JP2582669B2 - Device for determining benign / malignant breast mass shadow - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Application Field) The present invention relates to a method for determining whether a shadow (tumor shadow) of a tumor in a breast, which appears in an X-ray image of a human breast, is a shadow of a benign tumor. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a benign / malignant breast mass shadow judging apparatus for automatically judging a malignant mass shadow.

(Prior art) Reading a recorded radiation image to obtain image data,
After subjecting the image data to appropriate image processing, an image is reproduced and recorded in various fields. For example, an X-ray image is recorded using an X-ray film having a low gamma value designed to be compatible with later image processing, and this X-ray image is recorded.
The X-ray image is read from the film on which the line image is recorded, converted into an electric signal, the electric signal (image data) is subjected to image processing, and then reproduced as a visible image in a copy photograph or the like, so that the contrast and sharpness are improved. A system capable of obtaining a reproduced image having good image quality performance such as graininess has been developed (see Japanese Patent Publication No. 61-5193).

In addition, the applicant has determined that radiation (X-ray, α-ray, β-ray,
Ray, electron beam, ultraviolet ray, etc.), a part of this radiation energy is accumulated, and then, when irradiated with excitation light such as visible light, stimulable fluorescent light that emits the amount of photostimulated light corresponding to the accumulated energy Using the body (stimulable phosphor)
A radiation image of a subject such as a human body is once photographed and recorded on a sheet-shaped stimulable phosphor, and the stimulable phosphor sheet is scanned with excitation light such as a laser beam to generate stimulable emission light. Radiation recording / reproducing systems have already been proposed in which the emitted light is photoelectrically read to obtain image data, and based on this image data, a radiation image of the subject is output as a visible image on a recording material such as a photographic material, a CRT, or the like. (Japanese Patent Laid-Open No. 55-12
Nos. 429, 56-11395, 55-0163472, 56-16464
No. 5, No. 55-116340).

This system has the practical advantage of being able to record images over a very wide radiation exposure area compared to conventional silver halide photography radiography systems. That is, it has been recognized that the amount of stimulating light emitted by excitation after accumulation with respect to the amount of radiation exposure is proportional over a very wide range, and thus the amount of radiation exposure varies considerably depending on various imaging conditions. However, the stimulated emission light emitted from the stimulable phosphor sheet is read, the gain is set to an appropriate value, read by photoelectric conversion means, and converted into an electric signal (image data). By outputting a radiation image as a visible image to a display device such as a photosensitive material or a CRT, a radiation image that is not affected by a change in radiation exposure can be obtained.

In recent years, in a system using the X-ray film or the stimulable phosphor sheet, particularly a system configured for medical diagnosis of a human body, a reproduced image having good image quality suitable for observation (diagnosis) is simply obtained. In addition, pattern recognition of images has been performed (for example,
No. 125481, Japanese Patent Application No. 1-162909).

Here, image pattern recognition refers to an operation of extracting a target pattern from a complex radiographic image by performing various processes on image data, such as a chest X-ray image and a breast X-ray image of a human body. From a very complicated image mixed with various linear and circular patterns, for example, a substantially circular shadow (tumor shadow) corresponding to a nodule or a substantially circular shadow (calcification) corresponding to a calcified portion Shadowing).

Such a complicated radiographic image (for example, the chest X of the human body)
A target pattern (for example, a tumor shadow) is extracted from a line image or a breast X-ray image, and a visible image in which the extracted pattern is specified is reproduced and displayed, thereby assisting the observer in the observation (for example, a doctor's diagnosis). Assistance).

(Problems to be Solved by the Invention) In a system for obtaining an X-ray image of a breast using a recording sheet such as the above-mentioned X-ray film or stimulable phosphor sheet, a tumor in the breast which appears in the obtained X-ray image is obtained. The shadow (tumor shadow) appears as a substantially circular pattern in both the benign tumor and the malignant tumor, which is whitish (low in density or high in brightness) compared to the surroundings. Here, typically, when this tumor shadow is a benign tumor shadow, there is a slight difference that the tumor shadow is more circular than the malignant tumor shadow, but it is a benign tumor shadow or a malignant tumor shadow. It is often difficult to determine whether a tumor shadow is a tumor shadow even if a doctor observes the X-ray image. Even more, the detected tumor shadow is a benign tumor shadow or a malignant tumor shadow. There has been a problem that it is very difficult to automatically determine whether there is any data with high accuracy.

In view of the above circumstances, the present invention provides a breast capable of automatically and highly accurately determining whether a tumor shadow appearing on an X-ray image of the breast is a benign tumor shadow or a malignant tumor shadow. It is an object of the present invention to provide a benign / malignant determination apparatus for a tumor shadow.

(Means for Solving the Problems) A first apparatus for judging benign / malignant shadow of a breast mass according to the present invention for solving the above-mentioned object comprises: a plurality of X-rays having different energy distributions with respect to a human breast; First tumor shadow search means for obtaining a tumor shadow on the X-ray image by using at least one of a plurality of X-ray image data representing each of the plurality of X-ray images formed by: Subtraction processing means for performing subtraction processing using the X-ray image data of the subtraction image data in which the tumor shadow is emphasized; and using the subtraction image data, the X-ray image on the subtraction image A second tumor shadow search means for obtaining a tumor shadow in a region corresponding to the above-mentioned region where the tumor shadow has been obtained; When a tumor shadow is determined in a region corresponding to each other on the X-ray image, the tumor shadow is determined to be a tumor shadow of a malignant tumor, and the X-ray image is determined despite the tumor shadow being determined for the X-ray image. Determining a tumor shadow as a tumor shadow of a benign tumor when a tumor shadow is not obtained in a region on the subtraction image corresponding to the above-described tumor shadow region. It is a feature.

In addition, the second breast mass shadow benign / malignant determination device of the present invention is the first breast mass shadow benign / malignant determination device, wherein the second mass shadow search unit is configured to detect the X-ray on the subtraction image. A tumor shadow is also obtained for an area other than the area corresponding to the area where the tumor shadow is obtained on the image, and the determination unit determines that the tumor shadow is obtained for the other area on the subtraction image. In this case, the shadow of the tumor is determined to be a shadow of a malignant tumor.

(Action) On a normal breast X-ray image, both a benign tumor and a malignant tumor appear as substantially circular patterns similar to each other. When an image in which the shadow of the tumor is enhanced by performing energy subtraction processing is generated, the image becomes It is known that only malignant masses are specifically highlighted above (eg, T.As
aga et al “Breast Imaging: Dual-Energy Projection
Radiography with Digital Radiography "Radiology vo
l.164 No.3 pp867-870, Sep., 1987 "," Maga et al., "Dual-Energy Subtraction Radiaography of the Breast", Journal of the Japanese Society of Surgery, 89th, June 1988).

Here, the energy subtraction processing refers to a specific portion of the subject, for example, a soft tissue and a bone tissue in the chest when the chest of the human body is set as the subject, or a mass of the mammary gland when the subject is set in the breast of the human body, for example. Utilizing the fact that X-rays having different energies have different X-ray absorptances for X-rays having different energies, a plurality of X-rays having different energy distributions with respect to the same subject (for example, a chest or breast of a human body) are used. It refers to an operation of obtaining a line image, appropriately weighting the plurality of X-ray images, and calculating a difference between the X-ray images to extract a specific portion of the subject (for example, only soft tissue in the chest or only a tumor in the breast). The present applicant has also proposed an energy subtraction process using a stimulable phosphor sheet (see JP-A-59-83486 and JP-A-60-225541).

The present invention has been made based on the above findings.
That is, the benign / malignant determination apparatus for the first and second tumor shadows of the present invention obtains a tumor shadow appearing on the image before X before performing the energy subtraction processing (first tumor shadow detection means), and As for the subtraction image obtained by performing the energy subtraction processing, a tumor shadow appearing on the subtraction image is obtained (second tumor shadow detecting means), and the positions of the tumor shadows obtained on both images are compared. It was done.

Here, in the first breast mass shadow benign / malignant determination device of the present invention, when a mass shadow is obtained on the original image, a region on the subtraction image corresponding to the mass area obtained on the original image is obtained. In the case where a tumor shadow is obtained from both images, it is determined as a tumor shadow of a malignant tumor, and when no tumor shadow is obtained from a subtraction image, this is determined as a tumor shadow of a benign tumor. Thus, benign / malignant breast tumor shadows can be determined with high accuracy.

Further, a real space filter or the like for obtaining a tumor shadow on the original image may not be perfect, and for example, a tumor shadow whose density difference is not so clear from the surroundings may not be extracted. The second benign tumor mass benign / malignant judging device of the present invention is an improvement on this point. A tumor image on an original image is obtained by utilizing the fact that only a tumor shadow of a malignant mass emerges in a subtraction image. The tumor shadow is also obtained for the region other than the region on the subtraction image corresponding to the extracted region, and the obtained tumor shadow is classified as a tumor shadow of a malignant tumor, whereby the tumor shadow is detected with higher accuracy. It is determined whether the detected shadow of the tumor is a shadow of a benign tumor or a shadow of a malignant tumor.

(Example) Hereinafter, an example of the present invention will be described with reference to the drawings. Here, an example in which the above-described stimulable phosphor sheet is used will be described. Here, the breast mass shadow typically appears as a whitish (low density) substantially circular pattern on the visible image as compared with the surroundings.

 FIG. 3 is a schematic view of an example of an X-ray imaging apparatus.

An X-ray 12 is emitted from an X-ray source 11 of the X-ray imaging apparatus 10 toward a breast 13a of a human body 13 and transmitted through the breast 13a.
Is irradiated on the first stimulable phosphor sheet 14, and the X-rays transmitted through the first stimulable phosphor sheet 14 are transmitted through the filter 15 which transmits only the high energy X-rays among the transmitted X-rays. The transmitted X-ray image of the breast 13a of the human body is accumulated and recorded on the sheets 14 and 16, respectively, by irradiating the second stimulable phosphor sheet 16 via the.

Note that the X-ray imaging apparatus 10 accumulates and records X-ray images on two recording sheets 14 and 16 in one imaging.
One image may be taken at each of two timings that are temporally successive. Alternatively, a phosphor layer may be arranged on both the front side and the back side of one sheet, and each X-ray image may be stored and recorded on each of these phosphor layers.

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

After imaging is performed by the X-ray imaging apparatus 10 shown in FIG.
The first and second stimulable phosphor sheets 14 and 16 are set one by one at a predetermined position of the X-ray image reading device 20. Here, the case of reading the first stimulable phosphor sheet 14 will be described.

First stimulable phosphor sheet 14 set at a predetermined position
Is transported (sub-scanning) in the direction of arrow Y by sheet transporting means 22 such as an endless belt driven by a motor 21. On the other hand, the light beam 24 emitted from the laser light source 23
Is reflected and deflected by a rotating polygon mirror 26 driven by a motor 25 and rotating at a high speed in the direction of the arrow, passes through a converging lens 27 such as an fθ lens, and then changes its optical path by a mirror 28 to enter the sheet 14 and enter the sub-scan The main scanning is performed in an arrow X direction substantially perpendicular to the arrow Y direction). From the portion of the sheet 14 irradiated with the excitation light 24, stimulated emission light 29 of an amount corresponding to the stored and recorded X-ray image information is diverged, and the stimulated emission light 29 is guided by the light guide 30. , Are photoelectrically detected by a photomultiplier (photomultiplier tube) 31. The light guide 30 is formed by molding a light-guiding material such as an acrylic plate, and is arranged so that a linear incident end face 30a extends along a main scanning line on the sheet 14, and is annular. The light receiving surface of the photomultiplier 31 is connected to the emission end face 30b formed at the end. The stimulated emission light 29 that has entered the light guide 30 from the incident end face 30a travels through the inside of the light guide 30 by repeating total reflection, exits from the exit end face 30b, is received by the photomultiplier 31, and is subjected to X-ray Photostimulated luminescence 29
Is converted into an electric signal by the photomultiplier 31.

The analog output signal S0 output from the photomultiplier 31 is logarithmically amplified by a logarithmic amplifier 32, and the A / D converter 3
Digitization is performed in step 3 to obtain first image data S1 as an electric signal.

The obtained first image data S1 is input to the computer system 40. This computer system 40 has a configuration including an example of a benign / malignant determination apparatus for a breast mass shadow according to the present invention. A main body section 41 containing a CPU and an internal memory, and a floppy disk as an auxiliary memory are inserted. The system comprises a drive unit 42 to be driven and operated, a keyboard 43 for an operator to input necessary instructions and the like to the computer system 40, and a CRT display 44 for displaying necessary information.

Next, as described above, the second stimulable phosphor sheet 16
The second image data S2 representing the second X-ray image stored and recorded in the computer system is obtained, and this second image data is also input to the computer system 40.

Here, the first X-ray image is an image based on X-rays having a relatively high energy component and the second X-ray image is an image based on an X-ray having a relatively high energy component. Are the original images before the subtraction processing.

As described above, the first and second image data S1, S2
Is input into the computer system 40, the two image data S
The relative positioning of each X-ray image carried by S1 and S2 is performed on the image data (see JP-A-58-163338). Thereafter, for each of the first and second X-ray images corresponding to each other, an addition process is performed in accordance with the equation S3 = (S1 + S2) / 2 (1) to thereby obtain the first and second X-ray images. Superimposed image (superimposed image data S
3) is obtained. This superimposed image is also the original image before the subtraction processing. Here, the extraction of a tumor shadow on this superimposed image is performed as described below. Note that it is also possible to use the first X-ray image or the second X-ray image instead of the superimposed image,
Since the superimposed image is an image in which two X-ray images are superimposed, the noise component is reduced as compared with any of these X-ray images, which is advantageous for the extraction processing of a tumor shadow.

FIG. 1 is a view showing an example of a real space filter for extracting the above-mentioned tumor shadow, in which a predetermined pixel P0 on a superimposed image (hereinafter simply referred to as “original image”) is centered on the image. FIG. It is determined whether the predetermined pixel P0 is a pixel within the tumor shadow. By scanning the original image using the filter as shown here, a tumor shadow on the original image is extracted.

As shown in FIG. 1, a plurality of (eight in this case) line segments Li extending from a predetermined pixel P0 in the original image to the periphery of the original image
(I = 1, 2,..., 8) and a circle R with a radius r centered on a predetermined pixel P0. Further, a central area Q0 including a predetermined pixel P0 and each pixel Pi (i = 1, 2,..., 8) at the intersection of each of the line segments Li (i = 1, 2,. Consider each surrounding area Qi. The radius r, the central region Q0, the area of each peripheral region Qi, the number of peripheral regions to be assumed, and the like are the ratio of the noise component included in the original image, the size of the target tumor shadow, the determination accuracy, It is appropriately determined in consideration of the calculation speed and the like. Here, each pixel Pi is assumed to be equidistant from the predetermined pixel P0 by an equal distance r. For example, when an elliptical tumor shadow having a major axis in the X direction in FIG. 1 is targeted,
Each pixel Pi may have a different distance from a predetermined pixel P0, such as selecting a pixel far from the pixel P0 as the pixels P1 and P5.

Average values Q0, Qi (i = 1, 2,..., 8) of a large number of image data corresponding to a large number of pixels in the central area Q0 and the peripheral areas Qi assumed as described above are obtained. Here, for simplicity, the same number is used for the symbol indicating each area Q0, Q1 (i = 1, 2,..., 8) and the symbol indicating the average value of the image data in each area. .

Next, each difference Δi (i = 1, 2,..., 8) between the average value Q0 of the central region and the average value Qi of each peripheral region is obtained as Δi = Qi−Q0 (2).

Here, a first characteristic value representing the difference Δi (i = 1, 2,..., 8), and the difference Δi (i = 1, 2,
, 8) as the second characteristic values representing the variations of the differences Δi Are used, and the average value Δ and the variance σ ² are obtained.

Then, these mean delta, variance sigma ² ratio C1 is, Obtained as this ratio C1 is compared with a predetermined threshold value Th1, the predetermined pixel P0 relatively average value for Δ is large variance sigma ² is small in the case of C1 ≧ Th1 is a pixel in the tumor shadow If C1 <Th1, the predetermined pixel P0 is determined to be a pixel outside the tumor shadow.

FIG. 2 is a diagram virtually drawn on a predetermined pixel P0 on the original image to explain another spatial filter for extracting a tumor shadow.

As shown in FIG. 2, a plurality (eight in this case) of line segments Li extending from a predetermined pixel P0 in the original image to the periphery of the original image
(I = 1, 2,..., 8), and three circles Rj (j = 1, 2,
Assume 3). The central region including the predetermined pixel P0 is defined as Q0, and each pixel Pij located at each intersection of each line segment Li and each circle Rj
(In FIG. 2, symbols for P11, P12, P13, P51, P52, and P53 are shown.) Each peripheral region is represented by Qij (i = 1, 2,..., 8;
j = 1,2,3) (However, in FIG. 2, Q0 and Q1 are explicitly specified.
Only 1, Q12, Q13, Q51, Q52 and Q53 are shown. ).

For each of the regions Q0 and Qij (i = 1, 2,..., 8; j = 1, 2, 3), the average of a large number of image data corresponding to a large number of pixels in each of the regions Q0, Qij Values Q0, Qij (i = 1, 2,..., 8; j = 1, 2,
3) is required. Here, for simplicity, each region Q0,
The same symbol is used for the symbol indicating Qij (i = 1, 2,..., 8; j = 1, 2, 3) and the signal indicating the average value of the image data in each area.

Next, each difference Δij (i = 1, 2,..., 8; j = 1, 2, 3) between the average value Q0 of the central region and the average value Qij of each peripheral region is represented by Δij = Qij−Q0. (4) The maximum value Δi of the difference Δij is obtained for each line segment Li.

Next, a first characteristic value U representing the maximum value Δi (i = 1 to 8) and a second characteristic value V representing the variation of the maximum value Δi (i = 1 to 8) are obtained. For this purpose, first, the characteristic values U1 to U4 and V1 to V4 are obtained according to the following arithmetic expressions.

U1 = (Δ1 + Δ2 + Δ5 + Δ6) / 4 (5) U2 = (Δ2 + Δ3 + Δ6 + Δ7) / 4 (6) U3 = (Δ3 + Δ4 + Δ7 + Δ8) / 4 (7) U4 = (Δ4 + Δ5 + Δ8 + Δ1) / 4 ... (8) V1 = U1 / U3 (9) V2 = U2 / U4 (10) V3 = U3 / U1 (11) V4 = U4 / U2 (12) Here, for example, the characteristic value U1 is obtained according to equation (5). The addition of two adjacent areas (Δ1 and Δ2, or Δ5 and Δ6) means smoothing, and the areas (Δ1 and Δ2) on the opposite sides of the pixel P0 are interposed.
The addition of (+ Δ2 and Δ5 + Δ6) is performed so that, for example, even if there is a tumor shadow in an area having a density gradient, the tumor shadow can be detected.

Also, for example, when the characteristic value V1 is determined according to the equation (9), the characteristic value U1 and the characteristic value U3 are characteristic values determined in directions orthogonal to each other. Therefore, the tumor shadow 7 shown in FIG. In this case, V1 = 1.0 and deviates from the circle, that is, V1 deviates from 1.0 when the pixel P0 is within a linear shadow such as a rib shadow.

As the first characteristic value U representing the maximum value Δi (i = 1 to 8) of the difference, the maximum value U = MAX (U1, U2, U3, U4) of U1 to U4 is adopted. The maximum value of V1 to V4, V = MAX (V1, V2, V3, V4), as the second characteristic value V representing the variation of the maximum value Δi (i = 1 to 8) of the difference is as follows. Adopted. Thus, the first and second characteristic values
When U and V are obtained, the ratio of these first and second characteristic values is used as a characteristic value C2 for determining whether the predetermined pixel P0 is a pixel in the tumor shadow. Is adopted, and the characteristic value C2 is compared with a predetermined threshold value Th2. Depending on whether C2 ≧ Th2 or C2 <Th2, the pixel value
It is determined whether or not P0 is a pixel in the tumor shadow.

As described above, in the computer system 40 shown in FIG. 4, by scanning the original image using the real space filter, a tumor shadow considered as a tumor shadow is extracted.

Note that, in each of the above filter examples, as shown in FIGS. 1 and 2, the image data corresponding to each peripheral area Qi (Qij) including the pixels Pi (Pij) on the eight line segments L1 to L8. Average value Qi
Although (Qij) is used, it is needless to say that this line segment does not need to be eight, and may be, for example, sixteen. In the filter described with reference to FIG. 2, r1, r2,
The calculation was performed for the three distances of r3, but this is not limited to the three distances as well, and in order to extract tumor shadows of various sizes with higher accuracy, the distances were calculated from r1 to r3.
The calculation may be continuously changed until the calculation.

Further, each of the above filters is merely an example, and the present invention is not limited to a specific filter as an abnormal shadow extraction filter, and various known filters can be employed.

When a tumor shadow is extracted on the original image (superimposed image) as described above, the first and second X-ray images are then subjected to subtraction processing. In this subtraction processing, for each pixel corresponding to each other in the first and second X-ray images, the equation S4 = W2 · S2-W1 · S1 + C (16) where W1 and W2 are weighting coefficients, and C is bias Represents minutes.

By performing weighted subtraction in accordance with
Here, by appropriately selecting the weighting coefficients W1 and W2 and performing the subtraction processing according to the expression (16), a subtraction image (subtraction image data S4) in which the tumor shadow of the malignant tumor is emphasized is obtained.

When a subtraction image is obtained in this way,
The tumor shadow appearing in the subtraction image is extracted by using the same real space filter as in the case of the original image described above. However, since the region where the tumor shadow exists on the original image has already been found, the above real space filter was used only for the region on the subtraction image corresponding to the region where the tumor shadow exists on the original image, saving computation time. Scanning is performed. The threshold values Th1 and Th2 used for determining whether or not the predetermined pixel P0 is a pixel in the tumor shadow in the above-described real space filter are obtained when the tumor shadow is obtained on the original image and when the tumor is displayed on the subtraction image. When looking for a shadow,
Naturally, different threshold values may be used to suit each image, and the original image and the subtraction image have different image characteristics such as a ratio of a signal component to a noise component. Since there are many cases, the real space filter itself for obtaining the tumor shadow may be changed so as to be suitable for each image between the case where the tumor shadow is obtained on the original image and the case where the tumor shadow is obtained on the subtraction image.

As described above, a tumor shadow is obtained for both the original image and the subtraction image, and when a tumor image is obtained by both, the tumor shadow is determined to be a tumor shadow of a malignant tumor, and a tumor shadow is displayed on the original image. If a shadow is found but no tumor shadow is found in the corresponding region of the subtraction image, the tumor shadow is determined to be a benign tumor shadow.

The real space filter for finding a tumor shadow is not perfect. For example, in the case of a tumor shadow having a small density difference (difference in image data value) from the surroundings on the original image, if the tumor shadow is not extracted Is also conceivable. Therefore, if the calculation time is acceptable, scan the entire subtraction image with the tumor shadow extraction filter, and if a tumor shadow is found in the region on the subtraction image corresponding to the region where no tumor shadow was found on the original image, The tumor shadow may be classified as a tumor shadow of a malignant tumor.

When the benign / malignant determination of the tumor shadow is made in this manner, the determination result is sent to the CRT display shown in FIG. 4 together with the image data representing the original image and the position information of the tumor shadow on the original image. A visible image in which the position of the tumor shadow and the determination result of benign / malignant are clearly displayed on the display screen.

Although the above embodiment is an example using a stimulable phosphor sheet, the present invention is not applied only to a system using a stimulable phosphor sheet. It can also be applied to a system using such a combination.

(Effects of the Invention) As described in detail above, the benign tumor mass benign / malignant determination device of the present invention obtains a tumor shadow for both the original image before the subtraction process and the subtraction image after the subtraction process. The benign / malignant determination of the tumor shadow is made based on whether or not a tumor shadow has been obtained at a position corresponding to each other, so that highly accurate determination is possible.

[Brief description of the drawings]

FIG. 1 is a diagram virtually drawn on a predetermined pixel P0 on an original image to explain a spatial filter for extracting a circular tumor shadow, and FIG. FIG. 3 is a diagram virtually drawn on a predetermined pixel P0 on an original image on the image to explain another spatial filter for extracting a tumor shadow of the X-ray image photographing apparatus. FIG. 4 is a perspective view showing an example of an X-ray image reading apparatus and an example of a computer system. 10 X-ray imaging device, 14 stimulable phosphor sheet 20 X-ray image reading device 23 Laser light source 26 Rotating polygon mirror 29 Photostimulated emission light 30, Light guide 31 …… Photomultiplier 40 …… Computer system

Claims (2)

(57) [Claims] 1. A plurality of X-rays formed by a plurality of X-rays having different energy distributions from a human breast.
A first tumor shadow search unit for obtaining a tumor shadow on the X-ray image by using at least one of the plurality of X-ray image data representing each of the line images; Subtraction processing means for obtaining subtraction image data in which the tumor shadow is emphasized, and a tumor shadow on the X-ray image is obtained on the subtraction image using the subtraction image data. A second tumor shadow search means for finding a tumor shadow in a region corresponding to the selected region, and, when a tumor shadow is found in a region corresponding to the X-ray image and the subtraction image, the tumor shadow is converted into a malignant tumor mass. It is determined that the shadow is a shadow, and although the tumor shadow is obtained for the X-ray image, it corresponds to the region of the tumor shadow on the X-ray image. Determining a tumor shadow as a benign tumor shadow when a tumor shadow is not found in a region on the subtraction image, wherein the benign / malignant breast tumor shadow is benign / malignant. Judgment device. 2. The method according to claim 1, wherein the second tumor shadow search means calculates a tumor shadow in an area other than the area corresponding to the area on the X-ray image on which the tumor shadow is obtained on the subtraction image. Wherein, when a tumor shadow is determined in the other area on the subtraction image, the determination unit determines that the tumor shadow is a tumor shadow of a malignant tumor. Item 7. The breast mass shadow benign / malignant determination device according to Item 1.