and projection data closest to the predetermined slice position and second closest projection data sandwiches the predetermined slice position, the closest projection. Virtual projection data at the predetermined slice position is obtained by performing interpolation using the data and the projection data closest to the second, and the projection data closest to the predetermined slice position and the projection data closest to the second slice position are the predetermined slice position. If not, the extrapolation interpolation is performed on the closest projection data and the second closest projection data. And interpolating means for obtaining virtual projection data of the predetermined slice position by computed tomography and means for reconstructing an image of the predetermined slice position on the basis of the virtual projection data acquired by the interpolation means. 2. The computed tomography according to claim 1, wherein the closest projection data and the second closest projection data are projection data of a predetermined projection angle and opposite beam data of the projection data. apparatus. 3. A computer tomography apparatus for performing a helical scan to collect projection data, and reconstructing an image at a predetermined slice position based on the collected projection data, wherein the projection data obtained by the helical scan and the projection Means for determining whether or not the projection data closest to the predetermined slice position and the projection data closest to the second of the projection data of the opposite beams existing in the rotation phase opposite to the data sandwich the predetermined slice position. When the projection data closest to the predetermined slice position and the projection data closest to the second slice position sandwich the predetermined slice position, the projection data of the predetermined slice position is interpolated by interpolating the closest projection data and the second closest projection data. The virtual projection data is acquired, and the projection data closest to the predetermined slice position and the second When the projection data close to is not sandwiching the predetermined slice position, the interpolation means for extrapolating the closest projection data and the second closest projection data to obtain the virtual projection data at the predetermined slice position, And a means for reconstructing an image at the predetermined slice position based on the virtual projection data acquired by the interpolation means. 4. The projection data collected by the helical scan is fan data obtained by a fan beam, further comprising conversion means for converting the fan data into parallel data, and the interpolation means converts the parallel data. The computer tomography apparatus according to any one of claims 1 to 3, wherein the interpolation processing is performed using the interpolation processing.