Registration method and device for mammary gland image and mammary gland X-ray machine
Through the calculation of tissue region segmentation and projection data, the optimal alignment offset is determined, which solves the problem of low accuracy in breast image registration and achieves high-precision breast image registration.
Patent Information
- Application Number
- CN202510112582.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-01-23
- Publication Date
- 2025-05-13
AI Technical Summary
In the prior art, in breast image registration, the accuracy of the registration results is low due to the large differences in breast morphology and different image acquisition standards.
By obtaining the mask image of the bilateral breast to be registered separately for tissue area segmentation, the projection data is obtained, and the similarity index coefficient of the projection data under different offsets is calculated, the optimal alignment offset is determined, and the high-precision registration of the breast image is achieved.
It effectively reduces interference factors during the registration process, improves the accuracy of registration results, and can find the best registration offset more accurately, thereby achieving high-precision registration of bilateral breast images.
Smart Images
Figure CN119991627A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of medical imaging technology, for example, to a registration method and device for breast images, and a breast X-ray machine. Background Art
[0002] Mammography, as a key medical imaging technology, plays a vital role in the early screening, diagnosis and subsequent treatment of breast diseases. This technology uses low-dose X-rays to penetrate breast tissue, capture and record the structural details inside the breast, and form high-resolution imaging data. These imaging data provide clinicians with intuitive visual references, enabling them to accurately identify abnormal changes in the breast, such as lumps, calcifications and other potential lesions, thus laying the foundation for the accurate diagnosis of breast diseases.
[0003] In breast X-ray image analysis, doctors will analyze the breast images on both sides at the same time, and aligning the breast images on both sides is a key step. Doctors manually move the breast images on both sides to align the left and right breast structures, allowing doctors to more accurately compare and analyze the differences between the two breasts, making it easier to detect subtle abnormal changes. However, due to individual differences in breast tissue and differences in image acquisition conditions, the shape, size and position of the left and right breast images on the image are often uncertain, which brings trouble to the doctor's operation. In related technologies, the left and right breasts are aligned by translating the breasts through the distance difference between the center of mass positions of the left and right breast images. The aligned breast images are convenient for doctors to compare and analyze.
[0004] In the process of implementing the embodiments of the present disclosure, it is found that there are at least the following problems in the related art:
[0005] When the left and right breasts are aligned by translating the breasts using the distance difference between the centroid positions of the left and right breast images, the accuracy of the final breast image registration result is low due to the large differences in actual breast morphology and different image acquisition standards.
[0006] It should be noted that the information disclosed in the above background technology section is only used to enhance the understanding of the background of the present application, and therefore may include information that does not constitute the prior art known to ordinary technicians in the field. Summary of the invention
[0007] In order to provide a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. The summary is not an extensive review, nor is it intended to identify key / critical components or delineate the scope of protection of these embodiments, but rather serves as a prelude to the detailed description that follows.
[0008] The embodiments of the present disclosure provide a method and device for breast image registration, and a breast X-ray machine, so as to improve the accuracy of breast image registration results.
[0009] In some embodiments, a registration method for breast images includes: obtaining a first mask image and a second mask image after tissue region segmentation of a first breast image to be registered and a second breast image to be registered in the bilateral breast images to be registered, respectively; obtaining first projection data corresponding to the first mask image, and obtaining second projection data corresponding to the second mask image; determining similarity index coefficients of the first projection data and the second projection data at different offsets, and determining a target offset corresponding to a target similarity index coefficient in the similarity index coefficients; taking the first breast image to be registered as a reference, displacing the second breast image to be registered according to the target offset to align the bilateral breast images.
[0010] In some embodiments, a registration apparatus for breast images includes a processor and a memory storing program instructions, and the processor is configured to perform the aforementioned registration method for breast images when executing the program instructions.
[0011] In some embodiments, a breast X-ray machine includes: a breast X-ray machine body; and the aforementioned registration device for breast images, which is installed on the breast X-ray machine body.
[0012] The breast image registration method and device, and breast X-ray machine provided by the embodiments of the present disclosure can achieve the following technical effects:
[0013] In the technical solution of the present application, after obtaining the first mask image and the second mask image after tissue region segmentation of the first breast image to be registered and the second breast image to be registered in the bilateral breast images to be registered, the first projection data corresponding to the first mask image and the second projection data corresponding to the second mask image are obtained, and then the similarity index coefficients of the first projection data and the second projection data under different offsets are determined, and the target offset corresponding to the target similarity index coefficient in the similarity index coefficient is determined, and then the first breast image to be registered is used as a reference, and the second breast image to be registered is displaced according to the target offset to align the bilateral breast images. Through the steps of tissue region segmentation, acquisition of projection data, calculation of similarity index coefficients, and displacement based on target offset, the interference factors in the registration process are effectively reduced, and the accuracy of the registration results is improved. In particular, through the calculation of projection data and similarity index coefficients, the optimal registration offset can be found more accurately, thereby achieving high-precision registration of bilateral breast images.
[0014] The above general description and the following description are exemplary and explanatory only and are not intended to limit the present application. BRIEF DESCRIPTION OF THE DRAWINGS
[0015] One or more embodiments are exemplarily described by corresponding drawings, which do not limit the embodiments. Elements with the same reference numerals in the drawings are shown as similar elements, and the drawings do not constitute a scale limitation, and wherein:
[0016] Figure 1 is a flow chart of a registration method for breast images provided by an embodiment of the present disclosure;
[0017] Figure 2 is a flow chart of another registration method for breast images provided by an embodiment of the present disclosure;
[0018] Figure 3 is a flow chart of another registration method for breast images provided by an embodiment of the present disclosure;
[0019] Figure 4 is a flow chart of another registration method for breast images provided by an embodiment of the present disclosure;
[0020] FIG5( a) is a schematic diagram of a first mask image and a second mask image provided by an embodiment of the present disclosure;
[0021] FIG5( b ) is a schematic diagram of calculating first projection data and second projection data provided by an embodiment of the present disclosure;
[0022] FIG5( c ) is a schematic diagram of calculating a similarity index coefficient based on an offset provided by an embodiment of the present disclosure;
[0023] FIG6( a ) is an unregistered bilateral breast image provided by an embodiment of the present disclosure;
[0024] FIG6( b ) is a registered bilateral breast image provided by an embodiment of the present disclosure;
[0025] Figure 7 is a structural schematic diagram of a registration device for breast images provided by an embodiment of the present disclosure;
[0026] Figure 8 Schematic diagram of a breast X-ray machine provided in an embodiment of the present disclosure. DETAILED DESCRIPTION
[0027] In order to be able to understand the features and technical contents of the embodiments of the present disclosure in more detail, the implementation of the embodiments of the present disclosure is described in detail below in conjunction with the accompanying drawings. The attached drawings are for reference only and are not used to limit the embodiments of the present disclosure. In the following technical description, for the convenience of explanation, a full understanding of the disclosed embodiments is provided through multiple details. However, one or more embodiments can still be implemented without these details. In other cases, to simplify the drawings, well-known structures and devices can be simplified for display.
[0028] The terms "first", "second", etc. in the specification and claims of the embodiments of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the terms used in this way can be interchanged where appropriate, so that the embodiments of the embodiments of the present disclosure described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions.
[0029] Unless otherwise specified, the term "plurality" means two or more than two. In the disclosed embodiment, the character " / " indicates that the objects before and after are in an "or" relationship. For example, A / B means: A or B. The term "and / or" is a description of the association relationship between objects, indicating that three relationships may exist. For example, A and / or B means: A or B, or, the three relationships of A and B. The term "corresponding" may refer to an association relationship or a binding relationship, and the correspondence between A and B means that there is an association relationship or a binding relationship between A and B.
[0030] Mammographic X-ray images, also known as breast radiographs or mammography, are based on the penetration of X-rays and the differences in the absorption of X-rays by human tissue. Breast tissue has a higher density than other human tissues, so it absorbs X-rays to a greater degree. When X-rays penetrate breast tissue, most of the X-rays are absorbed, and only a small amount of X-rays can penetrate and be projected onto the photosensitive film or digital detector. Due to the different structures and density distributions within the breast tissue, the resulting images will also be different.
[0031] The craniocaudal position (CC position) and the medial-lateral oblique position (MLO position) are two common positions for breast X-ray imaging. The CC position focuses on showing the inner and posterior breast tissue, especially the inner part of the breast, which helps doctors find lesions such as breast hyperplasia, fibroids, cystic lesions, and malignant tumors. As a supplement to the medial-lateral oblique position, the CC position can also show the inner and posterior breast tissue, including the outer and posterior tissue. The MLO position is the best position for displaying unilateral breast tissue, which can show most of the breast tissue from the chest wall to the nipple and extend to the axillary area. This position helps doctors better evaluate different parts of the breast, especially the upper outer quadrant, which is the most common site of breast cancer.
[0032] The embodiment of the disclosure is described by taking the CC position breast X-ray image as an example, and setting the horizontal direction in the breast image as the x direction and the vertical direction as the y direction. For breast images in other positions, the processing method is the same.
[0033] Combination Figure 1 As shown, the present disclosure provides a method for registration of breast images, comprising the following steps:
[0034] S101, respectively obtaining a first mask image and a second mask image after performing tissue region segmentation on a first breast image to be registered and a second breast image to be registered in the bilateral breast images to be registered.
[0035] Here, the bilateral breast images to be registered include a first breast image to be registered and a second breast image to be registered. The first breast image to be registered and the second breast image to be registered can be left breast images or right breast images. When the first breast image to be registered is a left breast image, the second breast image to be registered is a right breast image. Conversely, when the first breast image to be registered is a right breast image, the second breast image to be registered is a left breast image. Of course, it is not limited to the case where multiple left (or right) breast images are taken for comparison under some special conditions. At this time, the first breast image to be registered and the second breast image to be registered are both left (or right) breast images.
[0036] The tissue region of interest is extracted from the bilateral breast images to be registered, and the corresponding mask image is generated. In some practical applications, a region-based image segmentation algorithm, such as a region growing algorithm, can be used to segment breast images. This algorithm selects an initial seed point and gradually expands the region according to a preset similarity criterion until the entire tissue region of interest is covered. After the segmentation is completed, the region of interest is set as the foreground (usually white or highlighted), and the rest is set as the background (usually black or low-brightness), thereby generating a mask image.
[0037] S102, obtaining first projection data corresponding to the first mask image, and obtaining second projection data corresponding to the second mask image.
[0038] Extract projection data from the mask image for subsequent similarity measurement. In some practical applications, radar projection, parallel projection, or perspective projection can be used to convert the two-dimensional mask image into one-dimensional projection data. The projection data usually represents the density or existence of tissue at different locations in the mask image, which can be used for subsequent similarity measurement.
[0039] S103, determining similarity index coefficients of the first projection data and the second projection data at different offsets, and determining a target offset corresponding to a target similarity index coefficient in the similarity index coefficients.
[0040] By comparing the similarity of the first projection data and the second projection data at different offsets, the best alignment offset is determined. In some practical applications, similarity measurement methods such as mutual information, correlation coefficient, and normalized mutual information can be used to calculate the similarity index coefficient of the first projection data and the second projection data at different offsets. By traversing all possible offsets, the offset with the largest similarity index coefficient is found, which is the target offset.
[0041] S104, taking the first breast image to be registered as a reference, shifting the second breast image to be registered according to the target offset, so as to align the breast images on both sides.
[0042] According to the target offset, the second breast image to be registered is displaced to achieve registration and alignment of the bilateral breast images. In some practical applications, an image transformation algorithm, such as affine transformation, rigid transformation or non-rigid transformation, can be used to displace the second breast image to be registered. After the displacement is completed, the registration result is verified by visual inspection or automatic evaluation method to ensure that the bilateral breast images have been accurately aligned.
[0043] The registration method for breast images provided by the embodiment of the present disclosure is adopted. First, the tissue region segmentation is performed on the bilateral breast images to be registered to generate the first mask image and the second mask image. This segmentation can remove unnecessary background information in the image and focus on the breast tissue itself. By focusing on the breast tissue, the influence of interference factors such as background noise and skin contour on the registration result can be reduced, thereby improving the accuracy of the registration. Secondly, the projection data corresponding to the first mask image and the second mask image are obtained. The projection data provides an opportunity to observe the image from another angle. Since the projection can capture the features of the breast tissue in different directions, it is easier to find the alignment relationship between the images. Again, by calculating the similarity index coefficient of the first projection data and the second projection data at different offsets, the best registration offset can be found. Taking the first breast image to be registered as the reference, the second breast image to be registered is displaced according to the target offset, thereby realizing the registration alignment of the bilateral breast images, ensuring that the registered images have the highest consistency in anatomical structure. In this way, through the steps of tissue region segmentation, projection data acquisition, similarity index coefficient calculation and displacement based on target offset, the interference factors in the registration process are effectively reduced and the accuracy of the registration results is improved. In particular, through the calculation of projection data and similarity index coefficient, the optimal registration offset can be found more accurately, thereby achieving high-precision registration of bilateral breast images.
[0044] In some embodiments, obtaining a first mask image after tissue region segmentation of the first breast image to be registered includes: determining a foreground region and a background region in the first breast image to be registered; setting the foreground region in the first breast image to be registered to 1, and setting the background region in the first breast image to be registered to 0, to obtain a first mask image.
[0045] In some embodiments, obtaining a second mask image after tissue region segmentation of the second breast image to be registered includes: determining a foreground region and a background region in the second breast image to be registered; setting the foreground region in the second breast image to be registered to 1, and setting the background region in the second breast image to be registered to 0, to obtain a second mask image.
[0046] Fig. 5(a) is a schematic diagram of a first mask image and a second mask image provided by an embodiment of the present disclosure. In conjunction with Fig. 5(a), tissue region segmentation masks M1 and M2 of two breast images are extracted by using a preset threshold.
[0047] Optionally, the foreground area of the bilateral breast images to be registered is determined as follows: for the bilateral breast images to be registered whose background area is a low value, the first pixel point corresponding to the first coordinate point in the bilateral breast images to be registered is determined, and when the first pixel point is greater than a first set threshold, it is determined that the first coordinate point belongs to the foreground area.
[0048] Optionally, the background area of the bilateral breast images to be registered is determined as follows: for the bilateral breast images to be registered whose background area is a low value, the second pixel point corresponding to the second coordinate point in the bilateral breast images to be registered is determined, and when the second pixel point is less than a first set threshold, it is determined that the second coordinate point belongs to the background area.
[0049] Among them, the first set threshold is 0. For breast images with low background areas, the image pixel value of the background part usually corresponds to the lowest value of the entire image, while the area corresponding to the foreground area is the human tissue structure included in the image acquisition range, and its pixel value will be greater than 0. Then, according to this set judgment rule, the first set threshold can be set to 0. For example, for any coordinate (x, y) on the input image corresponding to the pixel point Ixy, if the value of Ixy is greater than the threshold, it can be regarded as the foreground area, otherwise it is regarded as the background area, thereby realizing the segmentation of the tissue structure area.
[0050] In some other possible implementations, the foreground area of the bilateral breast images to be registered is determined as follows: for the bilateral breast images to be registered in which the background area is a high value, the first pixel point corresponding to the first coordinate point in the bilateral breast images to be registered is determined, and when the first pixel point is less than a second set threshold, it is determined that the first coordinate point belongs to the foreground area.
[0051] In some other possible implementations, the background area of the bilateral breast images to be registered is determined as follows: for the bilateral breast images to be registered in which the background area is a high value, the second pixel point corresponding to the second coordinate point in the bilateral breast images to be registered is determined, and when the second pixel point is greater than a second set threshold, it is determined that the second coordinate point belongs to the background area.
[0052] Among them, since the foreground and background areas of breast images are highly distinguishable, the second threshold value can be specifically determined according to the image output by the acquisition setting. For images with high background areas, the second threshold value is the highest value in the image. If the value of Ixy is less than the second threshold value, it can be regarded as a foreground area, otherwise it is regarded as a background area. Of course, it can also be achieved using commonly used automatic threshold segmentation algorithms such as the Otsu method.
[0053] Combination Figure 2 As shown, in some embodiments, a registration method for breast images is provided, comprising the following steps:
[0054] S201, determining a foreground area and a background area in a first breast image to be registered in the bilateral breast images to be registered.
[0055] S202, setting the foreground area in the first breast image to be registered to 1, and setting the background area in the first breast image to be registered to 0, to obtain a first mask image.
[0056] S203, determining a foreground area and a background area in a second breast image to be registered in the bilateral breast images to be registered.
[0057] S204, setting the foreground area in the second breast image to be registered to 1, and setting the background area in the second breast image to be registered to 0, to obtain a second mask image.
[0058] S205: Obtain first projection data corresponding to the first mask image, and obtain second projection data corresponding to the second mask image.
[0059] S206, determining similarity index coefficients of the first projection data and the second projection data at different offsets, and determining a target offset corresponding to a target similarity index coefficient in the similarity index coefficients.
[0060] S207 , using the first breast image to be registered as a reference, shifting the second breast image to be registered according to the target offset, so as to align the bilateral breast images.
[0061] In the disclosed embodiment, by determining the foreground area (i.e., breast tissue part) and the background area (background or other non-breast tissue part) of the breast image, a clear data basis is provided for subsequent processing. This precise image segmentation helps to reduce interference factors in the registration process and improve the accuracy of the registration. At the same time, the foreground area is set to 1 and the background area is set to 0 to generate a mask image. This binarization process greatly simplifies the image data, making the subsequent processing steps more efficient, and the mask image also retains key breast tissue information to ensure the effectiveness of the registration. In this way, not only the accuracy and efficiency of breast image registration are improved, but also strong support is provided for the diagnosis and treatment of breast diseases.
[0062] In some embodiments, obtaining first projection data corresponding to the first mask image includes: projecting the first mask image in a horizontal direction to a vertical direction to obtain the first projection data.
[0063] In some embodiments, obtaining second projection data corresponding to the second mask image includes: projecting the second mask image in a horizontal direction to a vertical direction to obtain the second projection data.
[0064] FIG5(b) is a schematic diagram of calculating the first projection data and the second projection data provided by an embodiment of the present disclosure. In conjunction with FIG5(b), in some practical applications, the tissue region masks M1 and M2 are projected onto the y axis along the horizontal x direction to obtain two sets of projection data P1 and P2. The projection here can be the cumulative projection or mean projection of the tissue region masks M1 and M2 in the x direction. Taking the cumulative projection as an example, for the tissue region mask M1, its resolution is R xn ×R yn , traverse all the segmentation mask pixel values row by row in the x direction and sum them up, such as for the first row M of the segmentation mask 00 , M 01 ,……,M 0Rxn , calculate the sum of all pixel values to get the projection value of the row, and after traversing all rows, get a set of length R yn The projection data P1 of the same length can be obtained by the same method. If the mean projection is calculated, it is only necessary to divide the projection data obtained by the cumulative projection by the x-direction length R xn That's it.
[0065] By projecting the two-dimensional mask image from the horizontal direction to the vertical direction, the complex two-dimensional image information can be simplified into one-dimensional projection data, which not only reduces the complexity of data processing but also improves the processing speed. Although the projection process simplifies the data, it still retains the key information of breast tissue. The projection data in the vertical direction can reflect the distribution and density changes of breast tissue in the horizontal direction, which is crucial for subsequent similarity evaluation and alignment.
[0066] In some embodiments, determining similarity index coefficients of first projection data and second projection data at different offsets includes: determining a target offset search range; taking the first projection data as a reference, calculating a similarity index coefficient of an overlapping area between the first projection data and the second projection data when the second projection data traverses all offsets within the target offset search range.
[0067] In some practical applications, the maximum vertical y-direction offset D is preset, and the target offset search range is [-D, D]. In order to adapt to the situation where the position deviation of the two sets of image tissue structures is large, the maximum offset D should not be set too small. One preset method is D = R yn / 4, which means that the compressed image is about 1+R in the y direction yn / 2. When the calculation is completed, 2D+1 similarity index coefficients S will be obtained.
[0068] Optionally, the similarity index coefficient is calculated as follows: determine the current offset within the target offset search range; obtain a first projection value corresponding to the current offset in the first projection data, and a second projection value corresponding to the current offset in the second projection data; and determine the similarity index coefficient corresponding to the current offset based on the projection difference between the first projection value and the second projection value.
[0069] Taking one set of projection data P1 as the benchmark, when the other set of projection data P2 traverses all offsets between -D and D, the similarity index coefficient S of the overlapping areas of the two sets of projection data is calculated one by one. The similarity index coefficient S can be an index coefficient of a method for measuring data similarity, such as the inverse of the Euclidean distance of the two sets of data, cosine similarity, etc.
[0070] Fig. 5(c) is a schematic diagram of calculating a similarity index coefficient based on an offset provided by an embodiment of the present disclosure. In conjunction with Fig. 5(c), in some possible implementations, the similarity index coefficient corresponding to the current offset is determined according to the projection difference between the first projection value and the second projection value, including: calculating the sum of the squares of the projection differences of all the first projection values and the second projection values under the current offset; and determining the similarity index coefficient corresponding to the current offset according to the sum of the squares of the projection differences.
[0071] In some specific embodiments, the similarity index coefficient is calculated according to the following formula:
[0072] S n =-(((P10-P2 -n ) 2 +(P1 n+1 -P2 1-n ) 2 +(P1n+2 -P2 2-n ) 2 +…+(P1 Ryn-n -P2 Ryn ) 2 ) 1 / 2 ), n<0;
[0073] S n =-(((P1 n -P20) 2 +(P1 n+1 -P21) 2 +(P1 n+2 -P22) 2 +…+(P1 Ryn -P2 Ryn-n ) 2 ) 1 / 2 ), n≥0.
[0074] When the maximum vertical y offset is D=2, the similarity index coefficient S is calculated in turn for five cases where the offset is between -D and D, i.e., -2, -1, 0, 1, and 2. -2 , S -1 , S0, S1, S2.
[0075] S -2 =-(((P10-P22) 2 +(P11-P23) 2 +(P12-P24) 2 +…+(P1 Ryn-2 -P2 Ryn ) 2 ) 1 / 2 );
[0076] S -1 =-(((P10-P21) 2 +(P11-P22) 2 +(P12-P23) 2 +…+(P1 Ryn-1 -P2 Ryn ) 2 ) 1 / 2 );
[0077] S0=-(((P10-P20) 2 +(P11-P21) 2 +(P12-P22) 2 +…+(P1 Ryn -P2 Ryn ) 2 ) 1 / 2 );
[0078] S1=-(((P11-P20) 2 +(P12-P21) 2 +(P13-P22) 2 +…+(P1 Ryn -P2 Ryn-1 ) 2 ) 1 / 2 );
[0079] S2=-(((P12-P20) 2 +(P13-P21) 2 +(P14-P22) 2 +…+(P1 Ryn -P2 Ryn-2 ) 2 ) 1 / 2 ).
[0080] In some other specific embodiments, the similarity index coefficient is calculated according to the following formula:
[0081] S n =-((P10-P2 -n ) 2 +(P1 n+1 -P2 1-n ) 2 +(P1 n+2 -P2 2-n ) 2 +…+(P1 Ryn-n -P2 Ryn ) 2 ), n<0;
[0082] S n =-((P1 n -P20) 2 +(P1 n+1 -P21) 2 +(P1 n+2 -P22) 2 +…+(P1 Ryn -P2 Ryn-n ) 2 ), n≥0.
[0083] When the maximum vertical y offset is D=2, the similarity index coefficient S is calculated in turn for the five cases where the offset is between -D and D, i.e., -2, -1, 0, 1, and 2. -2 , S -1 , S0, S1, S2.
[0084] S -2 =-((P10-P22) 2 +(P11-P23) 2+(P12-P24) 2 +…+(P1 Ryn-2 -P2 Ryn ) 2 );
[0085] S -1 =-((P10-P21) 2 +(P11-P22) 2 +(P12-P23) 2 +…+(P1 Ryn-1 -P2 Ryn ) 2 );
[0086] S0=-((P10-P20) 2 +(P11-P21) 2 +(P12-P22) 2 +…+(P1 Ryn -P2 Ryn ) 2 );
[0087] S1=-((P11-P20) 2 +(P12-P21) 2 +(P13-P22) 2 +…+(P1 Ryn -P2 Ryn-1 ) 2 );
[0088] S2=-((P12-P20) 2 +(P13-P21) 2 +(P14-P22) 2 +…+(P1 Ryn -P2 Ryn-2 ) 2 ).
[0089] Combination Figure 3 As shown, in some embodiments, a registration method for breast images is provided, comprising the following steps:
[0090] S301 , respectively obtaining a first mask image and a second mask image after performing tissue region segmentation on a first breast image to be registered and a second breast image to be registered in the bilateral breast images to be registered.
[0091] S302 : Project the first mask image in the horizontal direction to the vertical direction to obtain first projection data.
[0092] S303 : Project the second mask image in the vertical direction along the horizontal direction to obtain second projection data.
[0093] S304, determining a target offset search range.
[0094] S305 , taking the first projection data as a reference, calculating a similarity index coefficient of an overlapping area between the first projection data and the second projection data when the second projection data traverses all offsets within a target offset search range.
[0095] S306 , using the first breast image to be registered as a reference, shifting the second breast image to be registered according to the target offset, so as to align the breast images on both sides.
[0096] In the disclosed technical solution, the first mask image and the second mask image are projected to the vertical direction respectively along the horizontal direction to obtain the first projection data and the second projection data. By projecting from the horizontal direction to the vertical direction, not only the key position information of the breast tissue is retained, but also the processing speed is significantly accelerated. Especially when processing high-resolution breast images, this dimensionality reduction process can significantly improve the operating efficiency of the registration algorithm, making it more suitable for clinical practice and large-scale data analysis. At the same time, the determination of the target offset search range provides a reasonable search boundary for the registration process, avoids unnecessary global search, and further improves the efficiency of the algorithm. Taking the first projection data as a benchmark, the similarity index coefficient of the second projection data at different offsets is calculated, which can accurately quantify the degree of alignment between the two images and ensure that the optimal registration parameters are found.
[0097] In some embodiments, the second breast image to be registered is displaced according to a target offset with the first breast image to be registered as a reference to achieve bilateral breast image registration and alignment, including: when the bilateral breast images to be registered are original breast images, the second breast image to be registered is displaced along a preset direction according to the target offset with the first breast image to be registered as a reference to achieve registration and alignment of the first breast image to be registered and the second breast image to be registered.
[0098] Here, the preset direction may be vertically upward or vertically downward.
[0099] Among all similarity index coefficients S, find the highest similarity index coefficient S max , and use this to confirm the corresponding target offset value d. That is, after obtaining 2D+1 similarity index coefficients, the maximum value S of these coefficients can be calculated. max Since these coefficients correspond to a total of 2D+1 offsets from -D to D, we get S max The corresponding offset is d. The meaning of d is that when the offset between the projection data P1 and P2 reaches d, the similarity between the two sets of data is the highest. Take image I1 as the reference image and shift the other image I2 in the y direction by d. nThe displacement is performed to achieve the registration and alignment of the two breast images.
[0100] In some embodiments, the second breast image to be registered is displaced according to a target offset with the first breast image to be registered as a reference to achieve bilateral breast image registration and alignment, including: when the bilateral breast images to be registered are downsampled breast images, the target offset is enlarged according to the downsampling sampling rate, and the first original breast image corresponding to the first breast image to be registered is used as a reference, and the second original breast image corresponding to the second breast image to be registered is displaced along a preset direction according to the enlarged target offset to register and align the first original breast image and the second original breast image.
[0101] The original breast images of the bilateral breast images to be registered include a first original breast image I1 and a second original breast image I2. The first original breast image I1 and the second original breast image I2 are downsampled by n times in equal proportion to obtain two low-resolution downsampled breast images, namely, the bilateral breast images to be registered, which are the first breast image I1 and the second breast image I2. n and the second breast image to be registered I 2 n The downsampling operation may be linear interpolation or equidistant sampling. For example, if the resolution of the first original breast image I1 and the second original breast image I2 are both R x ×R y , downsample it by n times, and the resolution is R xn ×R yn Downsampled breast image I1 n , I 2 n Among them, R xn ,R yn Respectively represent R x / n,R y / n is the integer part. Taking image I1 as the reference image, another image I2 is offset in the y direction by n×d n The displacement is performed to achieve the registration and alignment of the two breast images.
[0102] Combination Figure 4 As shown, in some embodiments, a registration method for breast images is provided, comprising the following steps:
[0103] S401 , respectively obtaining a first mask image and a second mask image after performing tissue region segmentation on a first breast image to be registered and a second breast image to be registered in the bilateral breast images to be registered.
[0104] S402: Obtain first projection data corresponding to the first mask image, and obtain second projection data corresponding to the second mask image.
[0105] S403, determining similarity index coefficients of the first projection data and the second projection data at different offsets, and determining a target offset corresponding to a target similarity index coefficient in the similarity index coefficients.
[0106] S404, when the bilateral breast images to be registered are original breast images, the second breast image to be registered is displaced along a preset direction according to a target offset amount with the first breast image to be registered as a reference, so as to align the first breast image to be registered and the second breast image to be registered.
[0107] S405 , when the bilateral breast images to be registered are downsampled breast images, amplify the target offset according to the downsampling sampling magnification.
[0108] S406, taking the first original breast image corresponding to the first breast image to be registered as a reference, displacing the second original breast image corresponding to the second breast image to be registered along a preset direction according to the amplified target offset, so as to align the first original breast image and the second original breast image.
[0109] The registration method for breast images provided by the disclosed embodiment directly applies the calculated target offset to the original image for displacement, which is simple and direct, and ensures the accuracy of registration. For the downsampled image, the target offset is cleverly enlarged according to the sampling ratio and then applied to the original image, which not only takes into account the information loss caused by downsampling, but also ensures that the registration parameters obtained from the low-resolution image can effectively guide the registration of the high-resolution image, reflecting the flexibility and practicality of the algorithm.
[0110] FIG6(a) is an unregistered bilateral breast image provided by an embodiment of the present disclosure, and FIG6(b) is a registered bilateral breast image provided by an embodiment of the present disclosure. Combining FIG6(a) and FIG6(b), it can be seen that the scheme for breast image registration provided by an embodiment of the present disclosure takes into account the distribution information of the tissue structure in the breast image, achieves more accurate registration, and considers more detailed global features of the image, avoids dependence on structural features at specific locations in the breast image, and has better stability and applicability.
[0111] Combination Figure 7As shown, the embodiment of the present disclosure provides a registration device 700 for breast images, including a processor 70 and a memory 71, and may also include a communication interface 72 and a bus 73. The processor 70, the communication interface 72, and the memory 71 may communicate with each other through the bus 73. The communication interface 72 may be used for information transmission. The processor 70 may call the logic instructions in the memory 71 to execute the registration method for breast images of the above embodiment.
[0112] In addition, the logic instructions in the above-mentioned memory 71 can be implemented in the form of software functional units and can be stored in a computer-readable storage medium when sold or used as an independent product.
[0113] The memory 71 is a computer-readable storage medium that can be used to store software programs and computer executable programs, such as program instructions / modules corresponding to the method in the embodiment of the present disclosure. The processor 70 executes the function application and data processing by running the program instructions / modules stored in the memory 71, that is, the registration method for breast images in the above method embodiment is implemented.
[0114] The memory 71 may include a program storage area and a data storage area, wherein the program storage area may store an operating system and an application required for at least one function; the data storage area may store data created according to the use of the terminal device, etc. In addition, the memory 71 may include a high-speed random access memory and may also include a non-volatile memory.
[0115] The registration device for breast images provided by the embodiment of the present disclosure is used to obtain the first mask image and the second mask image after the tissue region segmentation of the first breast image to be registered and the second breast image to be registered in the bilateral breast images to be registered, respectively, and then obtain the first projection data corresponding to the first mask image and the second projection data corresponding to the second mask image, and then determine the similarity index coefficients of the first projection data and the second projection data under different offsets, and determine the target offset corresponding to the target similarity index coefficient in the similarity index coefficient, and then take the first breast image to be registered as a reference, and shift the second breast image to be registered according to the target offset to align the bilateral breast images. Through the steps of tissue region segmentation, acquisition of projection data, calculation of similarity index coefficients, and displacement based on target offset, the interference factors in the registration process are effectively reduced, and the accuracy of the registration results is improved. In particular, through the calculation of projection data and similarity index coefficients, the optimal registration offset can be found more accurately, thereby achieving high-precision registration of bilateral breast images.
[0116] Combination Figure 8 As shown, the embodiment of the present disclosure provides a breast X-ray machine 80 , including a breast X-ray machine body 800 , and the aforementioned registration device 700 for breast images, which is installed on the breast X-ray machine body 800 .
[0117] An embodiment of the present disclosure provides a computer-readable storage medium storing computer-executable instructions, wherein the computer-executable instructions are configured to execute the above-mentioned registration method for breast images.
[0118] An embodiment of the present disclosure provides a computer program product, which includes a computer program stored on a computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer executes the above-mentioned registration method for breast images.
[0119] The computer-readable storage medium mentioned above may be a transient computer-readable storage medium or a non-transitory computer-readable storage medium.
[0120] The technical solution of the embodiment of the present disclosure can be embodied in the form of a software product, which is stored in a storage medium and includes one or more instructions for enabling a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in the embodiment of the present disclosure. The aforementioned storage medium may be a non-transient storage medium, including: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes, or a transient storage medium.
[0121] The above description and the accompanying drawings fully illustrate the embodiments of the present disclosure so that those skilled in the art can practice them. Other embodiments may include structural, logical, electrical, process and other changes. The embodiments represent only possible changes. Unless explicitly required, individual components and functions are optional, and the order of operation may vary. The parts and features of some embodiments may be included in or replace the parts and features of other embodiments. The scope of the embodiments of the present disclosure includes the entire scope of the claims, and all available equivalents of the claims. When used in this application, although the terms "first", "second", etc. may be used in this application to describe each element, these elements should not be limited by these terms. These terms are only used to distinguish one element from another element. For example, without changing the meaning of the description, the first element can be called the second element, and similarly, the second element can be called the first element, as long as all occurrences of the "first element" are renamed consistently and all occurrences of the "second element" are renamed consistently. The first element and the second element are both elements, but may not be the same element. Moreover, the words used in this application are only used to describe the embodiments and are not used to limit the claims. As used in the description of the embodiments and claims, unless the context clearly indicates, the singular forms "a", "an" and "the" are intended to include the plural forms as well. Similarly, the term "and / or" as used in this application refers to any and all possible combinations of one or more associated listings. In addition, when used in this application, the term "comprise" and its variants "comprises" and / or comprising refer to the presence of stated features, wholes, steps, operations, elements, and / or components, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or groups of these. In the absence of further restrictions, the elements defined by the sentence "including one..." do not exclude the presence of other identical elements in the process, method or device including the elements. In this article, each embodiment may focus on the differences from other embodiments, and the same and similar parts between the embodiments may refer to each other. For the methods, products, etc. disclosed in the embodiments, if they correspond to the method part disclosed in the embodiments, then the relevant parts can refer to the description of the method part.
[0122] Those skilled in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software may depend on the specific application and design constraints of the technical solution. The technicians may use different methods for each specific application to implement the described functions, but such implementations should not be considered to exceed the scope of the embodiments of the present disclosure. The technicians may clearly understand that, for the convenience and simplicity of description, the specific working processes of the systems, devices and units described above may refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here.
[0123] In the embodiments disclosed herein, the disclosed methods and products (including but not limited to devices, equipment, etc.) can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units can be only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the coupling or direct coupling or communication connection between each other shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to implement this embodiment. In addition, each functional unit in the embodiment of the present disclosure may be integrated in a processing unit, or each unit may exist physically alone, or two or more units may be integrated in one unit.
[0124] The flowchart and block diagram in the accompanying drawings show the possible architecture, function and operation of the system, method and computer program product according to the embodiment of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of the code, and the module, the program segment or a part of the code contains one or more executable instructions for realizing the specified logical function. In some alternative implementations, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. In the description corresponding to the flowchart and the block diagram in the accompanying drawings, the operations or steps corresponding to different boxes can also occur in a different order from the order disclosed in the description, and sometimes there is no specific order between different operations or steps. For example, two consecutive operations or steps can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, which can depend on the functions involved. Each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may be implemented by a dedicated hardware-based system that performs the specified functions or actions, or may be implemented by a combination of dedicated hardware and computer instructions.
Claims
1. A registration method for breast images, characterized in that: include: Respectively obtaining a first mask image and a second mask image after performing tissue region segmentation on a first breast image to be registered and a second breast image to be registered in the bilateral breast images to be registered; Obtaining first projection data corresponding to the first mask image, and obtaining second projection data corresponding to the second mask image; Determine similarity index coefficients of the first projection data and the second projection data at different offsets, and determine a target offset corresponding to a target similarity index coefficient in the similarity index coefficients; Taking the first breast image to be registered as a reference, the second breast image to be registered is shifted according to the target offset to align the bilateral breast images.
2. The registration method according to claim 1, characterized in that: Obtaining a first mask image after performing tissue region segmentation on the first breast image to be registered, including: Determine a foreground area and a background area in a first breast image to be registered; The foreground region in the first breast image to be registered is set to 1, and the background region in the first breast image to be registered is set to 0, to obtain a first mask image; and / or, Obtaining a second mask image after performing tissue region segmentation on the second breast image to be registered, including: Determining a foreground area and a background area in a second breast image to be registered; The foreground area in the second breast image to be registered is set to 1, and the background area in the second breast image to be registered is set to 0, to obtain a second mask image.
3. The registration method according to claim 2, characterized in that: The foreground area of the bilateral breast images to be registered is determined as follows: For the bilateral breast images to be registered whose background area is a low value, determining a first pixel point corresponding to a first coordinate point in the bilateral breast images to be registered, and determining that the first coordinate point belongs to the foreground area when the first pixel point is greater than a first set threshold; For the bilateral breast images to be registered whose background area is a high value, determining a first pixel point corresponding to a first coordinate point in the bilateral breast images to be registered, and determining that the first coordinate point belongs to the foreground area when the first pixel point is less than a second set threshold; The background area of the bilateral breast images to be registered is determined as follows: For the bilateral breast images to be registered whose background area is a low value, determining a second pixel point corresponding to a second coordinate point in the bilateral breast images to be registered, and determining that the second coordinate point belongs to the background area when the second pixel point is less than a first set threshold; For the bilateral breast images to be registered with high background areas, the second pixel point corresponding to the second coordinate point in the bilateral breast images to be registered is determined. When the second pixel point is greater than a second set threshold, it is determined that the second coordinate point belongs to the background area.
4. The registration method according to claim 1, characterized in that: Obtaining first projection data corresponding to the first mask image, comprising: Along the horizontal direction, projecting the first mask image to the vertical direction to obtain first projection data; and / or, Obtaining second projection data corresponding to the second mask image, comprising: Along the horizontal direction, the second mask image is projected to the vertical direction to obtain second projection data.
5. The registration method according to claim 1, characterized in that: Determining similarity index coefficients of the first projection data and the second projection data at different offsets includes: Determine the target offset search range; Taking the first projection data as a reference, a similarity index coefficient of an overlapping area between the first projection data and the second projection data is calculated when the second projection data traverses all offsets within a target offset search range.
6. The registration method according to claim 5, characterized in that: The similarity index coefficient is calculated as follows: Determine the current offset within the target offset search range; Obtaining a first projection value corresponding to the current offset in the first projection data and a second projection value corresponding to the current offset in the second projection data; A similarity index coefficient corresponding to the current offset is determined according to a projection difference between the first projection value and the second projection value.
7. The registration method according to claim 6, characterized in that: Determining a similarity index coefficient corresponding to the current offset according to a projection difference between the first projection value and the second projection value includes: Under the current offset, calculate the sum of squares of the projection differences between all first projection values and second projection values; The similarity index coefficient corresponding to the current offset is determined according to the sum of the squares of the projection differences.
8. The registration method according to any one of claims 1 to 7, characterized in that: Taking the first breast image to be registered as a reference, the second breast image to be registered is shifted according to the target offset to achieve bilateral breast image registration and alignment, including: In the case where the bilateral breast images to be registered are original breast images, the second breast image to be registered is displaced along a preset direction according to a target offset amount with the first breast image to be registered as a reference, so as to align the first breast image to be registered with the second breast image to be registered; In the case where the bilateral breast images to be registered are downsampled breast images, the target offset is enlarged according to the downsampling sampling rate, and the first original breast image corresponding to the first breast image to be registered is used as a reference, and the second original breast image corresponding to the second breast image to be registered is displaced along a preset direction according to the enlarged target offset to align the first original breast image and the second original breast image.
9. A registration device for breast images, comprising a processor and a memory storing program instructions, characterized in that: The processor is configured to perform the registration method for breast images according to any one of claims 1 to 8 when executing the program instructions.
10. A breast X-ray machine, characterized in that: include; Mammography machine body; The registration device for breast images as described in claim 9 is installed on a breast X-ray machine body.