JPS6214269A - X-ray picture processing device - Google Patents

Abstract

PURPOSE:To obtain a subtraction image of a purposed part which is not influenced by the noise even in case of a picture with a large noise component by converting respective picture element data by a special converting means concerning inspecting object area picture data of two types of partially different specimen. CONSTITUTION:Respective picture element data (x) of the X-ray picture of the inspecting object area obtained by an X-ray photographing device are received, and the converting means is provided which converts these respective picture element data by using the converting formula: f(x)={x+(y<2>/a)}<1/2> provided that (y) is a thermal noise effective value and (a) is the value of the total gain from an image pick-up means to an image pick-up equipment output stage. By converting respective picture element data of respective pictures by the converting means concerning the X-ray picture data of the inspecting object area before and after dosing the contrast medium obtained by an X-ray photographing device, the level of the noise of respective picture elements can be equalized, and the noise of the subtraction image of two pictures can be suppressed to the extent without hindrance for forming the binary image.

Description

[Detailed description of the invention] [Technical field of invention] The present invention relates to an improvement in an X-ray image processing device.

[Technical background of the invention and its problems] In recent years, when interpreting an X-ray image of a subject to whom a contrast agent has been administered, a method has been used to extract only the area of interest or the region to be diagnosed. There is.

In this method, for example, an X-ray image of a subject is obtained as digital data. By binarizing an image based on data at a certain density threshold level (threshold value), an image of a region of interest or a region to be diagnosed is extracted.

That is, when a contrast agent is administered, the contrast agent distribution area is
Since the amount of radiation transmitted is small, the image density increases, and the difference between the contrast agent non-distribution area and the X-ray transmission value in the vicinity of the contrast agent distribution area increases, resulting in a large difference in the density level. Therefore, if you want to obtain an image of a region of interest or a region to be diagnosed, such as the stomach or blood vessels, you can obtain an X-ray image of the region before contrast agent administration and an XSS image after contrast agent administration, and combine both image data. By taking the difference, a difference image (subtraction image) of the region of interest and the region to be diagnosed can be obtained. Since this subtraction image is an image of the area where the contrast agent was administered, if there is no noise component, if it is binarized using an appropriate threshold, the target area can be clearly distinguished from other areas, and the morphology of the target area can be clearly grasped. This makes diagnosis easier.

However, in carrying out such processing, there are drawbacks similar to conventional methods. In other words, the conventional process for obtaining a subtraction image is as shown in Figure 3 (
In a), the X-ray image of the region including the region of interest and the diagnostic target f [(under IX, these are collectively referred to as the region of interest) ROl before contrast agent administration is 1q, and then the same as shown in (b). Obtain a post-contrast x-ray image of the area. Then, by taking the difference between the two image data in (C), a subtraction image of the target region ROI can be obtained as shown in (d). Here, the vertical axis in FIGS. 3(a) and (b) is the image density,
Further, the horizontal axis indicates the spatial position, and Ea indicates a region with high X-ray absorption, and Eb indicates a region with low X-ray absorption. In the area Eb where the X-ray absorption is small, the amount of transmitted X-rays is large, so the noise is small, and in the area Ea where the X-ray absorption is large, the transmitted x weight is small, so the noise is large. Further, EC in FIG. 3(b) is a contrast agent administration region.

When subtractimine is performed on such an image, a high noise part En remains in addition to the contrast agent distribution area EC, as shown in FIG. Even if a threshold value S is set for this purpose, since the noise level exceeds the threshold value, a noisy image is generated in areas other than the target area (contrast agent distribution area), resulting in an image that is very difficult to see.

Devices that collect X-ray images as digital data use image carrier devices, but because they use electronic circuits, thermal noise is unavoidable. In addition, the X-rays used for imaging include
There is a fluctuation in the photons of X-rays. Therefore, in general, the noise components are centered around photon fluctuations of these X-rays and thermal noise generated in the electronic circuit of the imaging device, and since these cannot be avoided, in the case of the conventional method, in the area where X-ray absorption is large, , the obtained subtraction image cannot be used for diagnosis due to noise.

[Object of the Invention] The present invention has been made in view of the above circumstances, and its purpose is to be able to obtain a subtraction image of a target portion that is not affected by noise even if the image has a large noise component. An object of the present invention is to provide an X-ray image processing device.

[Summary of the Invention] In order to achieve the above object, the present invention subtracts two types of partially different image data of an object examination target region obtained by an X-ray imaging device to obtain a subtraction image of both images. In an X-ray image processing system in which images of different parts of the two images are extracted by binarizing this subtraction image using a predetermined threshold value, X of
Receiving each pixel data of the line image A converting means is provided for converting using a total gain value, and each pixel data of each image is converted by the converting means for X-ray image data of partially different subject images obtained by the X-ray imaging device. Then, subtraction images are obtained for both of the converted images, and a binarized image is obtained from this subtraction image to extract and expand different portions of the two images. That is, by performing the conversion using the conversion means, the noise level is made constant and corrected to a level that does not affect feature extraction. As a result, for example, when subtraction images of both images are obtained by subtracting the X-ray image data of the examination target area before contrast agent administration and during invasion obtained by an X-ray imaging device, it is possible to obtain subtraction images of both images. There is a clear difference in the level of noise in the area, and therefore, the influence of noise is also noticeable when extracting an image of the diagnosis target area by binarizing this subtraction image using a predetermined threshold that separates the contrast agent distribution area from other areas. This makes it possible to obtain a clear extracted image by removing the .

[Embodiment of the Invention] Hereinafter, an embodiment of the present invention will be described with reference to FIGS. 1 and 2.

First, the principle of the present invention will be explained.

The core of the present invention is to create a concentration conversion function f for each image in which photon noise and thermal noise are superimposed so that the amount of noise is constant at any location regardless of the amount of X-rays in each part. It is to find (x).

"Derivation of density conversion function f(x)" The signal S of the density (vixel value, ie, pixel value) of an X-ray image and the noise N around it can be expressed as follows. The signal noise discussed here refers to that at the output stage of the imaging equipment.

5=aX...(1)N-47
]) Hima ... (2) Here, a is the total gain from the photon detector (X-ray detector; imaging means) to the output stage of the imaging device, X is the number of photons per pixel, and Y is the imaging device. This is the effective value of thermal noise output from itself.

Now, when considering a certain concentration conversion function f(x), it is approximated as follows.

f(x±Δx)=f(x)±f−(x)ΔX...(3
) Here, if we consider that ΔX represents the noise of signal
...(4) =f (aX)±t-<ax> 4 completed) Four horses...
(5) It becomes. Here, f (aX) is the concentration conversion value of the signal S,
Further, f-(aX)4][tau] is the density conversion value of noise. In other words, if the density conversion 11 of f(x) is performed, the signal value will naturally be f (S) - f (aX)... (
6), but the noise component is f (N) − f ′ (aX) ・6 pairs]) penalty − (7
) is approximated by The initial objective of keeping the amount of noise constant is f'(aX).]Go=C...(8) However,
Needless to say, C is a constant, and if you solve this, f (X)
-(2C/a)-48])70B-(2C/, ff)
・x0(Y2/a)+B...(9) is obtained. Note that B is a constant. Here, the constant C1B corresponds to the gain and bias as a result of density conversion, and has nothing to do with the essential conversion operation. In other words, as a form of concentration conversion, f(x) = x + (Y2/a)
・It is essential that it has the form (10), and the constants C and B do not need to be included in f(x) because they are completely equivalent to performing image linear operation after converting the dark 9 once. .

That is, by performing density conversion using the above equation (10), it is possible to suppress noise and obtain a binarized image of a desired region of interest or region to be diagnosed.

Based on this principle, the present invention performs X-ray image processing and
The aim is to obtain high-quality binarized images that are unaffected by noise even for areas with large linear absorption.

Figure 1 shows the configuration of this device. In the diagram, 1 is the original
Memory for storing line image data, 2-1.2-2.・
. . . is a processed image storage memory that stores processed image data, and 3 is a memory for storing processed images, and 3 is a memory that performs the above equation (1o) on the stored data for each pixel in the memory 1 to perform density conversion while keeping the amount of noise constant. This is a density conversion unit that converts data into density value data corresponding to X-ray transmission. The image data processed by the density conversion unit 3 is stored in the processed image storage memory 2-1 or 2-2. It is stored and saved in... In the case of this apparatus, at least one side stores a processed image of an X-ray image before contrast agent administration, and the other side stores a processed image of an X-ray image after contrast agent administration. 4 is an X-ray gain controller, which supplies thermal noise effective values Y2.4 output from the imaging equipment itself to the concentration conversion unit 3. and the value of the total gain a from the photon detector to the output stage of the imaging device is set. The concentration conversion unit 3 receives Y2. After receiving the value of a, the above equation (10) is calculated. 5 is a processor, and the processed image storage memories 2-1, 2-2. Subtraction is performed on the processed image data of the image data before and after contrast agent administration stored in ... to obtain subtraction images of both images. This method obtains a binarized image of the area of interest to which the contrast medium has been administered.

Next, the operation of this device having the above configuration will be explained.

First, an X-ray image of the area to be diagnosed before administration of a contrast agent is obtained using an xmm imaging device (not shown), which is converted into digital data and sent to a memory that stores data of the original X-ray image. The X-ray gain controller 4 is preliminarily supplied with noise power Y2. and the total gain a from the photon detector to the output stage of the imaging device are set, and the values of Y2 and a are given to the density conversion unit 3. The image data stored in the memory 1 is sequentially read out and sent to the density conversion unit 3, which uses the given values of Y2 and 8 to perform the density conversion of the image data using equation (10). Perform calculations. As a result,
Image data that contains noise components of different levels depending on the amount of X-ray transmission, as shown in (a) of Figure 2, is converted to a noise level that is approximately equal as shown in (b) of Figure 2 by density conversion calculation. It can be arranged. In this way, the image data stored in the memory 1 is sequentially read out to the density conversion unit 3.
Processed image storage memory 2-1
is stored in

Next, the xis imaging device obtains an xsit image of the area to be diagnosed after the contrast agent has been administered, converts it into digital data, and sends it to a memory that stores data of the original x-ray image. The image data stored in the memory 1 is sequentially read out and sent to the density conversion unit 3, and this density conversion unit 3 applies the density conversion of the equation (10) to this image data using the given values of Y2 and a. Perform calculations. As a result, the image data after contrast agent administration, which contains noise components of different levels depending on the amount of xm transmission, as shown in (C) in Figure 2, is converted to the image data as shown in (d) in Figure 2 by concentration conversion calculation. The noise levels are almost equal. In this way, memory 1
The image data is sequentially read out and subjected to density conversion processing.
The processed image is stored in the processed image storage memory 2-2.

In this way, the processed image data of the image data before and after contrast agent administration is stored in the processed image storage memories 2-1 and 2-2.
．． It is stored and saved in...

Next, the processed image storage memories 2-1 and 2-2°...
The processed image data of the image data before and after contrast agent administration stored in is sequentially read out at the same pixel position and sent to the processor 5, where subtraction is performed. As a result, the subtraction images of both images, which are less affected by noise components, as shown in FIG. 2(e), are cleared. Further, the processor 5 binarizes this subtraction image using a given threshold value SL as shown in FIG. 2(e) to obtain a binarized image of the region of interest to which the contrast agent has been administered. Thereby, an image of the portion to which the contrast medium has been administered can be extracted with high precision.

In this way, this device subtracts the X-ray image data of the inspection target area before and after contrast agent administration obtained by the X-ray imaging device, obtains a subtraction image of both images, and uses this subtraction image with the contrast agent. In an X-ray image processing device that extracts an image of a diagnosis target area by binarizing it using a predetermined threshold value that distinguishes a distribution area from other areas, Receiving each pixel data of an image A converting means is provided, and the converting means converts each pixel data of each image regarding the X-ray image data of the inspection target area before and after contrast agent administration obtained by the X-ray imaging apparatus, and converts the converted data. A subtraction image is obtained for both images, and a binarized image is obtained from this subtraction image to obtain an extracted image of the diagnosis target portion. That is, regarding the X-ray images of the diagnosis target area before and after contrast agent administration, the previous Inc! Perform the calculation of equation (10),
Concentration conversion is performed in a noise pseudo-constant manner to convert the data into concentration value data corresponding to the amount of X-ray transmission, and subtraction is performed on both the image data before and after administration of the contrast agent, which have undergone this concentration conversion process. Subtraction images of both images are obtained, and the subtraction images are binarized using a given threshold value to obtain a binarized image of the region of interest to which the contrast agent has been administered. Therefore, it is possible to make the noise level of the portion with a large amount of X-ray transmission and the portion with a small amount of X-ray transmission substantially constant, and therefore, it is possible to obtain a clear extracted image of the portion to be diagnosed that is not affected by noise components.

It should be noted that the present invention is not limited to the embodiments described above and shown in the drawings (it goes without saying that the present invention can be carried out with appropriate modifications within the scope of the gist; for example, in the above embodiments, a contrast agent We have explained the case of administering a contrast agent and extracting the image of the area where the contrast agent has been administered without being affected by noise. It can also be applied to detect defects such as size, cracks, and nests, and as an X-ray image, it can be used for images such as XICT (computer tomography) in addition to fluoroscopic images.

[Effects of the Invention] As detailed above, according to the present invention, noise components are superimposed on image data due to the influence of differences in the amount of X-ray transmission, thermal noise in the electronic circuit of the detector system, and moreover,
In images where noise levels differ depending on the pixel, it is possible to equalize the noise level of each pixel, thereby reducing the noise in the subtraction image of the two images so that it does not interfere with binarized image formation. Since this can be suppressed to an extent that there is no such problem, it is possible to provide an X-ray image processing apparatus having features such as being able to extract images of characteristic parts with high accuracy.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIG. 2 is a diagram for explaining the operation of the apparatus of the present invention, and FIG. 3 is a diagram for explaining a conventional example. DESCRIPTION OF SYMBOLS 1...Memory, 2-1.2-2...Memory for storing processed images, 3...Density conversion unit, 4...X-ray gain controller, 5...Processor. Applicant's agent Patent attorney Takehiko Suzue 171

Claims (1)

[Scope of Claims] Two types of partially different object inspection target area image data obtained by an X-ray imaging device are subtracted from each other to obtain a subtraction image of both images, and this subtraction image is subjected to a predetermined threshold value. In an X-ray image processing device that extracts images of different parts of the two images by binarizing each pixel data x of the X-ray image of the subject obtained by the X-ray imaging device. Then, convert each pixel data using the formula f(x)=√[x+(Y^2/a)], where Y is the effective value of stationary noise, and a is the value of the total gain from the imaging means to the output stage of the imaging device. The conversion means converts each pixel data of each image with respect to X-ray image data of partially different images of the subject obtained by the X-ray imaging device, and converts the converted data using the conversion means. An X-ray image processing apparatus characterized in that it obtains subtraction images for both images, and obtains a binarized image from the subtraction images to obtain extracted images of different portions of the two images.