A breast tumor recognition system based on multi-modal ultrasound imaging

Abstract

This invention relates to the field of computer-aided diagnosis, and in particular to a breast tumor identification system based on multimodal ultrasound imaging. The breast tumor identification system provided by this invention performs the following steps: acquiring a first ultrasound modality breast image, and locating a suspected first breast tumor region in the first ultrasound modality breast image; acquiring a second ultrasound modality breast image, and aligning the second ultrasound modality breast image with the first ultrasound modality breast image; based on the alignment result, locating a second suspected breast tumor region in the second ultrasound modality breast image corresponding to the suspected first breast tumor region; and identifying the confidence level of the breast tumor in the suspected second breast tumor region based on the distribution of microvessels in the suspected second breast tumor region. The breast tumor identification system based on multimodal ultrasound imaging provided by this invention, combining breast images from both the first and second ultrasound modalities, achieves more accurate breast tumor identification and localization, improving detection accuracy.

Description

Technical Field

[0001] This invention relates to the field of computer-aided diagnosis, and in particular to a breast tumor identification system based on multimodal ultrasound imaging. Background Technology

[0002] Breast health is an important part of women's health, and breast tumors are the most common type of breast disease. Breast tumors include both benign and malignant lesions, with breast cancer being the most concerning because it can pose a serious threat to a patient's life. Early detection of breast tumors is crucial for successful treatment and improved patient survival rates.

[0003] Traditional single-modal breast imaging is limited in providing detailed information and cannot comprehensively and accurately reflect the nature and distribution of tumors. The introduction of multimodal ultrasound imaging technology has provided a revolutionary tool for the accurate detection and assessment of breast tumors. Multimodal ultrasound imaging technology combines different ultrasound modalities, enabling physicians to gain a more comprehensive understanding of the characteristics of breast tumors, thereby better planning diagnosis and treatment. Summary of the Invention

[0004] Based on the shortcomings of existing technologies and the needs of practical applications, this invention provides a breast tumor identification system based on multimodal ultrasound imaging, aiming to achieve more accurate breast tumor identification by utilizing multimodal ultrasound imaging.

[0005] The breast tumor identification system based on multimodal ultrasound imaging provided in this invention includes a memory for storing a computer program. The computer program includes program instructions, which, when executed by a processor, cause the processor to perform the following steps: acquiring a first ultrasound modal breast image; locating a first suspicious area of ​​breast tumor in the first ultrasound modal breast image, wherein the first ultrasound modal breast image is a grayscale breast image; acquiring a second ultrasound modal breast image and aligning the second ultrasound modal breast image with the first ultrasound modal breast image, wherein the second ultrasound modal breast image is an ultramicrovascular breast image; locating a second suspicious area of ​​breast tumor corresponding to the first suspicious area of ​​breast tumor in the second ultrasound modal breast image based on the alignment result; and identifying the confidence level of breast tumor in the second suspicious area of ​​breast tumor based on the distribution of microvessels in the second suspicious area of ​​breast tumor. The breast tumor identification system based on multimodal ultrasound imaging provided by this invention achieves more accurate identification and localization of breast tumors through grayscale breast images and ultramicrovascular breast images, improving the accuracy of detection. It is expected to have a positive impact on the early detection and treatment of breast diseases, providing patients with better medical services and better treatment outcomes.

[0006] Optionally, acquiring the first ultrasound modality breast image includes the following steps: providing a B-mode ultrasound device and using the B-mode ultrasound device to scan the target breast to acquire the first ultrasound modality breast image. This option provides a simple and effective way to acquire the first ultrasound modality breast image, laying the foundation for subsequent multimodal imaging and recognition.

[0007] Optionally, acquiring the second ultrasound modality breast image includes the following steps: providing a color Doppler ultrasound device and scanning the target breast using the color Doppler ultrasound device to obtain the second ultrasound modality breast image. This option provides a simple and effective way to obtain the second ultrasound modality breast image, laying the foundation for subsequent multimodal imaging and recognition.

[0008] Optionally, locating the first suspected breast tumor region in the first ultrasound modal breast image includes the following steps: setting a target resolution, and coordinate-coding the first ultrasound modal breast image based on the target resolution, and obtaining the coordinates and grayscale values ​​of each pixel in the coordinate-coded first ultrasound modal breast image; setting the size of the initial positioning box, and defining different judgment regions by adjusting the center coordinates of the initial positioning box; locating the first suspected breast tumor region in the first ultrasound modal breast image according to the grayscale value distribution of the ultrasound image in the judgment region. This option provides a precise and efficient method for breast tumor localization, by setting a target resolution, coordinate-coding the breast image, and then locating the suspected tumor region according to the grayscale value distribution, so as to more accurately determine the location of the patient's breast tumor.

[0009] Optionally, locating the first suspicious area of ​​a breast tumor in the first ultrasound modality breast image further includes the following step: defining different judgment areas by adjusting the size of the initial positioning box. This option allows for the customization of different judgment areas according to specific circumstances during tumor identification. This is crucial for handling breast images and tumor characteristics from different patients, as each case may have unique features. By dynamically adjusting the judgment areas, it is possible to better adapt to different situations, improving the accurate localization and identification of breast tumors.

[0010] Optionally, locating the first suspicious area of ​​a breast tumor in the first ultrasound modality breast image based on the grayscale value distribution of the ultrasound breast image in the judgment area includes the following steps: obtaining the global grayscale mean of the first ultrasound modality breast image, wherein the global grayscale mean... Satisfy the following formula: Where m represents the maximum x-coordinate value, n represents the maximum y-coordinate value, and p (i,j) The grayscale value of the pixel at coordinates (x, y) is represented; a threshold for suspicious location coordinate pixels is set based on the global grayscale mean, and the threshold for suspicious location coordinate pixels satisfies the following formula: in, K represents the threshold coefficient, P max This represents the maximum grayscale value in the first ultrasound modality breast image. This indicates that the value exceeds the global grayscale mean in the first ultrasound modality breast image. The most frequently occurring grayscale value is identified; based on the grayscale value distribution of the ultrasound breast image in the judgment area, the coordinates of target pixels in the judgment area whose grayscale values ​​are greater than the threshold of the suspicious location coordinates are located; the target pixel coordinates are summarized to form one or more initial suspicious regions, wherein the initial suspicious region is a closed region formed by the edges of multiple adjacent target pixels; a suspicious region location model is constructed, and the first center coordinates of the initial suspicious region are located using the suspicious region location model, wherein the suspicious region location model satisfies the following formula: Among them, (x k ,y k Sk represents the first center coordinate of the k-th initial suspicious region in the judgment region, x a Let x represent the minimum x-coordinate in the k-th initial suspicious region. b Let y represent the maximum x-coordinate in the k-th initial suspicious region. c Let y represent the minimum ordinate in the k-th initial suspicious region. d This represents the maximum ordinate in the k-th initial suspicious region. Using the first center coordinates of the initial suspicious region, the first suspicious breast tumor region in the first ultrasound modality breast image is located. Each first suspicious breast tumor region includes all pixel coordinates from the corresponding initial suspicious region. This option combines global grayscale mean and pixel threshold, accurately identifying regions in breast images with grayscale values ​​significantly higher than the average level, thereby locating potential tumors. Simultaneously, by constructing a suspicious region localization model, the center coordinates of the initial suspicious region can be accurately calculated, improving the accuracy of localization.

[0011] Optionally, locating the second suspected breast tumor region corresponding to the first suspected breast tumor region in the second ultrasound modality breast image based on the alignment result includes the following steps: marking the second central coordinate corresponding to the first central coordinate in the second ultrasound modality breast image based on the alignment result, and marking the pixel coordinates associated with the second central coordinate; generating one or more suspected second breast tumor regions based on the central coordinate in the second ultrasound modality breast image and the pixel coordinates associated with the central coordinate, wherein each suspected second breast tumor region includes a second central coordinate and all pixel coordinates associated with the second central coordinate. This option, by marking the location of the first breast tumor in the second ultrasound modality breast image and then generating the suspected second breast tumor region based on these markings, can more accurately locate the patient's breast tumor.

[0012] Optionally, locating the second suspicious breast tumor region corresponding to the first suspicious breast tumor region in the second ultrasound modality breast image based on the alignment result further includes the following step: setting a tumor area threshold, and filtering the second suspicious breast tumor region using the suspicious tumor area threshold. This optional tumor area threshold reduces false positive detection results and improves the detection accuracy of breast tumors.

[0013] Optionally, identifying the credibility of a breast tumor in the suspected area of ​​the second breast tumor based on the distribution of microvessels in the suspected area includes the following steps: distinguishing between main blood vessels and microvessels in the suspected area of ​​the second breast tumor, and counting the number of main blood vessels and microvessels respectively; constructing a breast tumor credibility model, and using the breast tumor credibility model in combination with the number of main blood vessels and microvessels to identify the credibility of a breast tumor in the suspected area of ​​the second breast tumor, wherein the breast tumor credibility model satisfies the following formula: in, Indicates the suspicious area S′ of the second breast tumor k Confidence level of breast tumors, Q represents the suspicious area S′ of a second breast tumor. k The number of main blood vessels in the middle, I i Indicates the suspicious area S′ of the second breast tumor k The mean curvature of the i-th main blood vessel, where α represents the weighting coefficient of the main blood vessel curvature. Indicates the suspicious area S′ of the second breast tumor k The option represents the number of microvessels corresponding to the i-th main blood vessel, where β represents the bifurcation weight coefficient of the main blood vessel. This option utilizes the number of main blood vessels and microvessels to assess the confidence level of breast tumors in a second suspected breast tumor region using a constructed breast tumor confidence model, thereby improving the diagnostic accuracy of breast tumors and facilitating earlier detection of patient lesions.

[0014] Optionally, the breast tumor identification system based on multimodal ultrasound imaging further includes a processor, an input device, and an output device, wherein the processor, input device, output device, and memory are interconnected. The memory stores a computer program, which includes program instructions, and the processor is configured to call and execute the program instructions. The multimodal ultrasound imaging breast tumor identification system provided by this invention combines different ultrasound imaging modes. With the aid of a computer program and processor, through multimodal imaging and computer-aided analysis, it can provide more accurate and reliable tumor diagnostic results. Doctors can more accurately determine the nature of the tumor, providing earlier diagnosis and more effective treatment plans. Attached Figure Description

[0015] Figure 1 The flowchart of the breast tumor identification method based on multimodal ultrasound imaging provided by the present invention is shown below.

[0016] Figure 2 The first and second ultrasound modal breast images provided by this invention;

[0017] Figure 3 This is a structural diagram of the breast tumor identification system based on multimodal ultrasound imaging provided by the present invention. Detailed Implementation

[0018] Specific embodiments of the present invention will now be described in detail. It should be noted that the embodiments described herein are for illustrative purposes only and are not intended to limit the invention. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the invention. However, it will be apparent to those skilled in the art that these specific details are not necessary to practice the invention. In other instances, well-known circuits, software, or methods have not been specifically described to avoid obscuring the invention.

[0019] Throughout this specification, references to "an embodiment," "an embodiment," "an example," or "an example" mean that a particular feature, structure, or characteristic described in connection with that embodiment or example is included in at least one embodiment of the invention. Therefore, the phrases "in an embodiment," "in an embodiment," "an example," or "an example" appearing in various places throughout the specification do not necessarily refer to the same embodiment or example. Furthermore, specific features, structures, or characteristics can be combined in one or more embodiments or examples in any suitable combination and / or sub-combination. Moreover, those skilled in the art will understand that the illustrations provided herein are for illustrative purposes and are not necessarily drawn to scale.

[0020] In an optional embodiment, please refer to Figure 1 , Figure 1 This is a flowchart of the breast tumor identification method based on multimodal ultrasound imaging provided by the present invention. Figure 1 As shown, the breast tumor identification method based on multimodal ultrasound imaging includes the following steps:

[0021] S01. Obtain a first ultrasound modal breast image, and locate a first suspicious area of ​​breast tumor in the first ultrasound modal breast image. The first ultrasound modal breast image is a grayscale breast image.

[0022] It is understood that the first ultrasound modal breast image obtained by this invention is a grayscale breast image that represents tissue density through different grayscale levels. Typically, the first ultrasound modal breast image can be generated using a B-mode ultrasound imaging (also known as 2D ultrasound imaging) device.

[0023] Specifically, in one or more embodiments, the acquisition of the first ultrasound modality breast image in step S01 includes the following steps: providing a B-mode ultrasound device and using the B-mode ultrasound device to scan the target breast to acquire a first ultrasound modality breast image. The first ultrasound modality breast image uses different grayscale values ​​to characterize the density of the breast tissue at its location, such as... Figure 2 As shown in Figure (A).

[0024] Furthermore, to more quickly and accurately identify areas where breast tumors may exist in the first ultrasound modality breast image, in an optional embodiment, locating the suspected area of ​​a first breast tumor in the first ultrasound modality breast image as described in step S01 includes the following steps:

[0025] S011. Set a target resolution, and coordinate the first ultrasound modal breast image based on the target resolution, and obtain the coordinates and grayscale value of each pixel in the coordinated first ultrasound modal breast image.

[0026] Since the acoustic attenuation coefficient of tissues within a living organism is linearly related to the ultrasound frequency—lower ultrasound frequencies result in longer wavelengths and smaller amplitude attenuation, leading to greater detection depth but lower resolution; higher ultrasound frequencies result in shallower detection depth but higher resolution—the specific target resolution set in step S011 can be determined based on specific medical needs, the performance of the ultrasound imaging equipment, and / or the required size of the target detection area.

[0027] In this embodiment, the target resolution is the optimal image accuracy level that the first ultrasound modality mammogram can achieve, and this optimal image accuracy level depends on the performance of the ultrasound equipment used to acquire the first ultrasound modality mammogram.

[0028] S012. Set the size of the initial positioning frame, and define different judgment areas by adjusting the center coordinates of the initial positioning frame.

[0029] To better select and locate the tumor region, in this embodiment, the initial positioning frame includes a rectangular frame, the size of which is characterized by length (L) and width (W). In one or more other embodiments, the initial positioning frame also includes regular or irregular closed wireframes such as circles or ellipses.

[0030] The initial positioning box center coordinates can be set according to image characteristics and the doctor's needs. For example, the center of the initial positioning box can be placed in the center of the breast region, and then the judgment areas at different locations can be created by changing the center coordinates of the box to find potential tumors.

[0031] Based on the above coordinate conversion results, in this embodiment, the initial positioning box center coordinates (X,Y) satisfy the following optional model: Where (x0, y0) represents the origin coordinates (lower left corner coordinates) of the first ultrasound modality breast image after coordinate transformation, (x m ,y m ) represents the endpoint coordinates (top right corner coordinates) of the first ultrasound modality breast image after coordinateization, L represents the length of the initial positioning box, specifically represented by the number of unit pixels, and W represents the width of the initial positioning box, specifically represented by the number of unit pixels.

[0032] In one or more alternative embodiments, in addition to creating different judgment regions by changing the center coordinates of the initial positioning frame as described in the above embodiments, different judgment regions can also be defined by adjusting the size of the initial positioning frame. For example, the center of the initial positioning frame can be placed at any position in the first ultrasound modality breast image, and then judgment regions of different sizes can be created by changing the size of the frame to locate potential tumors.

[0033] S013. Based on the gray value distribution of the ultrasound breast image in the judgment area, locate the first suspicious area of ​​breast tumor in the first ultrasound modality breast image.

[0034] In another optional embodiment, step S013, which involves locating the first suspicious area of ​​a breast tumor in the first ultrasound modality breast image based on the grayscale value distribution of the ultrasound breast image in the judgment area, includes the following steps:

[0035] S0131. Obtain the global grayscale mean of the first ultrasound modality breast image.

[0036] In this embodiment, the global grayscale mean Satisfy the following formula: Where m represents the maximum x-coordinate value, n represents the maximum y-coordinate value, and p (i,j) This represents the grayscale value of the pixel at coordinates (x, y).

[0037] Furthermore, the global grayscale mean represents the density distribution of breast tissue throughout the image. For example, if the size of the first ultrasound modality breast image is 400x400 pixels, the global grayscale mean can be obtained by iterating through all pixels and calculating their grayscale values. The value is 110.

[0038] S0132. Set a threshold for suspicious location coordinate pixels based on the global grayscale mean.

[0039] In this embodiment, the threshold value of the suspicious location coordinate pixels satisfies the following formula: in, K represents the threshold coefficient, P max This represents the maximum grayscale value in the first ultrasound modality breast image. This indicates that the value exceeds the global grayscale mean in the first ultrasound modality breast image. And the grayscale value that appears most frequently.

[0040] In step S0132, a threshold P for suspicious location coordinate pixels is set using the global grayscale mean. H Then, examine the grayscale value of each pixel in the judgment area and filter out those greater than P. H These pixels are considered to potentially belong to breast tumors.

[0041] S0133. Based on the grayscale value distribution of the ultrasound breast image in the judgment area, locate the target pixel coordinates in the judgment area whose grayscale value is greater than the suspected positioning coordinate pixel threshold.

[0042] In this embodiment, the threshold coefficient K = 1.2, and the maximum gray value P max =255. If the global grayscale mean is... Then the pixel threshold P of the suspicious location coordinates H =1.2 × 110 = 132. Iterate through each pixel in the judgment area, and if its grayscale value is greater than 132, it is selected as the target pixel coordinate.

[0043] S0134. Summarize the target pixel coordinates to form one or more initial suspicious regions, wherein the initial suspicious region is a closed region formed by the edges of multiple adjacent target pixels.

[0044] In this embodiment, if a set or more sets of consecutive pixels with a gray value greater than 132 are formed in a judgment area, they are summarized into one or more initial suspicious areas; if there is no set of consecutive pixels with a gray value greater than 132 in the judgment area, the initial suspicious area of ​​the next judgment area is determined.

[0045] S0135. Construct a suspicious area positioning model and use the suspicious area positioning model to locate the first center coordinates of the initial suspicious area.

[0046] In this embodiment, the suspicious area location model satisfies the following formula: Among them, (x k ,y k )Sk Let x represent the first center coordinate of the k-th initial suspicious region in the judgment region. a Let x represent the minimum x-coordinate in the k-th initial suspicious region. b Let y represent the maximum x-coordinate in the k-th initial suspicious region. c Let y represent the minimum ordinate in the k-th initial suspicious region. d This represents the maximum ordinate in the kth initial suspicious region.

[0047] S0135. Using the first center coordinates of the initial suspicious region, locate the first breast tumor suspicious region in the first ultrasound modality breast image, wherein any first breast tumor suspicious region includes all pixel coordinates in the corresponding initial suspicious region.

[0048] In this embodiment, any initial suspicious area is identified as a possible breast tumor location. Therefore, the center coordinates of these areas, as well as all pixel coordinates (grayscale values ​​greater than 132) related to the center coordinates, are used as the final localization results for the subsequent localization of the second suspicious breast tumor area.

[0049] S02. Acquire a second ultrasound modal breast image and align the second ultrasound modal breast image with the first ultrasound modal breast image. The second ultrasound modal breast image is an ultramicrovascular breast image.

[0050] It is easy to understand that the aforementioned ultra-microvascular breast image is a special type of medical ultrasound image used to observe and analyze the distribution and blood flow of microvessels in breast tissue, such as... Figure 2 As shown in Figure (B). This type of image is typically obtained using color Doppler ultrasound imaging, and its main feature is the ability to display the location, shape, and blood flow velocity of blood vessels.

[0051] In an optional embodiment, the acquisition of the second ultrasound modality breast image in step S02 includes the following steps: providing a color Doppler ultrasound device and using the color Doppler ultrasound device to scan the target breast to obtain the second ultrasound modality breast image.

[0052] Since the first and second ultrasound modal images of the breast are typically obtained at different times or using different parameters, their positions may not be perfectly aligned. Therefore, a common feature or reference point, such as a landmark structure in the breast, can be selected from the two images. Then, the position of the second ultrasound modal image can be adjusted by stretching, rotating, and repositioning to align these reference points.

[0053] In one or more other embodiments, image processing software can be used to automatically identify and match corresponding features in two images to achieve alignment.

[0054] S03. Based on the alignment results, locate the second suspicious area of ​​breast tumor in the second ultrasound modality breast image, which corresponds to the first suspicious area of ​​breast tumor.

[0055] In order to locate the suspected area of ​​the first breast tumor in the second ultrasound modality breast image, in an optional embodiment, step S03, which involves locating the suspected area of ​​the second breast tumor in the second ultrasound modality breast image corresponding to the suspected area of ​​the first breast tumor based on the alignment result, includes the following steps:

[0056] S031. Based on the alignment result, mark the second center coordinates corresponding to the first center coordinates in the second ultrasound modality breast image, and mark the pixel coordinates associated with the second center coordinates.

[0057] In the second ultrasound modality breast image, the second central coordinates corresponding to the suspected area of ​​the first breast tumor are first marked. These central coordinates are typically the geometric center of the suspected area of ​​the first breast tumor or other key coordinates indicating its location. Then, the pixel coordinates associated with the second central coordinates are marked, and these pixel coordinates constitute the boundary or outline of the suspected area of ​​the second breast tumor.

[0058] S032. Based on the center coordinates in the second ultrasound modality breast image and the pixel coordinates associated with the center coordinates, generate one or more second breast tumor suspicious regions. Each second breast tumor suspicious region includes a second center coordinate and all pixel coordinates associated with the second center coordinate.

[0059] Based on the second center coordinates and associated pixel coordinates marked in step S031, one or more suspected regions of second breast tumor are generated. These suspected regions of second breast tumor are typically composed of pixels with the same or similar characteristics, constituting candidate regions of second breast tumor.

[0060] In one or more other embodiments, step S03, which involves locating the second suspicious breast tumor region in the second ultrasound modality breast image corresponding to the first suspicious breast tumor region based on the alignment result, further includes the following steps:

[0061] S033. Set a tumor area threshold and use the suspected tumor area threshold to screen for suspected areas of second breast tumors.

[0062] The tumor area threshold is typically set by a professional physician based on practical experience, and is measured in pixels. It is used to screen for areas of suspicion of secondary breast tumors that have a high probability of harboring a breast tumor. Specifically, if the area of ​​a suspected secondary breast tumor is greater than or equal to the set tumor area threshold, it will be retained; otherwise, it will be excluded.

[0063] S04. Based on the distribution of microvessels in the second suspected breast tumor area, identify the credibility of the breast tumor in the second suspected breast tumor area.

[0064] In an optional embodiment, identifying the confidence level of a breast tumor in the suspected area of ​​the second breast tumor based on the distribution of microvessels in the suspected area includes the following steps:

[0065] S041. Distinguish between the main blood vessels and microvessels in the suspected area of ​​the second breast tumor, and count the number of the main blood vessels and the number of microvessels respectively.

[0066] The main vessels are relatively large vessels in the suspected area of ​​a second breast tumor, typically exhibiting pronounced curvature and branching. The microvessels are small vessels branching off from the main vessels in the suspected area of ​​a second breast tumor.

[0067] S042. Construct a breast tumor credibility model, and use the breast tumor credibility model in combination with the number of the main blood vessels and the number of the micro-blood vessels to identify the credibility of breast tumors in the second suspected breast tumor area.

[0068] In this embodiment, the breast tumor confidence model satisfies the following formula: in, Q represents the confidence level of a breast tumor in the suspected second breast tumor region S′k, and Q represents the number of main blood vessels in the suspected second breast tumor region S′k. i This represents the mean curvature of the i-th main blood vessel in the suspected second breast tumor region S′k, where α represents the weighting coefficient of the main blood vessel curvature. β represents the number of microvessels corresponding to the i-th main blood vessel in the suspected second breast tumor region S′k, and β represents the bifurcation weight coefficient of the main blood vessel. Furthermore, the confidence level for identifying breast tumors in the suspected second breast tumor region can be calculated using the aforementioned breast tumor confidence model.

[0069] In one or more other embodiments, a breast tumor confidence threshold can be set to screen out a second breast tumor suspicious area with a sufficiently high breast tumor confidence, which can then be used as a breast tumor area for further detection or treatment.

[0070] The breast tumor identification method based on multimodal ultrasound imaging provided by this invention combines grayscale breast images and ultramicrovascular breast images, achieving more accurate breast tumor identification and localization, improving detection accuracy, and is expected to have a positive impact on the early detection and treatment of breast diseases, providing patients with better medical services and better treatment outcomes.

[0071] In an optional embodiment, please refer to Figure 3 , Figure 3 This is a structural diagram of the breast tumor identification system based on multimodal ultrasound imaging provided by the present invention. Figure 2 As shown, the present invention also provides a breast tumor identification system based on multimodal ultrasound imaging, including a processor, an input device, an output device, and a memory, wherein the processor, input device, output device, and memory are interconnected, wherein the memory is used to store a computer program, the computer program includes program instructions, and the processor is configured to call the program instructions to execute the breast tumor identification method based on multimodal ultrasound imaging as described above.

[0072] The processor is used to execute various computational and analytical tasks within the system, and to coordinate data exchange and information transmission between different functional modules. Specifically, the processor can be a high-performance computer or an embedded processor.

[0073] The input device is used to acquire a first ultrasound modality breast image and a second ultrasound modality breast image. Typically, the input device is a color Doppler ultrasound device used to acquire image data from different ultrasound modalities. For example, the first and second ultrasound modality breast images are obtained by scanning the breast with the ultrasound probe of a color Doppler ultrasound device.

[0074] The output device is used to present results to doctors or operators, such as the location of suspicious areas, tumor confidence levels, and other relevant information. Specifically, the output device may be a computer display screen, a printer, or other graphical user interface device.

[0075] The memory is used to store image data, computer programs, and intermediate results. This includes storing raw breast images (first and second ultrasound modal breast images), processed images, various models, and related data. The memory capacity typically needs to be large enough to store a large number of high-resolution images.

[0076] The multimodal ultrasound imaging breast tumor identification system provided by this invention combines different ultrasound imaging modes. With the help of computer programs and processors, through multimodal imaging and computer-aided analysis, it can provide more accurate and reliable tumor diagnosis results. Doctors can more accurately determine the nature of the tumor and provide earlier diagnosis and more effective treatment plans.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the claims and specification of the present invention.

Claims

1. A breast tumor identification system based on multimodal ultrasound imaging, characterized in that, The system includes a memory for storing a computer program, the computer program comprising program instructions that, when executed by a processor, cause the processor to perform the following steps: A first ultrasound modal breast image is acquired, and a first suspicious area of ​​breast tumor is located in the first ultrasound modal breast image. The first ultrasound modal breast image is a grayscale breast image. A second ultrasound modal breast image is acquired and aligned with the first ultrasound modal breast image. The second ultrasound modal breast image is an ultramicrovascular breast image. Based on the alignment results, the second suspicious area of ​​breast tumor, corresponding to the first suspicious area of ​​breast tumor, is located in the second ultrasound modality breast image; The credibility of the breast tumor in the suspected area of ​​the second breast tumor is identified based on the distribution of microvessels in that area. The step of locating the first suspicious area of ​​a breast tumor in the first ultrasound modality breast image includes the following steps: Set a target resolution, and coordinate the first ultrasound modal breast image based on the target resolution, and obtain the coordinates and grayscale value of each pixel in the coordinated first ultrasound modal breast image; Set the size of the initial positioning frame, and define different judgment areas by adjusting the center coordinates of the initial positioning frame; Based on the gray value distribution of the ultrasound breast image in the judgment area, the first suspicious area of ​​breast tumor in the first ultrasound modality breast image is located; Locating the suspected area of ​​a first breast tumor in the first ultrasound modality breast image further includes the following steps: Different judgment areas can be defined by adjusting the size of the initial positioning box; The step of locating the first suspicious area of ​​breast tumor in the first ultrasound modality breast image based on the gray value distribution of the ultrasound breast image in the judgment area includes the following steps: Obtain the global grayscale mean of the first ultrasound modality breast image, the global grayscale mean... Satisfy the following formula: ,in, This represents the maximum x-coordinate value. This represents the maximum ordinate value. Indicates coordinates as The pixel grayscale value; A threshold for suspicious location coordinate pixels is set based on the global grayscale mean, and the threshold for suspicious location coordinate pixels satisfies the following formula: ,in, , Represents the threshold coefficient. This represents the maximum grayscale value in the first ultrasound modality breast image. This indicates that the value exceeds the global grayscale mean in the first ultrasound modality breast image. And the grayscale value that appears most frequently; Based on the grayscale value distribution of the ultrasound breast image in the judgment area, locate the coordinates of the target pixel in the judgment area whose grayscale value is greater than the suspected location coordinate pixel threshold; The target pixel coordinates are aggregated to form one or more initial suspicious regions, wherein the initial suspicious region is a closed region formed by the edges of multiple adjacent target pixels; A suspicious area location model is constructed, and the first center coordinates of the initial suspicious area are located using the suspicious area location model. The suspicious area location model satisfies the following formula: ,in, Indicates the first in the judgment region The first center coordinates of the initial suspicious area Indicates the first The minimum x-coordinate in the initial suspicious region Indicates the first The maximum x-coordinate in the initial suspicious region Indicates the first The minimum ordinate in the initial suspicious region Indicates the first The maximum ordinate in the initial suspicious region; The first suspicious area of ​​the breast tumor in the first ultrasound modality breast image is located using the first center coordinates of the initial suspicious area. Any first suspicious area of ​​the breast tumor includes all pixel coordinates of the corresponding initial suspicious area.

2. The breast tumor identification system based on multimodal ultrasound imaging according to claim 1, characterized in that, The process of acquiring the first ultrasound modality breast image includes the following steps: providing a B-mode ultrasound device and using the B-mode ultrasound device to scan the target breast to acquire the first ultrasound modality breast image.

3. The breast tumor identification system based on multimodal ultrasound imaging according to claim 1, characterized in that, The acquisition of the second ultrasound modality breast image includes the following steps: providing a color Doppler ultrasound device and using the color Doppler ultrasound device to scan the target breast to obtain the second ultrasound modality breast image.

4. The breast tumor identification system based on multimodal ultrasound imaging according to claim 1, characterized in that, The step of locating the second suspicious breast tumor region in the second ultrasound modality breast image, corresponding to the first suspicious breast tumor region, based on the alignment result, includes the following steps: Based on the alignment results, the second center coordinates corresponding to the first center coordinates are marked in the second ultrasound modality breast image, and the pixel coordinates associated with the second center coordinates are also marked. Based on the center coordinates in the second ultrasound modality breast image and the pixel coordinates associated with the center coordinates, one or more second breast tumor suspicious regions are generated. Each second breast tumor suspicious region includes a second center coordinate and all pixel coordinates associated with the second center coordinate.

5. The breast tumor identification system based on multimodal ultrasound imaging according to claim 4, characterized in that, The step of locating the second suspicious breast tumor region in the second ultrasound modality breast image, corresponding to the first suspicious breast tumor region, based on the alignment result, further includes the following steps: A tumor area threshold is set, and a second suspicious area of ​​breast tumor is screened based on the suspicious tumor area threshold.

6. The breast tumor identification system based on multimodal ultrasound imaging according to claim 4, characterized in that, The step of identifying the credibility of a breast tumor in the suspected area of ​​a second breast tumor based on the distribution of microvessels in that area includes the following steps: Distinguish between the main blood vessels and microvessels in the suspected area of ​​the second breast tumor, and count the number of the main blood vessels and the number of microvessels respectively; A breast tumor confidence model is constructed, and the confidence level of breast tumors in a second suspected breast tumor region is identified by combining the number of main blood vessels and the number of microvessels with the breast tumor confidence model. The breast tumor confidence model satisfies the following formula: ,in, Indicates a suspicious area of ​​a second breast tumor. Credibility of breast tumors in the middle Indicates a suspicious area of ​​a second breast tumor. The number of main blood vessels in the middle, Indicates a suspicious area of ​​a second breast tumor. The Middle Mean curvature of the main blood vessels, This represents the weighting coefficient for the curvature of the main blood vessel. Indicates a suspicious area of ​​a second breast tumor. The Middle The number of microvessels corresponding to each main blood vessel. This represents the weighting coefficient for the bifurcation of the main blood vessel.

7. The breast tumor identification system based on multimodal ultrasound imaging according to any one of claims 1-6, characterized in that, It also includes a processor, an input device, an output device, and a memory interconnected thereto. The memory is used to store a computer program, which includes program instructions. The processor is configured to call and execute the program instructions.

Images