JPS58206726A - Image treating 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 The present invention relates to an image processing apparatus, in particular, to produce a large amount of projection data by performing X-ray projection from multiple directions on a specific cross section of a subject, and to perform reconstruction calculation processing based on this data. The present invention relates to image processing in a CT apparatus that obtains image information corresponding to the X-ray absorption rate inside a specimen.

The reconstruction algorithms of the xtsc'r device can be divided into the following two types. The calculation time is long due to the large number of repetitions, and in the case of the number of projection data of 180 to 360, which is common in current CT equipment,
) is not used because it is less accurate than the analytical method of
ion: filtered back projection) method is used. However, when the number of projection data is small, the successive approximation method is more accurate, and especially when there is noise, the MEM method (Maximum Entropy Me
thod: Maximum entropy method) has been shown to be effective [Book theory (D) VO6, J63-D
49, Image Reconstruction from Projection by Maximum Entropy Method].

When reconstructing a cross-sectional image including the heart using these methods, only a blurred image of the moving heart is obtained. This is because one scan takes 2.5 to 6 seconds. Then the heart),! Several reconstruction algorithms have been proposed to utilize the d-phase movement to reconstruct the shape of a specific phase of the heart.

First, in (1), the FBP method is applied using only projection data of a specific synchronized phase of the heart. One cycle of the heart is divided into 7 equal parts, and the projection data and the data of 1 part of the interval corresponding to the corresponding phase are 2/7 in total.
This data is considered and used as a synchronized arc. There is not enough data with this amount, so let's smooth out the phase.
This method uses 2/7 of the data, which is 4/7 of the data in total, by scanning twice.

Second, even with normal CT reconstruction methods, parts other than the heart can be reconstructed with high accuracy. After first reconstructing the whole using the FBP method, projection data of only the heart is obtained using data other than the heart. Next, there is a method of reconstructing the image using FBP using the same projection data of only the heart.

(1), ■) In either method, the number of projection data is small to use the FBP method, and the number of projected arcs is m'
There is noise due to V other than the arc synchronized with r-itc (using the pitch near the peak), and the reproduced Id! I could not say that the negative value was sufficient.

The present invention was made in order to overcome the above-mentioned drawbacks, and after reconstructing the whole using the FBP method, only the heart is reconstructed using the successive approximation method using projection data of the heart, which has a small number of projection data and a lot of noise. The F 13 P method is used for the still image part with a large number of projections, and the successive approximation method is used for the noisy part with a small number of projections to eliminate each other's shortcomings. In other words, since there is little projection data for the moving image part including the heart, we use the successive approximation method, which is effective when the number of projection data is small and there is a lot of noise. It is intended to be applicable only to moving image parts. The increase in gttT time, which is a drawback of the successive approximation method, can be dealt with by reducing the amount of calculation to t7C since the application area of this method is 1i inversion.

The present invention will be explained more specifically in the following seven examples.

FIG. 1 is a block diagram showing the configuration of a device according to the present invention. 1 is a data output unit for collecting projection data, and 2 is an electrocardiograph that obtains 6 points to determine which phase of the heart the projection data from each 1:1 device belongs to. ,3
41 is a computer that compiles Yuken's calligraphy and compiles projection data based on the collected data, and L is used to display the calculation results.
! 11 is a disability display device; 5 is a storage device for storing images and calculation results; and 6 is a human-powered device for specifying each of the 41 parameters and areas.

For the Higashimerada projection data in 1, based on the electrocardiogram obtained by 1 in 2, the calculator 3 calculates t, t, for each class with the same phase.
IG (name tag) will be attached.

That is, as shown in Fig. 2, the area between 11 and 1L in the electrocardiogram is divided into n (n is an integer) equal parts, each of which is 1, 2.3...t.
agi, and data projected onto the same tag section are assumed to have the same phase.

The operation of the present invention in the above configuration will be described above with reference to FIGS. 3 and 4. A certain range L at every predetermined angle (for example, every 1 degree)! Normal filtered back projection (FBP) is performed using projection data (shown in FIG. 4A) taken at an angle of 180° (for example, 180°) to perform mys reconstruction. The image obtained at this point is correct in the area excluding the heart, but since the heart is moving, only an averaged and blurred image is obtained (see Figure 4 (B)). (shaded area).

Next, select an appropriate region containing the heart (hereinafter referred to as ROI region)
is specified using the manual device 6, and projection data of only that ROI region is calculated. Now from projection data from each direction! T'
By subtracting the projection data excluding the heart part calculated by 1' (7] ) is obtained.

Since each projection arc has a tag in advance indicating which phase of the heart it was projected at, only the tOI area projection data of the phase to be reconstructed is selected. If it is 1 Sugyan, projection data of I/n of all data will be selected. Here, n is the number of divisions of one heartbeat and is determined in advance! (n-4 to 10).

Increasing the number of scans increases the number of projection data that can be used, but it also increases the scan time and introduces noise due to patient movement, so simply increasing the number of scans is not sufficient.

Finally, the algorithm of the successive approximation method (least square method 1
According to the algebraic overlapping element method, maximum entropy method, etc., only the ROI region including the heart region is reconstructed (as shown in FIG. 4(E)). What should be noted at this time is that if there is no noise in the projection data, both the algebraic reconstruction method and the maximum entropy method can reproduce the data with approximately the same accuracy; however, if the projection data contains noise, This means that the maximum entropy method is very effective. For example, in the maximum entropy method, the following function is maximized by Il! Determine the II image.

k-1t-1θ-θ1r-1 where, f, L: Estimated image no (r): Estimated value of noise included in projection data h e (r): Actual measured value of projection data, Heg, ( r) : Estimated value of ' U : Noise power θ : Projection angle r : Sample position of projection data λ1. Since it is not possible to find the solution to equation (1) analytically using λ2 Nilagran and Gier's constant, numerical calculation is performed using the maximum slope method or the like.

According to the present invention, for still images other than the heart, F
For images comparable to those of the BP method, and for moving images of the heart, FL
It is possible to obtain more accurate images than with the SP method, and the number of scans is also less, so the exposure line cover can be minimized. In addition, the successive approximation method generally has the disadvantage of requiring processing time, but in the present invention, the processing time is reduced because it is used only for a very limited partial screen, such as the heart, rather than overusing the entire screen. It is possible to reduce it significantly.

[Brief explanation of the drawing]

FIG. 1 is a block diagram showing the configuration of an embodiment of the present invention, FIG. 2 is a diagram showing an example of the relationship between projection data and an electrocardiogram, and FIG.
The figure is a flowchart showing an example of the m-image processing procedure of the present invention, and FIG. 4 is a diagram explaining data obtained corresponding to each procedure of FIG. 3. 1... Projection data result section, 2... Electrocardiograph, 3...
Calculator, 4...C)? ,T,5...Storage device,6.
...Input device 3゜Representative Patent attorney Oro Shinsada Hakata 1 Figure 2 12 Figure 7? 7? I-Ordinance No. 3 Continuation of figure 1 page 0 Author: Hiroshi Ihara - Hitachi Systems Development Laboratory, Hitachi, Ltd., 1099 Ozenji, Tama-ku, Kawasaki City ■Applicant: Hitachi Medeico Co., Ltd., 1 Uchikanda-chome, Chiyoda-ku, Tokyo No. 14

Claims (1)

[Claims] X-ray projection is performed on the subject from multiple directions to produce a large amount of projection data, and reconstruction processing based on this data yields an image corresponding to the X-ray absorption rate inside the fracture/organ.
In an image processing device consisting of a line CT device and an electrocardiograph connected to it, the FlJP method (
Il! by filtered back projection method) lI葎P+Configuration'
After performing Ir, projection data of only a specific area that refers to the heart is calculated from CT values other than the heart, and only projection data of the same phase is selected from the data obtained from the electrocardiograph, using the successive approximation method. An image processing device characterized by performing image reconstruction of an arbitrary phase of a specific region.