A breast ultrasound scanning quality control and blind area monitoring system and method

Abstract

The present application relates to the technical field of ultrasonic imaging, in particular to a breast ultrasonic scanning quality control and blind area monitoring system and method, which can acquire a breast ultrasonic image sequence scanned by a probe in real time, record position information and attitude information of the probe, identify and segment key anatomical structures in the image in real time, construct a three-dimensional breast model according to the position, attitude information of the probe and the key anatomical structures, map the real-time scanning section of the ultrasonic probe to the three-dimensional breast model, dynamically and in real time identify the scanned area and the blind area, calculate the coverage rate of each preset standard scanning partition occupied by the current scanned area, and generate guiding prompt information according to the coverage rate and the blind area. The technical scheme can dynamically and quantitatively evaluate the scanning coverage integrity, generate guiding prompts, greatly reduce the risk of missed diagnosis, and improve the integrity and accuracy of breast ultrasonic scanning.

Description

Technical Field

[0001] This invention relates to the field of ultrasound imaging technology, specifically to a breast ultrasound scanning quality control and blind zone monitoring system and method. Background Technology

[0002] Ultrasound examination, due to its non-invasive, radiation-free, and real-time characteristics, has become an important means of screening and diagnosing breast diseases. However, traditional handheld breast ultrasound scans are highly dependent on the operator's experience and skill level, and have drawbacks such as incomplete scan coverage, lack of real-time quality control feedback, and reliance on subjective judgment for blind spot identification.

[0003] Although some techniques exist to assist operators in performing ultrasound scans, these techniques still have the following main drawbacks: First, the lack of real-time dynamic monitoring of the scanning process makes it difficult for operators to grasp the scanning coverage in real time; second, operators cannot accurately assess the scanning coverage rate, making it impossible to objectively measure the scanning quality; and third, the identification of scanning blind spots relies on the operator's subjective experience, making it difficult to accurately detect missed areas caused by operating habits or the complexity of breast structure.

[0004] Therefore, existing breast ultrasound scans lack completeness and accuracy, resulting in a high rate of missed lesions. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a breast ultrasound scanning quality control and blind zone monitoring system and method to solve the problems of insufficient completeness and accuracy and high lesion missed rate in the existing breast ultrasound scanning.

[0006] According to a first aspect of the present invention, a breast ultrasound scanning quality control and blind zone monitoring system is provided, comprising:

[0007] The image acquisition unit is used to acquire the breast ultrasound image sequence collected during the ultrasound probe scanning process in real time, and simultaneously record the position and orientation information of the ultrasound probe in three-dimensional space. The anatomical structure recognition unit is used to identify and segment key anatomical structures in the breast ultrasound image in real time using a first trained deep learning model. The three-dimensional model construction unit is used to construct a three-dimensional breast model based on the position and orientation information of the ultrasound probe and the key anatomical structures; and to obtain the real-time scanning section of the ultrasound probe based on the position and orientation information of the ultrasound probe. The scanning coverage analysis unit is used to map the real-time scanning section onto the three-dimensional breast model, dynamically and in real time identify the scanned areas and unscanned blind areas of the three-dimensional breast model, and calculate the coverage rate of the currently scanned area to each preset standard scanning partition. The blind spot monitoring and feedback unit is used to generate guidance prompts when there is a coverage rate below a preset threshold, or when there are unscanned blind spots.

[0008] Preferably, the scanning coverage analysis unit is further configured to obtain the scanning path of the ultrasonic probe based on the position and orientation information of the ultrasonic probe in three-dimensional space; determine whether the scanning path conforms to a preset standard scanning path; and generate feedback information of path abnormality if there is a deviation.

[0009] Preferably, the breast ultrasound scanning quality control and blind spot monitoring system further includes: The suspected lesion analysis unit is used to detect in real time whether there are suspected lesions in the breast ultrasound image using a trained second deep learning model. If a suspected lesion is found, it is determined whether the suspected lesion is within the preset range of the blind area. If so, a key scanning instruction for the blind area is generated.

[0010] Preferably, the key anatomical structures identified and segmented by the anatomical structure recognition unit include at least the boundaries of the breast region, the boundaries of the pectoral muscles, the nipple location, and the axillary tail region markers.

[0011] Preferably, the breast ultrasound scanning quality control and blind spot monitoring system further includes: Based on the identified breast region boundaries and nipple location, the breast is divided into six standard scanning zones, including: upper inner region, lower inner region, upper outer region, lower outer region, posterior nipple region, and axillary tail region.

[0012] Preferably, each standard scanning zone is provided with a corresponding scanning coverage requirement, and the preset threshold is set according to the scanning coverage requirement; The blind spot monitoring and feedback unit is also used to mark the standard scanned area corresponding to the coverage rate as an unqualified area when there is a coverage rate lower than a preset threshold; and to generate guidance prompt information based on the unqualified area and the unscanned blind spot area.

[0013] Preferably, the blind spot monitoring and feedback unit generates guidance prompts, including: The three-dimensional breast model is displayed on the ultrasound display interface; On the three-dimensional breast model, unqualified zones or blind areas that have not been scanned are marked with colors. The direction of movement of the ultrasonic probe can be determined based on the color-coded areas and the real-time position of the ultrasonic probe. The real-time position of the ultrasound probe is displayed on the three-dimensional breast model, and the direction of movement of the ultrasound probe is marked. Based on unqualified zones or blind spots that have not been scanned, generate and broadcast voice prompts.

[0014] Preferably, after mapping the real-time scanning section onto the three-dimensional breast model, the scanning coverage analysis unit further includes: Based on the scanning status of the real-time scanning section, a heat map or overlay grid of the scanned area is dynamically generated.

[0015] Preferably, the breast ultrasound scanning quality control and blind spot monitoring system further includes: a quality control report generation unit, used to generate a quality control report containing coverage, blind spot area location distribution and scanning path map after the scanning is completed.

[0016] According to a second aspect of the present invention, a method for quality control and blind zone monitoring of breast ultrasound scanning is provided, comprising: The system acquires real-time sequences of breast ultrasound images during ultrasound probe scanning and simultaneously records the position and orientation information of the ultrasound probe in three-dimensional space. Using the first trained deep learning model, key anatomical structures in the breast ultrasound images are identified and segmented in real time. A three-dimensional breast model is constructed based on the position and orientation information of the ultrasound probe and the key anatomical structures; the real-time scanning section of the ultrasound probe is obtained based on the position and orientation information of the ultrasound probe. The real-time scanning section is mapped onto the three-dimensional breast model to dynamically and in real time identify the scanned areas and unscanned blind areas of the three-dimensional breast model; the coverage rate of the currently scanned area to each preset standard scanning partition is calculated respectively. When there is a coverage rate below a preset threshold, or when there are unscanned blind spots, a guidance prompt message is generated.

[0017] The technical solution provided by this invention may include the following beneficial effects: It is understood that the technical solution presented in this invention can acquire breast ultrasound image sequences scanned by the probe in real time and record the probe's position and orientation information; identify and segment key anatomical structures in the images in real time; construct a three-dimensional breast model based on the probe's position, orientation information, and key anatomical structures, and map the real-time scanning section of the ultrasound probe onto the three-dimensional breast model to dynamically identify scanned areas and unscanned blind areas in real time; calculate the coverage rate of the currently scanned area in each preset standard scanning partition; and generate guidance prompts based on the coverage rate and unscanned blind areas. This technical solution can dynamically and quantitatively evaluate the integrity of the scan coverage and generate guidance prompts, significantly reducing the risk of missed diagnoses and improving the integrity and accuracy of ultrasound scans.

[0018] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit the invention. Attached Figure Description

[0019] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0020] Figure 1 This is a schematic block diagram illustrating a breast ultrasound scanning quality control and blind spot monitoring system according to an exemplary embodiment; Figure 2 This is a schematic diagram of a breast ultrasound scanning quality control and blind zone monitoring device according to an exemplary embodiment; Figure 3 This is a schematic diagram illustrating the usage process of the device according to an exemplary embodiment; Figure 4 This is a schematic diagram of a scanning coverage heatmap visualization interface according to an exemplary embodiment; Figure 5 This is a schematic diagram of a visual interface displaying a guide path, according to an exemplary embodiment. Figure 6 This is a schematic diagram illustrating the steps of a breast ultrasound scanning quality control and blind zone monitoring method according to an exemplary embodiment. Detailed Implementation

[0021] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numerals in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Rather, they are merely examples of apparatuses and methods consistent with some aspects of the invention as detailed in the appended claims.

[0022] In one embodiment, Figure 1 This is a schematic block diagram illustrating a breast ultrasound scanning quality control and blind zone monitoring system according to an exemplary embodiment. See also... Figure 1 A breast ultrasound scanning quality control and blind spot monitoring system is provided, comprising: The image acquisition unit 101 is used to acquire breast ultrasound image sequences collected during ultrasound probe scanning in real time, and simultaneously record the position and orientation information of the ultrasound probe in three-dimensional space. Specifically, the ultrasound probe is a high-frequency probe with a frequency of 7.5MHz to 15MHz.

[0023] In practice, when the operator holds the ultrasound probe to perform a breast scan on the patient, the image acquisition unit 101 will collect the breast ultrasound image sequence generated during the probe scan in real time. At the same time, it will accurately record the position and posture changes of the probe in three-dimensional space, and save the ultrasound image data and probe spatial posture data simultaneously, providing complete raw data support for subsequent scan quality control.

[0024] The anatomical structure recognition unit 102 is used to identify and segment key anatomical structures in the breast ultrasound image in real time using a first deep learning model that has been trained.

[0025] During the operator's continuous scanning with the probe, the anatomical structure recognition unit 102 calls the first deep learning model that has been trained to quickly infer the real-time transmitted breast ultrasound images and automatically identify and segment the key anatomical structures of the breast.

[0026] The key anatomical structures mentioned include at least the boundaries of the breast region, the boundaries of the pectoral muscles, the location of the nipple, and the axillary tail area. These key anatomical structures help the system clearly define the standard scanning range of the breast, avoiding the impact of unclear anatomical structures on scanning judgment.

[0027] For the first deep learning model, a lightweight real-time segmentation network is used, such as U-Net+MobileNetV3 backbone or DeepLabV3+with MobileNetV2, to ensure an inference speed of ≥30 fps on the embedded system of the ultrasound equipment.

[0028] The input is a single-frame ultrasound image (grayscale, size 256×256 or 512×512). The output is a multi-class segmentation mask, including: breast region boundaries (overall glandular extent), pectoral muscle boundaries (used to identify the retromammary space), nipple location (as a key anatomical landmark), and axillary tail region (the portion of the breast extending into the axilla).

[0029] During the training of the first deep learning model, no fewer than 5000 ultrasound images were collected from different devices, different breast types (dense, fatty, and mixed), and different scanning angles (longitudinal, transverse, and oblique sections). Pixel-by-pixel annotation was performed by experienced ultrasound physicians to ensure annotation quality. Simultaneously, data augmentation was performed using random rotation, scaling, brightness and contrast adjustments, and elastic deformation to improve the model's generalization ability. Ultimately, the model achieved a mean intersection-over-union (mIoU) of 0.92 on the validation set.

[0030] For post-processing and real-time inference optimization of the model, morphological opening and closing operations are performed on the segmentation results output by the model to remove noise, and connected component analysis is used to extract the largest breast region. For model quantization, INT8 quantization is used, deployed on the embedded GPU or AI acceleration chip of the ultrasound host. Simultaneously, Kalman filtering is used to temporally smooth the segmentation results of consecutive frames to avoid anatomical boundary jumps caused by probe jitter.

[0031] The three-dimensional model construction unit 103 is used to construct a three-dimensional breast model based on the position and orientation information of the ultrasound probe and the key anatomical structures; and to obtain the real-time scanning section of the ultrasound probe based on the position and orientation information of the ultrasound probe.

[0032] This unit combines the real-time position and orientation information of the probe in the operator's hand with the identified key anatomical structures of the breast to quickly construct a three-dimensional breast model that fits the patient's breast shape, with the nipple as the origin. At the same time, based on the real-time position, orientation, and scanning field of view of the probe, it accurately calculates the real-time scanning section corresponding to the probe, completing the conversion from two-dimensional ultrasound image to three-dimensional scanning range.

[0033] In constructing the 3D breast model, an ellipsoidal or freeform surface model is established using the nipple location as the origin and the identified breast boundaries. The model is represented by a voxel grid with a resolution of 2mm×2mm×2mm. Each voxel records whether it has been scanned and the number of scans.

[0034] The scanning coverage analysis unit 104 is used to map the real-time scanning section onto the three-dimensional breast model, dynamically and in real time identify the scanned areas and unscanned blind areas of the three-dimensional breast model, and calculate the coverage rate of the currently scanned area to each preset standard scanning partition.

[0035] This unit maps the real-time scanning section of the operator's probe onto the three-dimensional breast model, dynamically distinguishing between scanned areas and unscanned blind spots in the model.

[0036] When mapping real-time scanning sections onto a 3D breast model, the corresponding fan-shaped section region in the 3D model is calculated based on the real-time position and orientation of the probe (6-DOF) (considering the probe's field of view depth and width). A ray casting method is used to mark the voxels covered by the section as "scanned," and the number of scans is accumulated.

[0037] In a preferred embodiment, based on the identified breast region boundaries and nipple location, the breast is divided into six standard scanning zones, including: upper inner region, lower inner region, upper outer region, lower outer region, posterior nipple region, and axillary tail region. The coverage rate of each zone relative to the standard scanning area is then calculated to quantify the completeness of the scan in real time.

[0038] When calculating coverage, dynamic coverage = number of scanned voxels / total number of voxels in the standard area × 100%.

[0039] In blind zone identification, morphological dilation is used to expand the scanned area (to compensate for the probe edge), and then the difference between the scanned area and the standard area is calculated to obtain the unscanned connected area, which is the blind zone area.

[0040] Preferably, a minimum number of scans (e.g., ≥2 times) can be set for each partition. Only when each voxel in the partition is scanned twice will the voxel be considered to belong to the scanned area.

[0041] In a preferred embodiment, the scanning coverage analysis unit 104 can also dynamically generate a heat map or coverage grid of the scanned area based on the scanning status of the real-time scanning section.

[0042] The blind spot monitoring and feedback unit 105 is used to generate guidance prompts when there is a coverage rate lower than a preset threshold, or when there are unscanned blind spots.

[0043] When an operator finds that the coverage of a certain area is lower than a preset threshold during a scan, or when the system identifies an unscanned blind area, the blind area monitoring and feedback unit 105 will immediately generate guidance prompts. For example, it can use a display screen to mark the blind area with color, arrows to guide the direction of probe movement, and voice broadcasts of the missing area to intuitively guide the operator to quickly fill in the blind area and ensure that no breast scan is missed.

[0044] In a preferred embodiment, each standard scanning zone is provided with a corresponding scanning coverage requirement, and a preset threshold is set according to the scanning coverage requirement. The blind spot monitoring and feedback unit 105 is further configured to mark the standard scanning zone corresponding to the coverage rate as an unqualified zone when there is a coverage rate lower than the preset threshold; and to generate guidance prompt information based on the unqualified zone and the unscanned blind spot area.

[0045] In a preferred embodiment, when the blind spot monitoring and feedback unit 105 generates guidance prompt information, First, relevant information can be displayed on the ultrasound equipment's monitor. For example, the three-dimensional breast model can be displayed on the ultrasound display interface. On the three-dimensional breast model, unqualified areas or blind areas that have not been scanned are marked with colors to form a scan coverage heatmap (green indicates sufficient coverage, yellow indicates a single scan, and red indicates no scan). At the same time, the movement direction of the ultrasound probe can be determined based on the color-marked areas and the real-time position of the ultrasound probe. The real-time position of the ultrasound probe is displayed on the three-dimensional breast model, and the direction of movement of the ultrasound probe is marked, that is, the path from the current probe position to the nearest blind zone center is indicated by directional arrows.

[0046] In addition to visual cues, voice prompts can be added. Voice prompts can be generated and broadcast based on unscanned or un-scanned blind spots, such as "Please scan the upper outer quadrant." Tactile vibrations can also be added, for example, by using a built-in vibration motor in the ultrasound probe to increase the vibration frequency as it approaches a blind spot.

[0047] In a preferred embodiment, the breast ultrasound scanning quality control and blind spot monitoring system further includes: The suspected lesion analysis unit is used to detect in real time whether there are suspected lesions in the breast ultrasound image using a trained second deep learning model; if there are suspected lesions (such as hypoechoic nodules, microcalcifications, etc.), it is determined whether the suspected lesion is within a preset range of the blind area; if so, a key scanning instruction for the blind area is generated.

[0048] After issuing a priority scanning instruction, the blind spot monitoring and feedback unit 105 can also set these suspected lesions as high priority, guiding the scan to these high-priority areas first. This technical solution can further reduce missed diagnoses.

[0049] In practical application scenarios, see Figure 2The device equipped with a breast ultrasound scanning quality control and blind spot monitoring system consists of a handheld ultrasound probe 11, an ultrasound host 10, and a display device 13. A spatial positioning module 12 is installed on the handheld ultrasound probe 11, while the ultrasound host 10 integrates an image acquisition card and a processor (not shown in the figure). The operator uses the handheld ultrasound probe 11 to perform breast scans on patients. The spatial positioning module 12 on the handheld ultrasound probe 11 can record the probe's position and posture information in three-dimensional space in real time. The breast ultrasound images scanned by the handheld ultrasound probe 11 are acquired by the image acquisition card inside the ultrasound host 10. The image acquisition card is connected to the processor, which internally deploys an anatomical structure recognition unit 102, a three-dimensional model construction unit 103, and a blind spot monitoring and feedback unit 105. The processor transmits the generated guidance prompts to the display device 13 to display these prompts, providing the operator with operational instructions, such as displaying real-time ultrasound images, a scan coverage heatmap, and guidance prompts.

[0050] In use such Figure 2 When referring to the device shown, see Figure 3 The operator holds an ultrasound probe to scan the patient's breast, while the system simultaneously acquires ultrasound images in real time and accurately records the probe's spatial movement trajectory. A built-in AI model (the first deep learning model) analyzes the images in real time, automatically identifying and segmenting key anatomical structures such as breast boundaries, nipples, and pectoral muscles. Based on probe pose and anatomical information, the system dynamically constructs a three-dimensional breast model and calculates the full breast scan coverage in real time. When the axillary tail area is detected as unscanned, the display device highlights the location in red and simultaneously provides a voice prompt, "Please scan the axillary tail area." The operator completes the supplementary scan as prompted, and once the coverage is updated to 100%, the system provides a voice prompt, "Scan complete." After the scan is finished, the system automatically generates a professional quality control report, fully including key information such as scan coverage, blind spot location distribution, and scan trajectory, making the scan quality quantifiable and traceable.

[0051] See Figure 4 The diagram shows a schematic of the heatmap visualization interface for ultrasound coverage. When the operator performs an ultrasonic scan, the interface is displayed as follows: Figure 4 As shown, it displays a real-time ultrasound image window and a view of a 3D breast model. On the 3D breast model, a heatmap of scan coverage is displayed, with different colors representing the number of scans: dark green represents adequate scanning (≥2 scans), light green represents 1 scan, and red represents no scan. The system also provides guidance and prompts to the operator. See also... Figure 5 It displays the current position of the scanning probe on a 3D breast model, marks the blind area, and shows the guide path from the scanning position to the center of the blind area, with arrow directions displayed on the interface.

[0052] In a preferred embodiment, the scanning coverage analysis unit 104 is further configured to obtain the scanning path of the ultrasonic probe based on the position and orientation information of the ultrasonic probe in three-dimensional space; determine whether the scanning path conforms to a preset standard scanning path; and generate feedback information of path abnormality if there is a deviation.

[0053] The scanning coverage analysis unit 104 shown in this embodiment continuously collects the real-time position and posture information of the probe in three-dimensional space throughout the entire process of the operator performing a breast scan with the handheld ultrasound probe. It strings together these continuous position and posture data to completely reconstruct the operator's current actual scanning path. The system then compares this actual scanning path with the preset standard breast scanning path (such as a standardized six-zone orderly scanning, a standard longitudinal / transverse scanning trajectory). Once a deviation in the operator's scanning path is detected, such as skipping the axillary tail area, incorrect scanning angle, failure to cover key areas according to the standard trajectory, path repetition, or omission, the unit will immediately determine that the path is abnormal and quickly generate corresponding path abnormality feedback information to promptly remind the operator that the scanning trajectory is not standardized and guide them back to the standard scanning path.

[0054] In a preferred embodiment, the breast ultrasound scanning quality control and blind spot monitoring system further includes: a quality control report generation unit, used to generate a quality control report containing coverage, blind spot area location distribution, and scanning path map after the scanning is completed.

[0055] Once the operator completes the full ultrasound scan of the patient's breast and ends the scan, the quality control report generation unit automatically summarizes all the data from the scan and quickly generates a complete breast ultrasound scan quality control report. The report clearly marks the scan coverage of the entire breast and six standard zones: upper inner, lower inner, upper outer, lower outer, behind the nipple, and axillary tail. It also presents the location distribution of all blind spots during the scan and fully reconstructs and plots the operator's scan path, visually demonstrating whether the scan trajectory is standardized and whether there are any path abnormalities. Ultimately, this forms an archiveable and traceable objective quality control document, providing detailed data support for scan quality assessment, medical quality management, and evidence collection in subsequent disputes.

[0056] In a preferred embodiment, the probe position information and attitude information are obtained through an electromagnetic positioning sensor, an optical positioning sensor, or an inertial measurement unit.

[0057] It is understood that the technical solution shown in this invention has the following technical effects: Real-time quality control during the scanning process: This invention combines probe spatial positioning with AI anatomical structure recognition, and for the first time realizes dynamic quantitative evaluation of the integrity of the scanning coverage during breast ultrasound scanning, filling the technical gap in quality control of the scanning process of existing ultrasound equipment.

[0058] Effectively reduces the rate of missed diagnoses: By identifying blind spots in real time and guiding operators to supplement the scan in a timely manner, it ensures that no breast tissue is missed. It is especially suitable for primary healthcare institutions, beginners, or large-scale breast cancer screening scenarios, and can significantly reduce the risk of missed diagnoses due to insufficient scanning.

[0059] Intelligent guidance improves inspection efficiency: The system can visually prompt unscanned areas, reducing the time cost of repeated scanning for operators and improving inspection efficiency.

[0060] Traceable scan quality: The generated quality control report provides an objective basis for quality evaluation of ultrasound examinations, which facilitates subsequent quality control management and evidence collection in medical disputes.

[0061] Significantly different from existing technologies: This invention is not limited to lesion detection or probe initial position calibration, but focuses on the complete quality control of the scanning process. It can be used in conjunction with existing lesion detection technologies to form a more complete intelligent ultrasound-assisted diagnostic system.

[0062] In another embodiment, see Figure 6 This invention provides a method for quality control and blind zone monitoring in breast ultrasound scanning, including: Step S11: Acquire the sequence of breast ultrasound images collected during the ultrasound probe scanning process in real time, and simultaneously record the position and orientation information of the ultrasound probe in three-dimensional space; Step S12: Using the trained first deep learning model, identify and segment key anatomical structures in the breast ultrasound image in real time; Step S13: Construct a three-dimensional breast model based on the position and orientation information of the ultrasound probe and the key anatomical structures; derive the real-time scanning section of the ultrasound probe based on the position and orientation information of the ultrasound probe. Step S14: Map the real-time scanning section onto the three-dimensional breast model, dynamically and in real-time identify the scanned areas and unscanned blind areas of the three-dimensional breast model; calculate the coverage rate of the currently scanned area to each preset standard scanning partition. Step S15: When there is a coverage rate lower than a preset threshold, or when there are unscanned blind areas, generate a guidance prompt message.

[0063] Understandably, this method can monitor the breast ultrasound scanning process in real time, automatically identify blind spots, and provide precise supplementary scanning guidance for the operator, thereby improving the completeness and accuracy of breast ultrasound scanning and reducing the rate of missed lesions. It solves the technical problems of existing breast ultrasound scanning techniques, such as the inability to monitor the completeness of scan coverage in real time and the difficulty in identifying blind spots, achieving intelligent quality control of the scanning process and reducing the rate of missed diagnoses.

[0064] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0065] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0066] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0067] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0068] Those skilled in the art will understand that all or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, the program includes one or a combination of the steps of the method embodiments.

[0069] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0070] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0071] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0072] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A breast ultrasound scanning quality control and blind zone monitoring system, characterized in that, include: The image acquisition unit is used to acquire the breast ultrasound image sequence collected during the ultrasound probe scanning process in real time, and simultaneously record the position and orientation information of the ultrasound probe in three-dimensional space. The anatomical structure recognition unit is used to identify and segment key anatomical structures in the breast ultrasound image in real time using a first trained deep learning model. The three-dimensional model construction unit is used to construct a three-dimensional breast model based on the position and orientation information of the ultrasound probe and the key anatomical structures; and to obtain the real-time scanning section of the ultrasound probe based on the position and orientation information of the ultrasound probe. The scanning coverage analysis unit is used to map the real-time scanning section onto the three-dimensional breast model, dynamically and in real time identify the scanned areas and unscanned blind areas of the three-dimensional breast model, and calculate the coverage rate of the currently scanned area to each preset standard scanning partition. The blind spot monitoring and feedback unit is used to generate guidance prompts when there is a coverage rate below a preset threshold, or when there are unscanned blind spots.

2. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 1, characterized in that, The scanning coverage analysis unit is also used to obtain the scanning path of the ultrasonic probe based on the position and orientation information of the ultrasonic probe in three-dimensional space. It determines whether the scanning path conforms to the preset standard scanning path. If there is a deviation, it generates feedback information indicating that the path is abnormal.

3. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 1, characterized in that, Also includes: The suspected lesion analysis unit is used to detect in real time whether there are suspected lesions in the breast ultrasound image using a trained second deep learning model. If a suspected lesion is found, it is determined whether the suspected lesion is within the preset range of the blind area. If so, a key scanning instruction for the blind area is generated.

4. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 1, characterized in that, The key anatomical structures identified and segmented by the anatomical structure recognition unit include at least the boundaries of the breast region, the boundaries of the pectoral muscles, the location of the nipple, and the axillary tail region markers.

5. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 4, characterized in that, Also includes: Based on the identified breast region boundaries and nipple location, the breast is divided into six standard scanning zones, including: upper inner region, lower inner region, upper outer region, lower outer region, posterior nipple region, and axillary tail region.

6. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 5, characterized in that, Each standard scanning zone has corresponding scanning coverage requirements, and the preset threshold is set according to the scanning coverage requirements; The blind spot monitoring and feedback unit is also used to mark the standard scanned area corresponding to the coverage rate as an unqualified area when there is a coverage rate lower than a preset threshold; and to generate guidance prompt information based on the unqualified area and the unscanned blind spot area.

7. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 6, characterized in that, The blind spot monitoring and feedback unit generates guidance prompts, including: The three-dimensional breast model is displayed on the ultrasound display interface; On the three-dimensional breast model, unqualified zones or blind areas that have not been scanned are marked with colors. The direction of movement of the ultrasonic probe can be determined based on the color-coded areas and the real-time position of the ultrasonic probe. The real-time position of the ultrasound probe is displayed on the three-dimensional breast model, and the direction of movement of the ultrasound probe is marked. Based on unqualified zones or blind spots that have not been scanned, generate and broadcast voice prompts.

8. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 1, characterized in that, After mapping the real-time scanning section onto the three-dimensional breast model, the scanning coverage analysis unit further includes: Based on the scanning status of the real-time scanning section, a heat map or overlay grid of the scanned area is dynamically generated.

9. The breast ultrasound scanning quality control and blind zone monitoring system according to claim 2, characterized in that, Also includes: The quality control report generation unit is used to generate a quality control report containing coverage, blind spot location distribution, and scan path map after the scan is completed.

10. A method for quality control and blind zone monitoring in breast ultrasound scanning, characterized in that, include: The system acquires real-time sequences of breast ultrasound images during ultrasound probe scanning and simultaneously records the position and orientation information of the ultrasound probe in three-dimensional space. Using the first trained deep learning model, key anatomical structures in the breast ultrasound images are identified and segmented in real time. A three-dimensional breast model is constructed based on the position and orientation information of the ultrasound probe and the key anatomical structures; the real-time scanning section of the ultrasound probe is obtained based on the position and orientation information of the ultrasound probe. The real-time scanning section is mapped onto the three-dimensional breast model to dynamically and in real time identify the scanned areas and unscanned blind areas of the three-dimensional breast model; the coverage rate of the currently scanned area to each preset standard scanning partition is calculated respectively. When there is a coverage rate below a preset threshold, or when there are unscanned blind spots, a guidance prompt message is generated.

Images