JP3721632B2 - X-ray image digital processing device - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to an apparatus for converting an X-ray image signal used in medical treatment into digital image data and processing it.
[0002]
[Prior art]
It has been widely practiced to take an X-ray fluoroscopic image of a subject (such as a patient's body) as an electrical image signal using an X-ray TV system, and display the image on a monitor device for observation. Furthermore, it is also popular to perform various digital image processing by converting this image signal into digital image data.
[0003]
By the way, since X-rays are scattered inside the substance when passing through the subject, the contrast of the X-ray fluoroscopic image is deteriorated or the sharpness (sharpness) becomes dull due to the influence of the scattered rays. That is, the image quality deteriorates. Therefore, at present, attempts have been made to suppress the scattered radiation at the collimator by changing the X-ray quality to make it difficult to scatter, or by devising the shape of the X-ray collimator.
[0004]
[Problems to be solved by the invention]
However, even if the X-ray quality is adjusted or the collimator is devised as in the prior art, it is essentially impossible to remove the image quality degradation due to the influence of scattered rays in the subject.
[0005]
In view of the above, an object of the present invention is to provide an X-ray image digital processing apparatus that is improved so as to remove the influence of scattered radiation by processing a digitized X-ray fluoroscopic image.
[0006]
[Means for Solving the Problems]
In order to achieve the above object, an X-ray image digital processing apparatus according to the present invention comprises an A / D conversion means for A / D converting an input analog X-ray image signal, and a low frequency from the digitized image signal. A means for extracting a component; a means for obtaining an average luminance in a small neighborhood of the pixel of interest and its peripheral portion from a digitized image signal; and a relationship between the average luminance in the small neighborhood and the average luminance in the peripheral portion. A means for obtaining a coefficient, which takes a positive value when the former is larger than the latter and a negative value when the former is smaller than the latter, and the above digitization after multiplying the low frequency component by this coefficient And a means for subtracting from the obtained image signal.
[0007]
The influence of the scattered radiation on the image is such that it oozes from a high luminance portion to a low luminance portion. Therefore, if a low-frequency component is extracted from the image signal, it is possible to extract the influence of the scattered radiation. However, since this low-frequency component also includes the low-frequency component of the true image, it is not safe to remove the influence of scattered radiation simply by subtracting this extracted low-frequency component from the original image signal. It is enough. As described above, the influence of scattered radiation oozes out from the high-luminance part of the image to the low-luminance part. For pixels included in a region, the brightness is increased by scattered radiation. Conversely, if the brightness of a small region is higher than the surroundings, the brightness of the pixels included in the small region is considered to be decreased by scattering. It is done. In other words, it is necessary to make a correction for lowering the brightness of the pixels like the former and increasing the brightness of the pixels like the latter. Therefore, a coefficient representing the relationship between the average luminance of the small area and the average luminance of the peripheral portion, and a coefficient that takes a positive value when the former is larger than the latter and takes a negative value when the former is smaller than the latter, was obtained. Above, if this coefficient is multiplied by the above low frequency component, a positive or negative signal that more accurately reflects the influence of scattered radiation can be obtained, and by subtracting this from the original image signal, the luminance can be reduced. It is possible to cope with both the direction of increasing and the direction of decreasing, and it is possible to appropriately remove the influence of scattered radiation.
[0008]
DETAILED DESCRIPTION OF THE INVENTION
Next, embodiments of the present invention will be described in detail with reference to the drawings. In FIG. 1, X-rays emitted from an X-ray tube 11 pass through a subject 10 and enter an image intensifier 12, and an X-ray transmission image is converted into an optical image. A TV camera 14 is coupled to the image intensifier 12 via an optical system 13, and an image signal of an output optical image of the image intensifier 12 is obtained. This image signal is converted into digital image data by the A / D converter 15 and then sent to the digital image processing device 16 to undergo various digital image processing, and then converted into an analog image signal by the D / A converter 17. Returned to the TV monitor device 18.
[0009]
The digital image processing apparatus 16 includes a low-frequency component extraction circuit 21, a small neighborhood average luminance calculation circuit 22, a peripheral average luminance calculation circuit 23, a scattered radiation influence calculation circuit 24, a multiplier 25, and a subtractor 26. And the image processing unit 27 for performing other image processing.
[0010]
The digital image data is first sent to the low frequency component extraction circuit 21 to extract the low frequency component IS of the image. The low-frequency component extraction circuit 21 is constituted by a low-pass spatial filter that performs a convolution operation using, for example, a NF × NF matrix template as shown in FIG. The digital image data is also sent to the small neighborhood average luminance calculation circuit 22 and the peripheral average luminance calculation circuit 23.
[0011]
Each of the small neighborhood average luminance calculation circuit 22 and the peripheral average luminance calculation circuit 23 is a spatial filter configured by a convolution operation using a template, for example. The small neighborhood average luminance calculation circuit 22 calculates an average luminance MS of a small area around the pixel of interest, and the peripheral average luminance calculation circuit 23 calculates an average luminance ML in a relatively wide area around the pixel. is there. Therefore, the template used in the former is composed of a small NS × NS matrix as shown in FIG. 2B, and the template used in the latter is composed of a relatively large NL × NL surrounding the template. These NS × NS and NL × NL templates all use a normalized weight of “1”. It is desirable to set the NS × NS matrix small to obtain the average luminance in the vicinity of the target pixel, and to set the NL × NL matrix large to obtain the average luminance in a relatively wide range.
[0012]
The scattered radiation influence calculation circuit 24 calculates the scattered radiation influence K for each pixel from the MS and ML obtained above. In this example, K is obtained from the value of MS-ML using a graph as shown in FIG. Then, the multiplier 25 multiplies the low-frequency component IS by the influence degree K (IS × K), sends this as a correction value IC to the subtractor 26, and subtracts the correction value IC from the original digital image data. As a result, the image data from which the influence of the scattered radiation has been removed is subjected to various digital image processing such as window conversion and edge enhancement in the image processing unit 27, and then sent to the D / A converter 17, where it is an analog image signal. Is returned to the TV monitor device 18 for display.
[0013]
Here, it is assumed that the true data profile on one line of the image is as shown in FIG. The narrow portion where the brightness sharply falls corresponds to, for example, a blood vessel portion in the case of angiography. However, the influence of scattered radiation means that if there is a low-luminance part around a high-luminance part, X-rays ooze out from the high-luminance part to the low-luminance part. The luminance in the high-intensity portion is low, and conversely, the luminance in the low-luminance portion is high. Therefore, due to the influence of the scattered radiation, the profile of one line of the actually obtained image data is as shown in FIG. 4B. When the low-frequency component IS is extracted from this image data, it becomes as shown in FIG. 4C, which includes the low-frequency component of the true data and the influence of scattering.
[0014]
On the other hand, the positive value of MS-ML means that the pixel value in the small neighborhood is larger than the peripheral part. In such a case, the pixel value is lower than the true value due to scattering. Conceivable. Conversely, the negative value of MS-ML means that the pixel value in the small neighborhood is smaller than that in the peripheral part. In such a case, the pixel value is higher than the true value due to scattering. Conceivable. Therefore, when the MS-ML value is positive, the influence of scattering acts on the negative side with respect to the pixel value. Conversely, when the MS-ML value is negative, the influence of scattering is positive on the pixel value. Work to the side. Therefore, it can be understood that K is obtained according to the value of MS-ML by using a curve (for example, a thick line) as shown in FIG. 3, multiplied by IS, and then subtracted from the original image data. In FIG. 3, the reason that the MS-ML value is near 0 and the K value is 0 is to provide a dead zone to avoid the influence of noise. The curve shown in FIG. 3 can be changed like a thin line or a dotted line shown in FIG. 3 according to the X-ray conditions (tube voltage, tube current, image intensifier type / field of view, etc.). The optimal curve is obtained through experiments and computer simulations. Although the difference between MS and ML is used here, K may be obtained from a similar curve using the ratio between MS and ML.
[0015]
The correction value IC obtained by multiplying the low frequency component IS by K is as shown in FIG. 4D, and an image as shown in FIG. 4E is obtained by subtracting this IC from the original image data. Data can be obtained. As described above, since the correction value IC reflects the influence of scattering more, it is possible to obtain data excluding the influence of scattering. As can be seen from the comparison between the original image data shown in FIG. 4B and the corrected data shown in FIG. 4E, the contrast and sharpness of a steep and narrow depression such as a blood vessel is reduced. Reproduced, blood vessels and the like are easier to see.
[0016]
In the above description, K is calculated for each pixel. However, K may be calculated for each small area, and K may be obtained for each pixel by interpolation. Further, although the small neighborhood average luminance MS is obtained by a process separate from the low frequency component extraction processing in the above, the small neighborhood average luminance MS is a kind of low-pass filter output. The obtained value IS may also be used. Furthermore, in the above description, the low-frequency component extraction circuit 21, the small neighborhood average luminance calculation circuit 22, the peripheral average luminance calculation circuit 23, the scattered radiation influence calculation circuit 24, and the like have been described as being configured by hardware. Of course, it is possible to obtain it by processing by software.
[0017]
【The invention's effect】
As explained above, according to the X-ray image digital processing apparatus of the present invention, the influence of scattered radiation is appropriately removed, the contrast and sharpness of the original subject are reproduced, and it is easier to see and has excellent image quality. An image can be obtained. Further, the S / N ratio can be improved without enhancing high frequency noise components such as quantum noise.
[Brief description of the drawings]
FIG. 1 is a block diagram showing an embodiment of the present invention.
FIG. 2 is a schematic diagram showing a template.
FIG. 3 is a graph showing a K value calculation curve;
FIG. 4 shows each data profile.
[Explanation of symbols]
DESCRIPTION OF SYMBOLS 10 Subject 11 X-ray tube 12 Image intensifier 13 Optical system 14 TV camera 15 A / D converter 16 Digital image processing device 17 D / A converter 18 TV monitor device 20 Scattered ray removal unit 21 Low frequency component extraction circuit 22 Small neighborhood average luminance calculation circuit 23 Peripheral average luminance calculation circuit 24 Scattered ray influence calculation circuit 25 Multiplier 26 Subtractor 27 Image processing unit

Claims (1)

  A / D conversion means for A / D converting the input analog X-ray image signal, means for extracting a low frequency component from the digitized image signal, and a small pixel of interest from the digitized image signal Means for determining the average brightness in the vicinity and its surrounding area, and the relationship between the average brightness of the small neighborhood and the average brightness in the surrounding area. X-ray image digital processing comprising: means for obtaining a negative value, a means for obtaining a coefficient; and means for multiplying the low-frequency component by the coefficient and subtracting it from the original digitized image signal apparatus.

Images