Radiation image processing device and its operating method

Abstract

To provide a radiation image processing device capable of improving accuracy in evaluating extensibility of a mammary gland, and to provide an operation method thereof.SOLUTION: A processor provided in a radiation image processing device executes control to: acquire two radiation images in each of which each breast of one subject is imaged; calculate a pixel mammary gland ratio, which is a ratio of a mammary gland of each pixel in each of the radiation images on the basis of the radiation images; calculate a region mammary gland ratio in a predetermined region of each of the radiation images on the basis of the pixel mammary gland ratio; evaluate extensibility of the mammary gland by comparing the two region mammary gland ratios calculated respectively using the two radiation images; and display an evaluation result of the extensibility.SELECTED DRAWING: Figure 10

Description

Technical Field

[0001] The present invention relates to a radiation image processing apparatus for evaluating the extensibility of a breast using a radiation image and an operating method thereof.

Background Art

[0002] In mammography, it is important to properly position the breast of a subject in order to obtain a radiation image in which reading by a doctor and image processing for assisting the reading are appropriately performed. One of the evaluation items for determining whether the positioning is appropriate is "extensibility of the breast".

[0003] As a method for evaluating the extensibility of the breast, a method of evaluating based on the area ratio of the breast region in one radiation image and the contrast within the breast is known (Patent Document 1). Also, a method of calculating the consistency of the shape of the breast by pattern matching between radiation image data of the same subject and determining whether the position and degree of compression of the breast are appropriate is known (Patent Document 2).

Prior Art Documents

Patent Documents

[0004]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0005] In the method using the area ratio of the breast region and the contrast within the breast for one radiation image, or the method using the consistency of the shape of the breast for two radiation images, the evaluation of the extensibility of the breast may be insufficient.

[0006] The present invention aims to provide a radiographic image processing apparatus and a method for operating the same that can improve the accuracy of evaluating the extensibility of mammary glands. [Means for solving the problem]

[0007] The radiographic image processing apparatus of the present invention comprises a processor, which acquires two radiographic images of the same subject's breast, calculates a pixel-level breast tissue ratio, which is the proportion of breast tissue per pixel in each radiographic image, calculates a regional breast tissue ratio in a predetermined area of ​​each radiographic image based on the pixel-level breast tissue ratio, evaluates the extensibility of the breast tissue by comparing the two regional breast tissue ratios calculated using the two radiographic images, and performs control to display the evaluation result of the extensibility.

[0008] In a radiographic image, it is preferable that the processor designates a region having pixels in which the pixel breast tissue ratio is equal to or greater than a first region threshold as a predetermined region.

[0009] The processor preferably calculates the amount of breast tissue for each pixel using the pixel breast tissue ratio, and defines a predetermined region in the radiographic image as a region containing pixels where the amount of breast tissue is equal to or greater than the second region threshold.

[0010] In a radiographic image, it is preferable that the processor defines a predetermined region as an area within a specific distance range from the pixel with the highest proportion of breast tissue.

[0011] The processor preferably evaluates extensibility based on the difference in the proportion of breast tissue in two regions, calculated using radiographic images.

[0012] The regional breast tissue percentage is preferably a statistical value of the pixel breast tissue percentage in a predetermined region. Alternatively, the regional breast tissue percentage is preferably the volume percentage of breast tissue in a predetermined region.

[0013] The processor evaluates extensibility using a degree of agreement calculated based on two regional breast tissue percentages calculated using radiographic images, and it is preferable that the regional breast tissue percentage is the distribution of pixel breast tissue percentages in a predetermined region.

[0014] Preferably, the radiographic image selected from the two radiographic images is a radiographic image of the subject's left and right breasts, respectively.

[0015] Preferably, the radiographic image selected from the two radiographic images is a radiographic image of the same subject's breast taken at different time points.

[0016] Preferably, the processor generates an operation instruction display based on the evaluation results of extensibility and the regional breast tissue percentage of the radiographic images selected from multiple radiographic images, and controls the display of the operation instruction display.

[0017] Preferably, the processor classifies the breast structure of each radiographic image based on the pixel-to-mammary gland ratio and controls the display of the breast structure.

[0018] The method for operating the radiographic image processing apparatus of the present invention comprises the steps of: acquiring two radiographic images of the breast of the same subject; calculating the pixel-based breast tissue ratio, which is the proportion of breast tissue per pixel in each radiographic image, based on the radiographic images; calculating the regional breast tissue ratio in a predetermined region of each radiographic image based on the pixel-based breast tissue ratio; evaluating the extensibility of the breast tissue by comparing the two regional breast tissue ratios calculated using the two radiographic images; and performing control to display the results of the extensibility evaluation. [Effects of the Invention]

[0019] According to the present invention, the accuracy of evaluating the extensibility of mammary glands can be improved. [Brief explanation of the drawing]

[0020] [Figure 1]It is a block diagram showing the functions of a radiation image processing apparatus. [Figure 2] It is an external view of a mammography apparatus. [Figure 3] It is an external view showing a mode of rotating a movable part. [Figure 4] It is an external view showing a mode of vertically moving a movable part. [Figure 5] It is an explanatory diagram showing a method of taking a radiation image by a mammography apparatus. [Figure 6] It is an image diagram showing an example of an area included in a radiation image. [Figure 7] It is an image diagram showing an example of a pixel included in a radiation image. [Figure 8] It is an explanatory diagram schematically showing adipose tissue and mammary gland tissue included in a breast corresponding to a pixel of a radiation image. [Figure 9] It is an explanatory diagram showing pixel mammary gland ratio. [Figure 10] It is an image diagram showing an example of a display image for comparing a right breast image and a left breast image and displaying an evaluation result as character information. [Figure 11] It is an image diagram showing an example of a display image for comparing a right breast image and a left breast image and displaying an evaluation result by making markers indicating a predetermined area different from each other. [Figure 12] It is an image diagram showing an example of a display image for comparing a right breast image and a left breast image and displaying an evaluation result by displaying a warning marker. [Figure 13] It is an image diagram showing an example of a display image for comparing a past image and a current image and displaying an evaluation result as character information. [Figure 14] It is an image diagram showing an example in the case where a predetermined area is an area within a specific distance range from a pixel having the maximum pixel mammary gland ratio. [Figure 15] It is an explanatory diagram showing an example of the volume of a breast. [Figure 16] It is an explanatory diagram showing an example of the volume of a mammary gland. [Figure 17] It is a block diagram showing the functions of a display control unit. [Figure 18]This image diagram shows an example of a display image used when displaying operation instructions. [Figure 19] This image shows an example of a display image used to show breast structure. [Figure 20] This is a flowchart showing how the radiation image processing device operates. [Modes for carrying out the invention]

[0021] As shown in Figure 1, the radiation image processing device 10 receives radiation images from a modality 11 that takes radiation images, or from a database 12 in which radiation images are stored. The radiation images are acquired by the image acquisition unit 30 of the radiation image processing device 10. The radiation image processing device 10, the modality 11, and the database 12 are connected to each other so as to be able to communicate with each other via wired or wireless connections over a network. The network is, for example, the Internet or a LAN (Local Area Network).

[0022] As shown in Figure 1, the radiographic image processing device 10 includes an image acquisition unit 30, a pixel-to-mammary tissue ratio calculation unit 40, a region determination unit 50, a region-to-mammary tissue ratio calculation unit 60, a stretchability evaluation unit 70, and a display control unit 80. The radiographic image processing device 10 evaluates stretchability by comparing multiple radiographic images and generates a display image that shows the evaluation result of stretchability. The display image is shown on a display 13 so that a user such as a radiologist can see it.

[0023] The radiation image processing device 10 is a computer equipped with a processor. The control unit (not shown), which is composed of the processor, operates programs related to various processing or control functions that are incorporated into the program storage memory (not shown) provided in the radiation image processing device 10, thereby realizing the functions of the image acquisition unit 30, the pixel breast tissue ratio calculation unit 40, the region determination unit 50, the region breast tissue ratio calculation unit 60, the extensibility evaluation unit 70, and the display control unit 80.

[0024] In this embodiment, modality 11 is a mammography apparatus 20. The mammography apparatus 20 is a radiography apparatus that uses radiation such as X-rays or gamma rays to image the breast of a patient. As shown in Figure 2, the mammography apparatus 20 comprises a support column 21 and a movable part 22. The movable part 22 comprises a radiation generating unit 23, an imaging table 24, a compression plate 25, and a lifting unit 26. The movable part 22 integrates the radiation generating unit 23 and the imaging table 24 which houses the radiography unit, and is also vertically movable while maintaining the relative positions of the radiation generating unit 23 and the imaging table 24, and is rotatable within a predetermined angular range, in order to adjust the imaging position when imaging a patient using radiation. The support column 21 supports the movable part 22.

[0025] The radiation generating unit 23 irradiates the radiography unit with radiation. When the radiation generating unit 23 generates X-rays, the radiation generating unit 23 is an X-ray tube or a monotank that integrates an X-ray tube with other circuits. The radiography unit is a radiation detector that uses the radiation that has passed through the subject to take an image of the subject. The radiography unit is, for example, an FPD (Flat Panel Detector).

[0026] The imaging table 24 is a stage on which the patient's breasts are positioned, and during imaging, the patient's breasts are held in place using a compression plate 25. The imaging table 24 is also equipped with a gripping part 27a, which the patient grasps with their right hand, and a gripping part 27b, which the patient grasps with their left hand. The gripping parts 27a and 27b are so-called armrests. The compression plate 25 compresses and flattens the patient's breasts placed on the imaging table 24. This is to reduce the overlap of normal mammary glands and to make it easier to clearly detect any lesions. When imaging is performed, a user such as a radiologist positions the patient's breasts and then compresses them in order to obtain a radiographic image suitable for interpretation. The lifting unit 26 raises and lowers the compression plate 25 relative to the imaging table 24. As a result, the lifting unit 26 supports the compression plate 25 approximately parallel to the imaging table 24 and at a distance corresponding to the thickness of the breast.

[0027] As shown in Figure 3, the movable part 22 can rotate within a predetermined angular range while maintaining the relative position and orientation of the radiation generating unit 23 and the imaging table 24. Therefore, the mammography device 20 can perform imaging with the imaging table 24 positioned horizontally or with the imaging table 24 tilted from horizontal. Specifically, as shown in Figure 3(A), the mammography device 20 can perform CC imaging (craniocaudal imaging), where the imaging table 24 is positioned horizontally and the breast is imaged from the craniocaudal direction. Also, as shown in Figure 3(B), the mammography device 20 can perform MLO imaging (medio-lateral oblique imaging), where the imaging table 24 is tilted and the breast is imaged from the medial-lateral oblique direction. Although not shown in the illustration, the mammography device 20 can rotate the imaging table 24, etc., in the opposite direction to that shown in Figure 3(B).

[0028] As shown in Figure 4, the movable part 22 can move vertically while maintaining the relative position and orientation of the radiation generating unit 23 and the imaging table 24. Therefore, the mammography device 20 can adjust the height of the imaging table 24 to match the height of the patient, allowing imaging to be performed in a comfortable posture. The height of the imaging table 24 is the height of the imaging table 24 relative to the floor surface 28 (the contact surface of the support column 21), and can be arbitrarily adjusted within the range of the minimum height H1 shown in Figure 4(A) and the maximum height H2 shown in Figure 4(B).

[0029] Radiation emitted from the radiation generating unit 23 passes through the patient's breast and enters the radiography unit, where it is detected and a radiographic image is taken. In mammography, typically, both a radiographic image of the patient's right breast and a radiographic image of the patient's left breast are obtained. The radiographic images are transmitted to the radiographic image processing device 10 and / or the database 12.

[0030] Database 12 is a storage device, file server, or cloud storage for storing radiographic images. Database 12 may also be part of a system that directly or indirectly interacts with the radiographic image processing device 10, such as a radiology information system (RIS), a hospital information system (HIS), or a picture archiving and communication system (PACS). Multiple radiographic images stored in database 12 are transmitted to the radiographic image processing device 10.

[0031] The following describes the process by which the radiographic image processing device 10 compares multiple radiographic images taken of the same subject's breast and displays the results of the extensibility evaluation. The image acquisition unit 30 acquires two radiographic images taken of the same subject's breast from the modality 11, the mammography device 20, or from the database 12. The two radiographic images may be selected by a user such as a radiologic technologist or a doctor, or the radiographic image processing device 10 may automatically select and acquire them using the patient number and date and time of acquisition attached to the radiographic images. The image acquisition unit 30 transmits the received radiographic images to the pixel breast tissue ratio calculation unit 40.

[0032] In this embodiment, the "radiological image" includes a preliminary image (pre-shot image) taken at a low dose during the examination to adjust the positioning. One of the two radiological images acquired by the image acquisition unit 30 may be used as the pre-shot image.

[0033] The pixel-based mammary gland ratio calculation unit 40 calculates the pixel-based mammary gland ratio for each pixel in the radiographic image. The pixel-based mammary gland ratio is the proportion of mammary gland tissue in each pixel of the radiographic image. The mammary gland ratio is the proportion of mammary gland tissue in the thickness direction of the breast in the breast region corresponding to each pixel that makes up the radiographic image.

[0034] The following explains the pixel-percentage of breast tissue. First, we will explain the radiographic image taken by the mammography device 20. In mammography, as shown in Figure 5, the breast B of the subject Ob, placed on the imaging table 24, is compressed and stretched with a compression plate 25, and then irradiated with radiation 23a. At this time, the distance between the compression plate 25 and the imaging table 24 becomes the thickness T of the breast. The radiography unit 24a built into the imaging table 24 detects the radiation that has passed through the breast B from the compression plate 25 toward the imaging table 24, and obtains a radiographic image 100 as shown in Figure 6.

[0035] The specific example of the radiographic image 100 shown in Figure 6 is an example of a radiographic image obtained by MLO imaging. The radiographic image 100 includes a direct region 101, which is a pass-through region where the radiation did not penetrate the subject Ob's breast B but was directly irradiated to the radiography area, and a pectoralis major region 102, which was penetrated by the subject Ob's pectoralis major muscle. The area remaining after removing the direct region 101 and the pectoralis major region 102 from the radiographic image 100 is the breast region 103 (shown by a dashed line).

[0036] Breast B, depicted as breast region 103 on the radiographic image, is mainly composed of adipose tissue and glandular tissue. Adipose tissue absorbs radiation relatively poorly and transmits it well. Therefore, in the radiographic image 100, areas containing a large amount of adipose tissue are displayed in black as low-absorption areas 104. On the other hand, glandular tissue absorbs radiation relatively well and transmits it poorly. Therefore, in the radiographic image 100, areas containing a large amount of glandular tissue are displayed in white as high-absorption areas 105.

[0037] In Figure 6, the breast region 103 shows a mixture of low-attenuation areas 104, which contain a large amount of adipose tissue, and high-attenuation areas 105, which contain a large amount of glandular tissue, in the planar direction. In Figure 6, the direct region 101, the pectoralis major region 102, and the low-attenuation area 104 are shown with different patterns, while the high-attenuation area 105 is not shown with a pattern, indicating that each region contains different absorbers or absorbers in different proportions. Furthermore, in Figure 6 and subsequent diagrams showing examples of radiographic images, the high-attenuation area 105 is labeled with a symbol to avoid cluttering the diagrams.

[0038] The pixel-by-pixel breast tissue ratio is calculated for each pixel 200a, 200b, 200c, 200d, and 200e that make up the radiographic image 100, as shown in Figure 7. Figure 7 shows the radiographic image 100 obtained by irradiating breast B with radiation 23a from the front to the back of the paper (see Figure 5). Also, in Figure 7, pixels 200a, 200b, 200c, 200d, and 200e are drawn larger for better visibility.

[0039] Figure 8 schematically shows the adipose tissue 201 and glandular tissue 202 contained in the breast B region corresponding to each pixel 200a, 200b, 200c, 200d, and 200e of the radiographic image 100 shown in Figure 7. As shown in Figure 8, in the breast B region corresponding to each pixel 200a, 200b, 200c, 200d, and 200e, adipose tissue 201 (shown as a dot pattern) and glandular tissue 202 (shown as an area without a pattern) are mixed in the direction of the breast thickness T. Note that in Figure 8 and Figure 9 (described later), the distance between pixels is drawn narrowly for better visibility of the figures.

[0040] The pixel-to-mammary gland ratio calculation unit 40 calculates the pixel-to-mammary gland ratio 203a for pixel 200a, 203b for pixel 200b, 203c for pixel 200c, 203d for pixel 200d, and 203e for pixel 200e, taking into account the attenuation by adipose tissue 201 and mammary gland tissue 202 to the radiation 23a irradiated in the direction of the thickness T of the breast, as shown in Figure 9.

[0041] The pixel-based mammary gland ratio calculation unit 40 calculates the extent to which mammary gland tissue 202 is contained within adipose tissue 201 in the direction of breast thickness T of the pixels constituting the radiographic image, using the attenuation coefficient of adipose tissue 201 and the attenuation coefficient of mammary gland tissue 202, as the pixel-based mammary gland ratio. Therefore, by calculating the pixel-based mammary gland ratio, it is possible to accurately represent the extent to which mammary gland tissue is present in the area occupied by each pixel corresponding to each breast region, compared to values ​​that do not adequately consider the overlap of mammary glands in the breast thickness direction, such as, for example, representing the grayscale of pixels in a radiographic image with 256 grayscale pixel values.

[0042] The pixel breast tissue percentage calculation unit 40 calculates, for example, that pixel breast tissue percentage 203a is 40%, pixel breast tissue percentage 203b is 90%, pixel breast tissue percentage 203c is 75%, pixel breast tissue percentage 203d is 20%, and pixel breast tissue percentage 203e is 10%, as values ​​between 0% and 100%.

[0043] The pixel glandular tissue ratio may be calculated by substituting the pixel values ​​of the direct region 101, the pixel values ​​of the adipose tissue pixels, and the pixel values ​​of each individual pixel into the formula ([Equation 1]) for calculating the "glandular tissue content" disclosed in Japanese Patent Application Publication No. 2010-253245. Adipose tissue pixels are pixels corresponding to the portion of the breast that is presumed to be composed solely of adipose tissue. Adipose tissue pixels may be the pixels in the breast region 103 that have the highest amount of radiation 23a transmitted through them, or they may be pixels with a pixel value equal to or greater than the threshold for determining adipose tissue pixels.

[0044] Furthermore, the pixel glandular tissue ratio may be calculated by substituting the value obtained from the dose received by the direct region 101, the value obtained from the dose received by the adipose tissue pixel, the value obtained from the dose received by the breast region 103 of each pixel, the attenuation coefficient for adipose tissue, and the attenuation coefficient for glandular tissue into the formula for calculating the "glandular tissue ratio" disclosed in Japanese Patent Application Publication No. 2020-370 (formula (1) or (5)).

[0045] The region determination unit 50 determines a predetermined region to be used to calculate the region of breast tissue using the pixel breast tissue ratio. The predetermined region is composed of multiple pixels and may be set by user input or automatically. The region determination unit 50 may set the breast region 103 as the predetermined region, or, as will be described later, a breast tissue concentration region having pixels with a particularly large pixel breast tissue ratio or breast tissue volume, as will be described later, may be set as the predetermined region. Furthermore, as will be described later, a local region which is an area within a certain range from a certain pixel may be set as the predetermined region.

[0046] The regional breast tissue percentage calculation unit 60 calculates the regional breast tissue percentage in a predetermined region using the pixel breast tissue percentage calculated for each radiographic image. The regional breast tissue percentage may be a statistical value of the pixel breast tissue percentage in a predetermined region, a volume ratio of breast tissue in a predetermined region, or a distribution of the pixel breast tissue percentage in a predetermined region, as will be described later. For example, the maximum value of the pixel breast tissue percentage in a predetermined region of a radiographic image of the subject's right breast (right breast image) is calculated as the regional breast tissue percentage to be "30%". Similarly, the maximum value of the pixel breast tissue percentage in a predetermined region of a radiographic image of the subject's left breast (left breast image) is calculated as the regional breast tissue percentage to be "60%".

[0047] The regional breast tissue percentage calculated for each radiographic image by the regional breast tissue percentage calculation unit 60 is transmitted to the stretchability evaluation unit 70. The stretchability evaluation unit 70 compares the regional breast tissue percentage in predetermined regions of two radiographic images, evaluates the stretchability, and outputs the evaluation result.

[0048] For example, the stretchability evaluation unit 70, when the statistical quantity of the pixel breast tissue ratio in a predetermined region is defined as the regional breast tissue ratio, calculates the difference in the regional breast tissue ratio in each predetermined region of two radiographic images. If the difference in the regional breast tissue ratio is less than the threshold for stretchability evaluation, it outputs an evaluation result of "good stretchability". If the difference in the regional breast tissue ratio is greater than or equal to the threshold for stretchability evaluation, it outputs "poor stretchability" for the predetermined region in the radiographic image with the higher regional breast tissue ratio.

[0049] Furthermore, the stretchability evaluation unit 70 may, when the distribution of the pixel breast tissue ratio in a predetermined region is defined as the regional breast tissue ratio, represent the distribution of the pixel breast tissue ratio in a predetermined region for each radiographic image using a histogram with the pixel breast tissue ratio on the horizontal axis and the frequency (number of pixels with the pixel breast tissue ratio belonging to each class of pixel breast tissue ratio) on the vertical axis, thereby defining the regional breast tissue ratio. The evaluation result of "good stretchability" or "poor stretchability" may be output based on the relationship between the degree of agreement obtained by comparing the two histograms and a threshold for evaluating the degree of agreement.

[0050] Preferably, the two radiographic images compared by the extensibility evaluation unit 70 are radiographic images taken of the left and right breasts of the same subject, or radiographic images taken of either the left or right breast of the same subject at different time points.

[0051] The following describes an example in which the stretchability evaluation unit 70 evaluates stretchability by comparing the regional breast tissue percentages calculated from radiographic images of the left and right breasts, respectively, when the maximum value of the pixel breast tissue percentage in a predetermined region is defined as the regional breast tissue percentage.

[0052] If the regional breast tissue ratio calculation unit 60 calculates that the maximum value of the pixel breast tissue ratio in a predetermined region of the right breast image is 30% and the maximum value of the pixel breast tissue ratio in a predetermined region of the left breast image is 60%, the stretchability evaluation unit 70 calculates the difference in regional breast tissue ratio between the right breast image and the left breast image as "30%". Here, if the stretchability evaluation threshold is set to "10%" in advance and it is set that "if the difference in regional breast tissue ratio is greater than or equal to the stretchability evaluation threshold, the radiographic image with the higher regional breast tissue ratio is evaluated as having poor stretchability", the stretchability evaluation unit 70 outputs an evaluation result of "the left breast image has poor stretchability (poor stretchability)". The outputted stretchability evaluation result is transmitted to the display control unit 80.

[0053] The stretchability evaluation threshold can be set to any value. Furthermore, if the regional breast tissue ratio calculation unit 60 calculates multiple types of regional breast tissue ratios for each radiographic image, such as "the maximum value of the pixel breast tissue ratio in a predetermined region" and "the volume ratio of breast tissue in a predetermined region," the user may pre-set multiple stretchability evaluation thresholds for each type of regional breast tissue ratio. Alternatively, the multiple stretchability evaluation thresholds may be automatically set by the stretchability evaluation unit 70 based on multiple types of stretchability evaluation thresholds used in the past.

[0054] The display control unit 80 generates a display image for displaying the results of the extensibility evaluation and controls the display of the display image on the display 13 connected to the radiographic image processing device 10. A specific example of the display image will be explained with reference to Figure 10. In the specific example of the display image 110 shown in Figure 10, the regional glandular tissue percentage within the range of the marker 111a indicating a predetermined area of ​​the right breast image 111 is calculated to be 30%, and the regional glandular tissue percentage within the range of the marker 112a indicating a predetermined area of ​​the left breast image 112 is calculated to be 60%, and these are displayed in the regional glandular tissue percentage display field 113. The example shown in Figure 10 is an example in which "if the extensibility evaluation threshold is 10% or more, the radiographic image with the higher regional glandular tissue percentage is evaluated as having poor extensibility," and the evaluation result display field 114 displays "Extensibility of left breast: poor."

[0055] Here, the evaluation result is output for the extensibility of the "specified region in the radiographic image" of the breast with a high proportion of regional breast tissue. However, the display of the evaluation result may also be the evaluation result of the extensibility of the "breast visible in the radiographic image" of the breast with a high proportion of regional breast tissue. Furthermore, it is preferable that the radiographic image includes information indicating which breast was photographed (left or right), as well as information indicating the time of shooting, such as the date and time of shooting, as described later for past and present images.

[0056] Markers 111a and 112a indicating predetermined regions may or may not be displayed on the display image 110. Furthermore, a threshold for evaluating extensibility may be displayed. By displaying a display image 110 as shown in Figure 10, when a left breast image is taken as a pre-shot image after a right breast image has been taken, it is possible to determine whether the extensibility of the mammary glands contained in the breast shown in the pre-shot image is appropriate. It can be evaluated. Furthermore, if the breast tissue shown in the pre-shot image has poor elasticity, the system can prompt the user to adjust their positioning. Also, if the pre-shot image shows good elasticity, the system can prompt the user to retake the previously taken radiographic image.

[0057] Furthermore, the display control unit 80 may display the results of the extensibility evaluation by changing the display mode of a marker indicating a predetermined area, or by displaying a warning. If the evaluation result "Poor extensibility of the left breast image" is output, for example, as illustrated in Figure 11, the display mode, such as the color or pattern of the marker 111a indicating a predetermined area in the right breast image 111 and the marker 112b indicating a predetermined area in the left breast image 112 may be made different from each other depending on the proportion of breast tissue in the area, and the legend 115 may be displayed.

[0058] Furthermore, if the evaluation result is "poor extensibility of the left breast image," a display image may be generated that shows markers 111a and 112a indicating predetermined areas, as illustrated in Figure 12, and also displays a warning marker 116 on the left breast image. As illustrated in Figures 11 and 12, the visibility of the evaluation result can be improved by displaying the extensibility evaluation result without using text information. Note that the method of displaying the extensibility evaluation result is not limited to this. For example, if "poor extensibility" is detected, a warning frame may be placed around the radiographic image with a high proportion of regional breast tissue.

[0059] Radiographic images of the left and right breasts of the same subject are usually symmetrical and very similar. Therefore, in mammography of the same subject under the same imaging conditions such as tube current value, irradiation time, and compression force on the breast, if there is a significant difference in the elasticity of the breast tissue when comparing the right and left breast images, it is considered that poor positioning of either the right or left breast has occurred. Accordingly, by evaluating elasticity using the right and left breast images, if either the right or left breast image is a pre-shot image, the user can be prompted to correct the positioning or retake the image.

[0060] Furthermore, breast size and the amount of mammary gland tissue vary greatly from person to person. Therefore, even if a radiographic image showing sufficiently stretched mammary gland tissue can be obtained for a particular subject, it is difficult to accurately evaluate the extensibility of the mammary gland by comparing the radiographic image of this subject's breast with radiographic images of a different subject's breast. With the above configuration, by comparing radiographic images of the same subject's breast, it is possible to evaluate the extensibility of the mammary gland while minimizing the influence of individual differences in breast size among subjects.

[0061] Next, we will describe an example in which the stretchability evaluation unit 70 evaluates stretchability by comparing the regional breast tissue percentages calculated from radiographic images of either the left or right breast of the subject taken at different time points, where the maximum value of the pixel breast tissue percentage in a predetermined region is defined as the regional breast tissue percentage.

[0062] For example, if the regional breast tissue ratio calculation unit 60 calculates that the maximum value of the pixel breast tissue ratio in a predetermined region of a previously taken radiographic image (past image) is 40% as the regional breast tissue ratio, and the maximum value of the pixel breast tissue ratio in a predetermined region of a pre-shot radiographic image (current image) taken in the current examination is 60% as the regional breast tissue ratio, then the extensibility evaluation unit 70 calculates the difference in regional breast tissue ratio between the past image and the current image as 20%. Here, if the extensibility evaluation threshold is set to "10%" in advance, and it is set that "if the difference in regional breast tissue ratio is greater than or equal to the extensibility evaluation threshold, it will be evaluated as poor extensibility," then the extensibility evaluation unit 70 will output an evaluation result of "the current image has poor extensibility (poor extensibility)."

[0063] Figure 13 shows an example of a display image 120 generated by the display control unit 80 when comparing radiographic images taken at different points in time. In the past image 121, which was taken on March 28, 2021, and is labeled with the date 123, the percentage of localized breast tissue within the range of the marker 121a indicating a predetermined area of ​​the right breast is calculated to be 40%, and in the current image 122, which was taken on March 28, 2022, and is labeled with the date 123, the percentage of localized breast tissue within the range of the marker 122a indicating a predetermined area of ​​the right breast is calculated to be 60%, and these are displayed in the localized breast tissue percentage display field 113. In addition, the evaluation result display field 114 displays "Current image extensibility: poor". This is how it is displayed in an example where the extensibility evaluation unit 70 outputs an evaluation result of poor extensibility when the extensibility evaluation threshold is 10% or more. Note that Figure 13 shows an example of comparing a right breast image taken in the past with a right breast image taken in the current examination.

[0064] In normal, non-pregnant adult women, the proportion of breast tissue generally does not change drastically over a period of 1-2 years. Therefore, if the time difference between past and present examinations is approximately 1-2 years, the proportion of breast tissue calculated from the radiographic images taken at the past and present points in time can be considered similar. Consequently, in mammograms of the same subject with the same imaging conditions (tube current value, irradiation time, and compression force on the breast), if there is a significant difference in breast tissue extensibility between past and present images, it is likely that poor breast positioning occurred in either the past or present imaging. Therefore, by comparing the regional breast tissue proportion between past and present images, the appropriateness of the positioning can be evaluated. Furthermore, if the current image is a pre-shot image, the user can be prompted to correct the positioning.

[0065] Furthermore, the stretchability evaluation unit 70 may compare radiographic images of the opposite breast taken at different times, such as comparing a radiographic image of the left breast taken in the past with a radiographic image of the right breast taken in the current examination, because these radiographic images are similar.

[0066] With the above configuration, by comparing the regional breast tissue ratio, which considers the amount of breast tissue contained in the thickness direction of the breast sandwiched between the compression plate and the imaging table for each predetermined region, it is possible to accurately and easily determine whether an ideal radiographic image has been taken. Furthermore, if the distensibility evaluation result is poor when comparing two radiographic images, one of which is a pre-shot image, it means that the positioning of the image with a higher regional breast tissue ratio is inappropriate. By displaying the distensibility evaluation result, the user can be prompted to correct the positioning. In addition to pre-shot images, it is also possible to check and review whether the positioning of radiographic images taken in the past by the radiologist was appropriate.

[0067] The following describes an example of a method for determining a predetermined region in this embodiment. As mentioned above, the region determination unit 50 can designate the breast region 103, the mammary gland concentration region, or a local region as the predetermined region.

[0068] In the first example where a breast tissue concentration region is defined as a predetermined region, the region determination unit 50 may define a region containing pixels in which the pixel breast tissue ratio is equal to or greater than the first region threshold as the breast tissue concentration region. The breast region of a radiographic image contains low-attenuation regions with a lot of adipose tissue, high-attenuation regions with a lot of glandular tissue, and regions that are between the low-attenuation and high-attenuation regions (intermediate-attenuation regions). By setting the breast tissue concentration region using the first region threshold, the regional breast tissue ratio can be calculated after excluding the low-attenuation regions, which do not significantly affect the evaluation of breast tissue extensibility, from the predetermined region. In other words, by making only the regions that particularly affect the evaluation of breast tissue extensibility (high-attenuation regions and intermediate-attenuation regions) the target of calculating the regional breast tissue ratio, the influence of noisy regions can be reduced and breast tissue extensibility can be evaluated more effectively than by comparing radiographic images on a pixel-by-pixel basis or across the entire radiographic image. The first region threshold can be set arbitrarily.

[0069] In the second example where the breast tissue concentration region is designated as a predetermined region, the region determination unit 50 may designate a region in the radiographic image that has pixels with a breast tissue volume equal to or greater than the second region threshold as the breast tissue concentration region. The breast tissue volume is an estimated value of the volume of breast tissue, which is the product of the pixel breast tissue ratio and the thickness of the breast corresponding to the distance between the compression plate and the imaging table. With the above configuration, the breast tissue volume is calculated for each pixel using the pixel breast tissue ratio, and the predetermined region is determined based on the amount of breast tissue volume. This makes it possible to designate a region with a breast tissue concentration region as the target for calculating the region breast tissue ratio, which is considered to greatly influence the evaluation of the extensibility of the breast tissue from the perspective of the volume of breast tissue. It is preferable to designate the breast tissue concentration region as a predetermined region because it is difficult to distinguish between normal breast tissue and lesions in the breast tissue concentration region, and therefore requires careful observation.

[0070] When a local area is defined as a predetermined area, the area determination unit 50 may define the predetermined area as the area within a specific distance range from the pixel with the highest pixel breast tissue ratio in the radiographic image. Specifically, as shown in Figure 14, the area within a specific distance 132 (indicated by an arrow) from the pixel 131 with the highest pixel breast tissue ratio in the radiographic image 130 is determined as the predetermined area 133 (indicated by a dot pattern). The specific distance can be arbitrarily set, such as within 10 pixels above and below the pixel 131 with the highest pixel breast tissue ratio. Furthermore, the specific distance range that constitutes the predetermined area is determined, for example, using the coordinate information of the pixel with the highest pixel breast tissue ratio and the specific distance. By setting a small specific distance, it is possible to focus on the area around the pixel with the highest pixel breast tissue ratio, which is a local area that is likely to have poor extensibility, calculate the regional breast tissue ratio, and use it to evaluate extensibility.

[0071] As described above, by designating a specific area as a concentrated breast tissue area or a localized area, the influence of noisy areas can be reduced, allowing for a more accurate evaluation of breast tissue extensibility. Furthermore, it is possible to compare only a portion of the two radiographic images where the breast tissue is compacted. Evaluating extensibility by comparing fine details of radiographic images, which are likely to have poor extensibility for the user, is difficult even for experienced radiographers. In particular, in routine screenings where both accuracy and speed are required, it is difficult to carefully observe fine details of radiographic images. With the above configuration, even in situations where positioning is considered to be reasonably good, the extensibility of breast tissue can be easily evaluated for localized areas of radiographic images. As a result, examinations can be performed smoothly, and the burden on the patient can be reduced.

[0072] The following describes an example of how to calculate the regional mammary gland ratio and how to output the results of the stretchability evaluation in this embodiment. In the first example of evaluating stretchability based on the difference in the regional mammary gland ratio, the statistical quantity of the pixel mammary gland ratio in a predetermined region, such as the breast region, the mammary gland concentration region, or the local region, is defined as the regional mammary gland ratio. The statistical quantity may be the maximum value (see the example shown in Figures 10-13), the mode, or the mean, but is not limited to these. In this case, the stretchability evaluation unit 70 outputs the stretchability evaluation result using the relationship between the difference in the regional mammary gland ratio and the threshold for stretchability evaluation. With the above configuration, the regional mammary gland ratio can be calculated in a way that reduces the computational burden on the processor.

[0073] In the second example of evaluating extensibility based on the difference in regional glandular ratio, the regional glandular ratio is defined as the volume ratio of glandular tissue in a predetermined region. In this case, the predetermined region is preferably a glandular tissue concentration region. In this example, the regional glandular ratio calculation unit 60 may calculate the volume of the breast in the predetermined region (a) and the volume of glandular tissue in the predetermined region (b), and calculate the volume ratio of glandular tissue as (b) / (a), which may be defined as the regional glandular ratio. For example, the volume of the breast in the predetermined region (a) 61 is calculated as the product of the predetermined region (unit: pixels) represented on a plane formed in the x and y directions, and the distance (unit: millimeters) between the imaging table 24 and the compression plate 25, which corresponds to the thickness of the breast, represented in the z direction, as shown in Figure 15. The volume of mammary gland (b)62 is calculated as shown in Figure 16 by determining the volume of mammary gland at each pixel by multiplying the distance (in millimeters) between the imaging table 24 and the compression plate 25, and the pixel mammary gland ratio, as shown in the z-axis direction. Furthermore, the volume of mammary gland is calculated for each pixel contained in a predetermined region represented on a plane formed in the x-axis and y-axis directions. If the predetermined region is defined as a mammary gland concentration region, the volume ratio of mammary gland (b) / (a) is the volume ratio of mammary gland in the mammary gland concentration region.

[0074] In this case, the extensibility evaluation unit 70 outputs an evaluation result of extensibility using the relationship between the difference in the volume ratio of mammary gland calculated for each radiographic image and the extensibility evaluation threshold. With the above configuration, it is possible to compare the volume of mammary gland in a predetermined three-dimensional region, taking into account the direction in which radiation is detected and the thickness direction of the breast. Furthermore, if the predetermined region is a mammary gland concentration region, by determining the volume ratio of mammary gland in the mammary gland concentration region, it is possible to perform the evaluation after eliminating areas that would be noise in the evaluation of extensibility.

[0075] When evaluating extensibility using the degree of agreement in the distribution of pixel breast tissue percentages, the regional breast tissue percentage calculation unit 60 generates a histogram of the distribution of pixel breast tissue percentages in a predetermined region of each radiographic image and uses it as the regional breast tissue percentage. In this histogram, the horizontal axis represents the pixel breast tissue percentage, and the vertical axis represents the frequency (the number of pixels with a pixel breast tissue percentage belonging to each class). The class width on the horizontal axis may be set arbitrarily.

[0076] In this case, the extensibility evaluation unit 70 calculates the degree of agreement between the histograms of each radiographic image using a known method for comparing histograms, such as the cross-eyed method or the use of Euclidean distance. Then, it outputs an evaluation result based on the relationship between the degree of agreement and the threshold for evaluating the degree of agreement.

[0077] When the stretchability evaluation unit 70 calculates the degree of agreement using the cross-eyed method, it normalizes the total number of pixels included in a predetermined region of the radiographic image so that the sum of the frequencies is 1, and substitutes this into the function used in the cross-eyed method to calculate the degree of agreement. In this case, the degree of agreement is calculated as an output value, and the closer it is to 1, the higher the degree of agreement. For example, the stretchability evaluation unit 70 sets the threshold for the degree of agreement evaluation to 0.9, and outputs an evaluation result of "good stretchability" if the degree of agreement is equal to or greater than the threshold for the degree of agreement evaluation, and outputs an evaluation result of "poor stretchability" if the degree of agreement is less than the threshold for the degree of agreement evaluation.

[0078] Furthermore, when the extensibility evaluation unit 70 calculates the degree of agreement using Euclidean distance, each histogram is represented as coordinates in a multidimensional space, and the distance between coordinates is calculated as the degree of agreement. In this case, the closer the distance between coordinates is to 0, the higher the degree of agreement. For example, the extensibility evaluation unit 70 sets the threshold for the degree of agreement evaluation to 0.1, and outputs an evaluation result of "good extensibility" if the degree of agreement is less than the threshold for the degree of agreement evaluation, and outputs an evaluation result of "poor extensibility" if the degree of agreement is equal to or greater than the threshold for the degree of agreement evaluation. With the above configuration, even when the thickness of the breast cannot be accurately calculated, the degree of agreement can be calculated using the pixel glandular tissue ratio of a predetermined area in each radiographic image.

[0079] The stretchability evaluation unit 70 may output multiple evaluation results, such as: a first evaluation result when the average value of the breast tissue ratio of pixels in a predetermined region is used as the regional breast tissue ratio; a second evaluation result when the volume ratio of breast tissue in a predetermined region is used as the regional breast tissue ratio; and a third evaluation result when the distribution of the breast tissue ratio of pixels in a predetermined region is used as the regional breast tissue ratio and the degree of agreement is calculated using the cross-eyed method. The final evaluation result may be calculated by combining these multiple evaluation results. By calculating multiple evaluation results, the accuracy of the stretchability evaluation can be improved.

[0080] In this embodiment, the display control unit 80 may further include a sub-information display control unit 140, as shown in Figure 17. The sub-information display control unit 140 further includes an operation instruction generation unit 150 and a breast structure classification unit 160. The sub-information display control unit 140 may also include either the operation instruction generation unit 150 or the breast structure classification unit 160.

[0081] The operation instruction generation unit 150 generates an operation instruction display according to the operation instruction conditions, using the regional breast tissue percentage and evaluation results. For example, as shown in Figure 18, if the regional breast tissue percentage of the right breast image 111 within the range of marker 111a, which indicates a predetermined region that is a breast tissue concentration area, is 30%, and the regional breast tissue percentage of the left breast image 112 within the range of marker 112a, which indicates a predetermined region that is a breast tissue concentration area, is 60%, and the operation instruction condition is "If extensibility is poor, instruct the camera to correct the imaging conditions for the radiographic image with a high regional breast tissue percentage," then the operation instruction display 152 "Please correct the positioning of the left breast" is generated and transmitted to the display control unit 80 and displayed on the display image 151. Alternatively, the operation instruction generation unit 150 may also generate an operation instruction display 152 that instructs the compression force of the compression plate 25 as an instruction to correct the imaging conditions, such as "Left breast compression force: A Newton." With the above configuration, the user can be prompted to perform operations such as positioning or compression force and retake the image.

[0082] The breast structure classification unit 160 classifies the breast structure of the breast as seen in the radiographic image using the pixel glandular percentage. The breast structure classification unit 160 calculates the breast structure classification area glandular percentage in the breast region using the pixel glandular percentage, and classifies the breast structure based on the relationship between the breast structure classification area glandular percentage and the breast structure classification threshold. For example, using the pixel glandular percentage of the right breast image 111, the breast structure classification area glandular percentage is calculated to be "70%". Similarly, using the pixel glandular percentage of the left breast image 112, the breast structure classification area glandular percentage is calculated to be "70%". The region on the radiographic image from which the breast structure classification area glandular percentage is calculated may be the breast region, a region pre-set by the user, or a region determined by setting a region threshold for determining the breast structure classification area. The breast structure classification unit 160 determines the region on the radiographic image from which the breast structure classification area glandular percentage is calculated.

[0083] As for setting thresholds for breast composition classification, for example, if the regional glandular percentage for breast composition classification is less than 10%, it can be classified as "fatty," if it is between 10% and less than 50%, it can be classified as "scattered glandular," if it is between 50% and less than 80%, it can be classified as "heterogeneously dense," and if it is 80% or more, it can be classified as "extremely dense." Also, if the regional glandular percentage for breast composition classification is 5 A breast density of 0% or more may be classified as "dense breast," while a density of less than 50% may be classified as "not dense breast." However, this is not the only method for setting thresholds for breast composition classification.

[0084] If it is pre-set that a breast is classified as "dense breast" when the threshold for breast composition classification is 50% or higher, and the breast composition classification unit 160 calculates that the breast tissue percentage in the breast composition classification area of ​​the right breast image 111 is 70% and the breast tissue percentage in the breast composition classification area of ​​the left breast image 112 is 70%, then the right breast image 111 and the left breast image 112 are classified as "dense breast".

[0085] In this embodiment, breast structures classified by the relationship between the breast tissue classification area and the breast tissue classification threshold are classified into "fatty," "scattered glandular tissue," "heterogeneously dense," and "extremely dense" depending on the amount of glandular tissue. In "fatty" breast tissue, most of the glandular tissue is replaced by fat, making lesion detection easy. In "scattered glandular tissue" breast tissue, the glandular tissue is replaced by fatty tissue, and only cord-like glands are visible, making lesion detection easy. In "heterogeneously dense" breast tissue, fat is mixed in with the glandular parenchyma, making lesion detection difficult. In "extremely dense" breast tissue, there is almost no fat mixed in with the glandular parenchyma, making lesion detection difficult. Of these, "heterogeneously dense" and "extremely dense" breasts are called dense breasts and require careful consideration in the differentiation of lesions.

[0086] The breast structure classification unit 160 transmits information on the results of classifying the breast structure of each radiographic image to the display control unit 80. The display control unit 80 may display the results of classifying the breast structure of the right breast image 111 and the left breast image 112 as a breast structure display field 162, as shown in Figure 19. In the example shown in Figure 19, the right breast image 111 and the left breast image 112 are displayed as "high-density breast". In addition, the evaluation result for extensibility is displayed as "Extensibility: Good".

[0087] As shown in the example in Figure 19, when the radiographic images being compared involve dense breasts, the regional breast tissue percentage calculated from each image will be high, potentially leading to an inaccurate assessment of stretchability. Therefore, the above configuration allows for the display of breast composition, prompting the user to consider whether the assessment of stretchability is accurate.

[0088] Furthermore, for dense breasts, increasing the compression pressure during a re-imaging can cause unnecessary discomfort to the patient. Additionally, it can encourage users to consider other modalities (ultrasound imaging devices) for examination.

[0089] Furthermore, in radiographic images of such dense breasts, because the breast region contains a large amount of glandular tissue, it is difficult to observe lesions such as calcifications or lumps, which appear as high-attenuation areas and appear white, similar to glandular tissue. For this reason, in order to inform the user that there is a high possibility that the assessment of extensibility has not been accurately performed, and that there is a high possibility that it is not possible to accurately distinguish whether or not a lesion is present in the radiographic image, the breast composition display area 162 may be displayed only when the photographed breast is dense.

[0090] The flow of how the radiographic image processing device 10 displays the results of the stretchability evaluation using multiple radiographic images of the same subject will be explained using the flowchart in Figure 20. First, the image acquisition unit 30 acquires two radiographic images from the modality 11 or the database 12 (step ST101). Next, the pixel breast tissue ratio calculation unit 40 calculates the pixel breast tissue ratio for each pixel of each radiographic image (step ST102). Next, the region determination unit 50 reads pre-set conditions and determines a predetermined region (step ST103), and the region breast tissue ratio calculation unit 60 calculates the region breast tissue ratio in the predetermined region using the pixel breast tissue ratio (step ST104). Then, the stretchability evaluation unit 70 compares the region breast tissue ratio between the two radiographic images of the same subject and outputs the evaluation result (step ST105). Finally, the display control unit 80 generates a display image to display the evaluation result (step ST106). The generated display image is displayed on the display 13.

[0091] In the above embodiment, the hardware structure of the processing unit that executes various processes such as the image acquisition unit 30, the pixel breast tissue ratio calculation unit 40, the region determination unit 50, the region breast tissue ratio calculation unit 60, the extensibility evaluation unit 70, the display control unit 80, and the sub-information display control unit 140 is the following type of processor. The types of processors include a CPU (Central Processing Unit), which is a general-purpose processor that executes software (programs) and functions as various processing units; a Programmable Logic Device (PLD), which is a processor whose circuit configuration can be changed after manufacturing, such as an FPGA (Field Programmable Gate Array); and a dedicated electrical circuit, which is a processor with a circuit configuration specifically designed to execute various processes.

[0092] A single processing unit may be composed of one of these various processors, or it may be composed of a combination of two or more processors of the same or different types (for example, multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, multiple processing units may be composed of a single processor. Examples of composing multiple processing units with a single processor include, firstly, a configuration where one or more CPUs and software are combined to form a single processor, and this processor functions as multiple processing units, as is typical of computers such as clients and servers. Secondly, a configuration using a processor that realizes the functions of the entire system, including multiple processing units, on a single IC (Integrated Circuit) chip, as is typical of a System on a Chip (SoC). Thus, various processing units are configured, in terms of hardware structure, using one or more of the above-mentioned various processors.

[0093] Furthermore, the hardware structure of these various processors is, more specifically, an electrical circuit formed by combining circuit elements such as semiconductor devices. The hardware structure of the memory unit is a storage device such as an HDD (hard disk drive) or SSD (solid state drive). [Explanation of Symbols]

[0094] 10. Radiation image processing device 11 Modalities 12 Databases 13 displays 20 Mammography equipment 21 Post 22 Moving parts 23 Radiation-generating section 23a Radiation 24 Shooting platform 24a Radiography Department 25 Compression plate 26 Lifting section 27a, 27b grip part 28 Floor surface 30 Image acquisition unit 40-pixel breast tissue ratio calculation unit 50 Area determination section 60-region breast tissue percentage calculation section 61. The volume of the breast 62. Volume of mammary glands 70 Extensibility evaluation section 80 Display Control Unit 100, 130 radiographic images 101 Direct area 102 Pectoralis major region 103 breast area 104 Low absorption region 105 High absorption region Images for displaying 110, 120, 151, and 161. 111 Right breast image 111a, 112a, 112b, 121a, 122a markers 112 Left breast image 113 Area breast tissue percentage display column 114 Evaluation Result Display Section 115 Legend 116 Warning Marker 121 Past Images 122 Current Image 123 Date of shooting 131, 200a, 200b, 200c, 200d, 200e pixels 132 Specific distance 133 Predetermined area 140 Sub-information display control unit 150 Operation instruction generation section 152 Operation instruction display 160 Breast composition classification section 162 Breast composition display field 201 Adipose tissue 202 Mammary gland tissue 203a, 203b, 203c, 203d, 203e Pixel breast tissue ratio B Breast Ob subject T. Breast thickness

Claims

1. Equipped with a processor, The aforementioned processor, Two radiographic images were obtained of the same subject's breast. Based on the aforementioned radiographic images, the pixel-by-pixel breast tissue ratio, which is the proportion of breast tissue in each pixel of each radiographic image, is calculated. Based on the aforementioned pixel breast tissue ratio, the regional breast tissue ratio in a predetermined area of ​​each of the aforementioned radiographic images is calculated. The extensibility of the breast tissue is evaluated by comparing the two breast tissue percentages in the region calculated using the two aforementioned radiographic images. A radiation image processing device that performs control to display the results of the extensibility evaluation.

2. The aforementioned processor, The radiographic image processing apparatus according to claim 1, wherein the region having the pixels in the radiographic image in which the pixel breast tissue ratio is equal to or greater than a first region threshold is defined as the predetermined region.

3. The aforementioned processor, Using the aforementioned pixel breast tissue ratio, the amount of breast tissue is calculated for each pixel. The radiographic image processing apparatus according to claim 1, wherein the region having the pixels in the radiographic image in which the amount of mammary gland tissue is equal to or greater than the second region threshold is defined as the predetermined region.

4. The aforementioned processor, The radiation image processing apparatus according to claim 1, wherein the predetermined region is defined as a region within a specific distance range from the pixel in the radiation image where the pixel proportion of mammary gland is maximum.

5. The aforementioned processor, A radiographic image processing apparatus according to any one of claims 1 to 4, which evaluates the extensibility based on the difference between the two regions of breast tissue calculated using the radiographic images.

6. The radiation image processing apparatus according to claim 5, wherein the region breast tissue ratio is a statistical quantity of the pixel breast tissue ratio in the predetermined region.

7. The radiographic image processing apparatus according to claim 5, wherein the regional breast tissue ratio is the volume ratio of breast tissue in the predetermined region.

8. The aforementioned processor, The extensibility is evaluated using the degree of agreement calculated based on the two regional breast tissue percentages calculated using the aforementioned radiographic images. The radiation image processing apparatus according to any one of claims 1 to 4, wherein the region breast tissue ratio is the distribution of the pixel breast tissue ratio in the predetermined region.

9. The radiographic image processing apparatus according to any one of claims 1 to 8, wherein the two radiographic images are radiographic images taken of the left and right breasts of the subject, respectively.

10. The radiographic image processing apparatus according to any one of claims 1 to 8, wherein the two radiographic images are radiographic images taken of the breast of the same subject at different time points.

11. The aforementioned processor, Based on the results of the extensibility evaluation and the percentage of breast tissue in the region of the radiographic image selected from the multiple radiographic images, an operation instruction display is generated. The radiation image processing apparatus according to any one of claims 1 to 10, which controls the display of the operation instruction display.

12. The aforementioned processor, Based on the aforementioned pixel ratio of mammary gland tissue, the breast structure of each of the aforementioned radiographic images is classified, A radiation image processing apparatus according to any one of claims 1 to 11, which controls the display of the breast structure.

13. The steps include obtaining two radiographic images of the same subject's breast, The steps include: calculating the pixel-by-pixel breast tissue ratio, which is the proportion of breast tissue in each pixel of the aforementioned radiographic image, based on the aforementioned radiographic image; The steps include: calculating the regional breast tissue ratio in a predetermined region of each of the aforementioned radiographic images based on the aforementioned pixel breast tissue ratio; A step of evaluating the extensibility of the breast tissue by comparing the two breast tissue percentages in the region calculated using the two aforementioned radiographic images, A method for operating a radiation image processing device, comprising the step of performing control to display the results of the extensibility evaluation.

Images