Image processing device, image processing method, and program

Abstract

To provide an image processing device, an image processing method, and a program capable of accurately determining a kind of shape of a calcified image.SOLUTION: An image processing device includes at least one processor. The processor executes: calcified image detection processing for detecting a calcified image on the basis of a plurality of tomographic images obtained from a series of a plurality of projection images obtained from tomosynthesis imaging of the breast; region-of-interest image group generation processing for generating a region-of-interest image group by cutting out a region including the calcified image detected by the calcified image detection processing from each of the plurality of projection images as a region-of-interest image; dispersion value calculation processing for calculating a dispersion value of a feature amount of each of the region-of-interest images constituting the region-of-interest image group; and kind-of-shape determination processing for determining the kind of shape of the calcified image on the basis of the dispersion value calculated by the dispersion value calculation processing.SELECTED DRAWING: Figure 5

Description

Technical Field

[0001] The present disclosure relates to an image processing apparatus, an image processing method, and a program.

Background Art

[0002] A technique for identifying tissue that may have calcification in a breast using a radiation image obtained by irradiating the breast with radiation is known. For example, Patent Document 1 discloses specifying a pixel region that may be calcified from a radiation image or the like, grouping the specified set of regions, and displaying the regions by color or luminance according to the number of pixel regions belonging to the group. This makes it possible to intuitively grasp the dense state of microcalcified tissue.

[0003] Furthermore, tomosynthesis imaging is known, in which a series of a plurality of projection images are acquired by irradiating a breast with radiation from a plurality of angles. By reconstructing a plurality of projection images obtained by tomosynthesis imaging, a plurality of tomographic images with little overlap of mammary glands can be obtained. Furthermore, it is known that by combining a plurality of tomographic images, one combined two-dimensional image with little overlap of mammary glands can be generated. In addition, Patent Document 2 discloses that instead of using a plurality of tomographic images, a projection image obtained at a radiation irradiation angle of around 0 degrees is input to a trained model, thereby generating a two-dimensional image corresponding to the combined two-dimensional image.

Prior Art Literature

Patent Literature

[0004]

Patent Document 1

Patent Document 2

Summary of Invention

Problem to be Solved by the Invention

[0005] In imaging diagnostics of breast calcifications, the shape of calcifications as seen in tomographic images and composite 2D images is important information. However, in tomographic images, calcifications are obscured by noise, reducing their visibility. Furthermore, composite 2D images generated from multiple tomographic images do not accurately represent the shape of calcifications.

[0006] Patent Document 1 considers the distribution of calcification patterns as useful information for imaging diagnosis of calcification, but does not consider the shape of the calcification patterns. Similarly, Patent Document 2 also does not consider the shape of the calcification patterns as useful information for imaging diagnosis of calcification. In imaging diagnosis of calcification, it is desirable to accurately determine the type of shape of the calcification patterns in order to determine whether the calcification patterns are malignant or benign.

[0007] The present invention aims to provide an image processing device, an image processing method, and a program that can accurately determine the type of shape of a calcification image. [Means for solving the problem]

[0008] To achieve the above objective, the image processing apparatus of the present disclosure comprises at least one processor, which performs: a calcification image detection process for detecting calcification images based on a plurality of tomographic images obtained from a series of plurality of projection images obtained by tomosynthesis imaging of the breast; a region of interest image group generation process for generating a group of region of interest images by cutting out a region containing the calcification image detected by the calcification image detection process from each of the plurality of projection images as a region of interest image; a variance value calculation process for calculating the variance value of each feature quantity of the region of interest images constituting the group of region of interest images; and a shape type determination process for determining the shape type of the calcification image based on the variance value calculated by the variance value calculation process.

[0009] When the processor detects multiple calcification images during the calcification image detection process, it is preferable that the processor generates a separate set of region of interest images for each of the multiple calcification images during the region of interest image generation process.

[0010] In the calcification detection process, the processor preferably detects only calcification images whose signal value is below a certain value.

[0011] In the shape type determination process, the processor preferably determines the shape of the calcification image based on the relationship between the variance value of a predetermined feature quantity and the shape type.

[0012] The feature is preferably the variance of the pixel values ​​contained in a single image of the region of interest.

[0013] The feature is preferably the variance of the pixel values ​​in one region of interest image relative to the average value of the pixel values ​​in the breast region in one projection image.

[0014] The feature is preferably the number of pixels in a single region of interest image that have a pixel value equal to or greater than a threshold.

[0015] Preferably, the processor further performs a display process to display the shape type determination result obtained from the shape type determination process on the display unit.

[0016] In the display processing, the processor preferably highlights calcification images having a specific shape based on the shape type determination result.

[0017] The image processing method of the present disclosure includes: a calcification image detection step in which calcification images are detected based on a plurality of tomographic images obtained from a series of plurality of projection images obtained by tomosynthesis imaging of the breast; a region of interest image group generation step in which a group of region of interest images is generated by cutting out a region corresponding to the calcification image detected in the calcification image detection step from each of the plurality of projection images as a region of interest image; a variance value calculation step in which the variance value of each feature quantity of the region of interest images constituting the group of region of interest images is calculated; and a shape type determination step in which the shape type of the calcification image is determined based on the variance value calculated in the variance value calculation step.

[0018] The program of the present disclosure causes a computer to execute: calcified image detection processing for detecting a calcified image based on a plurality of tomographic images obtained from a series of a plurality of projection images acquired by tomosynthesis imaging of a breast; region-of-interest image group generation processing for generating a region-of-interest image group by cutting out, from each of the plurality of projection images, a region corresponding to the calcified image detected by the calcified image detection processing as a region-of-interest image; variance value calculation processing for calculating a variance value of feature amounts of each region-of-interest image constituting the region-of-interest image group; and shape type determination processing for determining a shape type of the calcified image based on the variance value calculated by the variance value calculation processing.

Effects of the Invention

[0019] According to the technology of the present disclosure, it is possible to provide an image processing apparatus, an image processing method, and a program capable of accurately determining the shape type of a calcified image.

Brief Description of Drawings

[0020] [Figure 1] It is a diagram showing an example of the overall configuration of a radiographic imaging system. [Figure 2] It is a diagram explaining an example of tomosynthesis imaging. [Figure 3] It is a block diagram showing an example of the configuration of an image processing apparatus. [Figure 4] It is a block diagram showing an example of functions implemented by a control unit of the image processing apparatus. [Figure 5] It is a diagram schematically showing a flow of processing by the image processing apparatus. [Figure 6] It is a diagram conceptually showing an example of region-of-interest image group generation processing. [Figure 7] An example of variance value calculation processing is conceptually shown. [Figure 8] It is a diagram showing an example of display processing. [Figure 9] It is a flowchart showing a flow of a series of processing by the image processing apparatus. [Figure 10] It is a block diagram showing functions implemented by a control unit of an image processing apparatus according to a modified example. [Figure 11] This is a block diagram that schematically shows the processing flow by the image processing device related to the modified example. [Modes for carrying out the invention]

[0021] Embodiments of this disclosure will be described in detail below with reference to the drawings.

[0022] Figure 1 shows an example of the overall configuration of the radiographic imaging system 2 according to this embodiment. The radiographic imaging system 2 comprises a mammography device 10, a console 12, a PACS (Picture Archiving and Communication Systems) 14, and an image processing device 16. The console 12, PACS 14, and image processing device 16 are connected via a network 17 by wired or wireless communication.

[0023] Figure 1 shows an example of the appearance of the mammography device 10. Note that Figure 1 shows an example of the appearance of the mammography device 10 when viewed from the left side of the patient.

[0024] The mammography device 10 operates in accordance with the control of the console 12 and is a radiography device that acquires radiographic images of the breast M of the subject by irradiating the breast M with radiation R (e.g., X-rays) from the radiation source 29.

[0025] The mammography device 10 has a function to perform normal imaging, in which the radiation source 29 is positioned at an irradiation position along the normal direction of the detection surface 20A of the radiation detector 20, and a function to perform tomosynthesis imaging, in which the radiation source 29 is moved to each of multiple irradiation positions for imaging.

[0026] As shown in Figure 1, the mammography apparatus 10 comprises an imaging table 24, a base 26, an arm section 28, and a compression unit 32. A radiation detector 20 is located inside the imaging table 24. As shown in Figure 2, when performing imaging with the mammography apparatus 10, the subject's breast M is positioned on the imaging surface 24A of the imaging table 24 by the user.

[0027] The radiation detector 20 detects radiation R that has passed through the subject, the breast M. More specifically, the radiation detector 20 detects radiation R that has entered the subject's breast M and the imaging table 24 and reached the detection surface 20A of the radiation detector 20, and generates a radiation image based on the detected radiation R. The radiation detector 20 outputs image data representing the generated radiation image. Hereinafter, the series of operations from irradiating with radiation R from the radiation source 29 to generating a radiation image by the radiation detector 20 may be referred to as "imaging". The radiation detector 20 may be an indirect conversion type radiation detector that converts radiation R into light and the converted light into electric charge, or it may be a direct conversion type radiation detector that directly converts radiation R into electric charge.

[0028] The compression unit 32 is equipped with a compression plate 30 used to compress the breast M during imaging. The compression plate 30 is moved in a direction toward or away from the imaging table 24 (hereinafter referred to as the "vertical direction") by a compression plate drive unit (not shown) provided on the compression unit 32. The compression plate 30 compresses the breast M between itself and the imaging table 24 by moving in the vertical direction.

[0029] The arm portion 28 is rotatable relative to the base 26 by the shaft portion 27. The shaft portion 27 is fixed to the base 26, and the shaft portion 27 and the arm portion 28 rotate together as a single unit. Gears are provided on the shaft portion 27 and the compression unit 32 of the imaging table 24, and by switching between the meshed and dismeshed states of these gears, it is possible to switch between a state in which the compression unit 32 of the imaging table 24 and the shaft portion 27 are connected and rotate together as a single unit, and a state in which the shaft portion 27 is separated from the imaging table 24 and rotates freely. Note that the switching of power transmission and non-transmission of the shaft portion 27 is not limited to the above gears, and various mechanical elements can be used. The arm portion 28 and the imaging table 24 are rotatable separately relative to the base 26 with the shaft portion 27 as the axis of rotation.

[0030] When performing tomosynthesis imaging with the mammography device 10, the radiation source 29 is sequentially moved to each of several irradiation positions with different irradiation angles by the rotation of the arm portion 28. The radiation source 29 has a radiation tube (not shown) that generates radiation R, and the radiation tube moves to each of the several irradiation positions in accordance with the movement of the radiation source 29.

[0031] Figure 2 illustrates an example of tomosynthesis imaging. Note that the compression plate 30 is omitted from the illustration in Figure 2. In this embodiment, the radiation source 29 is moved to irradiation positions Pk (k=1,2,...7) where the irradiation angle differs by a constant angle β. That is, the radiation source 29 is sequentially moved to multiple positions where the irradiation angle of the radiation R with respect to the detection surface 20A of the radiation detector 20 differs. Note that although the number of irradiation positions Pk is set to 7 in Figure 2, the number of irradiation positions Pk is not limited and can be changed as appropriate.

[0032] At each irradiation position Pk, radiation R is irradiated from the radiation source 29 toward the breast M, and a radiation image is generated when the radiation detector 20 detects the radiation R that has passed through the breast M. In the radiation imaging system 2, when tomosynthesis imaging is performed by moving the radiation source 29 to each irradiation position Pk and generating a radiation image at each irradiation position Pk, seven radiation images are obtained in the example shown in Figure 2.

[0033] In the following, when referring to radiographic images taken at each irradiation position Pk in tomosynthesis imaging, they will be called "projection images" to distinguish them from tomographic images, and multiple projection images taken in a single tomosynthesis scan will be called "a series of multiple projection images." Furthermore, when referring to projection images without distinguishing them from tomographic images, they will simply be called "radiographic images."

[0034] Furthermore, as shown in Figure 2, the irradiation angle of radiation R is defined as the angle α between the normal CL of the detection surface 20A of the radiation detector 20 and the radiation axis RC. The radiation axis RC is the axis connecting the focal point of the radiation source 29 at each irradiation position Pk to a predetermined position. The detection surface 20A of the radiation detector 20 is a plane that is approximately parallel to the imaging surface 24A. The radiation R emitted from the radiation source 29 is a cone beam with the focal point as its apex and the radiation axis RC as its central axis.

[0035] On the other hand, in the mammography device 10, when normal imaging is performed, the position of the radiation source 29 is fixed at irradiation position P4 where the irradiation angle α is 0 degrees. According to the instructions of the console 12, radiation R is emitted from the radiation source 29, and a radiation image is generated when the radiation detector 20 detects the radiation R that has passed through the breast M.

[0036] The mammography device 10 and the console 12 are connected by wired or wireless communication. The radiographic image generated by the radiation detector 20 in the mammography device 10 is output to the console 12 via a communication interface (not shown) by wired or wireless communication.

[0037] The console 12 comprises a control unit 40, a storage unit 42, a user interface 44, and a communication interface 46. As described above, the control unit 40 has the function of controlling radiographic imaging by the mammography device 10. The control unit 40 is composed of a computer system, for example, a CPU (Central Processing Unit), ROM (Read Only Memory), and RAM (Random Access Memory).

[0038] The memory unit 42 stores information related to radiography, radiographic images acquired from the mammography device 10, etc. The memory unit 42 is a non-volatile storage device such as an HDD (Hard Disk Drive) or SSD (Solid State Drive).

[0039] The user interface 44 includes input devices such as buttons and switches operated by users, such as technicians, in relation to the acquisition of radiographic images, as well as display devices such as lamps and displays that show information related to the acquisition and radiographic images obtained from the acquisition.

[0040] Communication I / F46 communicates various data, such as radiographic information and radiographic images, with the mammography device 10 via wired or wireless communication. Furthermore, communication I / F46 communicates various data, such as radiographic images, with the PACS 14 and image processing device 16 via the network 17, using either wired or wireless communication.

[0041] Furthermore, the PACS 14 includes a storage unit 50 (see Figure 1) for storing a group of radiographic images 52. The group of radiographic images 52 includes projection images acquired from the console 12 via the network 17.

[0042] The image processing device 16 has a function to support a physician's diagnosis by making a determination regarding the diagnosis of a lesion when a physician (hereinafter simply referred to as "physician") makes a diagnosis regarding a lesion of the breast M using radiographic images.

[0043] Figure 3 shows an example of the configuration of the image processing device 16. The image processing device 16 comprises a control unit 60, a storage unit 62, a display unit 70, an operation unit 72, and a communication interface 74. The control unit 60, storage unit 62, display unit 70, operation unit 72, and communication interface 74 are connected to each other via a bus 79, such as a system bus or control bus, enabling the exchange of various types of information.

[0044] The control unit 60 controls the overall operation of the image processing device 16. The control unit 60 is composed of a computer system equipped with a CPU 60A, ROM 60B, and RAM 60C. ROM 60B pre-stores various programs, data, etc., for control by the CPU 60A. RAM 60C temporarily stores various data.

[0045] The memory unit 62 is a non-volatile storage device such as an HDD or SSD. The memory unit 62 stores programs 63 and other files that cause the control unit 60 to execute various processes.

[0046] The display unit 70 is a display that shows radiographic images, various information, etc. The operation unit 72 is used to input instructions and various information for the diagnosis of breast lesions using radiographic images by a physician. The operation unit 72 may include, for example, various switches, a touch panel, a stylus, and a mouse.

[0047] The communication interface 74 communicates various types of information with the console 12 and PACS 14 via the network 17 using wireless or wired communication.

[0048] Figure 4 shows an example of the functions realized by the control unit 60 of the image processing device 16. The CPU 60A of the control unit 60 realizes various functions by executing processing based on the program 63 stored in the storage unit 62. The control unit 60 functions as a tomographic image generation unit 80, a calcification image detection unit 81, a region of interest image group generation unit 82, a variance value derivation unit 83, a shape type determination unit 84, and a display control unit 85.

[0049] The tomography image generation unit 80 has the function of generating multiple tomography images 90 (see Figure 5) from a series of multiple projection images 100. Based on instructions to diagnose a lesion, the tomography image generation unit 80 acquires a series of multiple projection images 100 from the console 12 of the mammography device 10 or from the PACS 14. From the acquired series of multiple projection images 100, the tomography image generation unit 80 generates multiple tomography images 90 at different heights from the imaging plane 24A. For example, the tomography image generation unit 80 generates multiple tomography images 90 by reconstructing a series of multiple projection images 100 using the back projection method. The back projection method can be the FBP (Filter Back Projection) method, the iterative reconstruction method, etc. The tomography image generation unit 80 outputs the generated multiple tomography images 90 to the calcification image detection unit 81.

[0050] Figure 5 schematically shows the processing flow by the image processing device 16. The processing by the calcification image detection unit 81, the region of interest image group generation unit 82, the variance value derivation unit 83, and the shape type determination unit 84 will be explained with reference to Figure 5.

[0051] The calcification image detection unit 81 performs calcification image detection processing to detect images of tissue in the breast M that are presumed to have calcified (hereinafter referred to as calcification images) based on a plurality of tomographic images 90 generated by the tomographic image generation unit 80. As the calcification image detection unit 81, a detector using a known CAD (Computer-Aided Diagnosis) algorithm can be used. In the CAD algorithm, the probability (likelihood) that a pixel in the tomographic image 90 is a calcification image is derived, and pixels whose probability is above a predetermined threshold are detected as calcification images.

[0052] Furthermore, the calcification image detection unit 81 is not limited to a detector using a CAD algorithm, but may also be composed of a machine-learned model that has undergone machine learning.

[0053] The detection result of the calcification image by the calcification image detection unit 81 is output, for example, as multiple mask images 91 representing the location of the calcification image. Each of the multiple mask images 91 is a binary image in which pixels included in the calcification image are represented as "1" and other pixels as "0". The calcification image detection unit 81 outputs multiple mask images 91 corresponding to each of the multiple tomographic images 90. By performing detection processing using multiple tomographic images 90, calcification images can be detected with high detection accuracy. In the example shown in Figure 5, three calcification images C1 to C3 are detected by the calcification image detection unit 81.

[0054] The region of interest image generation unit 82 performs region of interest image generation processing to generate a group of region of interest images (hereinafter referred to as ROI (Region of Interest) image group) based on a series of multiple projection images 100 used for reconstruction processing by the tomographic image generation unit 80, the results of calcification image detection by the calcification image detection unit 81, and the position information of the radiation tube at the time each of the series of multiple projection images 100 was captured.

[0055] Figure 6 conceptually illustrates an example of the region of interest image group generation process by the region of interest image group generation unit 82. Based on a plurality of mask images 91, the region of interest image group generation unit 82 generates a group of ROI images containing multiple ROI images by extracting the region containing the calcification image as an ROI image from each of a series of plurality of projection images 100. Furthermore, if multiple calcification images are detected in the calcification image detection process, the region of interest image group generation unit 82 generates a separate ROI image group for each of the multiple calcification images. In the example shown in Figure 6, a separate ROI image group is generated for each of the three calcification images C1 to C3. This generates ROI image group G1 containing calcification image C1, ROI image group G2 containing calcification image C2, and ROI image group G3 containing calcification image C3.

[0056] The variance value derivation unit 83 performs a variance value calculation process to calculate the variance value of each feature of the ROI images that make up the ROI image group.

[0057] Figure 7 conceptually illustrates an example of the variance calculation process by the variance value derivation unit 83. Specifically, Figure 7 shows the variance calculation process for one ROI image group G. The ROI image group G consists of seven ROI images R1 to R7. ROI image Rk is an image extracted from a projection image acquired at the irradiation position Pk. First, the variance value derivation unit 83 calculates the feature quantity Fk of ROI image Rk based on the following equation (1).

[0058]

number

[0059] Here, r(x,y) is the pixel value of the pixel at coordinates x,y within the ROI image Rk. a is the average value of the pixel values ​​r(x,y) contained in the ROI image Rk. n is the number of pixels contained in the ROI image Rk.

[0060] In this embodiment, feature vector Fk is the variance of the pixel values ​​r(x,y) contained in the ROI image Rk. Seven feature vectors F1 to F7 are calculated from seven ROI images R1 to R7. The larger the feature vector Fk, the more pixels corresponding to calcifications (e.g., high-luminance pixels) are considered to be contained in the ROI image Rk.

[0061] Then, the variance value derivation unit 83 calculates the variance value D based on the following equation (2).

[0062]

number

[0063] Here, F a The value of F1 to F7 is the average value of the features F1 to F7. The variance D represents the degree of variability among the features F1 to F7. A larger variance D indicates a greater change in shape due to differences in irradiation position Pk. For example, a smaller variance D indicates that the calcification image is closer to a circular shape, while a larger variance D indicates that the calcification image is closer to a linear shape.

[0064] In the example shown in Figure 5, the variance value derivation unit 83 outputs the variance values ​​D1 to D3 generated based on the ROI image group G1 to G3 to the shape type determination unit 84.

[0065] The shape type determination unit 84 performs a shape type determination process to determine the shape type of the calcification image based on the variance value calculated by the variance value derivation unit 83. The shape type determination unit 84 holds information that defines the relationship between a predetermined variance value and the shape type of the calcification image, and determines the shape type of the calcification image based on this information. In other words, the shape type determination unit 84 is a rule-based determination model based on the relationship between a predetermined variance value and the shape type of the calcification image.

[0066] In the example shown in Figure 5, the shape type determination unit 84 determines the shape type of the calcification images C1 to C3 included in the ROI image group G1 to G3 based on the variance values ​​D1 to D3. The shape type determination unit 84 outputs a shape type determination result 84A that represents the shape type of the calcification image. The shape type determination result 84A includes determination results A1 to A3. Determination result A1 indicates that the shape type of calcification image C1 is "micro-circular". Determination result A2 indicates that the shape type of calcification image C2 is "circular". Determination result A3 indicates that the shape type of calcification image C3 is "fine linear".

[0067] In this embodiment, the classification of calcification patterns includes not only differences in shape but also differences in size. Furthermore, the classification of calcification patterns is not limited to the examples described above. Preferably, the classification of calcification patterns is categorized in a way that allows for the determination of whether the calcification is benign or malignant.

[0068] Furthermore, the shape type determination unit 84 may perform shape type determination processing using a machine learning model that has learned the relationship between the above-mentioned variance value and the shape type of the calcification image.

[0069] The display control unit 85 performs display processing to display the shape type determination result 84A obtained from the shape type determination process on the display unit 70. Specifically, the display control unit 85 highlights calcification images having a specific shape based on the shape type determination result 84A.

[0070] Figure 8 shows an example of display processing by the display control unit 85. For example, the display control unit 85 displays the shape type determination result 84A on the display unit 70 along with the tomographic image 90 as a clinical image. In the example shown in Figure 8, the display control unit 85 highlights calcification images with a highly malignant shape (e.g., linear) by surrounding them with a frame 110, based on the shape type determination result 84A. The tomographic image 90 on which the frame 110 is displayed is, for example, one tomographic image 90 selected from multiple tomographic images 90 via the operation unit 72.

[0071] In the example shown in Figure 8, only calcification images with a highly malignant shape are enclosed in the frame 110, but all calcification images may be enclosed in the frame 110. In this case, the color of the frame 110, the line type of the frame 110, etc. may be different based on the shape type determination result 84A. For example, the frame 110 surrounding calcification images with a highly malignant shape may be red, and the frame 110 surrounding calcification images with a less malignant shape may be blue.

[0072] Furthermore, the calcification image is not limited to being enclosed in a frame 110; the calcification images may be colored differently based on the shape type determination result 84A. For example, only calcification images with a highly malignant shape may be highlighted by coloring them. Additionally, characters, symbols, etc., representing the shape type determination result 84A may be displayed on the display unit 70.

[0073] Next, with reference to Figure 9, a series of processes performed by the image processing device 16 will be described. First, in step S10, the tomographic image generation unit 80 acquires a series of multiple projection images 100 from the console 12 of the mammography device 10 or from the PACS 14.

[0074] In step S11, the tomographic image generation unit 80 generates a series of tomographic images 90 based on a series of projection images 100 acquired in step S10.

[0075] In step S12, the calcification image detection unit 81 detects calcification images from the multiple tomographic images 90 generated in step S11 and generates multiple mask images 91 as a result of the detection.

[0076] In step S13, the region of interest image group generation unit 82 generates a group of ROI images by using the multiple mask images 91 generated in step S12 to extract the region containing the calcification image from each of the series of multiple projection images 100 as an ROI image.

[0077] In step S14, the variance value derivation unit 83 calculates the variance value of each feature of the ROI images that make up the ROI image group generated in step S13.

[0078] In step S15, the shape type determination unit 84 determines the shape type of the calcification image based on the variance value calculated in step S14 and outputs the shape type determination result 84A.

[0079] In step S16, the display control unit 85 performs display processing to display the shape type determination result 84A obtained in step S15 on the display unit 70. Specifically, the display control unit 85 highlights the calcification images that have a highly malignant shape.

[0080] Generally, the shape of calcifications is often not accurately represented in composite two-dimensional images generated from multiple tomographic images, and in tomographic images, they are often obscured by noise, resulting in poor visibility. In contrast, the technology of this disclosure generates a group of ROI images by extracting regions containing calcifications from each of a series of multiple projection images, and determines the type of shape of the calcification based on the variance value of the feature quantities of each ROI image constituting the ROI image group, thereby enabling accurate determination of the type of shape of the calcification.

[0081] Furthermore, in the above embodiment, the calcification detection unit 81 detects calcifications from multiple tomographic images 90. The calcification detection unit 81 may also detect only calcifications whose signal value is below a certain value (so-called faint calcifications). This is because faint calcifications do not accurately represent their shape, and it is difficult to determine the type of shape on the tomographic image 90, which is a clinical image displayed on the display unit 70.

[0082] [Differentiation] A modified version of the above embodiment will be described below. Figure 10 shows the functions realized by the control unit 60 of the modified image processing apparatus 16. This modified version differs from the above embodiment in that the control unit 60 functions as a good / bad determination unit 86 in addition to the tomographic image generation unit 80, calcification image detection unit 81, region of interest image group generation unit 82, variance value derivation unit 83, shape type determination unit 84, and display control unit 85.

[0083] Figure 11 schematically shows the processing flow of the image processing device 16 in the modified example. In this modified example, the shape type determination result 84A output from the shape type determination unit 84 is input to the benign / malignant determination unit 86. Based on the shape type determination result 84A, the benign / malignant determination unit 86 determines whether the calcification image is benign or malignant, or determines the degree of malignancy of the calcification image. In the example shown in Figure 11, the benign / malignant determination unit 86 determines whether the calcification image C1 is benign or malignant based on determination result A1, determines whether the calcification image C2 is benign or malignant based on determination result A2, and determines whether the calcification image C3 is benign or malignant based on determination result A3.

[0084] Furthermore, additional information 93 other than the shape type determination result 84A may be input to the good / bad determination unit 86. The additional information 93 may include, for example, the distribution of calcification patterns and the calcification patterns in the composite two-dimensional image. By using the additional information 93 in addition to the shape type determination result 84A, the good / bad determination unit 86 can make a good / bad determination with high accuracy.

[0085] In this modified example, the display control unit 85 performs display processing to display the good / bad judgment result 86A output from the good / bad judgment unit 86 on the display unit 70. For example, the display control unit 85 highlights the calcification image that has been determined to be malignant based on the good / bad judgment result 86A. The display control unit 85 may also display characters, symbols, etc., representing the good / bad judgment result 86A on the display unit 70.

[0086] As the good / bad judgment unit 86, for example, a judge using a known CAD algorithm can be used. The good / bad judgment unit 86 may also consist of a machine-trained model that has undergone machine learning.

[0087] [Other variations] In the above embodiment and modified examples, the feature quantity Fk calculated by the variance value derivation unit 83 is the average value r of the pixel values ​​included in the ROI image Rk, as shown in equation (1) above. a This is the variance of the pixel values ​​contained in the ROI image Rk relative to the mean value of the pixel values ​​in the breast region M (hereinafter referred to as the breast region) in one of the series of multiple projection images 100. a This could be the average value of pixel values ​​within the breast region in a single projection image (for example, a projection image from which the ROI image Rk has been extracted). In this case, the feature Fk is represented as the variability of pixel values ​​relative to normal values ​​within the breast region.

[0088] Furthermore, the feature quantity Fk calculated by the variance value derivation unit 83 may be the number of pixels among the multiple pixels included in the ROI image Rk that have a pixel value equal to or greater than a threshold. In this case, by setting the threshold to the lower limit of the values ​​that the calcification image can take, the number of pixels with a pixel value equal to or greater than the threshold can be made to correspond to the number of pixels that constitute the calcification image.

[0089] The above embodiments and their respective modifications can be combined as appropriate, as long as no inconsistencies arise.

[0090] Furthermore, in the above embodiments and their respective modifications, the hardware structure of the processing unit that performs various processes, such as the tomographic image generation unit 80, the calcification image detection unit 81, the region of interest image group generation unit 82, the variance value derivation unit 83, the shape type determination unit 84, the display control unit 85, and the good / bad determination unit 86, can be the following types of processors. These types of processors include CPUs as well as GPUs (Graphics Processing Units). Moreover, these types of processors are not limited to general-purpose processors such as CPUs that execute software (programs) and function as various processing units, but also include programmable logic devices (PLDs) such as FPGAs (Field Programmable Gate Arrays) whose circuit configuration can be changed after manufacturing, and dedicated electrical circuits such as ASICs (Application Specific Integrated Circuits) which have circuit configurations specifically designed to perform specific processes.

[0091] A single processing unit may consist of one of these various processors, or it may consist of a combination of two or more processors of the same or different types (for example, a combination of multiple FPGAs, or a combination of a CPU and an FPGA). Alternatively, multiple processing units may be composed of a single processor.

[0092] Examples of configuring multiple processing units with a single processor include, firstly, a configuration where one or more CPUs and software combine to form a single processor, as exemplified by client and server computers, and this processor functions as multiple processing units. 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 exemplified by System-on-a-Chip (SoC). Thus, various processing units are configured, in terms of hardware structure, using one or more of the above-mentioned processors.

[0093] Furthermore, the hardware structure of these various processors can more specifically utilize electrical circuits, which are combinations of circuit elements such as semiconductor devices.

[0094] Furthermore, although the above embodiments and their variations describe a configuration in which the program 63 is pre-stored in the storage unit 62, the invention is not limited thereto. The program 63 may be provided in a form that is not temporarily recorded on a recording medium such as a CD-ROM (Compact Disc Read Only Memory), DVD-ROM (Digital Versatile Disc Read Only Memory), or USB (Universal Serial Bus) memory. Alternatively, the program 63 may be downloaded from an external device via a network.

[0095] The descriptions and illustrations presented above are detailed explanations of the technical aspects of this disclosure and are merely examples of the technical aspects. For example, the above descriptions of the structure, function, operation, and effect are examples of the structure, function, operation, and effect of the technical aspects of this disclosure. Therefore, it goes without saying that you may delete unnecessary parts, add new elements, or replace elements in the descriptions and illustrations presented above, as long as you do not deviate from the essence of the technical aspects of this disclosure. Furthermore, in order to avoid confusion and facilitate understanding of the technical aspects of this disclosure, explanations of common technical knowledge and the like that do not require special explanation to enable the implementation of the technical aspects of this disclosure have been omitted from the descriptions and illustrations presented above.

[0096] All documents, patent applications, and technical standards described herein are incorporated by reference to the same extent as if each individual document, patent application, and technical standard were specifically and individually noted to be incorporated by reference. [Explanation of Symbols]

[0097] 2. Radiation imaging system 10. Mammography equipment 12 Consoles 14 PACS 16 Image Processing Device 17 Network 20 Radiation detectors 20A detection surface 24 Shooting platform 24A Imaging surface 26 base 27 Shaft section 28 Arm section 29 Radiation source 30 Compression plate 32 Compression Units 40 Control Unit 42 Storage section 44 User Interface 46 Communication I / F 50 Storage section 52 Radiological images 60 Control Unit 62 Storage section 63 Programs 70 Display section 72 Operation section 74 Communication I / F 79 Bus 80 Tomographic Image Generation Unit 81 Calcification image detection unit 82 Region of Interest Image Swarm Generation Unit 83. Variance Value Derivation Section 84 Shape Type Determination Unit 84A Shape Type Determination Result 85 Display Control Unit 86 Good / Bad Judgment Section 86A Good / bad judgment result 90 Fault images 91 Masked Images 93 Additional Information 100 Projection Images 110 slots α Irradiation angle β angle A1,A2,A3 Judgment results C1,C2,C3 Calcification image CL normal D,D1,D2,D3 dispersion value Fk special quantity G, G1, G2, G3 ROI Image Group M-shaped breasts Pk Irradiation Position Rk ROI portrait R radiation RC Radiation Axis

Claims

1. Equipped with at least one processor, The aforementioned processor, A calcification detection process that detects calcifications based on multiple tomographic images obtained from a series of multiple projection images obtained by tomosynthesis imaging of the breast, A region of interest image generation process generates a group of region of interest images by extracting a region containing the calcification image detected by the calcification image detection process from each of the multiple projection images, as a region of interest image. A variance calculation process that calculates the variance value of each feature quantity of the images in the region of interest that constitute the group of images in the region of interest, A shape type determination process that determines the shape type of the calcification image based on the variance value calculated by the variance value calculation process, An image processing device that performs the following: The aforementioned feature quantities are The variance of pixel values ​​included in one of the aforementioned region images, The variance of pixel values ​​in the region of interest image relative to the average value of pixel values ​​in the breast region in one of the projection images, and The number of pixels among a plurality of pixels included in one of the aforementioned region of interest images that have a pixel value greater than or equal to a threshold. It is one of the following: Image processing device.

2. The aforementioned processor, If multiple calcification images are detected in the calcification image detection process, the region of interest image group generation process generates the region of interest image group individually for each of the multiple calcification images. The image processing apparatus according to claim 1.

3. The aforementioned processor, In the shape type determination process, the shape of the calcification image is determined based on the relationship between the predetermined variance value and the shape type. The image processing apparatus according to claim 1 or claim 2.

4. The aforementioned processor, Further, a display process is performed to display the shape type determination result obtained by the shape type determination process on the display unit. The image processing apparatus according to any one of claims 1 to 3.

5. The aforementioned processor, In the display process described above, based on the shape type determination result, the calcification image having a specific shape is highlighted. The image processing apparatus according to claim 4.

6. A calcification detection step in which calcification images are detected based on multiple tomographic images obtained from a series of multiple projection images obtained by tomosynthesis imaging of the breast, A region of interest image generation step generates a group of region of interest images by extracting a region corresponding to the calcification image detected in the calcification image detection step from each of the multiple projection images as a region of interest image, A variance calculation step, which involves calculating the variance value of each feature quantity of the images in the region of interest that constitute the group of images in the region of interest, A shape type determination step in which the shape type of the calcification image is determined based on the variance value calculated in the variance value calculation step, An image processing method that performs the following: The aforementioned feature quantities are The variance of pixel values ​​included in one of the aforementioned region images, The variance of pixel values ​​in the region of interest image relative to the average value of pixel values ​​in the breast region in one of the projection images, and The number of pixels among a plurality of pixels included in one of the aforementioned region of interest images that have a pixel value greater than or equal to a threshold. It is one of the following: Image processing methods.

7. A calcification detection process that detects calcifications based on multiple tomographic images obtained from a series of multiple projection images obtained by tomosynthesis imaging of the breast, A region of interest image generation process generates a group of region of interest images by extracting a region corresponding to the calcification image detected by the calcification image detection process from each of the multiple projection images, as a region of interest image. A variance calculation process that calculates the variance value of each feature quantity of the images in the region of interest that constitute the group of images in the region of interest, A shape type determination process that determines the shape type of the calcification image based on the variance value calculated by the variance value calculation process, A program that causes a computer to execute, The aforementioned feature quantities are The variance of pixel values ​​included in one of the aforementioned region images, The variance of pixel values ​​in the region of interest image relative to the average value of pixel values ​​in the breast region in one of the projection images, and The number of pixels among a plurality of pixels included in one of the aforementioned region of interest images that have a pixel value greater than or equal to a threshold. It is one of the following: program.

Images