Ultrasound image processing method, electronic device, and storage medium

Abstract

The application provides an ultrasonic image processing method, an electronic device and a storage medium. The ultrasonic image processing method comprises: acquiring an ultrasonic image of a target object; performing cropping on the ultrasonic image based on the ultrasonic image and a first focus position in the ultrasonic image to obtain a focus image; and inputting the focus image and the first focus position into an image segmentation model to obtain a target region in the focus image. According to the above scheme of the application, on the one hand, the ultrasonic image can be cropped based on the first focus position to obtain a focus image which is more focused by a user, and the accuracy of the target region obtained based on the focus image is higher; on the other hand, the image segmentation model can pay more attention to image features related to the first focus position, so that the accuracy of the obtained target region is improved.

Description

Technical Field

[0001] This invention relates to the field of medical image processing technology, and more specifically to a method for processing ultrasound images, an electronic device, a storage medium, and a computer program product. Background Technology

[0002] In the medical field, ultrasound imaging is one of the important methods for diagnosing various diseases. It uses sound waves to generate images of the inside of the human body, which can be used to help doctors determine the condition of organs and blood vessels.

[0003] In existing technologies, doctors can manually identify target areas using ultrasound images. With the continuous development of artificial intelligence, users can also directly obtain the target area by simply inputting ultrasound images into a detection model within the relevant technology. However, because ultrasound images typically have low resolution, detection models often struggle to accurately determine the target area. Therefore, how to determine the target area with higher accuracy using a detection model is a technical problem that urgently needs to be solved by those skilled in the art. Summary of the Invention

[0004] The present invention was proposed in view of the above-mentioned problems. The present invention provides a method for processing ultrasound images, an electronic device, a storage medium, and a computer program product.

[0005] According to one aspect of the present invention, a method for processing ultrasound images is provided, the method comprising:

[0006] Acquire ultrasound images of the target object;

[0007] Based on the ultrasound image and a first location of interest in the ultrasound image, the ultrasound image is cropped to obtain a location of interest image, wherein the location of interest image includes the initial pixel in the ultrasound image located at the first location of interest.

[0008] The image of interest and the first position of interest are input into the image segmentation model to obtain the target region in the image of interest. The first position of interest is located in the target region. The image segmentation model is trained based on the training image of the training object, the target region in the training image of interest, and the training position of interest in the ultrasound training image of the training object. The training image of interest is obtained by cropping the ultrasound training image based on the ultrasound training image and the training position of interest.

[0009] For example, the ultrasound image is cropped based on the ultrasound image and a first location of interest within the ultrasound image to obtain an image of interest, including:

[0010] The ultrasound image and the first location of interest are input into the location of interest aggregation model to obtain the aggregation result corresponding to the ultrasound image. The location of interest aggregation model is trained based on the ultrasound training image, the location of interest and the aggregation training result of the ultrasound training image. The aggregation result includes the correlation value between each initial pixel in the ultrasound image and the initial pixel located at the first location of interest.

[0011] Initial pixels with correlation values ​​greater than the first correlation threshold in the aggregation result are used as target pixels. Based on the position information of the target pixels in the ultrasound image, the ultrasound image is cropped to obtain the image of interest.

[0012] For example, initial pixels in the aggregation result that are greater than a first correlation threshold are used as target pixels, including:

[0013] Determine the first total number of initial pixels in the aggregation result whose correlation value is greater than the second correlation threshold, wherein the second correlation threshold is less than the first correlation threshold;

[0014] The initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold corresponding to the first total number are taken as target pixels, wherein the value of the first correlation threshold is positively correlated with the value of the first total number.

[0015] For example, initial pixels in the aggregation result whose correlation value is greater than a first correlation threshold are selected as target pixels, including:

[0016] Determine the second total number of initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold;

[0017] If the second total number is less than the total number threshold, the first association degree threshold is reduced based on the adjustment step size, and the second total number of the initial pixels with an association degree value greater than the first association degree threshold in the aggregation result is re-executed based on the reduced first association degree threshold.

[0018] If the second total number is greater than or equal to the total number threshold, the initial pixels in the aggregation result whose correlation degree value is greater than the target correlation degree threshold are taken as target pixels, wherein the target correlation degree threshold is the first correlation degree threshold that is determined to be greater than or equal to the second total number threshold.

[0019] For example, acquiring an ultrasound image of a target object includes:

[0020] Obtain the ultrasound image sequence of the target object;

[0021] The above methods also include:

[0022] Based on the target region, a second location of interest is determined in a number of other images adjacent to the ultrasound image in the ultrasound image sequence, based on a threshold number of such images. Then, based on the second location of interest, the target region in the other images is determined.

[0023] For example, acquiring an ultrasound image of a target object includes:

[0024] Obtain the ultrasound image sequence of the target object;

[0025] The above methods also include:

[0026] For each ultrasound image in the ultrasound image sequence, based on the first location of interest in the ultrasound image, the target region in the ultrasound image is determined, and the two-dimensional size information of the target region is determined.

[0027] Based on the two-dimensional dimension information of each target region, the three-dimensional dimension information of the region where the part of interest is located is determined.

[0028] For example, the above method further includes: displaying an ultrasound image of the target object;

[0029] In response to a user's click on the displayed ultrasound image, the click location is selected as the first point of focus.

[0030] According to another aspect of the present invention, an electronic device is also provided. The electronic device includes a processor and a memory. The memory stores a computer program; the processor executes the computer program to implement the above-described method for processing ultrasound images.

[0031] According to another aspect of the present invention, a storage medium is also provided. This medium stores a computer program / instructions that, when executed, perform the aforementioned ultrasound image processing method.

[0032] According to another aspect of the present invention, a computer program product is also provided. The computer program instructions, when run by a processor, are used to execute the above-described method for processing ultrasound images.

[0033] According to the above-described scheme of the present invention, ultrasound data of a target object can be acquired. Then, based on the ultrasound image and a first location of interest within the ultrasound image, the ultrasound image is cropped to obtain an image of interest. Finally, the image of interest and the first location of interest are input into an image segmentation model to obtain the target region within the image of interest. On one hand, the above scheme can crop the ultrasound image based on the first location of interest to obtain an image of interest that the user is more concerned about, resulting in higher accuracy of the target region obtained based on this image of interest. On the other hand, the first location of interest is usually a location that the user is more concerned about. By inputting both the image of interest and the first location of interest into the image segmentation model, the above scheme allows the image segmentation model to focus more on image features related to the first location of interest, thereby improving the accuracy of the obtained target region. Attached Figure Description

[0034] The above and other objects, features, and advantages of the present invention will become more apparent from the more detailed description of the embodiments of the invention in conjunction with the accompanying drawings. The drawings are provided to further illustrate the embodiments of the invention and form part of the specification. They are used together with the embodiments of the invention to explain the invention and do not constitute a limitation thereof. In the drawings, the same reference numerals generally represent the same parts or steps.

[0035] Figure 1 A schematic flowchart of a method for processing ultrasound images according to an embodiment of the present invention is shown;

[0036] Figure 2 A schematic diagram of a method for processing ultrasound images according to an embodiment of the present invention is shown;

[0037] Figure 3 A schematic diagram of a method for processing ultrasound images according to an embodiment of the present invention is shown;

[0038] Figure 4 A schematic block diagram of an ultrasound image processing apparatus according to an embodiment of the present invention is shown; and

[0039] Figure 5 A schematic block diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0040] To make the objectives, technical solutions, and advantages of the present invention more apparent, exemplary embodiments according to the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are merely a part of the embodiments of the present invention, and not all of the embodiments of the present invention. It should be understood that the present invention is not limited to the exemplary embodiments described herein. Based on the embodiments of the present invention described herein, all other embodiments obtained by those skilled in the art without inventive effort should fall within the protection scope of the present invention.

[0041] In existing technologies, detection models are typically trained directly using ultrasound training images of the training object and the target region within those images. However, since the target region usually occupies a small area in an ultrasound image, the trained detection model is prone to pixel-level detection imbalance. Specifically, in this detection model, because the target region has fewer pixels compared to other regions in the ultrasound image, the detection model may be significantly affected by the pixels in those other regions, which are usually healthy areas – a situation that clearly does not reflect real-world detection practices.

[0042] To at least partially solve the above problems, embodiments of the present invention provide a method for processing ultrasound images. Figure 1 A schematic flowchart illustrating a method for processing ultrasound images according to an embodiment of the present invention is shown. Figure 1 As shown, the processing method may include steps S110 to S130.

[0043] In step S110, an ultrasound image of the target object is acquired.

[0044] The aforementioned ultrasound images can be generated as follows: ultrasound data of the target object can be acquired using an ultrasound probe. The target object can be the subject whose ultrasound data needs to be acquired. The ultrasound data can be data obtained by acquiring ultrasound echoes from cavities or other parts of the target object. The cavity can be any cavity of the subject that needs to be observed, such as the thoracic cavity, esophagus, or blood vessels.

[0045] Taking the acquisition of ultrasound data from a target cavity as an example, an ultrasound probe can be inserted into the subject's cavity, then emit ultrasound waves and receive the echoes to determine ultrasound data. An ultrasound image of the cavity can then be generated based on this data. Here, a blood vessel is used as an example cavity. For instance, an angiography catheter can be inserted through the subject's femoral or radial artery and extended to the coronary ostium of the heart. A guiding catheter is then inserted through the angiography catheter into the target blood vessel to assist in guiding the ultrasound catheter (or IVUS catheter, IVUS stands for Intravascular Ultrasound, which may include an ultrasound probe). The ultrasound catheter can be advanced along the guiding catheter to the target blood vessel, continuously or at preset time intervals emitting ultrasound waves during retraction and rotation along the guiding catheter, and continuously or at preset time intervals receiving the echoes of the ultrasound waves as part of the ultrasound data. It should be understood that different carriers (such as the catheter mentioned above) can be used to mount the ultrasound probe to meet the actual needs of different cavities.

[0046] Based on the aforementioned ultrasound data, ultrasound images can be generated using image reconstruction algorithms in related technologies. Details of this embodiment will not be elaborated upon here. It should be understood that if the ultrasound data includes echoes from multiple times, an ultrasound image is generated for each echo at that time, resulting in an ultrasound image sequence comprising ultrasound images from multiple times.

[0047] In step S120, the ultrasound image is cropped based on the ultrasound image and the first position of interest in the ultrasound image to obtain the image of interest.

[0048] An initial region of interest can be obtained using detection models or algorithms in related technologies, and then the aforementioned first location of interest can be determined within this initial region of interest. For example, the location information (e.g., represented by coordinates) of any pixel in the initial region of interest within the ultrasound image can be used as the aforementioned first location of interest. Alternatively, the location information of a more representative pixel in the ultrasound image within the initial region of interest can be used as the aforementioned first location of interest, specifically, for example, the center pixel or pixels near the center pixel in the initial region of interest.

[0049] Optionally, the first location of interest can be represented by a binary image. The pixels in the binary image can correspond to the initial pixels in the ultrasound image, for example, in a one-to-one correspondence. In the binary image, the pixel value corresponding to the first location of interest can be 1, and the pixel value of other pixels can be 0. Alternatively, the first location of interest can also be represented directly by the position information of the initial pixels in the ultrasound image, for example, by coordinates.

[0050] The image of interest includes an initial pixel located at a first position of interest in the ultrasound image. Further, the image of interest may include the initial pixel at the first position of interest and other initial pixels surrounding it. For example, a predetermined number of initial pixels may be added around the first position of interest. This predetermined number may be related to the total number of initial pixels in a row or column of the ultrasound image; for example, the predetermined number may be 15% of the total number of initial pixels in a row of the ultrasound image. It should be understood that users or developers may use the first position of interest as the center, boundary position, etc., of the image of interest according to actual needs, and this embodiment of the invention does not impose any limitations.

[0051] In step S130, the image of interest and the first location of interest are input into the image segmentation model to obtain the target region in the image of interest.

[0052] The target area mentioned above can be a region that is visually different from other regions in the image of interest, allowing the user to focus more on the target area. For example, the target area could be the region where the lesion is located, or the region where a specific structure is located.

[0053] The first focus position can be located within the target region. The image segmentation model can be trained based on the focus training image of the training object, the target region in the focus training image, and the focus training position in the ultrasound training image of the training object. The network architecture of the above image segmentation model is not limited in this embodiment of the invention; for example, a convolutional neural network architecture, a U-Net architecture, etc., can be used. The training method of the image segmentation model can be as follows: Obtain the ultrasound training image of the training object (which may be the same as or different from the target object mentioned above). Then, based on the ultrasound training image, generate the focus training image. Specifically, the above focus training image is obtained by cropping the ultrasound training image based on the ultrasound training image and the focus training position. For example, the relevant steps in step S120 above can be used to determine the above focus training position and crop the ultrasound training image; this embodiment of the invention will not elaborate further here. The above focus training image may include training pixels located at the focus training position in the ultrasound training image.

[0054] In one example, the training location can be treated as a textual cue, and its features can be extracted to obtain the textual features corresponding to the training location. Then, these textual features can be fused with the features extracted from the training image to obtain the training image features, and a prediction can be made based on these training image features to obtain the target prediction region in the training image. In another example, the aforementioned training location can also be represented by the binary image described above, where each pixel in the binary image corresponds to a training pixel in the ultrasound training image. In this binary image, the pixel value corresponding to the training location can be 1, and the pixel value of other pixels can be 0. The binary image can then be cropped to obtain a processed binary image. The pixels in the processed binary image and the training pixels in the training image have the same positional information in the ultrasound training image. Specifically, there can be a one-to-one correspondence between the pixels in the processed binary image and the training pixels in the training image. The processed binary image and the training image can then be superimposed to obtain a superimposed image. The total number of channels in the superimposed image can be equal to the sum of the total number of channels in the processed binary image and the total number of channels in the training image. Then, feature extraction processing can be performed on the superimposed image to obtain the features of the training image, and prediction can be made based on these features to obtain the target prediction region in the training image.

[0055] The image segmentation model to be trained can determine a loss value based at least on the difference between the predicted target region and the target region, in order to adjust the model parameters of the image segmentation model to be trained. It should be understood that the image segmentation model to be trained can also use other types of loss values ​​from related technologies to adjust the model parameters, and this embodiment of the invention does not impose any limitations on this. By iteratively training on multiple training objects, the image segmentation model used in step S130 can be obtained.

[0056] According to the above-described scheme of the present invention, ultrasound data of a target object can be acquired. Then, based on the ultrasound image and a first location of interest within the ultrasound image, the ultrasound image is cropped to obtain an image of interest. Finally, the image of interest and the first location of interest are input into an image segmentation model to obtain the target region within the image of interest. On one hand, the above scheme can crop the ultrasound image based on the first location of interest to obtain an image of interest that the user is more concerned about, resulting in higher accuracy of the target region obtained based on this image of interest. On the other hand, the first location of interest is usually a location that the user is more concerned about. By inputting both the image of interest and the first location of interest into the image segmentation model, the above scheme allows the image segmentation model to focus more on image features related to the first location of interest, thereby improving the accuracy of the obtained target region.

[0057] For example, the above processing method may further include steps S140 and S150.

[0058] In step S140, an ultrasound image of the target object is displayed.

[0059] Ultrasound images can be preprocessed using related technologies (e.g., noise reduction, contrast adjustment, etc.) before being displayed, and this embodiment of the invention does not impose any limitations on these methods.

[0060] In step S150, in response to the user's click operation on the displayed ultrasound image, the click location is designated as the first focus location.

[0061] After identifying the ultrasound image of the target object, the image can be displayed for the user to view. The user can click on the ultrasound image using the input device, and the clicked location can be selected as the first point of focus.

[0062] In one example, an initial region of interest can be obtained using a detection model or algorithm in related technologies, and then the first position of interest can be selected within that region. For instance, the position information of the center point of the region of interest in the ultrasound image can be used as the first position of interest. The representativeness of the position information of this center point is higher than that of other position information in the region of interest, so the position information of this center point in the ultrasound image can be used as the first position of interest. Alternatively, the first position of interest can be displayed so that the user can select an initial pixel within the region of interest, and the position information of the selected initial pixel can be used as the first position of interest. It should be understood that the region of interest selected by the detection model or algorithm in related technologies can serve as an aid to the user in selecting the first position of interest, so the accuracy of the output region of interest does not need to be too high. For example, the trained detection model obtained by directly using the ultrasound training image of the training object and the target region in the ultrasound training image can be used.

[0063] See Figure 2 , Figure 2 A schematic diagram of a method for processing ultrasound images according to an embodiment of the present invention is shown. (In conjunction with...) Figure 2 Users can click on any location in the ultrasound image, and the target area mask will be automatically generated.

[0064] According to the above-described scheme of the present invention, an ultrasound image of the target object can be displayed. Then, in response to a user's click operation on the displayed ultrasound image, the click location is designated as the first location of interest. This scheme allows the user to view the ultrasound image and, based on their practical experience, select the first location of interest within the ultrasound image, thereby improving the representativeness of the first location of interest and enhancing the accuracy of the obtained target area.

[0065] For example, step S120, which involves cropping the ultrasound image based on the ultrasound image and a first location of interest in the ultrasound image to obtain an image of interest, may include steps S121 and S122.

[0066] In step S121, the ultrasound image and the first location of interest are input into the location of interest aggregation model to obtain the aggregation result corresponding to the ultrasound image.

[0067] The aggregation result can include the correlation value between each initial pixel in the ultrasound image and the initial pixel located at the first interest position. This correlation value should be positively correlated with the correlation between the initial pixel and the initial pixel at the first interest position; the specific value can be determined based on the developer's actual needs. For example, the aggregation result can be a heatmap, and the correlation value can be the heat value of the pixel in the heatmap. Pixels in this heatmap can have a correspondence with the initial pixels in the ultrasound image; for example, a one-to-one correspondence (which can be adjusted as needed). In one example, the heat value of each pixel in the heatmap (which can correspond to pixel values; higher heat values ​​indicate a redder pixel, and lower heat values ​​indicate a bluer pixel) can be normalized to 0 to 1. Pixels with a heat value of 1 have the highest correlation between their initial pixel in the ultrasound image and the initial pixel at the first interest position. Pixels with a heat value of 0 have the lowest correlation between their initial pixel in the ultrasound image and the initial pixel at the first interest position. As another example, the aggregation result can also be an adjacency matrix. Each point in the adjacency matrix can correspond to an initial pixel in the ultrasound image, for example, in a one-to-one correspondence (which can be adjusted as needed). The value of each point can be positively correlated with the degree of association. It should be understood that the above aggregation result can simply represent the degree of association between each initial pixel in the ultrasound image and the initial pixel located at the first position of interest; the specific data format is not limited in this embodiment of the invention.

[0068] The attention location aggregation model can be trained based on ultrasound training images of the training object, attention training locations within the ultrasound training images, and the aggregated training results of the ultrasound training images. The network architecture of the attention location aggregation model described above is not limited in this embodiment of the invention; for example, a convolutional neural network architecture, a U-Net architecture, etc., can be used. The training method for the attention location aggregation model is as follows: obtain ultrasound training images of the training object (which may be the same as or different from the target object mentioned above) and attention training locations within the ultrasound training images.

[0069] In one example, the training location of interest can be treated as a textual cue, and its features can be extracted to obtain the textual features corresponding to the training location. Then, these textual features can be fused with the ultrasound training image features obtained from feature extraction on the ultrasound training image to perform prediction based on the ultrasound training image features, resulting in an aggregated prediction result for the ultrasound training image. In another example, the aforementioned training location of interest can also be represented by the binary image described above, where each pixel in the binary image corresponds to each initial pixel in the ultrasound training image. In this binary image, the pixel value corresponding to the training location of interest can be 1, and the pixel value of other pixels can be 0. Then, this binary image can be superimposed with the ultrasound training image to obtain a superimposed image. The total number of channels in the superimposed image can be equal to the sum of the total number of channels in the binary image and the total number of channels in the ultrasound training image. Then, feature extraction processing can be performed on the superimposed image to obtain ultrasound training image features, and prediction based on these features can yield an aggregated prediction result for the ultrasound training image.

[0070] The attention location aggregation model to be trained can determine a loss value based at least on the difference between the aggregation prediction result and the aggregation training result, in order to adjust the model parameters of the attention location aggregation model to be trained. The aggregation training result may include the correlation value between each training pixel in the ultrasound training image and the training pixels located at the attention training location.

[0071] It should be understood that the attention location aggregation model to be trained can also use other types of loss values ​​from related technologies to adjust the model parameters, and this embodiment of the invention does not impose any limitations on this. By iteratively training on multiple training objects, the attention location aggregation model used in step S121 can be obtained.

[0072] In step S122, the initial pixels with a correlation degree value greater than the first correlation degree threshold in the aggregation result are taken as target pixels, and the ultrasound image is cropped based on the position information of the target pixels in the ultrasound image to obtain the image of interest.

[0073] Optionally, the target pixel can be used as a pixel in the image of interest, and the ultrasound image can be directly cropped accordingly to obtain the image of interest. Alternatively, the maximum and minimum position information of the target pixel can be determined based on the position information of each target pixel in the ultrasound image. Then, the circumscribed rectangle is determined using the maximum and minimum position information. In one example, the image corresponding to the circumscribed rectangle in the ultrasound image can be used as the image of interest. In another example, a threshold for the expansion ratio of the circumscribed rectangle can be set, and the image corresponding to the expanded circumscribed rectangle in the ultrasound image can be used as the image of interest.

[0074] The specific values ​​of the first correlation threshold and the proportion threshold are not limited in this embodiment of the invention, and can be determined by the developers or users according to the actual situation.

[0075] See Figure 3 , Figure 3 A schematic diagram of an ultrasound image processing method according to an embodiment of the present invention is shown, in conjunction with... Figure 3 The primary location of interest can be represented by a binary image. This binary image can be overlaid with an ultrasound image and input into a location of interest aggregation model to output the aggregation result. Figure 3 (This is represented by a heatmap). The aggregation result can be cropped to obtain the image of interest mentioned above, or the binary image can be cropped. Finally, the image of interest and the cropped binary image can be input into the image segmentation model to obtain the target region. Figure 3 (This is represented in the form of a masked image).

[0076] According to the above-described scheme of the present invention, an ultrasound image and a first location of interest can be input into a location of interest aggregation model to obtain an aggregation result corresponding to the ultrasound image. Then, initial pixels with a correlation value greater than a first correlation threshold in the aggregation result are used as target pixels, and the ultrasound image is cropped based on the position information of the target pixels in the ultrasound image to obtain a location of interest image. The location of interest aggregation model in the above scheme is trained based on the ultrasound training image of the training object, the location of interest training in the ultrasound training image, and the aggregation training result of the ultrasound training image. Therefore, after training, it can combine the first location of interest in the ultrasound image, improving the representativeness of the aggregation result and helping to improve the accuracy of determining the target region.

[0077] For example, step S122, which selects the initial pixel whose correlation value in the aggregation result is greater than the first correlation threshold as the target pixel, may include steps S1221a and S1222a.

[0078] In step S1221a, the first total number of initial pixels in the aggregation result whose correlation value is greater than the second correlation threshold is determined.

[0079] The second correlation threshold can be less than the first correlation threshold. The specific value of the second correlation threshold is not limited in this embodiment of the invention and can be determined by the developer or user based on actual circumstances. It should be understood that the second correlation threshold can also be 0 or a decimal close to 0.

[0080] In step S1222a, the initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold corresponding to the first total number are taken as target pixels.

[0081] In one example, each first total number can correspond to its own first correlation threshold. In another example, each total number interval can correspond to its own first correlation threshold. Multiple total number intervals can be the numerical ranges where the first total number may be distributed. The total number interval in which the first total number is located can be determined first, and then the first correlation threshold corresponding to that total number interval can be used as the first correlation threshold corresponding to the aforementioned first total number.

[0082] The value of the first correlation threshold can be positively correlated with the value of the first total number. In practical terms, a smaller first total number indicates a smaller total number of initial pixels associated with the initial pixel located at the first location of interest. In other words, the actual region of interest (e.g., the actual lesion) may appear smaller in the ultrasound image (especially submucosal tumors, which are smaller in the early stages). Therefore, the value of the first correlation threshold can be positively correlated with the value of the first total number, so that the first correlation threshold decreases as the actual region of interest decreases, thereby identifying areas in the ultrasound image where smaller actual regions of interest might exist, thus serving as the image of interest.

[0083] Referring to the example of the aggregation result being a heatmap in step S121, for instance, the second correlation threshold can be 0 (or a value close to 0), thus determining the first total number of initial pixels corresponding to pixels with heat values ​​greater than 0 in the heatmap. The initial pixels corresponding to pixels with heat values ​​greater than 0 are the initial pixels associated with the initial pixels located at the first interest position. Then, the initial pixels corresponding to pixels with heat values ​​greater than the first correlation threshold corresponding to the first total number in the heatmap can be used as the aforementioned target pixels.

[0084] Referring to the example of the adjacency matrix obtained from the aggregation result in step S121, for instance, the second association threshold can be 0 (or a value close to 0). This allows us to determine the first total number of initial pixels corresponding to points in the adjacency matrix with values ​​greater than 0. The initial pixels corresponding to these points are those associated with the initial pixel located at the first point of interest. Then, the initial pixels corresponding to the points in the adjacency matrix with values ​​greater than the first total corresponding to the first association threshold can be used as the target pixels.

[0085] According to the above-described scheme of the present invention, a first total number of initial pixels in the aggregation result whose correlation value is greater than a second correlation threshold can be determined. Then, the initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold corresponding to the first total number are taken as target pixels. The above scheme can first determine the first total number, and then use the first correlation threshold corresponding to the first total number to determine whether the initial pixel is a target pixel. Since the first correlation threshold is flexibly determined according to the first total number, it can be adapted to situations where the actual region of interest is small in the ultrasound image. Specifically, when the actual region of interest is small, the above scheme can also use a smaller first correlation threshold to find pixels that are associated with the initial pixel located at the first point of interest, thereby avoiding missing the region of interest where the small actual region of interest is located.

[0086] For example, step S122, which selects the initial pixel whose correlation value in the aggregation result is greater than the first correlation threshold as the target pixel, includes steps S1221b and S1222b.

[0087] In step S1221b, a second total number of initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold is determined.

[0088] Referring to the example in step S121 where the aggregation result is a heatmap, the total number of initial pixels corresponding to pixels whose heat values ​​in the heatmap are greater than the first correlation threshold can be used as the aforementioned second total. Referring to the example in step S121 where the aggregation result is an adjacency matrix, the total number of initial pixels corresponding to points in the adjacency matrix whose values ​​are greater than the first correlation threshold can be used as the aforementioned second total.

[0089] In step S1222b, if the second total number is less than the total number threshold, the first correlation threshold is reduced based on the adjustment step size, and the second total number of initial pixels with correlation values ​​greater than the first correlation threshold in the aggregation result is re-executed based on the reduced first correlation threshold. If the second total number is greater than or equal to the total number threshold, the initial pixels with correlation values ​​greater than the target correlation threshold in the aggregation result are taken as target pixels.

[0090] In one example, the reduced first correlation threshold can be obtained by directly subtracting the adjustment step size from the first correlation threshold. In another example, the adjustment step size can be multiplied by the weights corresponding to the number of adjustments (the number of times the first correlation threshold is reduced) to obtain a new adjustment step size. Then, the reduced first correlation threshold is obtained by subtracting the new adjustment step size from the first correlation threshold. For example, the weights can be positively correlated with the number of adjustments.

[0091] The target association threshold is a first association threshold where the determined second total number is greater than or equal to the total threshold. In practical scenarios, if the second total number is less than the total threshold, it can be considered that the number of initial pixels associated with the initial pixel located at the first interest position is small, possibly due to an excessively high first association threshold. If the region image is directly determined based on this first association threshold, the region image may be too small (potentially misidentification), diminishing the significance of target region identification. Therefore, the first association threshold can be reduced, and step S1221b can be re-executed based on the reduced first association threshold. When the second total number is greater than or equal to the total threshold, it can be considered that the number of initial pixels associated with the initial pixel located at the first interest position is large, making target region identification more significant. The first association threshold where the determined second total number is greater than or equal to the total threshold can be used as the aforementioned target association threshold, and initial pixels in the aggregation result with an association value greater than the target association threshold can be used as target pixels.

[0092] According to the above-described scheme of the present invention, a second total number of initial pixels in the aggregation result whose correlation values ​​are greater than a first correlation threshold can be determined. Then, if the second total number is less than a total threshold, the first correlation threshold is reduced based on an adjustment step size, and the determination of the second total number of initial pixels in the aggregation result whose correlation values ​​are greater than the first correlation threshold is re-executed based on the reduced first correlation threshold. If the second total number is greater than or equal to the total threshold, the initial pixels in the aggregation result whose correlation values ​​are greater than a target correlation threshold are taken as target pixels. The above scheme can improve the representativeness of the image of interest, thereby improving the accuracy of determining the target region.

[0093] For example, step S110, acquiring ultrasound images of the target object, includes: acquiring a sequence of ultrasound images of the target object.

[0094] In this example, the ultrasound data may include echo data received at multiple times. Therefore, based on this ultrasound data, the image reconstruction algorithm described above can be used to obtain ultrasound images at multiple times, that is, an ultrasound image sequence.

[0095] The above method may further include: determining a second location of interest in a number of other images adjacent to the ultrasound image in the ultrasound image sequence based on the target region, and determining the target region in the other images based on the second location of interest.

[0096] After identifying the target region in an ultrasound image, the location information of a pixel within that target region can be used as a second location of interest (ROI) in other images. The ROI determined based on this target region is more representative, which helps improve the accuracy of the identified target region in other images. In practical scenarios, adjacent ultrasound images in a sequence typically target the same actual region of interest; therefore, the actual target regions in adjacent ultrasound images usually do not differ significantly. Thus, the location information of pixels within the identified target region in an ultrasound image can be used as the ROI in adjacent images to adapt to the distribution of the actual region of interest.

[0097] In one example, the location information of the center point of the target region can be used as a second location of interest in other images. Since the representativeness of this center point's location information is higher than that of other locations in the target region, its location information in the ultrasound image can be used as a first location of interest. In another example, pixels in the target region can be filtered to determine the location information of more representative pixels in the ultrasound image as a second location of interest in other images; this embodiment of the invention does not impose any limitations on this.

[0098] The aforementioned quantity threshold can be determined based on the actual needs of developers or users. The larger the quantity threshold, the higher the efficiency of determining the second location of interest, but the lower the representativeness of the second location of interest may be. For example, the quantity threshold can be 1, meaning that the second location of interest in an ultrasound image where the second location of interest has not yet been determined is determined based on adjacent ultrasound images where the second location of interest or the first location of interest has been determined. The second location of interest determined in this way is more representative.

[0099] Based on the second location of interest, and according to the relevant steps S120 and S130 above, the target region in the other images can be determined. Specifically, based on the other images and the second location of interest in those other images, the other images can be cropped to obtain the image of interest corresponding to those other images. The specific process can be referred to in step S120, and will not be repeated here in this embodiment of the invention. Then, the image of interest corresponding to the other image and the second location of interest can be input into the image segmentation model to obtain the target region in the image of interest corresponding to the other image. The specific process can be referred to in step S130, and will not be repeated here in this embodiment of the invention.

[0100] According to the above-described scheme of the present invention, an ultrasound image sequence of a target object can be acquired. Then, based on the target region, a second position of interest is determined in a threshold number of other images adjacent to the ultrasound image in the ultrasound image sequence, and based on the second position of interest, the target region in the other images is determined. The above scheme can automatically determine the second position of interest in adjacent other images from the target region in the ultrasound image. On the one hand, the above scheme eliminates the need for the user to manually select the second position of interest in other images, especially when the ultrasound image sequence typically includes a large number of ultrasound images, thus greatly reducing labor costs; on the other hand, the second position of interest determined by the above scheme is based on the target region in adjacent ultrasound images, therefore the second position of interest is more representative, which helps improve the accuracy of the target region obtained in other images.

[0101] For example, step S110, acquiring ultrasound images of the target object, may include: acquiring a sequence of ultrasound images of the target object.

[0102] In this example, the ultrasound data may include echo data received at multiple times. Therefore, based on this ultrasound data, the image reconstruction algorithm described above can be used to obtain ultrasound images at multiple times, that is, an ultrasound image sequence.

[0103] The above method may also include steps S180 and S190.

[0104] In step S180, for each ultrasound image in the ultrasound image sequence, a target region in the ultrasound image is determined based on a first location of interest in the ultrasound image, and the two-dimensional size information of the target region is determined.

[0105] The specific process for determining the target region in the ultrasound image based on the first location of interest in the ultrasound image can be referred to the relevant processes of steps S120 and S130, which will not be elaborated here in this embodiment of the invention.

[0106] The two-dimensional dimensional information of the target area mentioned above may include the maximum horizontal dimension and the maximum vertical dimension of the target area. It should be understood that the two-dimensional dimensional information may also include the horizontal and vertical dimensions at different locations within the target area to improve the representativeness of the two-dimensional dimensional information for the shape of the target area. The aforementioned two-dimensional dimensional information can be automatically measured based on the target area using relevant measurement algorithms; this will not be elaborated upon in the embodiments of the present invention.

[0107] In step S190, the three-dimensional dimension information of the region where the part of interest is located is determined based on the two-dimensional dimension information of each target region.

[0108] In practical scenarios, the target region in an ultrasound image is typically two-dimensional, while the actual region of interest (ROI) within the target object is three-dimensional. Therefore, the target region in a single ultrasound image often loses its dimensional information along the ultrasound probe retraction direction. However, ultrasound images in a sequence are arranged sequentially and can be considered as continuously distributed along the ultrasound probe retraction direction. Therefore, the three-dimensional dimensional information of the region containing the ROI can be determined based on the two-dimensional dimensional information of each target region. For example, the physical interval between ultrasound images can be a length threshold. The specific formula can be found as: (total number of ultrasound images - 1) * length threshold, which can also be adjusted by developers or users according to the actual situation. If two ultrasound images are used to determine the three-dimensional dimensional information, the dimensional information of the actual ROI along the ultrasound probe retraction direction can be considered as one times this length threshold.

[0109] According to the above-described scheme of the present invention, an ultrasound image sequence can be acquired. Then, for each ultrasound image in the sequence, a target region is determined based on a first location of interest in the ultrasound image, and the two-dimensional size information of the target region is determined. Finally, based on the two-dimensional size information of each target region, the three-dimensional size information of the region containing the region of interest is determined. The above scheme can reconstruct the three-dimensional size information of the target region in each ultrasound image from the two-dimensional size information of the ultrasound image sequence into the three-dimensional size information of the region containing the region of interest. This allows for the effective representation of the actual three-dimensional size of the region of interest through the three-dimensional size information, which is beneficial for assisting doctors in qualitatively identifying the region of interest by combining this three-dimensional size information.

[0110] This invention also provides an apparatus for processing ultrasound images. Figure 4 A schematic block diagram of an ultrasound image processing apparatus 200 according to an embodiment of the present invention is shown. (In conjunction with...) Figure 4 As shown, the ultrasound image processing device 200 may include an image acquisition module 210, a focus image generation module 220, and a target region generation module 230.

[0111] The image acquisition module 210 is used to acquire ultrasound images of the target object.

[0112] The focus image generation module 220 is used to crop the ultrasound image based on the ultrasound image and a first focus position in the ultrasound image to obtain a focus image, wherein the focus image includes the initial pixels in the ultrasound image located at the first focus position.

[0113] The target region generation module 230 is used to input the image of interest and the first position of interest into the image segmentation model to obtain the target region in the image of interest. The first position of interest is located in the target region. The image segmentation model is trained based on the training image of the training object, the target region in the training image of interest, and the training position of interest in the ultrasound training image of the training object. The training image of interest is obtained by cropping the ultrasound training image based on the ultrasound training image and the training position of interest.

[0114] For example, the image generation module 220 includes an aggregation result generation module and a first generation module.

[0115] The aggregation result generation module is used to input the ultrasound image and the first interest position into the interest position aggregation model to obtain the aggregation result corresponding to the ultrasound image. The interest position aggregation model is trained based on the ultrasound training image, the interest training position and the aggregation training result of the ultrasound training image. The aggregation result includes the correlation value between each initial pixel in the ultrasound image and the initial pixel located at the first interest position.

[0116] The first generation module is used to take the initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold as target pixels, and to crop the ultrasound image based on the position information of the target pixels in the ultrasound image to obtain the image of interest.

[0117] For example, the first generation module includes a first determination module and a first target pixel determination module.

[0118] The first determining module is used to determine the first total number of initial pixels in the aggregation result whose correlation degree value is greater than the second correlation degree threshold, wherein the second correlation degree threshold is less than the first correlation degree threshold.

[0119] The first target pixel determination module is used to select initial pixels in the aggregation result whose correlation degree value is greater than the first correlation degree threshold corresponding to the first total number as target pixels, wherein the value of the first correlation degree threshold is positively correlated with the value of the first total number.

[0120] For example, the first generation module includes: a second determination module, a re-execution module, and a second target pixel determination module.

[0121] The second determining module is used to determine the second total number of initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold.

[0122] The re-execution module is used to reduce the first correlation threshold based on the adjustment step size when the second total number is less than the total threshold, and to re-execute the determination of the second total number of initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold based on the reduced first correlation threshold.

[0123] The second target pixel determination module is used to select initial pixels in the aggregation result whose correlation degree value is greater than the target correlation degree threshold as target pixels when the second total number is greater than or equal to the total number threshold. The target correlation degree threshold is the first correlation degree threshold that is determined to be greater than or equal to the second total number threshold.

[0124] For example, the image acquisition module 210 includes: a first image sequence acquisition module.

[0125] The first image sequence acquisition module is used to acquire ultrasound image sequences of the target object.

[0126] The ultrasound image processing device 200 also includes an additional target area determination module.

[0127] The other target region determination module is used to determine, based on the target region, a second location of interest in a number of other images adjacent to the ultrasound image in the ultrasound image sequence, based on a threshold number of other images, and to determine the target region in the other images based on the second location of interest.

[0128] For example, the image acquisition module 210 includes: a second image sequence acquisition module.

[0129] The second image sequence acquisition module is used to acquire ultrasound image sequences of the target object.

[0130] The ultrasound image processing device 200 also includes a two-dimensional dimension information determination module and a three-dimensional dimension information determination module.

[0131] The two-dimensional size information determination module is used to determine the target region in each ultrasound image in the ultrasound image sequence based on the first location of interest in the ultrasound image, and to determine the two-dimensional size information of the target region.

[0132] The 3D dimension information determination module is used to determine the 3D dimension information of the region where the part of interest is located based on the 2D dimension information of each target region.

[0133] For example, the ultrasound image processing device 200 further includes an ultrasound image display module and a click response module.

[0134] The ultrasound image display module is used to display ultrasound images of the target object.

[0135] The click response module is used to respond to user clicks on the displayed ultrasound image, and the click location is used as the first point of focus.

[0136] According to another aspect of the present invention, an electronic device is also provided. Figure 5A schematic block diagram of an electronic device 300 according to an embodiment of the present invention is shown. Figure 5 As shown, the electronic device 300 includes a processor 310 and a memory 320. The memory 320 stores a computer program, and the computer program instructions are executed by the processor 310 to perform the aforementioned ultrasound image processing method. Exemplarily, the electronic device can be various ultrasound imaging devices, such as an ultrasound diagnostic instrument or an ultrasound imaging workstation. An ultrasound diagnostic instrument may include a probe, a processor, and a memory for performing ultrasound imaging examinations. Specifically, an ultrasound diagnostic instrument may have different probe types to examine different areas, such as the heart, abdomen, gynecological organs, and breasts. An ultrasound imaging workstation may be a device integrating functional modules such as patient registration, image acquisition, diagnostic editing, report printing, image post-processing, medical record retrieval, and statistical analysis. The ultrasound imaging workstation can be communicatively connected to the ultrasound diagnostic instrument, for example, through any wired or wireless communication method. The ultrasound diagnostic instrument can transmit the acquired ultrasound echo signals to the ultrasound imaging workstation for processing, analysis, and storage.

[0137] Furthermore, according to another aspect of the present invention, a storage medium is provided, on which program instructions are stored. When the program instructions are executed by a computer or processor, the computer or processor performs corresponding steps of the ultrasound image processing method described in the embodiments of the present invention, and is used to implement corresponding modules in the ultrasound image processing apparatus or the electronic device described in the embodiments of the present invention. The storage medium may, for example, include a memory card of a smartphone, a storage component of a tablet computer, a hard disk of a personal computer, a read-only memory (ROM), an erasable programmable read-only memory (EPROM), a portable compact disc read-only memory (CD-ROM), a USB memory, or any combination of the above storage media. The computer-readable storage medium may be any combination of one or more computer-readable storage media. According to yet another aspect of the present invention, a computer program product is also provided, including computer program instructions. When the computer program instructions are executed by a computer or processor, the computer or processor performs corresponding steps of the ultrasound image processing method described above.

[0138] Those skilled in the art can understand the specific implementation scheme of the above-mentioned electronic device and storage medium by reading the relevant description of the ultrasound image processing method. For the sake of brevity, it will not be described in detail here.

[0139] Although exemplary embodiments have been described herein with reference to the accompanying drawings, it should be understood that the above exemplary embodiments are merely illustrative and are not intended to limit the scope of the invention thereto. Various changes and modifications can be made therein by those skilled in the art without departing from the scope and spirit of the invention. All such changes and modifications are intended to be included within the scope of the invention as claimed in the appended claims.

[0140] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementations should not be considered beyond the scope of this invention.

[0141] In the several embodiments provided by this invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another device, or some features may be ignored or not executed.

[0142] Numerous specific details are set forth in the specification provided herein. However, it will be understood that embodiments of the invention may be practiced without these specific details. In some instances, well-known methods, structures, and techniques have not been shown in detail so as not to obscure the understanding of this specification.

[0143] Similarly, it should be understood that, in order to streamline the invention and aid in understanding one or more of the various aspects of the invention, features of the invention are sometimes grouped together in a single embodiment, figure, or description thereof in the description of exemplary embodiments of the invention. However, this approach should not be construed as reflecting an intention that the claimed invention requires more features than are expressly recited in each claim. Rather, as reflected in the corresponding claims, its inventive point lies in solving the corresponding technical problem with fewer features than all of those in a single disclosed embodiment. Therefore, the claims following the detailed description are hereby expressly incorporated into that detailed description, wherein each claim itself is a separate embodiment of the invention.

[0144] Those skilled in the art will understand that, apart from the mutual exclusion of features, all features disclosed in this specification (including the accompanying claims, abstract, and drawings) and all processes or units of any method or apparatus so disclosed can be combined in any combination. Unless otherwise expressly stated, each feature disclosed in this specification (including the accompanying claims, abstract, and drawings) may be replaced by an alternative feature that serves the same, equivalent, or similar purpose.

[0145] Furthermore, those skilled in the art will understand that although some embodiments described herein include certain features but not others included in other embodiments, combinations of features from different embodiments are intended to be within the scope of the invention and form different embodiments. For example, in the claims, any of the claimed embodiments can be used in any combination.

[0146] The various component embodiments of the present invention can be implemented in hardware, or as software modules running on one or more processors, or a combination thereof. Those skilled in the art will understand that microprocessors or digital signal processors (DSPs) can be used in practice to implement some or all of the functions of some modules in the ultrasound image processing apparatus according to embodiments of the present invention. The present invention can also be implemented as an apparatus program (e.g., a computer program and computer program product) for performing some or all of the methods described herein. Such programs implementing the present invention can be stored on a computer-readable medium or can be in the form of one or more signals. Such signals can be downloaded from an Internet website, provided on a carrier signal, or provided in any other form.

[0147] It should be noted that the above embodiments are illustrative of the invention and not restrictive, and that those skilled in the art can devise alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses should not be construed as limiting the claims. The word "comprising" does not exclude the presence of elements or steps not listed in the claims. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several different elements and by means of a suitably programmed computer. In the unit claims enumerating several means, several of these means may be embodied by the same item of hardware. The use of the words first, second, and third, etc., does not indicate any order. These words can be interpreted as names.

[0148] The above description is merely a specific embodiment of the present invention or an explanation of that embodiment. The scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. The scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A method for processing ultrasound images, characterized in that, The method includes: Acquire ultrasound images of the target object; Based on the ultrasound image and a first point of interest in the ultrasound image, the ultrasound image is cropped to obtain a point of interest image, wherein the point of interest image includes an initial pixel in the ultrasound image located at the first point of interest position; The image of interest and the first position of interest are input into the image segmentation model to obtain the target region in the image of interest, wherein the first position of interest is located in the target region. The image segmentation model is trained based on the training image of the training object, the target region in the training image of interest, and the training position of interest in the ultrasound training image of the training object. The training image of interest is obtained by cropping the ultrasound training image based on the ultrasound training image and the training position of interest.

2. The method as described in claim 1, characterized in that, The step of cropping the ultrasound image based on the ultrasound image and a first position of interest in the ultrasound image to obtain an image of interest includes: The ultrasound image and the first attention position are input into the attention position aggregation model to obtain the aggregation result corresponding to the ultrasound image. The attention position aggregation model is trained based on the ultrasound training image, the attention training position and the aggregation training result of the ultrasound training image. The aggregation result includes the correlation value between each initial pixel in the ultrasound image and the initial pixel located at the first attention position. The initial pixels with a correlation value greater than a first correlation threshold in the aggregation result are taken as target pixels, and the ultrasound image is cropped based on the position information of the target pixels in the ultrasound image to obtain the image of interest.

3. The method as described in claim 2, characterized in that, The step of using initial pixels in the aggregation result that are greater than the first correlation threshold as target pixels includes: Determine the first total number of initial pixels in the aggregation result whose correlation degree value is greater than the second correlation degree threshold, wherein the second correlation degree threshold is less than the first correlation degree threshold; The initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold corresponding to the first total number are taken as the target pixels, wherein the value of the first correlation threshold is positively correlated with the value of the first total number.

4. The method as described in claim 2, characterized in that, The step of using initial pixels with correlation values ​​greater than a first correlation threshold in the aggregation result as target pixels includes: Determine the second total number of initial pixels in the aggregation result whose correlation value is greater than the first correlation threshold; If the second total number is less than the total number threshold, the first correlation threshold is reduced based on the adjustment step size, and the second total number of the initial pixels whose correlation value in the aggregation result is greater than the first correlation threshold is re-executed based on the reduced first correlation threshold. If the second total number is greater than or equal to the total number threshold, the initial pixels in the aggregation result whose correlation value is greater than the target correlation threshold are taken as target pixels, wherein the target correlation threshold is the first correlation threshold that is determined to be greater than or equal to the total number threshold.

5. The method as described in claim 1, characterized in that, The acquisition of ultrasound images of the target object includes: Obtain the ultrasound image sequence of the target object; The method further includes: Based on the target region, a second position of interest is determined in a threshold number of other images adjacent to the ultrasound image in the ultrasound image sequence, and based on the second position of interest, a target region in the other images is determined.

6. The method as described in claim 1, characterized in that, The acquisition of ultrasound images of the target object includes: Obtain the ultrasound image sequence of the target object; The method further includes: For each ultrasound image in the ultrasound image sequence, based on the first location of interest in the ultrasound image, the target region in the ultrasound image is determined, and the two-dimensional size information of the target region is determined. Based on the two-dimensional dimension information of each target region, the three-dimensional dimension information of the region where the part of interest is located is determined.

7. The method as described in claim 1, characterized in that, The method further includes: displaying an ultrasound image of the target object; In response to a user's click on the displayed ultrasound image, the click location is designated as the first location of interest.

8. An electronic device, characterized in that, The device includes a memory and a processor, wherein: the memory is used to store a computer program; and the processor is used to execute the computer program to implement the ultrasound image processing method as described in any one of claims 1-7.

9. A storage medium storing a computer program / instructions, characterized in that, The computer program / instructions, when running, are used to perform the ultrasound image processing method as described in any one of claims 1-7.

10. A computer program product comprising computer program instructions, characterized in that, The computer program instructions, when executed by a processor, are used to perform the ultrasound image processing method as described in any one of claims 1-7.

Images