JPS59214435A - Computer tomographic apparatus - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION [Technical Field of the Invention] The present invention belongs to the technical field of tomography apparatuses, and relates to a computed tomography apparatus having an interpolation calculation device.

[Technical background of the invention and its problems]

Computed tomography apparatus (hereinafter also referred to as CT apparatus).

) For example, image reconstruction in an X-ray CT device involves rotating an X-ray tube and a detector, which are placed facing each other with the subject in between, while rotating the subject with the subject at the center. For example, a set of projection data (data obtained by one X1ill irradiation) obtained by a detector that detects the image is processed by a computer and converted into a so-called CT value. It will be done.

As shown in FIG. 1, projection data is obtained at every fixed rotation angle Δθ of the X-ray tube 1.

The smaller 80 is, the more projection data can be obtained, which improves the quality of the reconstructed image.

However, if there is too much projection data, it will take time to reconstruct the image. Therefore, conventionally, this Δθ is set to 0.
6 degrees or 1.2 degrees, in which case 600 projections or 300 degrees depending on each Δθ.
Project data has been obtained. Note that 2 is a detector.

When Δθ is 1.2 degrees (300 projections), 3
Shower-like artifacts may appear in re-created images obtained using 00 projection data, so in order to prevent this and improve image quality, each We create hypothetical projection data that would be obtained if the X-ray tube 1 emits X-rays in the middle of Δθ, and based on the projection data actually obtained by detection and the hypothetical projection data, Image reconstruction is performed using data from 600 projections. (Hereinafter, such processing will be referred to as projection interpolation.)

Furthermore, projection interpolation will be specifically explained. As shown in FIG. 3, the data collection unit 3 collects projection data obtained for each XI exposure by the rotating XM tube 1 into a central processing unit (JJ, sometimes abbreviated as CPU) 4. sequentially through the mass storage unit 5.
Store in. After all the projection data obtained when the X-ray tube 1 rotates around the subject is stored in the mass storage section 5, the CPU 4 sequentially reads out each projection data from the mass storage section 5, for example. (k is an integer from 1 to n, n indicates the number of X-ray irradiation) projection data, 1 + 1 projection data, and 2 fictitious projection data are created. The projection data and the fictitious projection data are outputted to the mass storage unit 5 again so that the fictitious projection data is stored after the projection data of . After the calculation to create the fictitious projection data is completed and all the projection data and fictitious projection data are stored in the mass storage unit 5, the C
The PLI 4 reads out the interpolated projection data from the mass storage unit 5 according to the storage type, and outputs it to the reconstruction unit 6 . The reconstruction unit 6 reconstructs a tomographic image of the subject based on the input projection data after interpolation, for example, by a superimposition method, a Fourier transform method, etc., and stores the reconstructed image data in an image memory 7. I'll keep it.

However, in the projection interpolation described above, calculations must be performed to create a fictitious projection in the CPL 14 before image reconstruction, and in addition, the actually collected projection data and the fictitious projection data must be A CPU for alternately sending the data to the reconfiguration unit 6 after storing it in the storage unit 5.
There was an excessive burden on the management of the programs within 4. Also,
Since calculation by the CPU 4 takes time, it takes a long time to reconstruct a tomographic image. Furthermore, the large capacity storage unit 5
is essential.

[Purpose of the invention]

This invention was made in view of the above circumstances, and
An arbitrary number of interpolated projection data can be created based on projection data obtained by radiation exposure, and images can be reconstructed in a short time using this interpolated projection data without using a large-capacity storage unit. The purpose is to provide a CT device.

[Summary of the invention]

The outline of the present invention for achieving the above object includes at least a data acquisition section that collects data obtained for each exposure of X-rays rotating around a subject, and a reconstruction section that reconstructs an image. In a combo-user tomography apparatus, a central processing unit transfers each projection data collected by the data collection unit, and a weighted average of the projection data transferred from the central processing unit and the next transferred projection data is calculated. The present invention is characterized by comprising an interpolation processing section that sequentially outputs an arbitrary number of averaged projection data to a reconstruction section.

[Embodiments of the invention]

First, the basic principle of this invention will be explained.

In the CT apparatus according to the present invention, an interpolation processing section is newly provided, and after the projection data from the data collection section transferred by the CPU is interpolated by the interpolation processing section, the interpolated data is input to the reconstruction section 6. . That is, as shown in FIG. 4, the projection data collected by the data collection unit 3 is once sent to the CPLI 4.

The CPLI 4 sends projection data as a transfer input to the interpolation processing section 8, as shown in FIG. The interpolation processing unit 8 takes a weighted average of the projection data transferred from the CPU 4 and the projection data sent at the immediately previous time, and sends it to the reconstruction unit 6, and the reconstruction unit 6 calculates the weighted average projection data. Image reconstruction is performed based on the data.

The basic principle will be explained in further detail.

Since the projection data consists of, for example, 320 channels of data (data equal to the number of detection elements), as shown in FIG. P (m, n) and P (m + 1.n) for each same channel in two projection data P (m, n), P (m + 1.n) obtained by X-ray irradiation by the ray tube 1.
) can be performed by taking a weighted average.

That is, the interpolation processing unit 8 executes the following calculation.

Q(k,n)=j!・P(J, rl)+(1-jri・P
(j-1, n) where k is the ordinal number of the projection after interpolation,
j is the ordinal number of the transferred projection data.

In this way, as shown in the time chart of FIG.
Projection data is continuously sent from the PU 4 as a transfer input to the interpolation processing unit 8 according to certain rules,
The interpolation process in the interpolation processing unit 8 is to perform weighted averaging of two consecutive projection data and to sequentially send out the weighted averaged projection data, which is the same projection interpolation as when using a mass storage device. It turns out.

Next, an example of the interpolation processing section 8 for realizing the basic principle will be described with reference to the drawings.

FIG. 7 is a block diagram showing the interpolation processing section 8. As shown in FIG.

In the same figure, the interpolation processing unit 8 has an input memory 8b that stores one projection data transferred from the CPU 4 at an address indicated by a first address counter 8a.
Then, one projection data stored in the input memory 8b is read out at high speed, and the second address counter 8C is read out.
a buffer memory 8d that stores one projection data read out for each channel at an address corresponding to the address in the input memory 8b specified by , and a buffer 7 memory 8d.
A first multiplier 8e that multiplies the data read out for each channel sequentially from channel 0 by a weight l, and a memory that temporarily stores and stores one projection data read out from the buffer 7 memory 8d for each channel. 8
f, a register 8Q that reads and temporarily stores one projection data stored in the memory 8f, and a weight (1- J
); an adder 8h that adds the data output from the first multiplier 8e and the data output from the second multiplier 8j; 2. An output memory 81 for storing the obtained projection data at the address indicated by the address counter 8C and outputting the obtained projection data to the reconstruction unit 6 at high speed, and first and second addresses for specifying addresses to each of the memories and registers. It is composed of counters 8a and 8c.

Next, the operation of the above-mentioned component device will be described.

For convenience of explanation, 300 types and 400 types are stored in the input memory 8b.
A set of projection data is input and output memory 8
A case will be described in which 1 to 400 sets of projection data are output.

While rotating the X-ray tube 1 about the body axis of the subject, the X-ray tube 1 irradiates the subject with X-rays at every predetermined rotation angle Δθ, for example, 1.2 degrees. The X-rays that have passed through the subject are detected by the detector 2, and the detected X-ray transmission data is sent to the data collection unit 3.
, and output to the CPU 4 as projection data.

The CPU 4 inputs the input projection data into the eighth
In accordance with the transfer sequence shown in the figure, the input memory 8
Transfer to b. That is, the first projection data PO is transferred twice in succession, and the second projection data P1 and third projection data P2 are transferred once each. Similarly, the un-1 <
However, (1-n-1=θn+3, (11=1> projection data Pn is transferred twice consecutively, and projection data Pn- other than cLn-1)
1 and Pn-2 are transferred once each. The first projection data PO transferred from the CPU 4 to the input memory 8b is stored in the buffer memory 8d by the first address counter 8a. Next, the first projection data PO stored in the buffer 7 memory 8d by the second address counter 8G is transferred to the first multiplier 8e and the memory 8f. When the first projection data PO is output from the buffer memory 8d, the first address counter 8a transfers the second projection data PO (same as the first projection data PO) from the input memory 8b to the buffer door memory 8d. content) is transferred. At this time, the first multiplier 8e applies the weight f generated by the constant generator 8j to the data of each channel in the first projection data PO.
Multiplyed by f1=1/4, multiplication result 1-PO is added to adder 8
It is output to h. The adder 8h receives the multiplication result! - Add PO and the output of the second multiplier 8j. However, since the output of the second multiplier 8j at this time is meaningless, the output data of the adder 8h is not used for image reconstruction. Multiplication result from the first multiplier 8e! -P
When O is output, the 1'th projection data PO stored in the memory 8f is transferred to the register 8g.
At the same time, the second projection data PO stored in the buffer memory 8d is output to the first multiplier 8e and the memory 8f. When the second projection data PO is output from the buffer memory 8d, the first address counter 8a transfers the third projection data P1 from the input memory 8b to the buffer memory 8d. At this time, the first multiplier 8e multiplies the data of each channel in the second projection data PO by weight loss -0/4, and outputs the multiplication result PO to the adder 8h. Also,
The first projection data PO stored in the register 8g is output to the second multiplier 8j, and the second multiplier 8j applies a weight (1) to the data of each channel in the first projection data PO. −jri=
'4/4 is multiplied, and the multiplication result (1-1)/PO is output to the adder 8h. The adder 8h adds the multiplication result PO by the first multiplier 8e and the multiplication result (1-1) PO by the second multiplier 8j, and outputs the addition result as interpolated projection data Q1. memory 81
Output to. The multiplication result from the first multiplier 8e is P
When O is output, the second projection data PO stored in the memory 8f is transferred to the register 8g, and the third projection data P1 stored in the buffer memory 8 is transferred to the first projection data PO. is output to the multiplier 8e and the memory 8f. Buffer memory 8
When the third projection data P1 is output from d, the first address counter 8a controls the input memory 8.
The fourth projection data P2 is transferred from b to the buffer memory 8d. At this time, the first multiplier 8
In e, the data of each channel in the third projection data P1 is multiplied by the weight l-3/4, and the multiplication result! - Pl is output to the adder 8h.

Further, the second projection data PO stored in the register 8q is output to the second multiplier 8j, and the second multiplier 8j weights the data of each channel in the second projection data PO. (1-
It is multiplied by j = 1/4, and the multiplication result (1-j)/PO is output to the adder 8h. The adder 8h multiplies the multiplication result Pl by the first multiplier 8e and the second multiplier 8.
The multiplication result (1-1) by j is added to PO, and the addition result is output to the output memory 81 as interpolated projection data Q2. When the first multiplier 8e outputs the multiplication result Pl, the third projection data 1 stored in the memory 8f is transferred to the register 8g.
At the same time, the fourth projection data P2 stored in the buffer 7 memory 8d is output to the first multiplier 8e and the memory 8f. Fourth projection data P2 from buffer memory 8d
is output, the first address counter 8a transfers the fifth projection data P3 from the input memory 8b to the buffer memory 8d. At this time, the first multiplier 8e transfers the fourth projection data P2.
The data of each channel in is multiplied by weight depletion=2/4, and the multiplication result 1-P2 is output to the adder 8h.

Further, the third projection data P1 stored in the register 8g is output to the second multiplier 8j, and the second multiplier 8j weights the data of each channel in the third projection data P1. (11
)=2/4, and the multiplication result (1-1)·Pl is output to the adder 8h. The adder 8h receives the multiplication result by the first multiplier 8e! - P2 and the second multiplier 8j
The multiplication result (1-1M Pl) is added and the addition result is output to the output memory 81 as interpolated projection data Q3. When the first multiplier 8e outputs the multiplication result P2, the memory The fourth projection data P2 stored in 8f is stored in register 8.
At the same time, the fifth projection data P3 stored in the buffer memory 8d is output to the first multiplier 8e and the memory 8f. When the fifth projection data P3 is output from the buffer memory 8d, 1. The sixth projection data P3 is transferred from the input memory 8b to the buffer memory 8d by the first address counter 8a. At this time, the first multiplier 8e outputs the fifth projection data P3.
The data of each channel in is multiplied by a weight Z=1/4, and the multiplication result P3 is output to the adder 8h.

Further, the fourth projection data P2 stored in the register 8g is output to the second multiplier 8j, and the second multiplier 8j weights the data of each channel in the fourth projection data P2. (11
>=3/4 is multiplied, and the multiplication result (1-1>・P2 is output to the adder 8h. Multiplier 8j
The multiplication result (1 to ri・P2 is added, and the addition result is interpolated projection data. 4 is output to the memory 81.
Output to.

Hereafter, interpolated projection data. The interception processing unit 8 follows the same data transfer sequence as generating 1 to Q4.
Data interpolation processing will be performed in . As a result, 400 interpolated projection data can be obtained from 300 pieces of projection data, and when an image is reconstructed using 400 pieces of interpolated projection data, the image is reconstructed using 300 pieces of projection data. accuracy can be increased.

Although one embodiment of the present invention has been described in detail above, this invention is not limited to the above-mentioned embodiment, and can be implemented with appropriate modifications within the scope of the gist of the invention. Not even.

For example, instead of transferring the same projection data from the CPU 4 to the input memory 8b twice in succession, the
Input projection data once from U4 memory 8b
is transferred to the input memory 8b, and stored in the buffer memory 8b.
The signal may be input to the first multiplier 8e via d. In this way, the transfer from the CPLI 4 can be further speeded up.

Further, the interpolation processing unit in the present invention can generate N projection data from N projection data, and in this case, the weight multiplied by the multiplier of cylinder 1 satisfies the following relational expression. It can be set as

K=k·Δ−n where n=[k−Δk], Δ=N/, and k is a sequence number.

〔Effect of the invention〕

According to this invention, the following effects can be achieved. That is, since the CPU only needs to transfer the projection data, the program contents can be simplified and software management can be made easier. Since a large-capacity storage unit is not required, the time required for image reconstruction can be significantly reduced. Therefore, in the CT apparatus according to the present invention, tomographic images can be converted into CRT images in a short time.
can be displayed. Furthermore, since the image is reconstructed using a larger number of interpolated projections than the number of input projections, it is possible to obtain an image with high precision and uniform quality.

[Brief explanation of drawings]

Figure 1 is an explanatory diagram showing the relationship between the X1ill irradiation position by a rotating X-ray tube and projection data, and Figure 2 shows the relationship between a fictitious X-ray irradiation created by projection interpolation and its fictitious projection data. 3 is a block diagram showing a conventional CT device, FIG. 4 is a block diagram showing a CT device according to the present invention, FIG. 5 is an explanatory diagram for explaining a data transfer sequence in the present invention, FIG. 6 is an explanatory diagram showing projection interpolation using data for each channel in this invention;
FIG. 7 is a block diagram showing an interpolation processing section according to an embodiment of the present invention, and FIG. 8 is an explanatory diagram showing a data transfer sequence in the interpolation processing section. 3... Data collection section, 4... CPU
. 6... Reconstruction unit, 8... Interpolation processing unit. Figure 7

Claims (1)

[Claims] In a computed tomography apparatus having at least a data collection unit that collects data obtained for each exposure of X-rays rotating around a subject, and a reconstruction unit that reconstructs an image, the data collected by the data collection unit A central processing unit that transfers each projection data takes a weighted average of the projection data transferred from the central processing unit and the next transferred projection data, and sequentially processes any number of projection data obtained by weighted averaging. A computerized tomography apparatus comprising: an interpolation processing section that outputs an output to a reconstruction section.