JPH0327484A - Method and device for deciding circular pattern - Google Patents

Abstract

PURPOSE:To decide the level of probability that a prescribed area is the area in a circular pattern by obtaining the respective mean values in two directions of the absolute value or the square value of the difference of picture data in a prescribed area designated on a radiation picture. CONSTITUTION:When the circular pattern of a tumor shadow, etc., and an area where the linear patterns of a blood vessel shadows, etc., crowd, for instance, on a breast X-ray picture are compared and observed, the tumor shadow has the comparatively less fine variation of density and is flat, but in the area where the blood vessel shadows. etc., crowd, the fine variation in one direction appears more frequently (the blood vessel shadow, etc., extends in a definite direction frequently). On the basis of this observed result, the prescribed area is decided whether it is the area where the blood vessel shadows, etc., crowd or not according to the means value of two directions different from each other. Thus, it can be decided by comparatively high probability whether the prescribed area is the area in the circular pattern or not.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to an image pattern recognition method and apparatus, and more particularly, the present invention relates to an image pattern recognition method and apparatus, and more particularly, the present invention relates to an image pattern recognition method and apparatus, and more particularly, to a method and an apparatus for recognizing image patterns. The present invention relates to a circular pattern determination method and apparatus for determining the height of a circular pattern.

(Prior Art) It is practiced in various fields to read a recorded radiation image to obtain image data, perform appropriate image processing on this image data, and then reproduce and record the image. For example, an X-ray image is recorded using an X-ray film with a low gamma value designed to be compatible with later image processing, and the X-ray image is read from the film on which the X-ray image was recorded and electrical signals ( By performing image processing on this image data and then reproducing it as a visible image in a copy photograph, etc., it is possible to obtain a reproduced image with good image quality performance such as contrast, sharpness, and graininess. (
(See Japanese Patent Publication No. 81-5193).

In addition, the applicant has proposed radiation (X-rays, α-rays, β-rays, γ-rays, etc.)
Ray, electron beam. When irradiated with ultraviolet rays, etc., a part of this radiation energy is accumulated, and then when excitation light, such as visible light, is irradiated, a stimulable phosphor (stimulable phosphor) that exhibits stimulated luminescence is produced according to the accumulated energy. The radiation image information of a subject such as a human body is recorded once on a sheet of stimulable phosphor, and this stimulable phosphor sheet is scanned with excitation light such as a laser beam to generate stimulated luminescence. A radiation image recording and reproducing system that photoelectrically reads the obtained stimulated luminescence light to obtain image data, and outputs a radiation image of the subject as a visible image to a recording material such as a photographic light-sensitive material, a CRT, etc. based on this image data. has already been proposed (Japanese Unexamined Patent Publication No. 55-1
No. 2429, No. 5B-11395, No. 55-1133
No. 472, No. 56-104845. 55-116
340 etc.).

This system has the practical advantage of being able to record images over a much wider range of radiation exposure than conventional radiographic systems using silver halide photography. In other words, for stimulable phosphors3 4 , it is recognized that the amount of emitted light that is stimulated and emitted due to excitation after accumulation is proportional to the amount of radiation exposure over an extremely wide range. Even if the amount of radiation exposure varies considerably, the amount of stimulated luminescence light emitted from the stimulable phosphor sheet is read by a photoelectric conversion means by setting the reading gain to an appropriate value and converted into an electrical signal. By using this electrical signal to output a radiation image as a visible image to a recording material such as a photographic light-sensitive material or a display device such as a CRT, a radiation image that is not affected by fluctuations in radiation exposure amount can be obtained.

In recent years, in systems using the above-mentioned X-ray film, stimulable phosphor sheets, etc., especially systems designed for medical diagnosis of the human body, it has become increasingly difficult to reproduce reproduced images with good image quality suitable for simple observation (diagnosis). In addition to image recognition, automatic recognition of images has also been carried out (for example, in Japanese Patent Application Laid-Open No.
(See Publication No. 5481).

Automatic image recognition here refers to the operation of extracting a desired pattern from a complex radiographic image by performing various processes on the image data. , an operation to extract, for example, a shadow corresponding to a tumor from a very complex image containing a mixture of circular patterns.

A target pattern (for example, a tumor shadow) in such a complex radiographic image (for example, a chest X-ray image of a human body)
By extracting the pattern and reproducing and displaying a visible image clearly showing the extracted pattern, it is possible to assist the observer in observation (for example, assist the doctor in diagnosis).

(Problems to be Solved by the Invention) In the above-mentioned Japanese Patent Application Laid-Open No. 82-125481, for example, in order to recognize a tumor shadow (an example of a circular pattern) on a chest X-ray image of a human body, A method is described in which a real space filter is used to scan an X-ray image and extract a circular pattern.

However, the above-mentioned filter does not necessarily have sufficient accuracy in extracting circular patterns, and often extracts areas where linear shadows, such as blood vessel shadows, are densely packed together as circular patterns.

In view of the above-mentioned circumstances, the present invention provides a method for determining whether a predetermined region designated on a radiographic image is a region where a linear pattern such as a blood vessel shadow is assembled, thereby determining whether the predetermined region is a region within a circular pattern. It is an object of the present invention to provide a circular pattern determination method and device for determining the level of a certain probability.

(Means for Solving the Problems) A circular pattern determination method of the present invention includes, based on image data representing a radiation image of a subject, differences between the image data corresponding to pixels that are close to each other in a predetermined region of the radiation image. are obtained in large numbers for each of two different directions on the radiation image, and for each direction, an average value of the absolute value or the square value of the large number of differences is obtained, and the average value in the two directions is obtained. The present invention is characterized in that the probability that the predetermined region is a region within a circular pattern constituting the radiation image is determined based on the value.

Further, the circular pattern determination device of the present invention calculates the difference between the image data corresponding to pixels close to each other in a predetermined area of the radiographic image based on the image data representing the radiographic image of the subject. difference calculation means for calculating a large number of differences in each of two different directions; average calculation means for calculating an average value of the absolute value or square value of the large number of differences for each of the directions; The present invention is characterized by comprising a determining means for determining, based on the value, the probability that the predetermined region is a region within a circular pattern constituting the radiographic image.

Here, the above-mentioned "average value" typically refers to a value obtained by an arithmetic mean calculation, but is not limited to this, for example, a geometric mean value, a median value, (maximum value - minimum value) / 2
It includes values obtained by various calculations that can be replaced with equal average values. However, for the sake of simplicity, these will be referred to as "average values" hereinafter.

7 8 Further, the above-mentioned "predetermined area" is, for example,
The predetermined region may be a region extracted as a circular pattern candidate by scanning a radiation image using the filter described in Publication No. 2-125481 or various filters described later, but the predetermined region is not necessarily a circular pattern candidate. For example, it does not have to be a region extracted as a circular pattern candidate using a filter, for example, each region separated by a large number of line segments virtually drawn vertically and horizontally on the target radiation image is extracted as a circular pattern candidate. Alternatively, the operator may manually designate a predetermined region on the radiographic image while observing the radiographic image. If the predetermined area specified on the radiographic image as described above is determined to have a high probability of being an area within a circular pattern according to the present invention, and if the probability is to be further increased, the predetermined area is within or near the predetermined area. For example, filter scanning as described above may be performed.

(Effect) For example, as a result of comparing and observing a circular pattern such as a tumor shadow on a chest X-ray image of a human body with an area where linear patterns such as blood vessel shadows are densely packed, it is found that there are relatively few small fluctuations in density within the tumor shadow. It has been found that, although the area is flat, there are many cases where there are small fluctuations in one direction (the blood vessel shadows and the like extend in a substantially constant direction) in a region where blood vessel shadows and the like are densely packed.

The present invention was made based on this observation,
With the above-mentioned features, it is possible to determine whether or not the predetermined area is an area where blood vessel shadows etc. are dense based on the average values in two different directions, and thereby it is possible to determine whether the predetermined area is an area within a circular pattern. It is possible to determine with a fairly high probability whether or not there is.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings. Here, the above-mentioned stimulable phosphor sheet was used,
An example of extracting the shadow of a tumor that typically occurs in a roughly spherical shape within the lungs of a human body will be described. This tumor shadow typically appears on a visible image as a whitish (low density) approximately circular pattern compared to the surrounding area.

FIG. 2 is a schematic diagram of an example of an X-ray imaging apparatus.

X-rays 12 are irradiated from the X-ray source 11 of the X-ray imaging device 10 toward the chest ISa of the human body 13, and the X-rays 12a that have passed through the human body 13 are irradiated to the stimulable phosphor sheet 14. , the transmitted X-ray image of the human chest 13a is
is accumulated and recorded.

FIG. 3 is a diagram schematically showing an example of a chest X-ray image accumulated and recorded on a stimulable phosphor sheet.

An X-ray image 15 of the lung is displayed on almost the entire surface of the stimulable phosphor sheet 14.
is being shaped.

In this embodiment, a large number of line segments drawn in the x and y directions of this sheet 14 are assumed, as shown by broken lines in the figure, and each area D. As the predetermined area referred to in the present invention, each of these areas D. The probability that the area is within the circular pattern is determined, and by repeating this determination over the entire X-ray image, a region that is considered to be a tumor shadow is extracted on the X-ray image.

FIG. 4 is a perspective view showing an example of an X-ray image reading device and a computer system incorporating an embodiment of the present invention.

A stimulable phosphor sheet 14 on which an X-ray image as shown in FIG. 3 has been recorded is set at a predetermined position in the X-ray image reading device 20. The stimulable phosphor sheet l4 set at a predetermined position is transported in the direction of arrow Y by a sheet transport means 22 such as an endless belt driven by a motor 21 (
sub-scanning). On the other hand, a light beam 24 emitted from a laser light source 23 is reflected and deflected by a rotating polygon mirror 26 that is driven by a motor 25 and rotates at high speed in the direction of the arrow.
After passing through a focusing lens 27 such as a θ lens, a mirror 28
The optical path is changed by , and the light enters the sheet 14 and is main-scanned in the direction of arrow X, which is substantially perpendicular to the direction of sub-scanning (direction of arrow Y).

Stimulated luminescence light 29 is emitted from the location where the excitation light 24 of the sheet 14 is irradiated, and the amount of stimulated luminescence light 29 corresponds to the accumulated and recorded X-ray image information. guided, photomultiplier (photomultiplier tube)
31 photoelectrically detected.

11 12 The light guide 30 is made by molding a light-guiding material such as an acrylic plate, and has a linear entrance end surface 30a aligned along the main scanning line on the stimulable phosphor sheet 14. A light-receiving surface of the photomultiplier 31 is coupled to an emitting end surface 30b which is arranged to extend and formed in an annular shape. Stimulated luminescent light 2 entering the light guide 30 from the incident end surface 30a
9 travels inside the light guide 30 by repeating total reflection, exits from the exit end surface 30b, and enters the photomultiplier 3.
The photomultiplier 31 converts the stimulated luminescence light 29, which is received by the photomultiplier 31 and represents an X-ray image, into an electrical signal.

The analog output signal SO output from the photomultiplier 31 is logarithmically amplified by the logarithmic amplifier 32, and the A/D
The data is digitized by a converter 33, and image data S1 as an electrical signal is obtained.

The obtained image signal S1 is input to the computer system 40. This computer system 40 constitutes an example of the present invention, and includes a main body part 41 in which a CPU and an internal memory are built-in, a drive part 42 into which a floppy disk as auxiliary memory is inserted and driven,
It is comprised of a keyboard 43 for an operator to input necessary instructions to the computer system 40, and a CRT display 44 for displaying necessary information.

Image data S1 manually input to the computer system 40
Based on this, for each region DI shown in FIG. 3, the probability that each region D1 is within the tumor shadow is determined as follows.

FIGS. 1A and 1B illustrate D., one of the many areas shown in FIG. D. is a region within a true tumor shadow and within a dense region such as a blood vessel, respectively. FIG. 2 is a diagram showing an X-ray image of FIG. In each figure, the area surrounded by the broken line 9 is one area D. Each graph is a profile (image data S1 is plotted) in the X direction and the y direction within the region D1.

The tumor shadow has a relatively flat profile with a valley near the center both in the
The profile has small fluctuations in other directions (y
direction) has a relatively flat profile. Therefore, here, by utilizing this difference in profile, each area D
Determine whether the eye is an area with dense blood vessels or the like or a tumor shadow. That is, pixels arranged in the X direction are IIl (m = 1.
2,...), pixels lined up in the y direction are n (n-1.2
．． ), and the image data of the pixel represented by (m, n) is f (IIl.n). At this time, as shown in the following equation, the area D. The average value of the squares of the primary difference values of the image data within is calculated.

zx−ΣΣ (f (IIl+1, n) − f (m
．． n) l 2/N(m.++)! -DI+
...(1) 2,1ΣΣ ( f (
m. n+1) − f (m.n) l 2/N(11
1.1+) this D mouth ・●・
(2) (where Σ Σ represents addition of the first-order difference value within the region D1, and N represents the number of pixels within the region D1) Next, this region DIj is a region with dense blood vessel shadows, etc. As the feature amount c1 for determining whether or not, the smaller value of Z1 and Zy is set as min(Zx+Zy), and the larger value is set as Illax(Z.Zy), then the following is calculated, and this feature The amount C1 is compared with a predetermined threshold value Thl, and when C1≧Thl, the region D1 is determined to be a region with a high probability of being a tumor shadow, and when C<Thl, the region D1 is determined to be a region with a high probability of being a tumor shadow. Therefore, the probability of being a tumor shadow is determined to be low.

Note that the feature amount C1 is not limited to that calculated using equation (3), and may be, for example, 15 16 2. +2, C2 − l Z, Zy I
-(5) etc. may be used. Also, in the above example, x,
First-order difference f (m+1, n) in two directions of y f (
IIl. n), f (m.n+1)f(m,n)
For example, in the diagonal direction. (a direction that is not perpendicular to either the x direction or the y direction) may be calculated. Further, in the embodiment described above, the average value of the square values of the first-order differences is determined, but instead of this, the average value of the absolute values of the first-order differences may be determined.

In the computer system 40 of FIG. 4, the above-described operations are performed for each region D shown in FIG. 3, thereby extracting regions that are highly likely to be tumor shadows.

After extracting a region with a high probability of being a tumor shadow as described above, when reproducing and displaying a visible image based on the image data S1 on a CRT display device, for example, by clearly indicating the extracted region, the viewer can You can have them provide assistance.

However, in this example, in order to further increase the probability that the region extracted as described above is a tumor shadow and to perform more accurate extraction, A filtering operation is performed using each pixel as a "predetermined pixel P", which will be described later, so that only the region with a higher probability of being a tumor shadow is extracted from the extracted region.

FIG. 5 shows a predetermined pixel P on an X-ray image in order to explain an example of this filtering operation. It is a diagram virtually drawn on the image with . It is determined whether the predetermined pixel Po is a pixel within the tumor shadow.

FIG. 6 is a diagram showing an example of the profile of an X-ray image in the direction in which line segments L1 and L in FIG. 5 extend (X direction), centered on the predetermined pixel Po. Here, it is assumed that the predetermined pixel Po is located approximately at the center of the tumor shadow 7 which is very close to the rib shadow 6.

Tumor shadow 7 typically appears as a nearly symmetrical profile, but as in this example, tumor shadow 7 appears as 17? In some cases, such as when it is located very close to the bone shadow 6, it may not be symmetrical. It is important to be able to extract this tumor shadow 7 even in such a case. Note that the broken line 8 in FIG. 6 is an example of a profile when there is no tumor.

As shown in FIG. 5, a plurality of (eight lines in this case) line segments L+ (1-1.2...... Assuming .8),
Further, each radius rl is centered around a predetermined pixel Po.
, r2, r3, a circle R, (j −1.2.l ) is assumed. Predetermined pixel P. The image data of is fo, and each pixel P++(
Figure 5 shows P. ■＋Pl■＋ P 13+ P 51+
Symbols are shown for P52+P53. ) image data f. shall be.

Here, image data fo of a predetermined pixel Po and each pixel P1
image data f. The difference Δ1 from

Δ11f,, -fo ... (6) (1 -1
．． 2. ......, 8;j -1.2.3)? For each line segment L+, the maximum value of the difference Δth obtained by equation (6) is obtained. That is, taking an example of line segment L,,1.5, for line segment Ll, pixel P ll+ P l
Each difference Δ++=f++ f corresponding to ■, P, and 3.

The maximum value of Δ12″″f, 2−fO Δ13″″ f 13 fQ is determined. In this example, as shown in FIG. 6, Δ, , < Δ12 < Δ11 O, so Δ1 is the maximum value. Moreover, for line segment L5, each difference Δ corresponding to pixel P51+P52+P53
The maximum value Δ,3 of 51"'f51 fO Δs2-fs2 fo Δ,3""f53 fO is determined.

In this way, a predetermined pixel P. By calculating the maximum value of the difference between the line segment L1 and a plurality of pixels for each line segment L1, it is possible to deal with tumor shadows of various sizes.

Next, two line segments extending in opposite directions from a predetermined drawing line Po are set as a set, that is, line segment L1 and line segment L.
5. Set each of line segment L2 and line segment L6, line segment L3 and line segment L7, and line segment L4 and line segment L8 as one set, and calculate the average value of the two maximum values for each set (respectively Ml, M26,
M3,, M48) are obtained.

For the pair of line segment Ll and line segment L5, the average value M1
5 is found as 2.

By treating two line segments extending in opposite directions from a predetermined pixel Po as a set, it is possible to detect when the tumor shadow 7 is located near the rib shadow 6, for example, and the image data of the tumor shadow 7, as shown in FIG. It is possible to reliably detect tumor shadows even if the distribution is asymmetric.

As above, the average value M,,, M26, M3,,
Once M4B is determined, these average values M, , M2
6, M 3 7 , M 4 B, as follows, predetermined ? Pixel P. A characteristic value C2 is determined to be used for determining whether or not the pixel is within the tumor shadow.

FIG. 7 is a diagram for explaining how to obtain this characteristic value C2. The horizontal axis is the average value M15 obtained as above,
M26, M37, M4B, the vertical axis is each evaluation value corresponding to these average values CI 5+ C 26 + C 37
+C48.

If the average value M,,,M26,M,7,M48 is smaller than a certain value Ml, the evaluation value is zero; if it is larger than a certain value M2, the evaluation value is l,Q, and in the middle between Ml-M2, the value is large. The evaluation value is a value between 0.0 and 1.0 depending on the value. In this way, each average value M , 5, M 26, M 3
Evaluation values C I corresponding to 7 and M 4 B, respectively
5+ C26+C37+ C4g is determined,
Sum of these evaluation values C1, C26, C37, and C4B C2 = C +s+C 26+C 3 ■ 10 C48
...(8) is taken as the characteristic value C2. That is, this characteristic value C2 has any value between a minimum value of 0.0 and a maximum value of 4.0.

This characteristic value C2 is compared with a predetermined threshold Th2, and C
It is determined whether the predetermined pixel Po is a pixel within the tumor shadow, depending on whether 2≧Th2 or C2×Th2.

By scanning each region D in the mouth that is highly likely to be a tumor shadow on the X-ray image using the above-mentioned real space filter, and in the vicinity of these regions, the regions D. By determining whether or not each pixel is a pixel within a tumor shadow by setting each pixel in the vicinity thereof as the predetermined pixel Po, a region with a higher probability of being a tumor shadow is identified from each region DI. Extracted.

The filter for extracting a region with a higher probability of being a tumor shadow from each region D is not limited to the above filter. Other examples of filters will be described below.

Each pixel P +4 (1 -L2,...8;
j −1.2.3) image data f-th gradient f
．． is required.

Here, the gradient refers to x of a certain pixel P on an X-ray image.
The y coordinates are (m, n), and the coordinates of pixels p' and p' adjacent to the pixel P in the X and y directions are respectively (Ill+1
, n), (m, n+1), and their pixels P,
The image data of P'P' are respectively f (m.n),
When f (m+1, n) and f (m, n+1), f (m, n) 1 ( f (IIl+l.
n) − f (vn), f (m.n+1)
− f (m, n) ) − (9).

FIG. 8 is a diagram showing the gradient described above and the calculation method described below.

After the gradients f1 are determined, the vector lengths of these gradients f1 are determined as l. Aligned to o. That is,
When the magnitude of the gradient f1 is f++l, a normalized gradient f+r/l V f +r 1 is obtained.

Next, the component of this normalized gradient f1/1f11 in the direction of the line segment L is determined. That is, when the unit vector from each pixel P1 toward a predetermined pixel Po is e, f. /1f0*e, (where * represents the inner product) is obtained.

Then, assuming that the inward direction (direction of the predetermined pixels 23 24 Po) is positive and the outward direction is negative for this component, each line segment L+
Each maximum value {Vf++/lVf++l *e+ ly(i −1
．． 2,...,8) are obtained, and an additional value Σ (Vf++/lVf++l*e+ly) is obtained by adding these maximum values (Vf++/lVf++)y. This added value is set as the feature amount c3, This feature amount C3 is compared with a predetermined threshold value Th3, and it is determined whether the predetermined pixel Po is a pixel within the tumor shadow depending on whether C3≧Till or C3<Th3. .

This filter uses a gradient f. Normalize the size f++l and calculate its direction (degree of difference in direction from line segment L1)
By focusing only on the feature c3, which has a large value due to the circular shape, regardless of the contrast with the surroundings, the tumor shadow can be extracted with high accuracy.

Next, another example of the real space filter will be described.

Qo and Q mouth (i-1, 2, ..., 8
;j = 1.2.3) (However, in Figure 5, QO and Q rl・Ql2・Q 13・
Only Q5, Q52, and Q53 are shown) are pixels P, respectively. and each pixel P ++ (
i = 1, 2. . . , 8; j −1.2.3).

Each region Qo and Q +r (1 −1.2.−,
8; j=1.2. l ), each region Qo +
The average value of a large number of image data corresponding to a large number of pixels in Q + +
., 8: j-1, L8) is obtained. For simplicity, each area Qo+QB (1 -1+2,
・・・． 8; j −1.2.3 ) and the symbol indicating the average value of the image data in each region are the same.

Next, the average value Qo of the central area and the average value Q+1 of each peripheral area
Each difference Δth (i −1.2...,8
;? −1.2.3) is obtained as Δ++=Q++ Qo −(10), and furthermore, for each line segment L, the maximum value Δ of the difference Δ1
．． is required. That is, to give an example regarding the line segment L1+L5, for the line segment L1, Δ1,,Δ1. , Δ13, the maximum value Δ1, and for line segment L5, Δ, 1, Δ5■,
The maximum value Δ5 of Δ, 3 is determined.

Next, a first characteristic value U representing the maximum value Δ+(i-1 to 8) and a second characteristic value V representing the variation in the maximum value Δ+(1=L to 8) are determined. For this purpose, first, each characteristic value U1-U4, V1-v4 is determined according to the following calculation formula.

U1-(Δl+Δ2+Δ5+Δ6)/4...(l1
) in U2 (Δ2+Δ3+Δ6+ΔT)/4...(1
2>U3-(Δ3+Δ4+Δ7+ΔB)/4...(
l3〉U4-(Δ4+Δ5+Δ6+Δ1)/4...
(i4) V1=U1/U3 ・
...(l5)V2-U2/Ua
”' (1B)V3-U3/
U1 − (17) V4 = l
=U4 /U2...
(18) Here, for example, according to equation (11), the characteristic value U
To explain the case of finding 1, adding two adjacent areas (Δl and Δ2 or Δ5 and Δ6) means smoothing, and adding up two adjacent areas (Δl and Δ2 or Δ5 and Δ6) means smoothing, and adding up two adjacent areas (Δ1 + Δ2 and Δ5 + Δ6) on opposite sides of pixel P0. Similar to the first filter described above, the purpose of addition is to enable tumor shadows to be detected even if the image data is asymmetrical, as shown in FIG.

Furthermore, for example, to explain the case where the characteristic value V1 is determined according to equation (15), the characteristic value U1 and the characteristic value U3 are characteristic values determined in mutually orthogonal directions, and therefore the tumor shadow 7 shown in FIG. 6 is circular. If v1→l.
When the value becomes Q and deviates from a circle, that is, when the pixel Po is within a linear shadow such as a rib shadow, V1 deviates from 1.0.

The first characteristic value U representing the maximum value Δ+(i-1 to 8) of the above difference is the maximum value U=MAX of U1 to U4.
(U1, U2, U3, Ua) −(19
)27 28 is adopted, and the second characteristic value V representing the variation of the maximum value Δ+(1-1 to 8) of the above difference is adopted as ■1 to V. The maximum value V-MAX (Vs * V2 . V3 , Va
) −(20) is adopted. When the first and second characteristic values U and V are determined in this way, these first and second characteristic values C4 are used to determine whether a predetermined pixel Po is a pixel within a tumor shadow. The ratio of the two characteristic values U C4 - (21) (2) is adopted, and this characteristic value C4 is compared with a predetermined threshold Th4, and depending on whether C4≧Th4 or Ca<Th4, It is determined whether each pixel Po is a pixel within the tumor shadow.

In addition, in each of the above filter examples, as shown in FIG. 5, pixels PI, average value Q, etc. on eight line segments L1 to L8 are used, but it is not necessary that there are eight line segments. It goes without saying that there may be, for example, 16 words. Also,
The distance from the predetermined pixel Po is also rl+'2.
Although calculations were performed for the three distances in '3, this is not limited to three distances; if the size of the tumor shadow to be extracted is approximately constant, one distance may be used; In order to extract tumor shadows of various sizes with higher accuracy, calculations may be performed while continuously changing the distance from r1 to r3. In addition to the above filters,
Further, different filters may be used.

As described above, when an area with a higher probability of being a tumor shadow is extracted by the filtering calculation process, a visible image clearly showing this extracted area is displayed on the CRT display 44 (see FIG. 4) and observed. served.

In the above embodiment, first, each area DI,
The probability that each region is within a tumor shadow was determined.First, the X-ray image was scanned using a real space filter such as the one described above to extract regions that were considered to be tumor shadows. Then, the region may be used as a predetermined region according to the present invention, and the probability that the region is a tumor shadow may be determined.

The above embodiment is an example of extracting a tumor shadow that typically appears as a circle in a chest X-ray image of a human body obtained using a stimulable phosphor, but the present invention is limited to the extraction of tumor shadows. The invention is not limited to chest X-ray images, nor is it limited to systems using stimulable phosphors, and is a circular pattern that constitutes a radiographic image based on image data representing a radiographic image of a subject. It can be widely used when extracting.

(Effects of the Invention) As described above in detail, the circular pattern determination method and apparatus of the present invention square Since the calculation is performed in each direction, the predetermined region can be determined based on these average values by determining that the predetermined region has a collection of linear patterns such as blood vessel shadows as an region with a low probability of being a tumor shadow. The level of probability, which is a region within the circular pattern, is determined.

[Brief explanation of drawings]

FIG. IA and FIG. 1B show the X-ray image of the region specified on the X-ray image, its X direction, y direction, and Figure 2 is a schematic diagram of an example of an X-ray imaging device; Figure 3 is a schematic diagram of an example of a chest X-ray image stored and recorded on a stimulable phosphor sheet. FIG. 4 is a perspective view showing an example of an X-ray image reading device and a computer system incorporating an embodiment of the present invention, and FIG. In order to explain an example of a real space filter that extracts a region with a higher probability, FIG. pixel P of
31 32 A diagram showing an example of the profile of an X-ray image in the direction (X direction) in which line segments Ll and L5 in FIG. FIG. 8 is a diagram for explaining how to obtain characteristic values used to determine whether or not a pixel is a pixel of image data f. The gradient f. FIG. 6... Rib shadow 10... X-ray imaging device 20... X-ray image reading device 23... Laser light source 2B... Rotating polygon mirror 29
... Stimulated luminescent light 30... Light guide 3l...
Photomultiplier 40...Computer system 7...Tumor shadow 14...Stormable phosphor sheet 2 Fig. 12a

Claims (2)

[Claims] (1) Based on image data representing a radiographic image of a subject, calculate a large number of differences between the image data corresponding to pixels close to each other within a predetermined region of the radiographic image in two different directions on the radiographic image. determining, for each of the directions, an average value of the absolute value or the square value of the large number of differences, and based on the average values in the two directions, the predetermined region constitutes the radiographic image. A circular pattern determination method characterized by determining the level of probability of an area within the circular pattern. (2) Based on the image data representing the radiographic image of the subject, calculate a large number of differences between the image data corresponding to pixels that are close to each other in a predetermined area of the radiographic image in two different directions on the radiographic image. difference calculation means for calculating the average value of the absolute value or square value of a large number of the differences for each of the directions; 1. A circular pattern determining device comprising determining means for determining the level of probability of a region within a circular pattern constituting the radiographic image.