Mammary gland prostrate ultrasonic imaging method, device and equipment and storage medium

Abstract

The invention relates to the technical field of medical imaging, and discloses a mammary gland prostrating ultrasonic imaging method, device and equipment and a storage medium, the method is applied to prostrating position mammary gland ultrasonic equipment, the method comprises the steps that multiple pre-scanning images of target breast tissue are acquired, and the multiple pre-scanning images comprise images collected from multiple different angles of the target breast tissue; skin boundaries in the multiple pre-scanning images are obtained through identification; based on the skin boundary identified by each pre-scanning image, respectively acquiring a target ultrasonic emission angle vertical to the skin boundary; during formal scanning imaging, based on the target ultrasonic transmitting angle, transmitting ultrasonic perpendicular to the target breast tissue, and acquiring an ultrasonic image. In the mammary gland prostrate ultrasonic imaging process, the ultrasonic transmitting direction is perpendicular to the target breast tissue as much as possible, and the ultrasonic image quality is improved.

Description

Technical Field

[0001] The present invention relates to the field of medical imaging technology, and particularly relates to a prone breast ultrasound imaging method, device, equipment and storage medium. Background Art

[0002] Prone breast ultrasound imaging is an imaging examination method for evaluating breast tissue. Compared with traditional supine ultrasound, when performing prone ultrasound examination, the patient takes a prone position and the breasts hang naturally. This can provide different perspectives to observe the breast tissue structure. Especially for lesions deep near the chest wall or lymph nodes in the axillary region, clearer images can be provided.

[0003] The advantages of this imaging method include but are not limited to: More natural breast position: In the prone position, the breasts are not pressed against the chest wall and can hang more naturally on the examination table, which helps to better display the internal structure of the breasts.

[0004] Reduced artifacts: Due to the effect of gravity, the distribution of breast tissue is more natural, which may help reduce artifacts caused by compression and improve image quality.

[0005] Good visualization of deep tissues: Especially for the area near the chest wall and axillary lymph nodes, the prone position can provide better vision and resolution.

[0006] The incident angle of ultrasound will directly affect image quality and diagnostic accuracy. When ultrasonic waves are incident vertically (90 degrees) on an interface, the strongest reflection signal can be obtained because the energy loss of the reflected wave is the smallest at this time. Vertical incidence can also simplify the analysis of the reflected wave because it reduces the complexity caused by angle changes, such as refraction and scattering effects. Additionally, in some cases, vertical incidence helps to improve image resolution.

[0007] During the process of prone breast ultrasound imaging, since the breast tissue hangs naturally when the patient is in the prone position, its shape and position are significantly different from those in the conventional supine position, resulting in difficulty in standardizing the incident angle of the ultrasound probe. That is to say, it is difficult to achieve vertical incidence of ultrasound during prone breast ultrasound imaging. Summary of the Invention

[0008] In view of this, the present invention provides a prone breast ultrasound imaging method, device, equipment and storage medium to solve the problem that it is difficult to achieve vertical incidence in prone breast ultrasound imaging.

[0009] In a first aspect, the present invention provides a prone breast ultrasound imaging method, which is applied to a prone breast ultrasound device. The prone breast ultrasound device includes an ultrasound probe, and the method includes: Obtain multiple pre-scanned images of the target breast tissue, where the multiple pre-scanned images include images acquired from multiple different angles of the target breast tissue; Identify the skin boundaries in the multiple pre-scanned images respectively; Based on the skin boundaries identified in each of the pre-scanned images, obtain the target ultrasonic emission angles perpendicular to the skin boundaries respectively; During formal scanning and imaging, based on the target ultrasonic emission angles, emit ultrasonic waves perpendicular to the target breast tissue and acquire ultrasonic images.

[0010] In an alternative embodiment, the identifying the skin boundaries in the multiple pre-scanned images respectively includes: Perform contrast enhancement processing on the pre-scanned image to obtain a first image; Perform edge processing on the first image to obtain a second image; the edge processing is edge enhancement filtering processing or edge extraction processing; Perform binarization on the second image to obtain a binary image; Extract connected regions in the binary image, and filter the extracted connected regions according to region size, position, and / or shape to obtain remaining connected regions; Perform boundary curve fitting on the boundary points of the remaining connected regions to obtain a continuous skin boundary line.

[0011] In an alternative embodiment, the identifying the skin boundaries in the multiple pre-scanned images respectively includes: Use a deep learning model to identify the breast region in the pre-scanned image; Extract the contour of the identified breast region to obtain a skin boundary curve.

[0012] In an alternative embodiment, the emitting ultrasonic waves perpendicular to the target breast tissue based on the target ultrasonic emission angles during formal scanning and imaging includes: Based on the target ultrasonic emission angles corresponding to each of the pre-scanned images, determine the ultrasonic probe angles; Adjust the ultrasonic probe according to the determined ultrasonic probe angles and emit ultrasonic beams.

[0013] In an alternative embodiment, the emitting ultrasonic waves perpendicular to the target breast tissue based on the target ultrasonic emission angles during formal scanning and imaging includes: Based on the target ultrasonic emission angles corresponding to the pre-scanned image, determine the beam incident angle in the first target stage during the formal scanning process; Among them, the first target stage is determined according to the scanning angle. The starting scanning angle of the first target stage is the scanning angle corresponding to the pre-scanned image, and the ending scanning angle is the scanning angle of the next frame of pre-scanned image of the pre-scanned image.

[0014] In an alternative embodiment, determining the beam incident angle of the first target stage in the formal scanning process based on the target ultrasonic emission angle corresponding to the pre-scanned image includes: Taking the target ultrasonic emission angle corresponding to the pre-scanned image as the beam reference angle of the first target stage; In the first target stage of the formal scanning process, obtain the first ultrasonic image collected at the first moment; Identify the skin boundary in the first ultrasonic image; Based on the skin boundary in the first ultrasonic image, finely adjust the beam reference angle to obtain the beam incident angle, which is used as the beam incident angle at the second moment of the first target stage; the second moment is after the first moment and before a new beam incident angle is obtained.

[0015] In an alternative embodiment, the breast prone ultrasonic imaging method further includes: Obtain a first target ultrasonic image; the first target ultrasonic image is a formal ultrasonic image collected at the third moment in the formal scanning process, or is the pre-scanned image; Obtain the breast region, skin boundary, and various tissue recognition results in the first target ultrasonic image; Starting from the skin boundary, divide the breast region into multiple depth partitions according to the depth; Based on the tissue recognition results, obtain the tissue distribution information in each depth partition; Respectively determine the corresponding gain according to the tissue distribution information of the depth partition, and form a target time gain compensation curve; In the second target stage of the formal scanning process, adjust the gain of ultrasonic signal reception according to the target time gain compensation curve until the next stage; Among them, if the first target ultrasonic image is the formal ultrasonic image collected at the third moment, the starting moment of the second target stage is after the third moment, and the ending moment is at or after the time when a new target time gain compensation curve is obtained; if the first target ultrasonic image is the pre-scanned image, the second target stage is determined according to the scanning angle. The starting scanning angle of the second target stage is the scanning angle corresponding to the pre-scanned image, and the ending scanning angle is the scanning angle of the next frame of pre-scanned image of the pre-scanned image.

[0016] In an alternative embodiment, in the second target stage during the formal scanning process, adjusting the gain of ultrasonic signal reception according to the target time gain compensation curve includes: If the first target ultrasonic image is a pre-scan image, determine the starting position of the gain according to the position of the skin boundary in the second ultrasonic image, where the second ultrasonic image is a formal ultrasonic image acquired before the current moment; Adjust the gain of ultrasonic signal reception according to the starting position of the gain and the target time gain compensation curve.

[0017] In an alternative embodiment, the breast prone ultrasonic imaging method further includes: Obtain a second target ultrasonic image and the probe pose; wherein, the second target ultrasonic image is a third ultrasonic image acquired at the fourth moment during the formal scanning process, and correspondingly, the probe pose is the probe pose during the formal scanning process; or the second target ultrasonic image is the pre-scan image, and correspondingly, the probe pose is the probe pose during the pre-scan process or the probe pose during the formal scanning process; Based on the skin boundary in the second target ultrasonic image and the probe pose, determine the distance between the probe and the skin; According to the preset subcutaneous target imaging depth and the distance, determine the transmit focus point parameters and receive focus point parameters in the third target stage during the formal scanning process; if the second target ultrasonic image is the third ultrasonic image, the third target stage includes one or more moments after the fourth moment; if the second target ultrasonic image is the pre-scan image, the third target stage is determined according to the scanning angle, the starting scanning angle of the third target stage is the scanning angle corresponding to the pre-scan image, and the ending scanning angle is the scanning angle of the next frame of the pre-scan image of the pre-scan image.

[0018] In a second aspect, the present invention provides a breast prone ultrasonic imaging device, which is applied to a prone breast ultrasonic device, and the prone breast ultrasonic device includes an ultrasonic probe. The device includes: A pre-scan image acquisition module, configured to acquire multiple pre-scan images of a target breast tissue, where the multiple pre-scan images include images acquired from multiple different angles of the target breast tissue; A first skin boundary recognition module, configured to respectively recognize the skin boundaries in the multiple pre-scan images; A target ultrasonic emission angle acquisition module, configured to respectively acquire a target ultrasonic emission angle perpendicular to the skin boundary based on the skin boundary recognized in each pre-scan image; An ultrasonic imaging module, configured to emit ultrasonic waves perpendicular to the target breast tissue based on the target ultrasonic emission angle during formal scanning and imaging, and acquire ultrasonic images.

[0019] In a third aspect, the present invention provides a prone breast ultrasonic imaging device, comprising: An imaging assembly, the imaging assembly including an ultrasonic probe; A breast cup, configured to accommodate breast tissue; A probe driving assembly, configured to drive the probe to adjust the angle; A support assembly, configured to support the imaging assembly, the breast cup and the probe driving assembly; A memory and a processor, which are communicatively connected to each other. The memory stores computer instructions, and the processor executes the computer instructions to execute the prone breast ultrasonic imaging method according to the first aspect or any corresponding embodiment thereof.

[0020] In a fourth aspect, the present invention provides a computer-readable storage medium, on which computer instructions are stored. The computer instructions are used to cause a computer to execute the prone breast ultrasonic imaging method according to the first aspect or any corresponding embodiment thereof.

[0021] In a fifth aspect, the present invention provides a computer program product, including computer instructions, which are used to cause a computer to execute the prone breast ultrasonic imaging method according to the first aspect or any corresponding embodiment thereof.

[0022] The prone breast ultrasonic imaging method, device, equipment and storage medium provided by the embodiments of the present invention pre-scan the target breast tissue to be examined to obtain the skin boundaries of the target breast tissue at multiple angles, that is, obtain the external shape structure of the target breast tissue, so as to determine the ultrasonic emission angle in the formal scanning stage, so that the ultrasonic emission direction is as perpendicular as possible to the target breast tissue, improve the quality of ultrasonic images, and further improve the diagnostic accuracy. Description of the Drawings

[0023] In order to more clearly illustrate the specific embodiments of the present invention or the technical solutions in the related art, the following will briefly introduce the drawings required for use in the description of the specific embodiments or the related art. Obviously, the following drawings are some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0024] Figure 1 It is a schematic diagram of the imaging process of prone breast ultrasonic imaging according to an embodiment of the present invention; Figure 2It is a schematic flowchart of a prone breast ultrasound imaging method according to an embodiment of the present invention; Figure 3 It is a schematic diagram of a pre-scanning imaging angle according to an embodiment of the present invention; Figure 4 It is a schematic diagram of an ultrasound image; Figure 5 It is according to an embodiment of the present invention for Figure 4 One of the schematic diagrams for boundary recognition of the shown ultrasound image; Figure 6 It is according to an embodiment of the present invention for Figure 4 Another schematic diagram for boundary recognition of the shown ultrasound image; Figure 7 It is a schematic diagram of a probe angle according to an embodiment of the present invention; Figure 8 One of the schematic diagrams for fine-tuning the beam incident angle according to an embodiment of the present invention; Figure 9 Another schematic diagram for fine-tuning the beam incident angle according to an embodiment of the present invention; Figure 10 It is a structural block diagram of a prone breast ultrasound imaging device according to an embodiment of the present invention; Figure 11 It is a schematic diagram of the hardware structure of a prone breast ultrasound imaging device according to an embodiment of the present invention. Detailed implementation manners

[0025] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Apparently, the described embodiments are some, but not all, of the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0026] As Figure 1 shown, in the imaging process of prone breast ultrasound imaging, the ultrasound probe 102 rotates around the breast tissue 101 for one circle, and multiple frames of ultrasound images are respectively acquired during the rotation process (103 indicates a single-frame ultrasound image).

[0027] According to an embodiment of the present invention, an embodiment of a prone breast ultrasound imaging method is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of executable computer instructions, and although the logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in a different order from here.

[0028] In this embodiment, a prone breast ultrasound imaging method is provided, which can be used for a prone breast ultrasound device. The prone breast ultrasound device includes an ultrasound probe. Figure 2 is a flowchart of the prone breast ultrasound imaging method according to an embodiment of the present invention. As Figure 2 shown, the process includes the following steps: Step S201, obtain multiple pre-scanned images of the target breast tissue. The multiple pre-scanned images include images collected from multiple different angles of the target breast tissue.

[0029] Specifically, as Figure 1 and Figure 3 shown, during the prone breast ultrasound imaging process, the breast tissue to be examined is placed in the cup 301, and the ultrasound probe 102 (abbreviated as the probe) rotates around the cup 301, that is, around the breast tissue. The images collected when the ultrasound probe 102 rotates to different angles are the images collected from the above different angles.

[0030] The prone breast ultrasound imaging method provided by the embodiment of the present invention has two imaging stages. One is the pre-scanning stage, and the other is the formal scanning stage. In the pre-scanning stage, after the precise positioning of the breast tissue is completed, the ultrasound probe is controlled to rotate around the breast tissue for one circle. During this process, a small number of pre-scanned images are collected by sparse sampling. The number of pre-scanned images can be 4, for example, the images collected from the four angles 302 as Figure 3 shown. The number of pre-scanned images can also be 8, 16, 32, etc. There is no limitation on the number of pre-scanned images here, and it can be specifically determined according to actual situations such as imaging accuracy and processing performance of the device. The purpose of collecting pre-scanned images is to obtain the structural information (outer contour information, internal tissue distribution information, etc.) of the target breast tissue of the target patient, so as to provide a basis for the personalized setting of scanning parameters in the subsequent formal scanning stage and improve the quality of the ultrasound images collected in the formal scanning stage.

[0031] Step S202, respectively identify the skin boundaries in the multiple pre-scanned images.

[0032] Specifically, a deep learning algorithm or a traditional filtering algorithm can be used to identify the skin boundaries in the pre-scanned images.

[0033] The following will respectively illustrate the use of a deep learning algorithm and a traditional filtering algorithm to identify the skin boundaries in the pre-scanned images.

[0034] In some optional specific embodiments, step S202, that is, the respectively identifying the skin boundaries in the multiple pre-scanned images, includes: In step S2021, use a deep learning model to identify the breast region in the pre-scanned image.

[0035] Specifically, the deep learning model can be, for example, U-Net or Attention U-Net. U-Net is a special convolutional neural network (CNN). Its name comes from the U-shaped structure of the network architecture. The main features of U-Net include: 1. Encoder-decoder structure: U-Net consists of a downsampling path (encoder) and an upsampling path (decoder). The encoder is used to capture the context information of the image, while the decoder is used for precise localization. 2. Skip connections: In U-Net, there are direct connections between the encoder and the decoder, which are called "skip connections". These connections allow the decoder layer to access high-resolution feature maps, which is very important for accurate pixel-level prediction. Attention U-Net is an improvement of the original U-Net architecture by introducing an attention mechanism to enhance the performance of the model. The attention mechanism helps the network focus on important regions and ignore irrelevant background information, thus improving the segmentation accuracy.

[0036] During the process of breast region identification, the pre-scanned image can be input into the deep learning model, and the deep learning model can output the pixel mask of the breast region.

[0037] In step S2022, extract the contour of the identified breast region to obtain the skin boundary curve.

[0038] As Figure 4 shown in a pre-scanned image, its boundary has relatively obvious features. Therefore, as Figure 5 and Figure 6 shown, the deep learning segmentation network (i.e., the deep learning model) can easily distinguish the skin boundary.

[0039] Specifically, findContours() or cv2.Canny + cv2.approxPolyDP in OpenCV (Open Source Computer Vision Library, whose Chinese name is "Open Source Computer Vision Library", abbreviated as "Open Source CV Library") can be used to extract the contour and obtain the skin boundary curve. The skin boundary is the uppermost contour of the breast region or the boundary line closest to the probe.

[0040] Before using the deep learning model for skin boundary recognition, it needs to be trained. During the training process, a large number of breast tissue ultrasound images (not necessarily of the target breast tissue or the target patient) need to be collected as samples. Then, using manual expert annotation or semi-automatic auxiliary tools, pixel-level segmentation annotation of the breast regions in a large number of breast tissue ultrasound images is performed to form breast region mask labels. Finally, the deep learning model is trained using the breast tissue ultrasound image samples, and Dice Loss + BCE Loss (where Dice Loss can be called dice loss or Dice loss, and BCE Loss is binary cross-entropy loss) is used as the loss function to quantify the deviation between the recognition result of the model and the breast region mask label.

[0041] A skin boundary recognition method provided by an embodiment of the present invention is based on a deep learning semantic segmentation model (such as the U-Net series). The breast tissue region is trained as the target category, and the model outputs a breast region mask. The skin boundary line segment is extracted through the upper edge of the mask or the edge closest to the probe direction. This method has strong robustness and is applicable to automatic recognition scenarios with diverse body types, different image qualities, and complex tissue structures.

[0042] Skin boundary recognition will be used as the basis for setting the imaging parameters of the ultrasound device in the subsequent formal scanning stage. Through the above method, the embodiment of the present invention can accurately recognize the skin boundary in the pre-scan image, providing a reliable basis for accurately setting the imaging parameters subsequently.

[0043] In some other optional specific embodiments, step S202, that is, separately recognizing the skin boundaries in the multiple pre-scan images, includes: Step S202a, performing contrast enhancement processing on the pre-scan image to obtain a first image.

[0044] Specifically, histogram equalization (HE) or contrast limited adaptive histogram equalization (CLAHE) can be used to enhance the contrast of the pre-scan image to highlight the differences between tissue structures and boundaries.

[0045] Step S202b, performing edge processing on the first image to obtain a second image; the edge processing is edge enhancement filtering processing or edge extraction processing.

[0046] Specifically, a high-pass filter or a Laplacian operator can be applied to strengthen the high-frequency edges in the image and highlight the structural mutation regions. Or algorithms such as Canny edge detection can be used to extract the edges of the image.

[0047] Step S202c: Binarize the second image to obtain a binary image.

[0048] Specifically, the Otsu (i.e., Otsu's method) adaptive threshold algorithm or local adaptive threshold can be used to binarize the second image (i.e., the edge map) to separate the highlighted structure from the background.

[0049] In other embodiments, before binarization, contrast projection (such as taking the maximum brightness for each column or row, or performing a weighting operation) can also be used to enhance the structural features of the image, highlight the highlighted area or boundary, thereby assisting in determining the candidate boundary area.

[0050] Step S202d: Extract connected components in the binary image, and filter the extracted connected components according to the region size, position, and / or shape to filter out pseudo-boundaries and obtain the remaining connected components, that is, the candidate boundary area.

[0051] Step S202e: Fit the boundary points of the remaining connected components to obtain a continuous skin boundary line. Specifically, the Hough line / curve detection algorithm or spline interpolation can be applied for boundary curve fitting.

[0052] In the embodiments of the present invention, a traditional filtering algorithm is provided to identify the skin boundary in the pre-scanned image. Specifically, by performing contrast enhancement, edge detection, and binarization processing on the ultrasonic image, and combining traditional image processing processes such as region screening and curve fitting, the highlighted boundary line at the junction of the skin and the ultrasonic coupling medium is extracted.

[0053] Step S203: Based on the skin boundary recognized from each pre-scanned image, respectively obtain the target ultrasonic emission angle perpendicular to the skin boundary. The perpendicular mentioned here does not require absolute perpendicularity, and being close to perpendicular is also acceptable.

[0054] Specifically, in the embodiments of the present invention, the incident optimal angle can be estimated based on the tangent direction or normal direction of the skin boundary (for example, it can be the tangent direction or normal direction corresponding to the point on the skin boundary closest to the probe). Specifically, the normal direction of the skin boundary in the pre-scanned image can be used as the target ultrasonic emission angle. This is only an example, and an angle determined by other geometric methods that can make the ultrasonic emission direction perpendicular to the target breast tissue can also be used as the target ultrasonic emission angle.

[0055] Step S204: During the formal scanning and imaging, based on the target ultrasonic emission angle, emit ultrasonic waves perpendicular to the target breast tissue and collect ultrasonic images.

[0056] Adjusting the ultrasonic emission angle can be achieved either by adjusting the angle of the ultrasonic probe or by adjusting the emission direction of the ultrasonic beam. Of course, it can also be a combination of adjusting the probe angle and the beam angle. The following will separately illustrate the adjustment of the probe angle and the beam angle of the ultrasonic probe.

[0057] In some optional specific embodiments, step S204, that is, when performing formal scanning and imaging, based on the target ultrasonic emission angle, emitting ultrasonic waves perpendicular to the target breast tissue, includes: Step S2041, based on the target ultrasonic emission angle corresponding to each pre-scanning image, determine the ultrasonic probe angle. For example, the average value or weighted average value of multiple target ultrasonic emission angles (determining the corresponding weight value according to the occurrence times of each target ultrasonic emission angle) can be calculated to obtain the ultrasonic probe angle. Additionally, the target ultrasonic emission angle with the most occurrences can also be selected as the ultrasonic probe angle, or the median value of the target ultrasonic emission angles can be used as the ultrasonic probe angle.

[0058] Step S2042, adjust the ultrasonic probe according to the determined ultrasonic probe angle and emit an ultrasonic beam.

[0059] In the embodiment of the present invention, as Figure 7 shown, based on the extracted skin boundary, by fitting the tangent direction 701 of the boundary curve, calculate the angle between the probe and the skin surface in each direction ( Figure 7 the angle between the tangent direction 701 of the skin boundary curve and the probe surface 702 in the figure), and accordingly derive a comprehensive "standard skin angle vector". This vector is used to calculate the globally optimal probe angle (corresponding to 703 in the figure) to achieve the maximum degree of perpendicular incidence, improve the acoustic coupling effect and the initial image quality.

[0060] Specifically, the ultrasonic device can automatically adjust to the calculated optimal probe angle through controlling the probe driving structure (for example, it can be a robotic arm) and lock it to ensure the stability and consistency of subsequent scanning. This angle setting serves as the reference incident angle for the entire inspection process and remains unchanged during all image acquisitions in the subsequent formal scanning stage, ensuring imaging consistency, comparability, and reconstruction accuracy.

[0061] In some optional specific embodiments, step S204, that is, when performing formal scanning and imaging, based on the target ultrasonic emission angle, emitting ultrasonic waves perpendicular to the target breast tissue, includes: Step S204a, based on the target ultrasonic emission angle corresponding to the pre-scanning image, determine the beam incident angle in the first target stage during the formal scanning process; Among them, the first target stage is determined according to the scanning angle. The starting scanning angle of the first target stage is the scanning angle corresponding to the pre-scanned image, and the ending scanning angle is the scanning angle of the next pre-scanned image of the pre-scanned image (the scanning angle of the first target stage may not include the scanning angle of the next pre-scanned image of the pre-scanned image).

[0062] Specifically, step S204a, that is, determining the beam incident angle of the first target stage in the formal scanning process based on the target ultrasonic emission angle corresponding to the pre-scanned image, includes: Step S204a1, taking the target ultrasonic emission angle corresponding to the pre-scanned image as the beam reference angle of the first target stage.

[0063] That is to say, during the formal scanning process, when the scanning angle is between the scanning angles corresponding to two adjacent pre-scanned images, the target ultrasonic emission angle corresponding to the earlier one of the two adjacent pre-scanned images is taken as the beam reference angle. In other embodiments, the beam reference angle corresponding to the current formal scanning angle can also be estimated based on the target ultrasonic emission angles corresponding to the adjacent pre-scanned images. For example, it can be estimated by interpolation.

[0064] Step S204a2, in the first target stage during the formal scanning process, acquiring the first ultrasonic image collected at the first moment.

[0065] Step S204a3, identifying the skin boundary in the first ultrasonic image.

[0066] Step S204a4, based on the skin boundary in the first ultrasonic image, finely adjusting the beam reference angle to obtain the beam incident angle, which is used as the beam incident angle at the second moment of the first target stage. For example, the normal direction of the point on the skin boundary closest to the probe in the first ultrasonic image can be used as the final beam incident angle. The second moment is after the first moment and before a new beam incident angle is obtained.

[0067] Specifically, the beam incident angle can be finely adjusted based on each frame of the formal scanning image to obtain the beam incident angle for the acquisition of the next frame of the image, or the beam incident angle can be finely adjusted every few frames, that is, the beam incident angle is finely adjusted based on the current frame of the formal scanning image to obtain the beam incident angle for the acquisition of the subsequent few frames of the image.

[0068] In the embodiments of the present invention, during the formal scanning process, to adapt to problems such as the complex breast shape, large skin curvature change, and strong directionality of the internal structure of the organ, the ultrasonic device integrates a fine-level beam angle dynamic fine-tuning algorithm to compensate for the limitation of the fixed physical probe angle and achieve more flexible directional scanning control.

[0069] Specifically, please refer to Figure 8 and Figure 9 , first, the target ultrasonic emission angle 802 estimated based on the pre-scanned image is used as the beam reference angle, which is nearly perpendicular to the emission surface 801 of the probe after angle adjustment. Then, the beam incident angle 803 is obtained after fine-tuning based on the formal scanned image. For the convenience of representation in the figure, the fine-tuning angle is relatively large, but the actual fine-tuning angle is generally small. In the embodiment of the present invention, the beam angle is automatically adjusted within a certain range to improve the imaging effect in different curvature regions. With the slight fluctuations in the geometric shape of each part of the breast, the ultrasonic device can fine-tune the beam emission angle within a small angle range of ±θ (the value range of θ can be, for example, 2° to 5°) near each scan line to ensure that the sound beam is always perpendicular or approximately perpendicular to the skin and the target structure, thereby improving the reflectivity and signal-to-noise ratio.

[0070] The prone breast ultrasonic imaging method provided in this embodiment obtains the skin boundaries of the target breast tissue at multiple angles by pre-scanning the target breast tissue to be examined, that is, obtains the external shape structure of the target breast tissue, so as to determine the ultrasonic emission angle in the formal scanning stage, making the ultrasonic emission direction as perpendicular as possible to the target breast tissue, improving the quality of the ultrasonic image, and further improving the diagnostic accuracy.

[0071] In the embodiment of the present invention, based on the skin boundary recognition result of the pre-scanned image, the ultrasonic emission angle in the subsequent formal scanning process is determined. In addition, in the formal scanning stage, dynamic gain adjustment (ATGC) and focus correction are also required based on the skin boundary recognition result. Therefore, after obtaining the skin boundary recognition result of the pre-scanned image, the ultrasonic device can cache and structurally store it for calling in the subsequent formal scanning process.

[0072] The following is an example to illustrate the method for gain determination and focus correction in the formal scanning stage.

[0073] In some optional specific embodiments, the prone breast ultrasonic imaging method further includes: Step 1, obtaining a first target ultrasonic image; the first target ultrasonic image is a formal ultrasonic image collected at the third moment during the formal scanning process, or is the pre-scanned image.

[0074] The third moment is one moment or multiple moments during the formal scanning process.

[0075] Step 2, obtaining the breast region, skin boundary, and various tissue recognition results in the first target ultrasonic image.

[0076] Specifically, if the first target ultrasound image is a pre-scan image, then after obtaining the pre-scan image through pre-scanning, the breast region, skin boundary, and various tissue recognition results therein can be identified. For the recognition of the breast region and skin boundary, please refer to the above embodiments and will not be elaborated here.

[0077] If the first target ultrasound image is a formal scan image during the formal scanning process, then after obtaining the formal scan image, it is recognized to obtain the breast region, skin boundary, and various tissue recognition results therein. For the method of recognizing the breast region and skin boundary of the formal scan image, please refer to the process of recognizing the pre-scan image in the above embodiments and will not be elaborated here.

[0078] Regarding tissue recognition, it can be specifically recognized based on gray-scale features. It can also be recognized by means of a deep learning model to generate a tissue distribution map. Specific tissue recognition results include fat, gland, and / or connective tissue, etc.

[0079] The following is an example to illustrate the recognition of tissues in the first target ultrasound image by means of a deep learning model.

[0080] First, preprocess the original first target ultrasound image (noise suppression and enhancement processing), including: Gaussian filtering or median filtering to remove speckle noise; contrast stretching and histogram equalization to enhance the tissue boundary contrast; normalization processing to unify the gray-scale range for subsequent processing by the deep learning model.

[0081] Secondly, use a tissue segmentation model for inference. Use a medical image segmentation model based on a convolutional neural network (CNN) to perform pixel-level classification on different types of tissues in the image. Typical network structures can include U-Net, ResUNet, DeepLabv3+, or nnU-Net; the input of the model is the preprocessed ultrasound image, and the output is a tissue probability map with multiple channels, and each channel represents the distribution probability of adipose tissue, glandular tissue, connective tissue, etc.; an argmax operation is performed on the output result to obtain the final tissue category mask.

[0082] Finally, according to the tissue mask image, perform quantitative statistics on the distribution of different tissue types in the image. Statistically calculate the depth range (pixel index / mm) occupied by each type of tissue in the image; mark the central region and boundary of each type of tissue; construct partitions (such as superficial, middle, and deep layers) according to the depth positions where each tissue type is located.

[0083] Step 3, starting from the skin boundary, divide the breast region into multiple depth partitions according to depth, for example, it can be divided into 5 or 8 partitions.

[0084] Step 4: Based on the tissue recognition result, obtain the tissue distribution information within each of the depth partitions.

[0085] Specifically, the main tissue type within a depth partition can be determined according to the image proportion within the depth partition.

[0086] Step 5: Determine the corresponding gain for each depth partition according to the tissue distribution information of the depth partition, and form a target time gain compensation curve.

[0087] Specifically, the gain corresponding to a depth partition can be determined according to the main tissue type within the depth partition. For example, the gain of various tissue types can be determined by referring to Table 1 below, and thus the gain of each partition can be determined: Table 1 Suggestions for Gain Adjustment of Each Tissue Type

[0088] Specifically, different tissue types have different degrees of acoustic wave attenuation, usually as follows: Adipose tissue: low attenuation, small required gain; Glandular tissue, medium attenuation, medium gain required; Connective tissue / dense gland, high attenuation, high gain required.

[0089] Of course, when determining the gain of each depth partition, it is still necessary to determine it in combination with the depth information. For example, the gain of each depth partition can be preset according to the depth information first, and then the final gain of the depth partition can be adjusted according to the tissue conditions within the depth partition.

[0090] After the gain of each depth partition is determined, the target time gain compensation curve can be generated. Specifically, the gain values of each depth partition can be interpolated into a complete TGC (Time-Gain Compensation) curve, that is, the target time gain compensation curve is obtained.

[0091] The goal of TGC is to compensate for the attenuation of the ultrasonic signal during propagation in tissue as the depth increases, so that the brightness of the entire image is relatively consistent at different depths. Traditional TGC adjustment often sets the gain of different depth segments through a manual slider or compensates through a fixed linear / exponential function. In the embodiment of the present invention, through the recognition of the tissue type in different depth regions of the image, starting from the "attenuation characteristics of the tissue itself", a personalized TGC curve is set for each depth partition. In other words, it is not simply linearly increasing according to the depth, but non-linearly preset according to the "tissue absorption coefficient / attenuation characteristics".

[0092] In other embodiments, the gain of the depth partition may not be determined according to the main tissue type within the depth partition. Instead, first determine the area ratio of each tissue type within the depth partition and the attenuation coefficient of each tissue type, then determine the weight value of each tissue type according to the area ratio, and perform weighted averaging of the attenuation coefficients according to the weight values. Finally, determine the gain of the depth partition according to the attenuation coefficient obtained by weighted averaging.

[0093] Step 6, in the second target stage during the formal scanning process, adjust the gain of the ultrasonic signal reception according to the target time gain compensation curve until the next stage.

[0094] Wherein, if the first target ultrasonic image is the formal ultrasonic image acquired at the third moment, the start moment of the second target stage is after the third moment, and the end moment is at or after the time when the new target time gain compensation curve is obtained. That is to say, in the embodiments of the present invention, during the formal scanning process, the target time gain compensation curve is updated once in each stage. Each stage may have a duration of only one frame or multiple frames. Specifically, during the formal scanning process, the target time gain compensation curve can be generated in real time based on the ultrasonic image of the current frame and used as the time gain compensation curve for the acquisition of the next frame of ultrasonic image. If there is not enough time, it can also be used as the time gain compensation curve for the acquisition of the next next frame of ultrasonic image, or even as the time gain compensation curve for the acquisition of the ultrasonic image after an interval of 2 frames. In addition, during the formal scanning process, the target time gain compensation curve can also be generated in real time based on the ultrasonic image of the current frame and used as the time gain compensation curve for the acquisition of subsequent consecutive multiple frames of ultrasonic images. In other words, a strategy of predicting the next n frames based on the previous frame is adopted.

[0095] If the first target ultrasonic image is the pre-scanning image, the second target stage is determined according to the scanning angle. The start scanning angle of the second target stage is the scanning angle corresponding to the pre-scanning image, and the end scanning angle is the scanning angle of the next frame of pre-scanning image of the pre-scanning image. For example, as Figure 3 shown, if only pre-scanning images are acquired at four scanning angles of up, down, left, and right during the pre-scanning process, then during the formal scanning process, if the probe turns to the upper angle, the ultrasonic image is acquired with the time gain compensation curve corresponding to the pre-scanning image at that angle, and during the subsequent scanning process, the time gain compensation curve is always used to acquire the ultrasonic image until the probe turns to the scanning angle of the next pre-scanning image, such as the right angle, and then the ultrasonic image is acquired with the time gain compensation curve corresponding to the pre-scanning image at that angle.

[0096] In summary, the embodiments of the present invention provide two methods for determining the ultrasonic reception gain. One is to determine the gain based on the pre-scan image, and the other is to adjust the gain based on the image collected during the formal scan. The method of adjusting the gain based on the image collected during the formal scan realizes the dynamic adaptive time gain compensation (ATGC) mechanism, that is, it realizes the dynamic adjustment of the gain curve of the scan line, and optimizes the gray-scale performance and tissue contrast of the entire image.

[0097] In other embodiments, if the gain is determined based on the pre-scan image, in addition to the above technical solution (when the formal scan angle is between the scan angles of two adjacent pre-scan images, the time gain compensation curve remains unchanged), the ultrasonic image between two adjacent scan images can also be calculated by interpolation or other methods, and then a new time gain compensation curve is determined based on the calculated ultrasonic image, so that when the formal scan angle is between the scan angles of two adjacent pre-scan images, the time gain compensation curve can also be updated, improving the imaging quality of the formal scan image.

[0098] In some specific embodiments, step 6 above, that is, in the second target stage during the formal scan, adjusting the gain of the ultrasonic signal reception according to the target time gain compensation curve, includes: Step 61, if the first target ultrasonic image is a pre-scan image, determine the starting position of the gain according to the position of the skin boundary in the second ultrasonic image, where the second ultrasonic image is a formal ultrasonic image collected before the current moment. Specifically, the second ultrasonic image can be the previous frame of the formal ultrasonic image or a frame of the formal ultrasonic image separated by one or more frame durations.

[0099] If the first target ultrasonic image is a pre-scan image, considering that the angle of the probe changes relative to the pre-scan stage during the formal scan stage, that is, the pose of the probe changes relative to the pre-scan stage, therefore, the distance between the probe and the skin boundary changes, so it is necessary to determine the starting position of the gain according to the skin interface in the ultrasonic image collected during the formal scan stage, so as to improve the accuracy of the starting position of the gain, and further improve the quality of the ultrasonic image.

[0100] Of course, in some other embodiments, the starting position of the gain can also be determined according to the position of the skin boundary in the pre-scan image.

[0101] Step 62, adjust the gain of the ultrasonic signal reception according to the starting position of the gain and the target time gain compensation curve.

[0102] In addition, if the first target ultrasound image is an official ultrasound image acquired at the third moment during the formal scanning process, then the starting point of the gain can be directly determined based on the position of the skin boundary in the official ultrasound image.

[0103] The above technical solution for determining the target time gain compensation curve not only effectively compensates for the attenuation caused by the increase in depth during the acoustic wave transmission process during imaging, but also takes into account the different reflection characteristics caused by tissue structure differences. In addition, the embodiment of the present invention also identifies the skin boundary recognition result based on real-time image analysis, thereby identifying the skin starting point of the scanning path and setting this position as the starting point of the TGC curve, that is, realizing dynamic adjustment of the starting point position of the gain, rather than using a unified distance from the probe surface to the breast tissue to fix the starting point position of the gain. The embodiment of the present invention can avoid excessive gain under the conditions of containing non-uniform water medium or the probe not being in contact with the skin.

[0104] In addition, in the embodiment of the present invention, the ultrasound device can also refer to the tissue distribution data (such as fat area, gland area, lesion area) obtained in the pre-scanning stage, as well as the real-time echo intensity distribution trend, and perform adaptive slope fine-tuning in the deep part of the image to realize the linkage adjustment of the gain curve slope and tissue type.

[0105] In some optional specific embodiments, the breast prone ultrasound imaging method further includes: Step I, obtaining a second target ultrasound image and the probe pose; wherein, the second target ultrasound image is the third ultrasound image acquired at the fourth moment during the formal scanning process, and correspondingly, the probe pose is the probe pose during the formal scanning process; or the second target ultrasound image is the pre-scanning image, and correspondingly, the probe pose is the probe pose during the pre-scanning process or the probe pose during the formal scanning process.

[0106] Step II, determining the distance between the probe and the skin based on the skin boundary in the second target ultrasound image and the probe pose. Specifically, this distance can be a vertical distance or a distance along the beam emission direction.

[0107] Step III, determining the transmit focus point parameter and the receive focus point parameter of the third target stage during the formal scanning process according to the preset subcutaneous target imaging depth and the distance.

[0108] Specifically, if the second target ultrasound image is the third ultrasound image, the third target stage includes one or more moments after the fourth moment. If the second target ultrasound image is the pre-scan image, the third target stage is determined according to the scanning angle. The starting scanning angle of the third target stage is the scanning angle corresponding to the pre-scan image, and the ending scanning angle is the scanning angle of the next frame of the pre-scan image (the third target stage may not include this ending scanning angle).

[0109] In the embodiments of the present invention, according to the probe pose and the skin boundary position, the distance between the probe and the skin is dynamically measured to obtain the real-time water medium thickness. This parameter directly determines the path required for propagation from the probe surface to the skin surface layer, providing a key reference for the subsequent calibration of the focusing depth. The embodiments of the present invention provide two schemes to determine the focusing point parameters, one is based on the pre-scan image, and the other is based on the formal scan image.

[0110] Specifically, regarding the preset subcutaneous target imaging depth, the ultrasound device can set the subcutaneous target imaging depth (such as 5 mm under the skin) based on the tissue layer position where the doctor expects to obtain the clearest image, and use this as a fixed focusing target. Then, based on the water thickness data (i.e., the distance), the focusing path from the probe to the preset subcutaneous target imaging depth can be automatically calculated, and the transmitting focus point and receiving focus point parameters can be dynamically set to ensure that the beam energy accurately converges in this area.

[0111] In the embodiments of the present invention, to balance image stability and calculation efficiency, the focus position can be updated frame by frame, or updated every n frames using a sliding window strategy. In areas where tissue changes are slow, a low-frequency update strategy is adopted to reduce resource consumption.

[0112] In the embodiments of the present invention, during the formal scanning stage, to ensure that the key structures are in the best imaging focusing area, the ultrasound device adjusts the dynamic focus position for each frame of the image. The goal is to always maintain the focus at a specific depth under the skin (such as 5 mm or 8 mm under the skin), so as to achieve stable focused imaging of the key shallow structures. To achieve this goal, the ultrasound device comprehensively considers factors such as the skin boundary position, the water medium thickness (i.e., the distance between the probe and the skin boundary), and the tissue structure distribution, and finely adjusts the beam focusing parameters.

[0113] The dynamic fine-tuning of beam direction and dynamic fine-tuning of focus depth mentioned in the above embodiments of the present invention can be used alone or in combination. That is to say, in a breast prone ultrasound imaging embodiment, only dynamic fine-tuning of beam direction can be applied, or only dynamic fine-tuning of focus depth can be applied, or both dynamic fine-tuning of beam direction and dynamic fine-tuning of focus depth can be applied simultaneously. In areas with significant curvature or edge structures, it is recommended to fine-tune the linkage beam emission angle and focusing parameters so that the focus always falls on the target depth, while offsetting the offset effect caused by the angle change to ensure imaging continuity and clarity.

[0114] Through the above method, the ultrasound device does not need to calculate the tissue depth or automatically determine the imaging layer. Instead, it always performs focus positioning and adaptive adjustment around the preset subcutaneous depth, ensuring high consistency of imaging quality in key areas in complex scanning paths and improving diagnostic reliability.

[0115] In summary, the prone breast ultrasound imaging method provided by the embodiment of the present invention, in the formal image acquisition stage, based on the parameter template established in the pre-scan stage, combined with the real-time feedback mechanism, dynamically controls the gain curve, beam direction, focusing depth, etc. in the image acquisition process to improve the imaging quality and consistency.

[0116] In this embodiment, a breast prone ultrasound imaging device is also provided, which is used to implement the above-mentioned embodiments and preferred embodiments, and the descriptions that have been made will not be repeated. As used below, the term "module" can implement a combination of software and / or hardware for a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, the implementation of hardware, or a combination of software and hardware, is also possible and conceivable.

[0117] This embodiment provides a breast prone ultrasound imaging device, which is applied to a breast ultrasound device in a prone position, such as Figure 10 As shown, the device comprises: A pre-scan image acquisition module 1001 is used to acquire multiple pre-scan images of target breast tissue, wherein the multiple pre-scan images include images acquired from multiple different angles of the target breast tissue; A first skin boundary recognition module 1002, used to respectively recognize skin boundaries in the plurality of pre-scanned images; A target ultrasound emission angle acquisition module 1003 is used to respectively acquire a target ultrasound emission angle perpendicular to the skin boundary based on the skin boundary identified in each of the pre-scan images; The ultrasound imaging module 1004 is used to transmit ultrasound perpendicular to the target breast tissue based on the target ultrasound transmission angle during formal scanning and imaging, and to acquire an ultrasound image.

[0118] In some alternative embodiments, the first skin boundary recognition module 1002 includes: An enhancement processing unit, configured to perform contrast enhancement processing on the pre-scanned image to obtain a first image; An edge processing unit, configured to perform edge processing on the first image to obtain a second image; the edge processing is edge enhancement filtering processing or edge extraction processing; A binarization unit, configured to binarize the second image to obtain a binary image; A connected region extraction and filtering unit, configured to extract connected regions in the binary image, and filter the extracted connected regions according to region size, position, and / or shape to obtain remaining connected regions; A curve fitting unit, configured to perform boundary curve fitting on the boundary points of the remaining connected regions to obtain a continuous skin boundary line.

[0119] In some alternative embodiments, the first skin boundary recognition module 1002 includes: An intelligent recognition unit, configured to use a deep learning model to perform breast region recognition on the pre-scanned image; A contour extraction unit, configured to extract a contour of the recognized breast region to obtain a skin boundary curve.

[0120] In some alternative embodiments, the ultrasonic imaging module 1004 includes: An ultrasonic probe angle determination unit, configured to determine an ultrasonic probe angle based on the target ultrasonic emission angle corresponding to each pre-scanned image; An ultrasonic probe angle adjustment unit, configured to adjust the ultrasonic probe according to the determined ultrasonic probe angle and emit an ultrasonic beam.

[0121] In some alternative embodiments, the ultrasonic imaging module 1004 is specifically configured to determine an incident angle of a beam in a first target stage during a formal scanning process based on the target ultrasonic emission angle corresponding to the pre-scanned image; Wherein, the first target stage is determined according to a scanning angle, a starting scanning angle of the first target stage is the scanning angle corresponding to the pre-scanned image, and an ending scanning angle is the scanning angle of the next frame of pre-scanned image corresponding to the pre-scanned image.

[0122] In some alternative embodiments, the ultrasonic imaging module 1004 includes: A beam reference angle determination unit, configured to use the target ultrasonic emission angle corresponding to the pre-scanned image as the beam reference angle of the first target stage; A first ultrasonic image acquisition unit, configured to acquire a first ultrasonic image collected at a first moment during the first target stage in the formal scanning process; An identification unit, configured to identify the skin boundary in the first ultrasonic image; A fine-tuning unit, configured to fine-tune the beam reference angle based on the skin boundary in the first ultrasonic image to obtain the beam incident angle, which is used as the beam incident angle at a second moment in the first target stage; the second moment is after the first moment and before a new beam incident angle is obtained.

[0123] In some alternative embodiments, the breast prone ultrasonic imaging device further includes: A first target ultrasonic image acquisition module, configured to acquire a first target ultrasonic image; the first target ultrasonic image is a formal ultrasonic image collected at a third moment in the formal scanning process or the pre-scanning image; An image recognition result acquisition module, configured to acquire the breast region, skin boundary, and various tissue recognition results in the first target ultrasonic image; A partitioning module, configured to divide the breast region into multiple depth partitions according to the depth starting from the skin boundary; A tissue distribution situation acquisition module, configured to acquire the tissue distribution information in each depth partition based on the tissue recognition result; A gain curve generation module, configured to determine the corresponding gain respectively according to the tissue distribution information of the depth partition and form a target time gain compensation curve; A gain adjustment module, configured to adjust the gain of ultrasonic signal reception according to the target time gain compensation curve during the second target stage in the formal scanning process until the next stage; Wherein, if the first target ultrasonic image is the formal ultrasonic image collected at the third moment, the starting moment of the second target stage is after the third moment, and the ending moment is at or after the time when a new target time gain compensation curve is obtained; if the first target ultrasonic image is the pre-scanning image, the second target stage is determined according to the scanning angle, the starting scanning angle of the second target stage is the scanning angle corresponding to the pre-scanning image, and the ending scanning angle is the scanning angle of the next frame of pre-scanning image of the pre-scanning image.

[0124] In some alternative embodiments, the gain adjustment module includes: A gain starting point determination unit, configured to determine the starting point position of the gain according to the position of the skin boundary in the second ultrasonic image if the first target ultrasonic image is a pre-scanning image, and the second ultrasonic image is a formal ultrasonic image collected before the current moment; A gain adjustment unit for adjusting the gain of ultrasonic signal reception according to the starting position of the gain and the target time gain compensation curve.

[0125] In some alternative embodiments, the breast prone ultrasonic imaging device further includes: An acquisition module for acquiring a second target ultrasonic image and a probe pose; wherein, the second target ultrasonic image is a third ultrasonic image acquired at a fourth moment during the formal scanning process, and correspondingly, the probe pose is the probe pose during the formal scanning process; or the second target ultrasonic image is the pre-scanning image, and correspondingly, the probe pose is the probe pose during the pre-scanning process or the probe pose during the formal scanning process; A distance determination module for determining the distance between the probe and the skin based on the skin boundary in the second target ultrasonic image and the probe pose; A focal depth determination module for determining the transmit focus point parameters and receive focus point parameters of the third target stage during the formal scanning process according to a preset subcutaneous target imaging depth and the distance; if the second target ultrasonic image is the third ultrasonic image, the third target stage includes one or more moments after the fourth moment; if the second target ultrasonic image is the pre-scanning image, the third target stage is determined according to the scanning angle, and the starting scanning angle of the third target stage is the scanning angle corresponding to the pre-scanning image, and the ending scanning angle is the scanning angle of the next frame of pre-scanning image of the pre-scanning image.

[0126] The further function descriptions of the above-mentioned various modules and units are the same as those in the corresponding embodiments above, and will not be elaborated here.

[0127] The breast prone ultrasonic imaging device in this embodiment is presented in the form of functional units. Here, the unit refers to an ASIC (Application Specific Integrated Circuit) circuit, a processor and a memory that execute one or more software or fixed programs, and / or other devices that can provide the above functions.

[0128] An embodiment of the present invention further provides a breast prone ultrasonic imaging device having the above Figure 10 shown breast prone ultrasonic imaging device.

[0129] Please refer to Figure 11 , Figure 11 which is a schematic structural diagram of a breast prone ultrasonic imaging device provided by an alternative embodiment of the present invention. As Figure 11 shown, the breast prone ultrasonic imaging device includes: An imaging component, and the imaging component includes an ultrasonic probe; A cup for accommodating breast tissue; A probe driving assembly for driving the probe to adjust the angle; A support assembly for supporting the imaging assembly, the cup, and the probe driving assembly; One or more processors 10, a memory 20, and an interface for connecting various components, including a high-speed interface and a low-speed interface. Each component is communicatively connected to each other using different buses and can be installed on a common motherboard or in other ways as needed. The processor can process instructions executed within the prone breast ultrasound imaging device, including instructions stored in the memory or on the memory to display graphical information of the GUI on an external input / output device (such as a display device coupled to the interface). In some alternative embodiments, multiple processors and / or multiple buses can be used together with multiple memories and multiple memories if needed. Figure 11 Taking one processor 10 as an example.

[0130] The processor 10 can be a central processing unit, a network processor, or a combination thereof. Among them, the processor 10 can further include a hardware chip. The above hardware chip can be an application-specific integrated circuit, a programmable logic device, or a combination thereof. The above programmable logic device can be a complex programmable logic device, a field-programmable gate array, a generic array logic, or any combination thereof.

[0131] Among them, the memory 20 stores instructions executable by at least one processor 10, so that the at least one processor 10 executes the method shown in the above embodiments.

[0132] The memory 20 can include a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the prone breast ultrasound imaging device, etc. In addition, the memory 20 can include a high-speed random access memory and can also include a non-transitory memory, such as at least one magnetic disk storage device, a flash memory device, or other non-transitory solid-state storage devices. In some alternative embodiments, the memory 20 can optionally include a memory remotely set relative to the processor 10, and these remote memories can be connected to the prone breast ultrasound imaging device through a network. Examples of the above network include but are not limited to the Internet, an enterprise intranet, a local area network, a mobile communication network, and combinations thereof.

[0133] The memory 20 can include volatile memory, such as random access memory; the memory can also include non-volatile memory, such as flash memory, a hard disk, or a solid-state drive; the memory 20 can also include a combination of the above types of memory.

[0134] The breast prone ultrasound imaging device further includes an input device 30 and an output device 40. The processor 10, the memory 20, the input device 30, and the output device 40 may be connected by a bus or other means. Figure 11 Taking connection by bus as an example.

[0135] The input device 30 can receive input digital or character information and generate key signal inputs related to the user settings and function controls of the breast prone ultrasound imaging device, such as a touch screen, a keypad, a mouse, a trackpad, a touchpad, a pointing stick, one or more mouse buttons, a trackball, a joystick, etc. The output device 40 may include a display device, an auxiliary lighting device (e.g., an LED), and a haptic feedback device (e.g., a vibration motor), etc. The above display device includes, but is not limited to, a liquid crystal display, a light-emitting diode, a display, and a plasma display. In some alternative embodiments, the display device may be a touch screen.

[0136] The breast prone ultrasound imaging device further includes a communication interface for the breast prone ultrasound imaging device to communicate with other devices or a communication network.

[0137] In addition, the breast prone ultrasound imaging device may further include some other structures, which can be specifically set according to actual needs and will not be listed in detail here.

[0138] The embodiment of the present invention also provides a computer-readable storage medium. The method according to the embodiment of the present invention can be implemented in hardware, firmware, or be implemented as computer code that can be recorded on a storage medium, or be implemented by downloading through a network and originally stored in a remote storage medium or a non-transitory machine-readable storage medium and will be stored in a local storage medium, so that the method described herein can be stored in such software processing on a storage medium using a general-purpose computer, a dedicated processor, or programmable or dedicated hardware. Among them, the storage medium can be a magnetic disk, an optical disc, a read-only memory, a random access memory, a flash memory, a hard disk, or a solid-state drive, etc.; further, the storage medium may further include a combination of the above types of memories. It can be understood that a computer, a processor, a microprocessor controller, or programmable hardware includes a storage component that can store or receive software or computer code, and when the software or computer code is accessed and executed by the computer, the processor, or the hardware, the method shown in the above embodiment is implemented.

[0139] A part of the present invention can be applied as a computer program product, such as computer program instructions, which, when executed by a computer, can invoke or provide the methods and / or technical solutions according to the present invention through the operations of the computer. Those skilled in the art should understand that the forms of existence of computer program instructions in a computer-readable medium include but are not limited to source files, executable files, installation package files, etc. Correspondingly, the ways for a computer to execute computer program instructions include but are not limited to: the computer directly executes the instructions, or the computer compiles the instructions and then executes the corresponding compiled program, or the computer reads and executes the instructions, or the computer reads and installs the instructions and then executes the corresponding installed program. Here, the computer-readable medium can be any available computer-readable storage medium or communication medium accessible by the computer.

[0140] Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, those skilled in the art can make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims.

Claims

1. A prone breast ultrasound imaging method, characterized in that, Applied to a prone breast ultrasound device, the method includes: Obtaining multiple pre-scanned images of a target breast tissue, where the multiple pre-scanned images include images collected from multiple different angles of the target breast tissue; Respectively identifying the skin boundaries in the multiple pre-scanned images; Based on the skin boundaries identified in each of the pre-scanned images, respectively obtaining target ultrasound emission angles perpendicular to the skin boundaries; During formal scanning and imaging, based on the target ultrasound emission angles, emitting ultrasound perpendicular to the target breast tissue and collecting ultrasound images.

2. The method according to claim 1, wherein The respectively identifying the skin boundaries in the multiple pre-scanned images includes: Performing contrast enhancement processing on the pre-scanned image to obtain a first image; Performing edge processing on the first image to obtain a second image; the edge processing is edge enhancement filtering processing or edge extraction processing; Performing binarization on the second image to obtain a binary image; Extracting connected regions in the binary image and filtering the extracted connected regions according to region size, position, and / or shape to obtain remaining connected regions; Performing boundary curve fitting on the boundary points of the remaining connected regions to obtain a continuous skin boundary line.

3. The method according to claim 1, characterized in that, The respectively identifying the skin boundaries in the multiple pre-scanned images includes: Using a deep learning model to identify the breast region in the pre-scanned image; Performing contour extraction on the identified breast region to obtain a skin boundary curve.

4. The method according to claim 1, wherein During the formal scanning and imaging, based on the target ultrasound emission angles, emitting ultrasound perpendicular to the target breast tissue includes: Based on the target ultrasound emission angle corresponding to each pre-scanned image, determining the ultrasound probe angle; Adjusting the ultrasound probe according to the determined ultrasound probe angle and emitting an ultrasonic beam.

5. The method according to claim 1 or 4, characterized in that, During the formal scanning and imaging, based on the target ultrasound emission angles, emitting ultrasound perpendicular to the target breast tissue includes: Based on the target ultrasound emission angle corresponding to the pre-scanned image, determining the beam incident angle in the first target stage during the formal scanning process; Wherein, the first target stage is determined according to the scanning angle, the starting scanning angle of the first target stage is the scanning angle corresponding to the pre-scanned image, and the ending scanning angle is the scanning angle of the next frame of pre-scanned image corresponding to the pre-scanned image.

6. The method according to claim 5, wherein The determining the beam incident angle in the first target stage during the formal scanning process based on the target ultrasound emission angle corresponding to the pre-scanned image includes: Taking the target ultrasound emission angle corresponding to the pre-scanned image as the beam reference angle for the first target stage; During the first target stage in the formal scanning process, obtaining a first ultrasound image collected at a first moment; Identifying the skin boundary in the first ultrasound image; Based on the skin boundary in the first ultrasound image, finely adjusting the beam reference angle to obtain the beam incident angle, which is used as the beam incident angle at a second moment in the first target stage; the second moment is after the first moment and before a new beam incident angle is obtained.

7. The method according to claim 1, wherein Also includes: Obtain a first target ultrasound image; the first target ultrasound image is an official ultrasound image collected at the third moment during the formal scanning process, or is the pre-scanning image; Obtain the breast region, skin boundary, and various tissue recognition results in the first target ultrasound image; Starting from the skin boundary, divide the breast region into multiple depth partitions according to depth; Based on the tissue recognition results, obtain the tissue distribution information within each depth partition; Respectively determine the corresponding gains according to the tissue distribution information of the depth partitions, and form a target time gain compensation curve; During the second target stage in the formal scanning process, adjust the gain of the ultrasound signal reception according to the target time gain compensation curve until the next stage; Among them, if the first target ultrasound image is the official ultrasound image collected at the third moment, the starting moment of the second target stage is after the third moment, and the ending moment is at or after the time when a new target time gain compensation curve is obtained; if the first target ultrasound image is the pre-scanning image, the second target stage is determined according to the scanning angle, the starting scanning angle of the second target stage is the scanning angle corresponding to the pre-scanning image, and the ending scanning angle is the scanning angle of the next frame of pre-scanning image of the pre-scanning image.

8. The method according to claim 7, wherein The adjusting the gain of the ultrasound signal reception according to the target time gain compensation curve during the second target stage in the formal scanning process includes: If the first target ultrasound image is a pre-scanning image, determine the starting position of the gain according to the position of the skin boundary in the second ultrasound image, and the second ultrasound image is an official ultrasound image collected before the current moment; Adjust the gain of the ultrasound signal reception according to the starting position of the gain and the target time gain compensation curve.

9. The method according to claim 1, characterized in that, It further includes: Obtain a second target ultrasound image and the probe pose; wherein, the second target ultrasound image is the third ultrasound image collected at the fourth moment during the formal scanning process, correspondingly, the probe pose is the probe pose during the formal scanning process; or the second target ultrasound image is the pre-scanning image, correspondingly, the probe pose is the probe pose during the pre-scanning process or the probe pose during the formal scanning process; Based on the skin boundary in the second target ultrasound image and the probe pose, determine the distance between the probe and the skin; According to the preset subcutaneous target imaging depth and the distance, determine the transmit focus point parameters and receive focus point parameters during the formal scanning process for the third target stage; if the second target ultrasound image is the third ultrasound image, the third target stage includes one or more moments after the fourth moment; if the second target ultrasound image is the pre-scanning image, the third target stage is determined according to the scanning angle, the starting scanning angle of the third target stage is the scanning angle corresponding to the pre-scanning image, and the ending scanning angle is the scanning angle of the next frame of pre-scanning image of the pre-scanning image.

10. A breast prone ultrasound imaging device, characterized in that, Applied to a prone breast ultrasound device, the device includes: A pre-scanning image acquisition module, configured to acquire multiple pre-scanning images of a target breast tissue, where the multiple pre-scanning images include images acquired from multiple different angles of the target breast tissue; A first skin boundary recognition module, configured to respectively recognize the skin boundaries in the multiple pre-scanning images; A target ultrasonic emission angle acquisition module, configured to respectively acquire target ultrasonic emission angles perpendicular to the skin boundaries based on the skin boundaries recognized in each of the pre-scanning images; An ultrasonic imaging module, configured to, when performing formal scanning and imaging, emit ultrasonic waves perpendicular to the target breast tissue based on the target ultrasonic emission angles and acquire ultrasonic images.

11. A breast prone ultrasound imaging device, characterized in that, Comprising: An imaging assembly, where the imaging assembly includes an ultrasonic probe; A cup, configured to accommodate breast tissue; A probe driving assembly, configured to drive the probe to adjust the angle; A support assembly, configured to support the imaging assembly, the cup, and the probe driving assembly; A memory and a processor, where the memory and the processor are communicatively connected to each other, the memory stores computer instructions, and the processor executes the computer instructions to perform the breast prone ultrasonic imaging method according to any one of claims 1 to 7.

Images