JPH119589A - X-ray ct tomograph and image recomposing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make possible to recompose an image continuously at high speed and reduce artifacts by dividing groups of parallel beam projection data at a prescribed angle of projection, and obtaining a regular recomposed image by multiplying multiple divided component images by a weighting factor by the synthesization of outputs of multiple multipliers. SOLUTION: The recomposition arithmetic unit 11 of an image processing unit 3 obtains a recomposed image by carrying out pretreatment with an arithmetic unit 21, collimation with a converter 22, filter correction with an arithmetic unit 23 and reverse projection arithmetic with an arithmetic unit 24. The restructuring arithmetic intended to obtain divided restructured images from collimated beam data with a partial angular width, a weighting image adder 12 aligns divided recomposition images obtained with an arithmetic unit 11 to memories #1 to #7 of an image memory one after another. The multiplier 13 makes weighting factors W1 to W7 to correspond to the memories #1 to #7, multiplies and weights divided images, and a synthesizer 25 obtains a single restructured image by carrying out summation.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an X-ray CT apparatus and an image reconstruction method suitable for obtaining a tomographic image with reduced artifacts at a high speed when continuously measuring substantially the same cross section.

[0002]

2. Description of the Related Art X-ray CT apparatuses are widely used, and various uses are made by users. Recently, CT has been used as a puncture guide when performing histological examination or treatment of a lesion percutaneously. It is regarded as effective because the operation time is reduced and the accuracy is improved by performing the procedure under the CT guide.

[0003] For continuous observation of such CT images,
High-speed image reconstruction is required. Currently, several proposals have been made. In Japanese Patent Publication No. 23136/1990, the difference between the projection data used to reconstruct the first image and the projection data necessary to obtain the second image is determined, and the difference image obtained by reconstructing the difference data is calculated. A second image having a time difference is obtained by adding the second image to the first image.

In Japanese Patent Application Laid-Open No. Hei 8-24252, after reconstructing a first image, the image is reconstructed with projection data (1 / N view: N is the number of divisions) that is less than one image. A new tomographic image is to be obtained from the new image. Japanese Patent Application Laid-Open No. Hei 8-24253 discloses a similar content.

[0005] Both of them make it possible to update an image at high speed by performing partial reconstruction. A normal scan is 360 ° / scan, and one tomographic image is reconstructed from data for 360 °. On the other hand, Japanese Patent Application Laid-Open No. H8-24252 discloses a method of obtaining a completed tomographic image by adding a required number of images (divided images) (for example, 12 images in the case of 30 °) obtained by reconstructing projection data at every 30 ° It is.

That is, FIG. 5A is a diagram showing a state of parallel projection data in two views (two views of 0 ° and 30 ° are shown in the figure). Here, the parallel projection data conversion is to mathematically convert the fan beam X-ray into a parallel beam, and FIG.
0 ° to 60 °, a memory storage of a projection data _{P (P 1, P 2 ...} ) which sorted into 30 ° unit represented by ....

FIG. 5C shows a divided reconstructed image g (g ₁ ,
g ₂ ...) and an example of memory storage of the normal reconstructed image r (r ₁ , r ₂ ...). Here, the divided reconstructed image g is 30 °
And that of the reconstructed images obtained from parallel projection data width divided the reconstructed image g _1, 0 ° ~30 ° parallel projection reconstruction image obtained from the data, a dividing reconstructed image g ₂ is 3
This is a reconstructed image obtained from parallel projection data of 0 ° to 60 °, and the same applies to other divided reconstructed images.

[0008] The normal reconstructed image r refers to a 360 ° reconstructed image obtained by combining (usually adding) a plurality of divided reconstructed images g. It is a normal CT image. 30 °
The divided reconstructed image g obtained from the parallel projection data having the width does not form an image due to insufficient data. So 30 °
The normal reconstructed image is obtained by adding the twelve divided reconstructed images g obtained in the width. Further, the addition of the divided reconstructed images having a width of 30 ° is performed sequentially. The advantage of sequential addition is that regular reconstructed images can be obtained one after another in real time. It is a well-known matter that reconfiguration can be performed by adding.

That is, in FIG. 5C, the first normal reconstructed image r ₁ is obtained by adding g ₁ to g ₁₂ . Reconstructed image r ₂ of the second normal is obtained by adding the g ₂ to g _13. The second image g ₂ to g ₁₂ for normalization of the reconstructed image r ₂ are those already previously asked when obtaining a reconstructed image r ₁ of the first normal, the newly required image g ₁₃ Only. This image g
₁₃ is 0-30 ° of 0 ° -360 ° of the next scan without waiting for completion of all next 360 ° scans.
The data acquisition time is 1/12 of the time for one scan, and a regular image r ₂ can be obtained only by adding the 30 ° divided reconstructed images. Was completed.

As described above, the second image r ₂ is further converted into g ₂ to g ₂ if the _thirteenth divided image g ₁₃ is reconstructed.
It can be obtained by adding the thirteenth image to the added divided image obtained by the addition of _twelve . Therefore, a new tomographic image can be obtained in the time Δt ₁ for reconstructing one divided image g ₁₃ and the time Δt ₂ for image addition. Since the reconstruction time is proportional to the number of backprojected views, one divided image can be reconstructed in a shorter time, about 1/12 compared to normal reconstruction, and high-speed reconstruction is possible. If Δt ₁ + Δt ₂ is 1/12 of the scan time, a tomographic image can be continuously obtained in real time in parallel with the measurement. The same applies to r ₃ and r _{4 for the third} and subsequent sheets.

[0011]

All of these algorithms have a basic idea of speeding up. In real-time CT, a moving object such as monitoring of a puncture needle or a measurement cross section at the time of a spiral scan is often taken as an imaging target.
However, these known examples do not consider suppression of an artifact peculiar to a CT apparatus which appears when capturing an image of a moving object.

It is well understood by those skilled in the art that the artifact of the CT apparatus has the most remarkable discontinuity of data at the start and end of the measurement.
At the time of reconstruction using data of °, the discontinuity of the data at 0 ° and 180 ° where the projection directions are opposite to each other becomes a problem. Such a discontinuity correction algorithm has been conventionally used, but it is difficult to implement in terms of calculation time, and at present, it has been omitted in real-time reconstructed images.

The present invention relates to an X-ray CT apparatus and an X-ray CT apparatus which enable a continuous image to be reconstructed at a high speed and reduce artifacts when imaging is performed continuously or continuously. It is intended to provide a reconstruction method.

[0014]

In order to achieve the above object, an X-ray CT apparatus according to the present invention divides a parallel beam projection data group obtained by converting a fan beam into a parallel beam at a predetermined projection angle. A memory for storing a plurality of divided constituent images corresponding to each of the divided sections, a plurality of multipliers for multiplying each of the plurality of divided constituent images by a predetermined weighting coefficient, and outputs of the plurality of multipliers And a synthesizer for obtaining a normal reconstructed image by synthesizing.

In the present invention, a plurality of continuous divided reconstructed images are sequentially selected so that some of the divided reconstructed images overlap, and the selected divided reconstructed image is multiplied by a predetermined weighting coefficient. I have to. At this time, for the selected plurality of divided reconstructed images, the weighting factor is set such that the sum of the weights of the divided reconstructed images whose view angles are opposite to each other is equal to the weight of the other divided reconstructed images. Is set.

When a reconstructed image is obtained by using a recursive filter by using a 180 ° divided reconstructed image, the sum of the weights of the divided reconstructed images whose view angles are opposed to each other is set to one. The weight of the recursive filter applied between images is determined.

Further, according to the present invention, in an X-ray CT apparatus in which a reconstructed image having a projection angle in a predetermined range is sequentially shifted by a predetermined angle to obtain a continuous reconstructed image, fan beam measurement data is converted into a parallel beam. Dividing the obtained parallel beam projection data at a predetermined projection angle to obtain a plurality of divided reconstructed images corresponding to the respective divided sections; and An image reconstruction method including a step of multiplying a weight coefficient and a step of combining the plurality of divided reconstructed images multiplied by the weight coefficient to obtain one reconstructed image is disclosed.

Further, according to the present invention, in an X-ray CT apparatus for sequentially reconstructing a reconstructed image having a projection angle within a predetermined range by a predetermined angle to obtain a continuous reconstructed image, the fan beam measurement data is converted into a parallel beam. Dividing the obtained 180 ° parallel beam projection data at a predetermined projection angle to obtain a plurality of divided constituent images respectively corresponding to each divided section, and a plurality of newly formed divided reconstructions A first weight multiplying step for assigning a first weight to an image, a second weight multiplying step for assigning a second weight to the immediately preceding image frame, and an immediately preceding image to which the first weight is assigned Composing a frame and a plurality of newly constructed divided reconstructed images given the second weight to form one regular reconstructed image.
The first and second weight multiplying steps include: setting the first and second weights so that the sum of the weights of the divided reconstructed images whose view angles are opposite to each other becomes 1;
An image reconstruction method is disclosed. Further, the present invention provides an X-ray CT apparatus which obtains a continuous reconstructed image by shifting a reconstructed image having a projection angle in a predetermined range by a predetermined angle, and is obtained by converting fan beam measurement data into a parallel beam. Obtaining, from the parallel beam projection data, m divided reconstructed images corresponding to a predetermined projection angle; and assigning a predetermined weight to each of the m divided reconstructed images;
The weighted m divided reconstructed images are added to obtain 1
Reconstructing a single image.

[0019]

FIG. 4 shows a fan beam X-ray CT apparatus according to the present embodiment. A host computer 1 that controls the whole, a scanner 2 equipped with an X-ray generation system, an X-ray detection system, etc., capable of continuous scanning of 1 second / 1 rotation by a slip ring, preprocessing, image reconstruction processing and various analysis It comprises an image processing device 3 in charge of processing, a high voltage generator 4 for supplying a high voltage to X-rays, a patient table 5, and a display device 6. The host computer 1 controls the high-voltage generator 4 and the table 5 according to the set imaging plan.
When the preparation for imaging is completed and measurement of the subject is started, the measured projection data is sequentially transferred to the image processing device 3 and reconstructed, and the reconstructed image is displayed on the display device 6.

FIG. 1 shows an example of the internal configuration of the image processing device 3 and the display device 6. The image processing device 3 includes a reconstruction arithmetic unit 11 and a weighted image adder 12, and the display device 6
Are input interface 26, LUT (lookup table) circuit 27, display image memory 28, CRT2
9 The reconstruction arithmetic unit 11 includes a projection data memory 2
0, a pre-processing calculator 21, a fan beam-to-parallel beam converter 22, a filter correction calculator 23, and a back projection calculator 24. The weighted image adder 12 includes seven image memories 10 (# 1 to # 7) and seven weight coefficient multipliers 13 (W ₁
ＷW ₇ are weighting factors), and an image synthesizer 25.

The image processing apparatus 3 shown in FIG. 1 is capable of reconstructing an image in less than one second, for example. this is,
Sequentially obtaining divided reconstructed images one after another with a width of 30 °;
Furthermore, one reconstructed image can be achieved by adding the latest 30 ° -width divided reconstructed image newly obtained to the previous one. For example, if the angle is updated at a width of 30 °, 1
Two reconstructed images per second can be obtained.

The reconstruction operation unit 11 shown in FIG. 1 includes a pre-processing unit 21, a parallel beam conversion by a converter 22, and an operation unit 2.
The reconstructed image is obtained by performing the filter correction processing by 3 and the back projection calculation by the arithmetic unit 24. This reconstruction operation is 3
This is an operation for obtaining a divided reconstructed image obtained from parallel beam data with a partial angular width (for example, a 30 ° width), instead of a batch reconstruction for 60 °. The weighted image adder 12 sequentially generates the divided reconstructed images (g ₁ , g ₂ ,
..) Are sequentially assigned to the memories # 1 to # 7 of the image memory 10. For example, dividing the reconstructed image g ₁ is the # 1, g ₂
Is to # 2, ..., g ₇ stores assigned to # 7. After that, the divided reconstructed images g ₈ , g ₉ ,... Are stored by assigning g ₈ as # 1 instead of g ₁ and g ₉ as # 2 instead of g ₂ . The multiplier 13 for each memory # 1 to # 7 of the image, performs weighting of each divided image by multiplying in association with weighting coefficients W ₁ to _W-7. The synthesizer 25 performs total addition to obtain one reconstructed image.

Here, the number of the memories 10 is set to seven.
Basically, if the width for obtaining the divided reconstruction is 30 °, the minimum is six. The reason why the number is set to a minimum of six is that one reconstructed image can be obtained by half scan (180 °). (180 ° = 30 ° minute ×
6). By using half scan, the memory capacity can be reduced and the number of additions in the combiner 25 can be reduced as compared with 360 °, so that high-speed reconstruction can be achieved. Rotation angle (180 °)
Therefore, there is an advantage that reconstruction can be performed by measurement in half the time as compared with 360 °. in this case,
In the case of a scanner with one rotation per second (also referred to as a one-second scan), it is possible to output twelve weighted and added reconstructed images per second (that is, a frame rate of 12).
Sheets / sec. The frame rate is the number of reconstructed images that can be obtained in a unit time). Note that 180 ° means 180 ° for a parallel beam, and 180 ° for a fan beam is insufficient for reconstructing data, so it is 180 ° + α on the fan beam. Further, the parallel beams may be either at regular intervals or irregular intervals.

In the present invention, the basic memory 10 (# 1
To # 7) and weight W₁~ W ₇Setting example
It is described below. (1), W₁= W_Two= ... = W₆= 1 and W₇= 0
Example. This example is the case of FIG. In FIG. 2A,
The parallel beam projection data P is a fan-parallel beam converter.
Output of each P₁, P_TwoReconstructed image from…
g₁, G_Two, ... g_n~ G_{n + 5}Weight coefficient W is
Reconstructed image r_nSegmentation image used to obtain
Statue g_n~ G_{n + 5}Is a weighting factor for. (A), first reconstructed image r₁The steps to get
It is on the street. In the memory 10 # 1, a 0 ° to 30 ° split
Composition image g₁Is stored. Next, a division of 30 ° to 60 °
Reconstructed image g_TwoIs stored in # 2 of the memory 10. As well
And image g_ThreeTo # 3, image g_FourTo # 4, image g_FiveTo # 5,
Image g₆Is stored in # 6. Nothing is stored in # 7.
After storing in # 1 to # 6, images of # 1 to # 6
Is taken out through the multiplier 13 and added by the combiner 25.
And a regular reconstructed image in a half scan of 0 to 180 °
Image r₁Get. (B) Then, the divided reconstructed image 7 of 180 ° to 210 ° is
Stored in # 1. At this time, the image g stored in (a)₁Is
By clearing in advance, the image g₁Picture on behalf of
g₇Is stored. Next, the images g of # 1 to # 6 again₇, G
_Two~ G₆Is read out, the multiplier 13 is obtained, and added by the synthesizer 25.
The next regular reconstructed image r_TwoGet. (C) Next, a divided reconstructed image g of 210 ° to 240 °₈
Is stored in # 2. Naturally, the image stored before # 2
Keep 2 clear. Similarly, the image is synthesized by the synthesizer 25.
g₇, G₈, G_Three~ G₆Of the next regular reconstructed image
Image r_ThreeGet. (D) Hereinafter, similarly, one reconstructed image r one after another
_Four, R_Five, ... By the above procedure, the second reconstructed image r
_TwoIs r₁From 180 ° to 210 ° for 30 ° flat
Immediately after obtaining the row data, image g₇Time to get
It can be calculated in a short time determined by the time and the addition time. g
_Three, G_Four, ... are the same.

(2) An example in which data having opposite view angles is averaged. This example is shown in FIGS. 3A and 9A. FIG. 9A shows a change in weight for the divided reconstructed image of FIG. 3A. (A), This example is an example in which the weights W _{1 to} W ₇ are changed every time a divided reconstructed image for 30 ° is obtained. How to change
Changes are made so that the sum of the weights becomes "1" for the images that are in a relationship facing each other. This weighting is performed in order to reduce edge artifacts by taking a weighted average by adding data in the opposite relationship to each other. (B) First, a first reconstructed image r of 0 ° to 180 °
_In order to obtain ₁ (this is different in content from r _{1 in} FIGS. 2A and 2B, the same symbol is used, the same applies hereinafter), W ₁ = W ₇
= １ ／, and W _{2 to} W ₆ are 1. In this case, g ₁ , g ₇
Are opposed to each other. Under this setting, 0 is stored in ° to 30 ° divided reconstructed image g ₁ to # 1. Below, 30 ° ~
The divided reconstructed images g ₂ , g _3, ..., G ₇ of 60 °, 60 ° to 90 °,..., 180 ° to 210 ° are represented by # 2, # 3,.
7 is stored. Then, a weighting multiplier in the multiplier 13 reads out these images are added in combiner 25 to obtain a reconstructed image r ₁ regular. In this image r _1, the image g ₁ and the image g ₇ is data in the opposite relationship to each other view angle, adds the counter data, a weighted average value by ized 1/2. Thus, the discontinuity at the end was reduced, and the artifact due to the discontinuity at the end could be reduced. (C) Then, to obtain the second reconstructed image r ₂ , W ₁ = W
₂ = 1/2 and W _{3 to} W ₇ are set to 1. In this case, g ₂ and g
₈ are facing each other. Under this setting 210 ° ~ 24
The 0 ° divided reconstructed images g ₈ stored in # 1. On this occasion,
# Image g _1, which had been stored in the 1 to clear. And
# Image g ₈ of _{1~ # 7, g 2, g} 3, ... the g _7, adding at the combiner 25 via the multiplier 13. At this time, the image g
And ₈ and g ₂ become opposite relationship, are averaged to obtain an image r ₂ with reduced discontinuities in the end. (D) To obtain an image r ₃ , W ₂ = W ₃ = １ ／, W ₁ ,
W _{4 to} W ₇ are set to 1. In this case, g ₃ and g ₉ are in opposed relationship. To obtain an image _{r 4 W 3 = W 4 =} 1/2, W 1, W
_2, W ₅ ~W ₇ is set to 1. Similarly, every time a reconstructed image is successively obtained, the weight is changed each time in a short time period of the switching.

(3) FIG. 2B shows an example in which one reconstructed image is obtained every time a two-divided image is obtained. W _{1 to} W ₆ are 1, W ₇ = 0
To (A) In order to obtain the first reconstructed image r ₁ , 0 ° to 30 °
An image g ₁ of ° # 1,30 ° to 60 ° image g ₂ and # 2,
..., and stores the image g ₆ of 0.99 ° to 180 ° to # 6.
Then, by reading and adding, an image r ₁ is obtained.
The procedure so far is the same as the example of FIG. (B) In order to obtain the second reconstructed image r ₂ , 180 °
An image g ₇ of 210 ° and stored in the # 1. In FIG. 2A, the addition for the reconstruction is performed after this, but in this example, the addition is not performed at this time, and the image g of 210 ° to 240 ° is not added.
₈ is stored in # 2. When the new images g ₇ and g ₈ are obtained, the images g ₇ , g ₈ and g _{3 to} g ₆ are added and the reconstructed image r
_{Get two} . (C) Hereinafter, similarly, each time two continuous divided images are obtained, addition for reconstruction is performed, and a reconstructed image r ₃ ,
r ₄ , ...

(4), FIG. 3 (b) and FIG. 9 (b) show examples in which the discontinuity at the end is further reduced. FIG. 9B shows a change in weight for the divided reconstructed image of FIG. 3B. In this example, the memory 10 is 8 (# 1 to # 8), and a multiplier 13 of the weight coefficients W ₁ to _W-8 is also eight prepared. (A) First, when the first reconstructed image r1 is obtained, W ₁ =
４, W ₇ = ３, W ₂ = １ ／, W ₈ = １ ／, W ₃
ＷW ₆ is set to 1. In this case, g ₁ and g ₇ , g ₂ and g ₈
Are opposed to each other. The 0 ° to 30 images g ₁ of ° # 1,30 ° to 60 ° image g ₂ and # 2, ..., 18
0 ° to 210 ° image g ₇ to # 7, 210 ° to 240
° pictures g ₈ to # 8, are stored. And the multiplier 13
The image r ₁ is obtained by reading out and adding via the. (B) When obtaining the second reconstructed image r2, W ₁ = W ₃ =
、, W ₂ １ ／, W ₈ = ３, and W _{4 to} W ₇ are set to 1. In this case, g ₂ and g _8, g ₃ and g ₉ are in opposed relationship respectively. Then, the image g ₉ of 240 ° to 270 °
Was stored in the # 1 adds, to obtain an image r _2. Similarly, r ₃ , r ₄ ,... Are obtained. As an example of the weight for the divided image data that is opposed to each other, various distributions can be considered while satisfying the condition that the sum is 1, but when real-time performance is important, the weight for the latest data is increased. It is desirable to set a large value.

(5) An example of a 60 ° section instead of a 30 ° section. This is an example shown in FIG. This is characterized by a 60 ° section. In order to add data in which the view angles are opposed, parallel projection data at opposed positions (for example, 180 ° to 240 ° for 0 ° to 60 °)
Is stored in the memory 10. However, four memories # 1 to # 4 may be used.

In the above examples, (2), (4), (5)
Needed to set a weighting factor for each frame. There are ways to achieve similar goals without doing so. This is an example of recursive filtering. FIG. 6 is a configuration diagram thereof. 1 in that the number of memories 10 is six (# 1 to # 6), the seventh memory # 7 is a previous frame memory 33, multipliers 30 and 31, and a combiner 32.
And the multiplier 13 on the output side of the memory 10
And that the memory 10 and the like are provided on the display device 6 side. The multiplier 30 is a means for simple averaging, this a simple averaging results, and to perform addition of the reconstructed image r _i immediately before at the synthesizer 32.
The output of the synthesizer 32 is sent to the display image memory 28 and, after the multiplier 31 data weighting (W ₂ ), is stored in the memory 33 for calculation of the next reconstructed image ri _{+ 1.} . The weights W ₁ and W ₂ have a relation of W ₁ + W ₂ = 1. The average of the previous image A larger value of W ₂ is increased becomes much residual image, while looking at the image, W ₁ and W ₂
Adjust the value of.

According to this configuration, there is an advantage that only _two weights, W ₁ and W ₂ , are required because the recursive connection is a simple average. The operation of the configuration of FIG. 6 will be described with reference to FIGS. 7A to 7C, the upper part is the memory 33.
And indicates the immediately preceding image frame. The lower stage is the output of the multiplier 30 and represents a new image frame composed of a plurality of newly formed divided reconstructed images. First, weights W ₁ = 1 and W ₂ = １ ／ are initially set. At the time of this initial setting, the memory 33 is cleared. Then, the measurement sequence starts. In the divided image memory 10, 0 ° to 3 are stored in # 1.
0 °, 30 ° to 60 ° for # 2, ..., 150 ° to 1 for # 6
The 80 ° divided reconstructed images g _{1 to} g ₆ are sequentially sent to the measurement sequence and stored in the calculated order. These images g _{1 to} g ₆ are read out, added by the synthesizer 25, and passed through the multiplier 30. Since W ₁ = 1, the added image of the images g _{1 to} g ₆ is input to the synthesizer 32 as it is. At this stage, since the contents of the memory 33 have been cleared, the output of the multiplier 30 is directly output from the synthesizer 32. This output is a regular reconstructed image r _1, which is sent to the memory 28 and stored in the memory 33 via the multiplier 31. Since W ₂ = １ ／, the reconstructed image r ₁ is halved and stored in the image memory 33. At the time of displaying the first image, a fixed weight of W ₁ = W ₂ = １ ／ may be used for reasons such as speeding up.

Next, the process for obtaining the second reconstructed image r ₂ is started. At this time, the weight coefficient W ₁ = W ₂ =＝ is set. First, the parallel beam data and the divided reconstructed image g ₇ thereunder of 180 ° to 210 °, reconstruction operation unit 1
1 and stored in # 1 of the memory 10. At this time, the image g ₁ of # 1 of the memory 10 which has been stored earlier than that you reset, only the new image g ₇ is stored in # 1. Then, 1/2 of a multiplier 30 in terms of the sum in the synthesizer 25 reads out the image g _7, g ₂ to g ₆ of # 1 to # 6. In the synthesizer 32, the output of the multiplier 30 and the memory 3
3 (# 7) output (a value of 1 half of the previous reconstructed image r ₁₎ and the sum was (see FIG. 7 (a)), and sends the second reconstructed image r ₂ as the memory 28 At the same time, 1
/ 2 phased on the memory 33 in (# 7), on behalf of the previous reconstructed image r ₁ stores for the next reconstructed image r ₃ calculated.
For calculation of a reconstructed image including the next reconstructed image r ₃ , W ₁ = W ₂ = １ ／ may be set, and the weight is not changed.

According to such an operation, the divided reconstructed images g _{2 to} g ₆ are commonly included in the first and second images, and therefore, as shown in FIG. Only ₁ and g ₇ are 1 /
This means that weighting of 2 has been performed. When obtaining the second reconstructed image r _2, as shown in FIG. 7B, g ₁ and g ₇ have a weight of ４ as in the case of FIG. when obtaining a structure image r ₃ becomes a weight of 1/8 for g ₁ and g ₇ as shown in FIG. 7 (c), slide into decay as slowly image. Image g ₂ and g _8, g ₃ and g ₉ even slide into attenuated similarly sequentially. Further, FIG.
In FIG. 7, g ₂ and g ₈ in the opposite relationship have a weight of 、, and similarly, in FIG. 7C, g ₃ and g ₉ in the opposite relationship have a weight of １ ／. Therefore, by adding the recursive filter to the 180-degree reconstructed image, it is possible to reduce artifacts and noise by averaging the facing data.

In FIG. 6, a multiplier for multiplying a weight coefficient is not provided on the output side of # 1 to # 6 of the memory 10.
All the output sides of # 1 to # 6 may be provided in the same manner as in FIG. The same applies to the output side of the memory 33 in FIG. 1 and 6, and also on the deformed side described above, the setting of each weight can be arbitrarily set by the effect of reducing noise and artifacts.

FIGS. 8A and 8B show the configuration of the weight setting unit. The weight setting unit includes the multipliers 13, 31, 32
It is a circuit for determining a weight coefficient such as. FIG. 8 (a)
Is an example of a single weight table 50, which is a memory that stores the division coefficient numbers as addresses and the weight coefficients W ₁ , W ₂ ,... Corresponding to the numbers as data. The division section numbers are 0 ° to 30 ° as numbers 1 and 30 ° to 60 °.
Is set as number 2,..., Which can be obtained directly from the control signal of the scanner or via an encoder or the like.

FIG. 8B shows weight tables 51 and 52,
53 is an example having a weight table according to the reconstruction mode. For example, the memory for performing the operation in FIG. 2A is the table 51, the memory for performing the operation in FIG. 2B is the table 52, the memory for performing the operation in FIG. ... is provided. Therefore, a reconstruction mode signal (this is an input setting signal, and any one of the selection designation signals shown in FIGS. 2A, 2B, 3A, 3B, and 3C) , One of the tables 51 to 53 is selected, and the division section number signal (in the case of division by 30 °, 0 to 30 °, 30 ° to 60 °,.
2, etc., and 0 ° to 6 if the section is 60 °
0 °, 60 ° to 120 ° are set as numbers 1, 2,...), And the corresponding weight coefficients W ₁ , W ₂ ,. Of course, in the example also the weighting factor 6, Toka only two W _1, W _2, in Figure 1, the W ₁ to _W-6 6
Individual and various.

[0036]

According to the present invention, it is possible to reconstruct a continuous image at a high speed when the image is continuously or continuously taken by the X-ray CT apparatus. It also reduces artifacts and noise during ultra-low-dose radiography such as medical examinations. Further, the frame rate (sheets / second) can be freely changed, and the frame rate can be freely selected and set in accordance with the purpose of display or diagnosis.

[Brief description of the drawings]

FIG. 1 is a structural example of a main part of an X-ray CT apparatus according to the present invention.

FIG. 2 is a diagram illustrating a setting example of a weight coefficient in the X-ray CT apparatus according to the present invention.

FIG. 3 is a diagram illustrating another example of setting a weight coefficient in the X-ray CT apparatus according to the present invention.

FIG. 4 is an overall configuration diagram of the X-ray CT apparatus of the present invention.

FIG. 5 is an explanatory diagram of a conventional example.

FIG. 6 is another configuration example of a main part of the X-ray CT apparatus of the present invention.

FIG. 7 is a diagram showing an example of another weighting factor of the X-ray CT apparatus of the present invention.

FIG. 8 is an example of a weight table according to the present invention.

FIG. 9 is a diagram showing a weight change for the divided reconstructed image of FIG. 3;

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Host computer 2 Scanner 3 Image processing device 4 High voltage generator 5 Patient table 6 Display device 10 Image memory 11 Reconstruction arithmetic unit 12 Weighted image processing device 13 Multiplier

Claims (19)

[Claims] 1. An X-ray CT for obtaining a continuous reconstructed image by sequentially shifting a reconstructed image having a projection angle in a predetermined range by a predetermined angle.
An X-ray CT apparatus that divides a group of parallel beam projection data obtained by converting fan beam measurement data into parallel beams at a predetermined projection angle,
A storage device for storing a plurality of divided reconstructed images respectively corresponding to the respective divided sections; a plurality of multipliers for multiplying each of the plurality of divided reconstructed images by a predetermined weight coefficient; and the plurality of multipliers An X-ray CT apparatus, comprising: a synthesizer that synthesizes an output of (1) to obtain a normal reconstructed image. 2. The X-ray CT apparatus according to claim 1, wherein the storage device sequentially selects and stores a plurality of continuous divided reconstructed images such that some divided reconstructed images overlap. X-ray CT apparatus including: 3. The X-ray CT apparatus according to claim 2, wherein the plurality of multipliers for multiplying the weighting coefficient are configured such that the sum of the weights of the divided reconstructed images whose view angles are opposite to each other is the other divided image. An X-ray CT apparatus including a weight coefficient set to be equal to a weight of a reconstructed image. 4. The X-ray CT apparatus according to claim 1, wherein the storage device includes a memory for storing a plurality of divided constituent images corresponding to 180 ° projection data.
T device. 5. A fan-parallel beam converter for continuously rotating the fan beam X-ray source at least two times or more and converting the fan beam measurement projection data sequentially obtained by the rotation into a parallel beam; An arithmetic unit that divides the output of the parallel beam converter and forms a divided reconstructed image for each section; and sequentially selects a plurality of continuous divided reconstructed images so that some of the divided reconstructed images overlap. A plurality of multipliers connected to the output side of the storage device for multiplying the stored plurality of divided reconstructed images by a weighting factor; and An X-ray CT apparatus, comprising: a synthesizer configured to form a reconstructed image. 6. The X-ray CT apparatus according to claim 5, wherein the computing unit includes a unit configured to form the divided reconstructed image for each of N divided sections of the projection data of 180 °. The plurality of multipliers include a weight coefficient set so that the sum of the weights of the divided reconstructed images whose view angles are opposite to each other is equal to the weight of the other divided reconstructed images. An X-ray CT apparatus including a unit configured to form the normal reconstructed image using the divided reconstructed image. 7. A 180 ° parallel beam projection data obtained by converting fan beam measurement data into a parallel beam is divided at a predetermined projection angle, and a plurality of divided reconstructed images respectively corresponding to each divided section. A frame memory for storing the immediately preceding image frame; a first multiplier for weighting a plurality of newly-constructed divided and reconstructed images; and a weighting unit for weighting the immediately preceding image frame. A second multiplier, and a combiner that combines the weighted image frame immediately before and the plurality of newly formed divided reconstructed images to form one normal reconstructed image. An X-ray CT apparatus, wherein the first and second multipliers include means for assigning weights so that the sum of the weights of the divided reconstructed images whose view angles are opposed to each other becomes one. CT device. 8. An X-ray CT for obtaining a continuous reconstructed image by sequentially shifting a reconstructed image having a projection angle in a predetermined range by a predetermined angle.
An X-ray CT apparatus for obtaining m divided reconstructed images corresponding to a predetermined projection angle out of parallel beam projection data obtained by converting fan beam measurement data into parallel beams. Weighting means for assigning weights to the m divided reconstructed images; and adding the weighted m divided reconstructed images to 1
An X-ray CT apparatus comprising: synthesizing means for reconstructing a single image; and means for displaying the reconstructed image, wherein the synthesizing means is configured to shift the projection angle by α ゜. An X-ray CT apparatus including means for obtaining a continuous reconstructed image from a constituent image. 9. An X-ray CT apparatus according to claim 8, wherein said weighting means includes means for changing a weighting factor and assigning weight to a predetermined divided reconstructed image. 10. The X-ray CT apparatus according to claim 8, wherein the means for obtaining the divided reconstructed image includes a means for setting an angle exceeding 180 ° as the predetermined projection angle.
apparatus. 11. An X-ray C that obtains a continuous reconstructed image by sequentially shifting a reconstructed image of a predetermined range of projection angles by a predetermined angle.
In the T device, the parallel beam projection data obtained by converting the fan beam measurement data into parallel beams is divided at a predetermined projection angle, and a plurality of divided reconstructed images corresponding to each divided section are obtained. Multiplying each of the plurality of divided reconstructed images by a predetermined weighting factor; and combining the plurality of divided reconstructed images multiplied by the weighting factor to obtain one reconstructed image. An image reconstruction method, comprising: 12. The image reconstructing method according to claim 11, wherein the step of obtaining the divided reconstructed images is performed such that a plurality of consecutive divided reconstructed images are successively arranged so that some of the divided reconstructed images overlap. An image reconstruction method comprising the step of selecting and passing to a step of multiplying by a weighting factor. 13. The image reconstructing method according to claim 12, wherein the step of multiplying the weighting coefficient is performed such that a sum of weights of the divided reconstructed images whose view angles are opposite to each other is equal to that of another divided reconstructed image. Image reconstructing method including the step of setting the weight to be equal to 14. The image reconstructing method according to claim 11, wherein the step of obtaining the plurality of divided reconstructed images comprises:
An image reconstruction method including a step of obtaining a plurality of divided reconstructed images corresponding to 80 ° of projection data. 15. The image reconstructing method according to claim 14, wherein the step of combining the plurality of divided reconstructed images to obtain one reconstructed image corresponds to 180 ° + α ° of projection data. An image reconstruction method including a step of combining a plurality of divided reconstructed images to obtain one reconstructed image. 16. An X-ray C that obtains a continuous reconstructed image by sequentially shifting a reconstructed image having a projection angle in a predetermined range by a predetermined angle.
In the T device, the parallel beam projection data of 180 ° obtained by converting the fan beam measurement data into a parallel beam is divided at a predetermined projection angle, and a plurality of divided configurations corresponding to each divided section. Obtaining an image, a first weight multiplying step of giving a first weight to the plurality of newly formed divided reconstructed images, and a second weight of giving a second weight to the immediately preceding image frame Multiplying the image frame immediately before the first weighted image and the plurality of newly constructed divided reconstructed images to which the second weight is applied to form one normal reconstructed image And a weighting multiplying step, wherein the first and second weight multiplying steps are performed such that the sum of weights of the divided reconstructed images whose view angles are opposed to each other becomes one. 1st, 2nd Image reconstructing method comprising the steps of: setting a weight, a. 17. An X-ray CT apparatus which obtains a continuous reconstructed image by shifting a reconstructed image having a projection angle in a predetermined range by a predetermined angle and converting the fan beam measurement data into a parallel beam. Obtaining, from the parallel beam projection data, m divided reconstructed images corresponding to a predetermined projection angle; and applying a predetermined weight to each of the m divided reconstructed images; Addition of m divided reconstructed images gives 1
Reconstructing a number of images. 18. The image reconstruction method according to claim 17, wherein the step of obtaining the divided reconstructed images includes the step of obtaining the m divided reconstructed images each time the projection angle is shifted by α ゜. Image reconstruction method. 19. The image reconstruction method according to claim 18, wherein the step of obtaining the divided reconstructed images comprises the step of reconstructing the m divided reconstructed images from parallel beam projection data corresponding to a projection angle exceeding 180 °. An image reconstruction method comprising the step of obtaining.