Image segmentation methods and related products

Abstract

The present application discloses an image segmentation method and related products. The method comprises: obtaining a target ultrasound image of a target tissue, a first foreground image and a background image, wherein the target ultrasound image is an image obtained by ultrasonic scanning of the target tissue, the first foreground image comprises an image of the target tissue in the target ultrasound image, and the background image comprises an image of an image area other than the target tissue in the target ultrasound image; determining target positioning information of the target tissue based on the first foreground image and the background image; determining an image area of ​​the target tissue in the target ultrasound image based on the target positioning information, and obtaining a first target image.

Description

Technical Field

[0001] The present application relates to the technical field of image segmentation, and in particular to an image segmentation method and related products. Background Art

[0002] With the advancement of computer technology, image segmentation technology is becoming more and more mature, but people's requirements for the accuracy of image segmentation are also getting higher and higher. In the medical field, medical images with low segmentation accuracy will lead to reduced accuracy of medical judgment based on the medical images, or even inability to make medical judgments based on the medical images. Therefore, how to improve the accuracy of image segmentation is a very important issue. Summary of the invention

[0003] The present application provides an image segmentation method and related products, wherein the related products include an image segmentation device, an electronic device, a computer-readable storage medium, and a computer program product.

[0004] In a first aspect, an image segmentation method is provided, the method comprising:

[0005] Acquire a target ultrasound image of a target tissue, a first foreground image, and a background image, wherein the target ultrasound image is an image obtained by performing ultrasound scanning on the target tissue, the first foreground image includes an image of the target tissue in the target ultrasound image, and the background image includes an image of an image area other than the target tissue in the target ultrasound image;

[0006] Determine target positioning information of the target tissue based on the first foreground image and the background image;

[0007] Based on the target positioning information, an image region of the target tissue in the target ultrasound image is determined to obtain a first target image.

[0008] In combination with any embodiment of the present application, the target positioning information includes a position of a first feature point in the target ultrasound image, and the first feature point includes a point at a boundary of the target tissue;

[0009] The step of determining an image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image includes:

[0010] Determining the first feature point from the target ultrasound image based on a position of the first feature point in the target ultrasound image;

[0011] Determine an image area surrounded by the first feature points in the target ultrasound image to obtain a first area;

[0012] The first target image is obtained based on the image of the first area in the target ultrasound image.

[0013] In combination with any embodiment of the present application, the target positioning information includes a position of a second feature point of an image of a diseased tissue in the target tissue in the target ultrasound image, where the second feature point is a point at a boundary of the image of the diseased tissue in the target ultrasound image;

[0014] Before obtaining the first target image based on the image of the first area in the target ultrasound image, the method further includes:

[0015] Determining the second feature point from the target ultrasound image based on a position of the second feature point in the target ultrasound image;

[0016] The step of obtaining the first target image based on the image of the first area in the target ultrasound image includes:

[0017] Determine an image area enclosed by the second feature points in the first target image to obtain a second area;

[0018] An image region in the first target image corresponding to the second region is marked to obtain a second target image.

[0019] In combination with any embodiment of the present application, the step of marking an image area in the first target image corresponding to the second area to obtain the second target image includes:

[0020] The image area corresponding to the second area in the first target image is deleted to obtain the second target image.

[0021] In combination with any embodiment of the present application, the determining the target positioning information of the target tissue based on the first foreground image and the background image includes:

[0022] Determine background tissue pixels whose grayscale values ​​of pixels in the background image are less than a first threshold;

[0023] Determine foreground tissue pixels whose grayscale values ​​of pixels in the first foreground image are greater than a second threshold;

[0024] Correcting the foreground tissue pixels based on the background tissue pixels to obtain target tissue pixels;

[0025] The target positioning information is obtained according to the position information of the target tissue pixel in the first foreground image.

[0026] In combination with any embodiment of the present application, the correcting the foreground tissue pixel based on the background tissue pixel to obtain the target tissue pixel includes:

[0027] Determining a first position of the background tissue pixel in the background image;

[0028] Determine a pixel at a first position in the first foreground image to obtain a candidate pixel;

[0029] Determine pixels among the foreground tissue pixels and pixels among the candidate pixels that are different from the foreground tissue pixels to obtain the first reference pixels;

[0030] Determine a pixel different from the first reference pixel within a preset distance from the first reference pixel in the first foreground image to obtain a second reference pixel;

[0031] The first reference pixel and the second reference pixel are used as the target tissue pixels.

[0032] In a second aspect, an image segmentation device is provided, the image segmentation device comprising:

[0033] an acquisition unit, configured to acquire a target ultrasound image of a target tissue, a first foreground image, and a background image, wherein the target ultrasound image is an image obtained by performing ultrasound scanning on the target tissue, the first foreground image includes an image of the target tissue in the target ultrasound image, and the background image includes an image of an image region other than the target tissue in the target ultrasound image;

[0034] a determining unit, configured to determine target positioning information of the target tissue based on the first foreground image and the background image;

[0035] The determination unit is further used to determine the image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image.

[0036] In combination with any embodiment of the present application, the target positioning information includes a position of a first feature point in the target ultrasound image, and the first feature point includes a point of a boundary of the target tissue;

[0037] The above-mentioned determination unit is specifically used for:

[0038] Determining the first feature point from the target ultrasound image based on a position of the first feature point in the target ultrasound image;

[0039] Determine an image area surrounded by the first feature points in the target ultrasound image to obtain a first area;

[0040] The first target image is obtained based on the image of the first area in the target ultrasound image.

[0041] In combination with any embodiment of the present application, the target positioning information includes a position of a second feature point of an image of a diseased tissue in the target tissue in the target ultrasound image, where the second feature point is a point at a boundary of the image of the diseased tissue in the target ultrasound image;

[0042] The above determination unit is further used for:

[0043] Determining the second feature point from the target ultrasound image based on a position of the second feature point in the target ultrasound image;

[0044] The image segmentation device also includes a marking unit, which is used to:

[0045] Determine an image area enclosed by the second feature points in the first target image to obtain a second area;

[0046] An image region in the first target image corresponding to the second region is marked to obtain a second target image.

[0047] In conjunction with any embodiment of the present application, the marking unit is specifically used for:

[0048] The image area corresponding to the second area in the first target image is deleted to obtain the second target image.

[0049] In conjunction with any embodiment of the present application, the above-mentioned determination unit is specifically used to:

[0050] Determine background tissue pixels whose grayscale values ​​of pixels in the background image are less than a first threshold;

[0051] Determine foreground tissue pixels whose grayscale values ​​of pixels in the first foreground image are greater than a second threshold;

[0052] Correcting the foreground tissue pixels based on the background tissue pixels to obtain target tissue pixels;

[0053] The target positioning information is obtained according to the position information of the target tissue pixel in the first foreground image.

[0054] In conjunction with any embodiment of the present application, the above-mentioned determination unit is specifically used to:

[0055] Determining a first position of the background tissue pixel in the background image;

[0056] Determine a pixel at a first position in the first foreground image to obtain a candidate pixel;

[0057] Determine pixels among the foreground tissue pixels and pixels among the candidate pixels that are different from the foreground tissue pixels to obtain the first reference pixels;

[0058] Determine a pixel different from the first reference pixel within a preset distance from the first reference pixel in the first foreground image to obtain a second reference pixel;

[0059] The first reference pixel and the second reference pixel are used as the target tissue pixels.

[0060] According to a third aspect, an electronic device is provided, comprising: a processor and a memory, wherein the memory is used to store computer program code, wherein the computer program code comprises computer instructions, and when the processor executes the computer instructions, the electronic device executes the first aspect and any one of its embodiments described above.

[0061] In a fourth aspect, another electronic device is provided, comprising: a processor, a sending device, an input device, an output device and a memory, wherein the memory is used to store computer program code, and the computer program code includes computer instructions. When the processor executes the computer instructions, the electronic device executes the first aspect and any one of its embodiments.

[0062] In a fifth aspect, a computer-readable storage medium is provided, in which a computer program is stored. The computer program includes program instructions. When the program instructions are executed by a processor, the processor is caused to execute the first aspect and any one of its embodiments.

[0063] In a sixth aspect, a computer program product is provided, wherein the computer program product comprises a computer program or instructions, and when the computer program or instructions are run on a computer, the computer is enabled to execute the above-mentioned first aspect and any embodiment thereof.

[0064] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present application.

[0065] In an embodiment of the present application, after the image segmentation device acquires a target ultrasound image of the target tissue, including a first foreground image of the image of the target tissue in the target ultrasound image and a background image of the image area other than the target tissue in the target ultrasound image, the image segmentation device determines the target positioning information of the target tissue based on the first foreground image and the background image, and then determines the image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image. After determining the target positioning information of the target tissue in the target ultrasound image, the image segmentation device segments the image of the target tissue in the target ultrasound image based on the target positioning information. Compared with directly segmenting the image of the target tissue in the target ultrasound image, the target positioning information of the target tissue can enable the image segmentation device to more accurately identify the target tissue during the segmentation process, thereby improving the accuracy of the first target image obtained by segmentation. BRIEF DESCRIPTION OF THE DRAWINGS

[0066] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the background technology, the drawings required for use in the embodiments of the present application or the background technology will be described below.

[0067] The drawings herein are incorporated into the specification and constitute a part of the specification. These drawings illustrate embodiments consistent with the present application and are used together with the specification to illustrate the technical solution of the present application.

[0068] Figure 1 A flowchart of an image segmentation method provided in an embodiment of the present application;

[0069] Figure 2 A schematic diagram of a target ultrasound image provided in an embodiment of the present application;

[0070] Figure 3 A schematic diagram of a first foreground image provided in an embodiment of the present application;

[0071] Figure 4 A schematic diagram of a first characteristic image provided in an embodiment of the present application;

[0072] Figure 5 A schematic diagram of a second characteristic image provided in an embodiment of the present application;

[0073] Figure 6 A schematic diagram of determining a first target image using a neural network model provided in an embodiment of the present application;

[0074] Figure 7 A schematic diagram of another method of determining a first target image using a neural network model provided in an embodiment of the present application;

[0075] Figure 8 A training flow chart of a neural network model provided in an embodiment of the present application;

[0076] Fig. 9 A schematic diagram of the structure of an image segmentation device provided in an embodiment of the present application;

[0077] Fig.10 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. DETAILED DESCRIPTION

[0078] In order to enable those skilled in the art to better understand the solution of the present application, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0079] The terms "first", "second", etc. in the specification and claims of this application and the above-mentioned drawings are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not limited to the listed steps or units, but optionally includes steps or units that are not listed, or optionally includes other steps or units inherent to these processes, methods, products or devices.

[0080] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present application. The appearance of the phrase in various locations in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0081] The execution subject of the embodiment of the present application is an image segmentation device, wherein the image segmentation device can be any electronic device that can execute the technical solution disclosed in the method embodiment of the present application. Optionally, the image segmentation device can be one of the following: a computer, a server.

[0082] It should be understood that the method embodiment of the present application can also be implemented by a processor executing a computer program code. The following describes the embodiment of the present application in conjunction with the drawings in the embodiment of the present application. Figure 1 , Figure 1 A flowchart of an image segmentation method provided in an embodiment of the present application.

[0083] 101. Acquire a target ultrasound image, a first foreground image, and a background image of a target tissue.

[0084] In the embodiment of the present application, the target ultrasound image is an image obtained by performing an ultrasound scan on the object to which the target tissue belongs, the first foreground image is an image of the target tissue in the target ultrasound image, and the background image is an image of the image area other than the target tissue in the target ultrasound image. Optionally, in the case where the target ultrasound image has a problem that the ribs block the target tissue, and / or the boundary of the target tissue is unclear, the first foreground image contains an image of non-target tissue (false positive), and / or the first foreground image lacks an image of part of the target tissue (false negative).

[0085] In a possible implementation scenario, the image segmentation device receives the target ultrasound image, the first foreground image and the background image input by an input component. The input component includes at least one of the following: a keyboard, a mouse, a touch screen, a touch pad, and an audio input device.

[0086] In another possible implementation scenario, after receiving the target ultrasound image input by the input component, the image segmentation device segments the target ultrasound image. Specifically, the image area of ​​the target tissue in the target ultrasound image is determined to obtain a first foreground image containing the target tissue, and the image other than the target tissue in the target ultrasound image is determined to obtain a background image. Optionally, after receiving the first ultrasound image input by the input component, the first ultrasound image is preprocessed to obtain the target ultrasound image to reduce the information in the target ultrasound image and speed up the processing of the target ultrasound image. For example, the first ultrasound image is cropped to obtain the target ultrasound image to reduce the number of pixels in the image. Optionally, after receiving the first ultrasound image input by the input component, the first ultrasound image is preprocessed to obtain the target ultrasound image to reduce the discreteness of the pixel information of the pixels in the target ultrasound image. For example, the pixel information of the pixels in the first ultrasound image is normalized to [0,1] to obtain the target ultrasound image, so as to avoid the pixel information of the pixels in the target ultrasound image being too high and / or too low, resulting in the need to judge the pixels with too high and / or too low pixel information when segmenting the target ultrasound image, thereby reducing the speed of segmenting the target ultrasound image. Optionally, the pixel information includes grayscale value, brightness value, saturation, etc.

[0087] In a possible implementation manner, a target tissue in a target ultrasound image is segmented using a threshold value to obtain a first foreground image containing the target tissue and a background image.

[0088] In yet another possible implementation, a region growing algorithm is used to determine an image of a target tissue in a target ultrasound image to obtain a first foreground image and a background image.

[0089] In another possible implementation, a neural network model is used to segment a target tissue in a target ultrasound image to obtain a first foreground image and a background image. Optionally, the neural network model includes a deeplab model and a U-net model, wherein the deeplab model includes a deeplab second generation (version 2, v2) model, a deeplab third generation (version 3, v3) model, and a deeplab third generation plus (version 3 plus, v3+) model.

[0090] 102. Determine target positioning information of the target tissue based on the first foreground image and the background image.

[0091] In an embodiment of the present application, the first foreground image is an image of the target tissue in the target ultrasound image, and the background image is an image of the image area other than the target tissue in the target ultrasound image. The positioning information of the target tissue in the target ultrasound image can be determined based on the first foreground image and the background image to obtain the target positioning information, wherein the first foreground image and the background image are both images obtained based on the target ultrasound image, so the image size of the first foreground image is the same as the image size of the background image. Optionally, the target positioning information may include the coordinate range of the target tissue in the target ultrasound image. The target positioning information may also include the position of the boundary of the target tissue in the target ultrasound image. The target positioning information may also include pixel information of the target tissue in the target ultrasound image.

[0092] In one possible implementation, the background image is used to correct the image area of ​​the target tissue in the first foreground image, and the position information of the corrected image area of ​​the target tissue in the first foreground image is used as the target positioning information. Specifically, background tissue pixels whose grayscale values ​​of pixels in the background image are less than a first threshold are determined, wherein the background image is an image of the image area other than the target tissue in the target ultrasound image. Therefore, the pixels in the image area of ​​the target tissue in the background image do not contain image information, while the pixels in the image area other than the target tissue in the background image contain image information, resulting in lower grayscale values ​​of pixels in the image area of ​​the target tissue in the background image, which in turn results in higher confidence that pixels whose grayscale values ​​are less than the first threshold (background tissue pixels) are pixels of the target tissue. For example, the grayscale value matrix of pixels in the target ultrasound image is: , where the gray value matrix of the pixels in the image area of ​​the target tissue is . Extract pixels of the image area other than the target tissue in the target ultrasound image to obtain a background image. The pixels of the image area of ​​the target tissue in the background image do not contain image information. Optionally, the grayscale value of the pixels of the image area of ​​the target tissue in the background image is 0, and the grayscale value matrix of the pixels in the background image is When the first threshold is 0.5, the grayscale value of the pixel in the second row and first column, the grayscale value of the pixel in the second row and second column, the grayscale value of the pixel in the third row and first column, and the grayscale value of the pixel in the third row and second column in the background image are less than the first threshold, and the confidence that the above four pixels are pixels of the target tissue is high.

[0093] Determine foreground tissue pixels whose grayscale values ​​of pixels in the first foreground image are greater than the second threshold, wherein the first foreground image is the image of the target tissue in the target ultrasound image, and therefore the pixels in the image area of ​​the target tissue in the first foreground image contain image information, and the pixels in the image area other than the target tissue in the first foreground image do not contain image information, resulting in a higher grayscale value of the pixels in the image area of ​​the target tissue in the first foreground image, which in turn results in a higher confidence that the pixels (foreground tissue pixels) whose grayscale values ​​are greater than the second threshold are pixels of the target tissue. For example, the grayscale value matrix of the pixels in the target ultrasound image is: , where the gray value matrix of the pixels of the target tissue image is . Extract pixels of the image of the target tissue in the target ultrasound image to obtain a first foreground image, where pixels in the image area other than the target tissue in the first foreground image do not contain image information. Optionally, the grayscale value of pixels in the image area other than the target tissue in the first foreground image is 0, and the grayscale value matrix of the pixels in the first foreground image is When the second threshold is 2, the grayscale value of the pixel in the second row and first column, the grayscale value of the pixel in the second row and second column, the grayscale value of the pixel in the third row and first column, and the grayscale value of the pixel in the third row and second column in the first foreground image are greater than the second threshold, and the confidence that the above four pixels are pixels of the target tissue is high.

[0094] After determining the background tissue pixels and the foreground tissue pixels, the foreground tissue pixels are corrected using the background tissue pixels to obtain the target tissue pixels, and the position information of the target tissue pixels in the first foreground image is used as the target positioning information. Since the confidence that the background tissue pixels and the foreground tissue pixels are pixels of the target tissue is high, correcting the foreground tissue pixels based on the background tissue pixels can improve the confidence that the target tissue pixels are pixels of the target tissue, thereby improving the accuracy of the target positioning information determined based on the target tissue pixels.

[0095] Optionally, the position of the background tissue pixel in the background image is determined to obtain the first position, and the pixel at the first position is determined in the first foreground image to obtain the candidate pixel. Since the confidence that the background tissue pixel is the pixel of the target tissue is high, and both the background image and the first foreground image are images obtained based on the target ultrasound image, the pixels at the same position in the first foreground image and the background image represent the same pixel in the target ultrasound image, so the confidence that the pixel at the first position in the first foreground image (candidate pixel) is the pixel of the target tissue is high. In the case where there are pixels different from the foreground tissue pixels in the candidate pixels, the pixels in the foreground tissue pixels and the pixels different from the foreground tissue pixels in the candidate pixels are determined as target tissue pixels, and the position information of the target tissue pixel in the first foreground image is used as the target positioning information. For example, the background tissue pixels include A1 pixels and A2 pixels, the A1 pixel is at the first pixel position in the background pixels, and the A2 pixel is at the second pixel position in the background pixels. The foreground tissue pixels include B1 pixels, B2 pixels, and B3 pixels. The pixel at the first pixel position in the first foreground image is B2 pixel, and the pixel at the second pixel position in the first foreground image is B4 pixel. The B2 pixel and the B4 pixel are taken as candidate pixels. Since there is a B4 pixel different from the foreground tissue pixel among the candidate pixels, and the confidence that the B4 pixel is the pixel of the target tissue is high, the foreground tissue pixel and the pixels different from the foreground tissue pixel among the candidate pixels are taken as target tissue pixels, that is, the target tissue pixels include the B1 pixel, the B2 pixel, the B3 pixel, and the B4 pixel. The position information of the B1 pixel, the B2 pixel, the B3 pixel, and the B4 pixel in the first foreground image is taken as the target positioning information.

[0096] Because the confidence that the background tissue pixel is the pixel of the target tissue is high, the confidence that the pixel at the first position in the first foreground image is the pixel of the target tissue is high, that is, the confidence that the candidate pixel is the pixel of the target tissue is high. Because the confidence that the foreground tissue pixel and the candidate pixel are the pixels of the target tissue are both high, when there are pixels in the candidate pixels that are different from the foreground tissue pixels, the target tissue pixels are determined based on the pixels in the foreground tissue pixels and the pixels in the candidate pixels that are different from the foreground tissue pixels. The target positioning information determined based on the target tissue pixels can improve the accuracy of the target positioning information.

[0097] Further optionally, pixels in the foreground tissue pixels and pixels in the candidate pixels that are different from the foreground tissue pixels are determined as first reference pixels, and pixels that are different from the first reference pixels within a preset distance from the first reference pixels in the first foreground image are determined to obtain second reference pixels, and target tissue pixels are obtained based on the first reference pixels and the second reference pixels. By using the first reference pixels and pixels within a preset distance from the first reference pixels as target tissue pixels, the number of pixels in the target tissue pixels is increased, and the possibility that the image of the target tissue pixels in the first foreground image covers the image of the target tissue is increased, thereby improving the accuracy of the target tissue pixels. For example, the grayscale value matrix of the first foreground image is , the gray value matrix of the image of the first reference pixel in the first foreground image is [ , the gray value matrix of the image of the target tissue in the first foreground image is [ . It can be seen from the gray value matrix of the image of the first reference pixel in the first foreground image and the gray value matrix of the image of the target tissue in the first foreground image that the first reference pixel does not include all the pixels in the target tissue, and the image of the first reference pixel in the first foreground image cannot cover the image of the target tissue in the first foreground image. When the preset distance is 1 pixel, the second reference pixel includes the pixel in the second row and second column, the pixel in the second row and fourth column, the pixel in the third row and first column, and the pixel in the third row and third column in the gray value matrix of the first foreground image. After determining the second reference pixel, the first reference pixel and the second reference pixel are used as target tissue pixels, and the gray value matrix of the target tissue pixels in the first foreground image is [ From the gray value matrix of the image of the target tissue pixels in the first foreground image and the gray value matrix of the image of the target tissue in the first foreground image, it can be seen that the target tissue pixels include all pixels in the target tissue, and the image of the target tissue pixels in the first foreground image covers the image of the target tissue in the first foreground image.

[0098] Optionally, the pixels in the foreground tissue pixels that are the same as the candidate pixels are determined as target tissue pixels, and the positions of the target tissue pixels in the first foreground image are used as target positioning information. The pixels in the target tissue pixels are pixels in the foreground tissue pixels and are also pixels in the candidate pixels, so the confidence that the pixels in the target tissue pixels are pixels of the target tissue is higher, and therefore the target positioning information determined based on the target tissue pixels is more accurate.

[0099] In another possible implementation, the background boundary of the image area of ​​pixels in the background image having grayscale values ​​less than a first threshold is determined, and then the foreground boundary of the image area of ​​pixels in the first foreground image having grayscale values ​​greater than a second threshold is determined. When the similarity between the background boundary and the foreground boundary is greater than the similarity threshold, the boundary point of the target tissue is determined based on the foreground boundary and the background boundary to obtain a first feature point, and the position of the first feature point in the target ultrasound image is used as target positioning information.

[0100] In another possible implementation method, the pixels at the first image position in the first foreground image and the pixels at the first image position in the background image are determined based on the pixels at the first image position in the target ultrasound image, so the number of pixels in the first foreground image is the same as the number of pixels in the background image, and the pixels in the first foreground image and the pixels in the background image correspond one to one, wherein the first image position generally refers to any position in the image. Determine x first foreground pixels in the first foreground image, and use a probability distribution function to determine the confidence that each of the x first foreground pixels is a pixel of the target tissue, and obtain x first confidences, wherein the probability distribution function includes: Gibbs distribution (GibbsDistribution), normalized exponential (Softmax) function. Determine x background pixels in the background image, and use a probability distribution function to determine the confidence that each of the x background pixels is a pixel of the target tissue, and obtain x second confidences. After determining x first confidences and x second confidences, determine the maximum value of the first confidence and the second confidence corresponding to the two corresponding pixels in the first foreground image and the background image as the third confidence, wherein the third confidence corresponds to the two corresponding pixels. Repeat the above steps for the corresponding x pairs of pixels in the first foreground image and the background image to determine x third confidences. After determining x third confidences, determine a third confidence greater than a confidence threshold to obtain y fourth confidences, and use the positions of the pixels corresponding to the y fourth confidences in the first foreground image in the first foreground image as target positioning information.

[0101] For example, there are two pixels in both the first foreground image and the background image. The first foreground image includes pixel B1 and pixel B2, wherein pixel B1 is located at position W1 in the first foreground image, and pixel B2 is located at position W2 in the first foreground image. The background image includes pixel B3 and pixel B4, wherein pixel B3 is located at position W1 in the background image, and pixel B4 is located at position W2 in the background image. There is a corresponding relationship between pixel B1 and pixel B3, and there is a corresponding relationship between pixel B2 and pixel B4. The first confidence of pixel B1 is 0.4, the first confidence of pixel B2 is 0.4, the second confidence of pixel B3 is 0.6, the second confidence of pixel B4 is 0.3, and the confidence threshold is 0.5. The maximum value of the first confidence (the first confidence of pixel B1 is 0.4) and the second confidence (the second confidence of pixel B3 is 0.6) corresponding to pixel B1 and pixel B3 at position W1 is determined, and the third confidence of the pixel at position W1 is 0.6. The maximum value of the first confidence (the first confidence of the B1 pixel is 0.4) and the second confidence (the second confidence of the B3 pixel is 0.3) corresponding to the B2 pixel and the B4 pixel at position W2 is determined, and the third confidence of the pixel at position W2 is 0.4. The third confidence greater than the confidence threshold is determined, that is, the third confidence of the pixel at position W1 is used as the fourth confidence, and the position W1 of the element corresponding to the fourth confidence in the first foreground image in the first foreground image is used as the target positioning information.

[0102] 103. Based on the target positioning information, determine an image region of the target tissue in the target ultrasound image to obtain a first target image.

[0103] In an embodiment of the present application, after determining the target positioning information, the image segmentation device determines the image area of ​​the target tissue in the target ultrasound image based on the target positioning information of the target tissue in the target ultrasound image, and segments to obtain a first target image.

[0104] In a possible implementation scenario, the target positioning information includes a position of a first feature point in a target ultrasound image, wherein the position of the first feature point in the target ultrasound image includes one or more of the following: coordinates, scale-invariant feature transform (SIFT) feature positions. The image segmentation device determines the first feature point based on the position of the first feature point in the target ultrasound image, determines the image area enclosed by the first feature point in the target ultrasound image, obtains the first area, and obtains the first target image based on the image of the first area in the target ultrasound image. Optionally, the position of the first feature point in the target ultrasound image includes a first coordinate of a pixel. After determining the first coordinate, the image segmentation device determines the pixel at the first coordinate in the coordinate system of the target ultrasound image to obtain the first feature point, and obtains the first target image based on the image enclosed by the first feature point in the target ultrasound image.

[0105] In another possible implementation scenario, the target positioning information includes the position information of the pixels of the target tissue in the target ultrasound image. The image segmentation device segments the image of the target tissue in the target ultrasound image based on the region growing method. Specifically, the image segmentation device determines the pixels of the target tissue in the target ultrasound image according to the position information of the pixels of the target tissue in the target ultrasound image, and uses the pixels as seed pixels. After determining the seed pixels, the image segmentation device determines the target connected domain including the seed pixels according to the similarity with the features of the seed pixels, wherein the similarity between the features of the pixels in the target connected domain and the features of the seed pixels is greater than or equal to the similarity threshold. After determining the target connected domain, the image segmentation device determines the image of the target connected domain in the target ultrasound image to obtain a first target image. Since the seed pixels are the pixels of the target tissue, and the features of the pixels in the target connected domain are all similar to the features of the seed pixels, the similarity is greater than or equal to the similarity threshold, that is, the pixels in the target connected domain have a high similarity with the seed pixels, so the confidence that the pixels in the connected domain are the pixels of the target tissue is high, and the confidence of the first target image obtained according to the image of the connected domain including the seed pixels in the target ultrasound image is high.

[0106] In an embodiment of the present application, after the image segmentation device acquires a target ultrasound image of the target tissue, including a first foreground image of the image of the target tissue in the target ultrasound image and a background image of the image area other than the target tissue in the target ultrasound image, the image segmentation device determines the target positioning information of the target tissue based on the first foreground image and the background image, and then determines the image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image. After determining the target positioning information of the target tissue in the target ultrasound image, the image segmentation device segments the image of the target tissue in the target ultrasound image based on the target positioning information. Compared with directly segmenting the image of the target tissue in the target ultrasound image, the target positioning information of the target tissue can enable the image segmentation device to more accurately identify the target tissue during the segmentation process, thereby improving the accuracy of the first target image obtained by segmentation.

[0107] In an embodiment of the present application, when there is a problem that the target tissue is blocked by the ribs in the target ultrasound image and the boundary of the target tissue is unclear, it is easy to cause the image area where the ribs block the target tissue and the image area where the boundary of the target tissue is unclear in the first foreground image, which in turn leads to the images of the above two image areas being determined as target tissues with low confidence and the first foreground image having low accuracy. Optionally, the image area of ​​the pixels of the target tissue in the first foreground image is extracted to obtain a first feature image to highlight the image area of ​​the target tissue in the first foreground image. Because the images of the above two image areas in the first foreground image are determined as target tissues with low confidence, the extracted pixels of the target tissue lack the pixels of the above two image areas, the first feature image does not include the images of the target tissue in the above two image areas, and the accuracy of the first feature image is low. In an implementation method for determining a first feature image, a probability distribution function is used to determine the probability that each pixel in a first foreground image is a target tissue, and then a pixel determination function is used to determine the pixel corresponding to the probability greater than a probability threshold in the above probability to obtain a first feature pixel, and an image of the first feature pixel in the first foreground image is used as the first feature image, wherein the pixel determination function includes: arguments of the maxima (argmax) function and not a number (NaN) argmax function.

[0108] See also Figure 2 , Figure 3 , Figure 4 ,in, Figure 2 A schematic diagram of a target ultrasound image provided in an embodiment of the present application, Figure 3 A schematic diagram of a first foreground image provided in an embodiment of the present application, Figure 4 A schematic diagram of a first characteristic image provided in an embodiment of the present application. Figure 2There is a problem of ribs blocking the target tissue in the position 1 marked in the figure. Figure 2 The target tissue boundary is unclear in the position 2 marked in the figure. Figure 3 The image area of ​​the target tissue that is blocked in the first foreground image and the image area where the boundary of the target tissue is unclear are determined to be the target tissue with low confidence. Specifically, Figure 3 A medium pixel brightness value represents the confidence that the pixel is the target tissue, and a low pixel brightness value represents a low confidence that the pixel is determined to be the target tissue. Figure 3 The image area of ​​the target tissue that is blocked and the image area where the boundary of the target tissue is unclear have low brightness values, so the confidence that the above two image areas are determined as target tissues is low. Since the confidence that the above two image areas in the first foreground image are determined as target tissues is low, the images of the above two image areas are determined as non-target tissues, resulting in Figure 4 The images of the above two image regions are missing in the first feature image, so the accuracy of the first feature image is low.

[0109] In the embodiment of the present application, the positioning information of the target tissue is determined by the first foreground image and the background image, and the image area of ​​the target tissue in the target ultrasound image is segmented based on the target positioning information to obtain the first target image, so that the image segmentation device can determine the part of the target tissue blocked by the ribs according to the target positioning information in the process of segmenting the target ultrasound image, and can also determine the boundary of the target tissue according to the target positioning information, thereby improving the accuracy of the determined first target image. Optionally, the image area of ​​the pixels of the target tissue in the first target image is extracted to obtain the second feature image to highlight the image area of ​​the target tissue in the first target image. Because the image segmentation device can determine the part of the target tissue blocked by the ribs according to the target positioning information, and can also determine the boundary of the target tissue according to the target positioning information, the image area of ​​the target tissue blocked by the ribs and the image area where the boundary of the target tissue is unclear in the first target image determined by the image segmentation device is determined to be the target tissue with high confidence, the pixels of the target tissue in the first target image include the pixels of the above two image areas, the second feature image includes the image of the target tissue in the above two image areas, and the second feature image has high accuracy. The second feature image obtained by the method of the embodiment of the present application has a recall rate of 3.1% higher than that of the first feature image. The Dice value of the second feature image obtained by the method of the embodiment of the present application can reach 94.6%, which is 1.2% higher than the Dice value of the first feature image. In an implementation method for determining the second feature image, the probability of each pixel in the first target image being the target tissue is determined using a probability distribution function, and then the pixel determination function is used to determine the pixel corresponding to the probability greater than the probability threshold in the above probability, to obtain the second feature pixel, and the image of the second feature pixel in the first target image is used as the second feature image.

[0110] See also Figure 5 , Figure 5 A schematic diagram of a second characteristic image provided in an embodiment of the present application. Figure 4 The first feature image of Figure 5 The second characteristic image includes the target tissue in the image area where the ribs block the target tissue and the target tissue in the image area where the boundary of the target tissue is unclear, and the accuracy of the second characteristic image is higher than the accuracy of the first characteristic image.

[0111] In a possible implementation scenario, a neural network model can be used to segment a target ultrasound image, including a first foreground image of the image of the target tissue in the target ultrasound image and a background image of the image area other than the target tissue in the target ultrasound image. The neural network model can also be used to determine target positioning information of the target tissue based on the first foreground image and the background image, and the neural network model can be used to determine the image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image. Specifically, an image segmentation device obtains a target ultrasound image of the target tissue, and inputs the target ultrasound image into the neural network model so that the neural network model outputs a first foreground image and a background image, and then inputs the first foreground image, the background image, and the target ultrasound image into the neural network model so that the neural network model outputs the image area of ​​the target tissue in the target ultrasound image as the first target image. To better understand this implementation scenario, please refer to Figure 6 , Figure 6 A schematic diagram of using a neural network model to determine the first target image is provided for this implementation scenario. Figure 6 In the embodiment, the structure of the neural network model is a dual deeplabv3 model structure, that is, the neural network model includes two deeplabv3 models, the first deeplabv3 model is used to extract the feature map of the target ultrasound image, and obtain the first foreground image and background image of the target tissue in the target ultrasound image, and the second deeplabv3 model is used to output the first target image according to the first foreground image, the background image and the target ultrasound image. Optionally, the first foreground image, the background image and the target ultrasound image can be merged and then input into the second deeplabv3 model, so as to reduce the computational overhead of merging different images in the second deeplabv3 model, and directly determine the correlation between different images according to the merged image, thereby accelerating the execution efficiency of the neural network model.

[0112] See also Figure 7 , Figure 7Another schematic diagram of determining the first target image using a neural network model is provided for this implementation scenario. After receiving the first ultrasonic image, the image segmentation device performs data processing on the first ultrasonic image to obtain a target ultrasonic image to reduce the information in the first ultrasonic image and / or reduce the discreteness of the pixel information of the pixels in the first ultrasonic image. After determining the target ultrasonic image, the target ultrasonic image is input into the neural network model, and the target tissue in the target ultrasonic image is segmented using the first deeplabv3 model in the neural network model to obtain a first foreground image and a background image. The first foreground image, the background image and the target ultrasonic image are merged, and the merged image is input into the second deeplabv3 model to determine the image area of ​​the target tissue in the target ultrasonic image to obtain the first target image. Optionally, the second deeplabv3 model can output the image area outside the target tissue in the target ultrasonic image to obtain the target background image. Optionally, the image area of ​​the pixels of the target tissue in the first target image is extracted to obtain a second feature image to highlight the image area of ​​the target tissue in the first target image.

[0113] When the neural network model in the scenario is obtained through training, the model to be trained can be trained based on the training data to obtain the neural network model, wherein the training data includes a training ultrasound image and a first label, and the training ultrasound image is an ultrasound image including a training tissue. The first label is an image of the training tissue in the training ultrasound image. Specifically, the training ultrasound image is input into the model to be trained so that the model to be trained outputs a first training image and a second training image, wherein the first training image is an image of the training tissue in the training ultrasound image, and the second training image is an image of an image area other than the training tissue in the training ultrasound image. After the first training image and the second training image are output, the training ultrasound image, the first training image, and the second training image are input into the model to be trained so that the model to be trained determines the positioning information of the training tissue based on the first training image and the second training image, and determines the image area of ​​the training tissue in the training ultrasound image based on the positioning information, and obtains the target training image. After obtaining the target training image, the loss of the model to be trained is determined according to the two outputs of the model to be trained. Specifically, the first loss is determined according to the first difference between the first training image, the second training image, and the image of the first label, and the second loss is determined according to the difference between the target training image and the image of the first label. The parameters of the model to be trained are updated based on the first loss and the second loss to obtain a neural network model. Optionally, the first branch in the model to be trained is used to segment the training ultrasound image to obtain the first training image and the second training image, and the second branch in the model to be trained is used to determine the positioning information of the training tissue based on the first training image and the second training image, and determine the image area of ​​the training tissue in the training ultrasound image based on the positioning information to obtain the target training image. After determining the first loss and the second loss, the first loss can be used to update the parameters of the first branch to improve the accuracy of the first training image and the second training image output by the first branch. The second loss can also be used to update the parameters of the first branch and the second branch to adjust all parameters in the model to be trained, thereby improving the accuracy of the target training image output by the model to be trained, and training the model to be trained as a neural network model.

[0114] See also Figure 8 , Figure 8The present invention is a training flowchart of a neural network model in this implementation scenario. Specifically, the training data is used to train the model to be trained, wherein the training data includes a training ultrasound image and a first label, the training ultrasound image includes a training tissue, the first label is an image of the training tissue in the training ultrasound image, and the structure of the model to be trained is a dual deeplabv3 model structure, that is, the model to be trained includes two deeplabv3 models. The training ultrasound image is input into the model to be trained so that the first deeplabv3 model in the model to be trained segments the training tissue in the training ultrasound image to obtain a feature map including the training ultrasound image, and the feature map of the training ultrasound image includes a first training image of the training tissue in the training ultrasound image, and a second training image of the image area outside the training tissue in the training ultrasound image. After determining the first training image and the second training image, the first training image, the second training image and the training ultrasound image are merged and used as the input of the second deeplabv3 model in the model to be trained, so that the second deeplabv3 model segments the image area of ​​the training tissue in the training ultrasound image based on the merged image to obtain the target training image. Based on the difference between the segmentation result of the first deeplabv3 model and the image of the first label, the first loss is determined, and the parameters of the first deeplabv3 model are updated by back propagation based on the first loss. Based on the difference between the segmentation result of the second deeplabv3 model and the image of the first label, the second loss is determined, and the parameters of the model to be trained are updated by back propagation based on the second loss, that is, the parameters of the first deeplabv3 model and the second deeplabv3 model in the model to be trained are updated, and the model to be trained is trained as a neural network model.

[0115] As an optional implementation, the target positioning information includes the position of a second feature point of the image of the diseased tissue in the target tissue in the target ultrasound image, wherein the second feature point is a point at the boundary of the image of the diseased tissue in the target ultrasound image, wherein the diseased tissue is an area in the target tissue where a lesion occurs. The image segmentation device performs the following steps before executing step 103:

[0116] 201. Determine the second feature point from the target ultrasound image based on a position of the second feature point in the target ultrasound image.

[0117] In the embodiment of the present application, the first target image is an image determined according to the image area of ​​the target tissue in the target ultrasound image, and the diseased tissue is the area where the lesion occurs in the target tissue, so the first target image includes the diseased tissue, and the first target image includes the second feature point of the diseased tissue. Optionally, the position of the second feature point in the target ultrasound image includes one or more of the following: coordinates, SIFT feature positions.

[0118] The image segmentation device performs the following steps before performing step 103:

[0119] 202. Determine an image region enclosed by the second feature points in the first target image to obtain a second region.

[0120] In the embodiment of the present application, since the diseased tissue is the area where the disease occurs in the target tissue, and the first target image is the image of the target tissue, the first target image includes the diseased tissue and the second feature point. The image area surrounded by the second feature point in the first target image is determined to obtain the second area.

[0121] 203. Mark an image region in the first target image corresponding to the second region to obtain a second target image.

[0122] In the embodiment of the present application, after determining the second area surrounded by the second feature points in the first target image, the image segmentation device marks the image area corresponding to the second area in the first target image to obtain the second target image. In one possible implementation, the image segmentation device deletes the image area corresponding to the second area in the first target image to obtain the second target image. In another possible implementation, the image segmentation device increases the brightness value of the image area corresponding to the second area in the first target image to obtain the second target image.

[0123] In this embodiment, before determining the image area surrounded by the first feature points in the target ultrasound image and obtaining the first target image, the image processing device determines the second feature points in the target ultrasound image based on the positions of the second feature points in the target ultrasound image, wherein the second feature points are points at the boundary of the diseased tissue, and the diseased tissue is the tissue in the target tissue where the disease occurs. Since the first target image is an image of the target tissue, the first target image also includes the diseased tissue and the second feature points. The image processing device determines the image area surrounded by the second feature points in the first target image to obtain the second area, and marks the image area corresponding to the second area in the first target image to obtain the second target image, wherein the image of the second area is an image of the diseased tissue, and therefore the image area of ​​the diseased tissue in the second target image is marked. The second target image determined by the above steps includes not only the image of the target tissue, but also the image of the marked diseased tissue, thereby highlighting the diseased tissue in the second target image. Optionally, after obtaining the second target image, the second target image is displayed so that relevant personnel can plan a puncture path for the diseased tissue based on the second target image. At this time, the second target image highlights the diseased tissue, which is helpful for relevant personnel to determine the diseased tissue from the second target image, and further helps to plan the puncture path for the diseased tissue.

[0124] In a possible implementation scenario, the position information of the diseased area in the target tissue can be determined according to the marked image area in the second target image, so as to provide the position information for the diagnosis and treatment of the diseased tissue.

[0125] Those skilled in the art will appreciate that, in the above method of specific implementation, the order in which the steps are written does not imply a strict execution order and does not constitute any limitation on the implementation process. The specific execution order of the steps should be determined by their functions and possible internal logic.

[0126] The method of the embodiment of the present application is described in detail above, and the device of the embodiment of the present application is provided below.

[0127] See also Fig. 9 , Fig. 9 This is a schematic diagram of the structure of an image segmentation device provided in an embodiment of the present application. The image segmentation device 1 includes: an acquisition unit 11 and a determination unit 12. Among them:

[0128] An acquisition unit 11 is used to acquire a target ultrasound image of a target tissue, a first foreground image, and a background image, wherein the target ultrasound image is an image obtained by performing ultrasound scanning on the target tissue, the first foreground image includes an image of the target tissue in the target ultrasound image, and the background image includes an image of an image area other than the target tissue in the target ultrasound image;

[0129] A determination unit 12, configured to determine target location information of the target tissue based on the first foreground image and the background image;

[0130] The determination unit 12 is further configured to determine an image region of the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image.

[0131] In combination with any embodiment of the present application, the target positioning information includes a position of a first feature point in the target ultrasound image, and the first feature point includes a point of a boundary of the target tissue;

[0132] The determination unit 12 is specifically used for:

[0133] Determining the first feature point from the target ultrasound image based on a position of the first feature point in the target ultrasound image;

[0134] Determine an image area surrounded by the first feature points in the target ultrasound image to obtain a first area;

[0135] The first target image is obtained based on the image of the first area in the target ultrasound image.

[0136] In combination with any embodiment of the present application, the target positioning information includes a position of a second feature point of an image of a diseased tissue in the target tissue in the target ultrasound image, where the second feature point is a point at a boundary of the image of the diseased tissue in the target ultrasound image;

[0137] The above-mentioned determination unit 12 is further used for:

[0138] Determining the second feature point from the target ultrasound image based on a position of the second feature point in the target ultrasound image;

[0139] The image segmentation device 1 further includes a marking unit 13, which is used to:

[0140] Determine an image area enclosed by the second feature points in the first target image to obtain a second area;

[0141] An image region in the first target image corresponding to the second region is marked to obtain a second target image.

[0142] In combination with any embodiment of the present application, the marking unit 13 is specifically used for:

[0143] The image area corresponding to the second area in the first target image is deleted to obtain the second target image.

[0144] In conjunction with any embodiment of the present application, the determination unit 12 is specifically configured to:

[0145] Determine background tissue pixels whose grayscale values ​​of pixels in the background image are less than a first threshold;

[0146] Determine foreground tissue pixels whose grayscale values ​​of pixels in the first foreground image are greater than a second threshold;

[0147] Correcting the foreground tissue pixels based on the background tissue pixels to obtain target tissue pixels;

[0148] The target positioning information is obtained according to the position information of the target tissue pixel in the first foreground image.

[0149] In conjunction with any embodiment of the present application, the determination unit 12 is specifically configured to:

[0150] Determining a first position of the background tissue pixel in the background image;

[0151] Determine a pixel at a first position in the first foreground image to obtain a candidate pixel;

[0152] Determine pixels among the foreground tissue pixels and pixels among the candidate pixels that are different from the foreground tissue pixels to obtain the first reference pixels;

[0153] Determine a pixel different from the first reference pixel within a preset distance from the first reference pixel in the first foreground image to obtain a second reference pixel;

[0154] The first reference pixel and the second reference pixel are used as the target tissue pixels.

[0155] In an embodiment of the present application, after the image segmentation device acquires a target ultrasound image of the target tissue, including a first foreground image of the image of the target tissue in the target ultrasound image and a background image of the image area other than the target tissue in the target ultrasound image, the image segmentation device determines the target positioning information of the target tissue based on the first foreground image and the background image, and then determines the image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image. After determining the target positioning information of the target tissue in the target ultrasound image, the image segmentation device segments the image of the target tissue in the target ultrasound image based on the target positioning information. Compared with directly segmenting the image of the target tissue in the target ultrasound image, the target positioning information of the target tissue can enable the image segmentation device to more accurately identify the target tissue during the segmentation process, thereby improving the accuracy of the first target image obtained by segmentation.

[0156] In some embodiments, the functions or modules included in the device provided in the embodiments of the present application can be used to execute the method described in the above method embodiments. The specific implementation can refer to the description of the above method embodiments. For the sake of brevity, it will not be repeated here.

[0157] Fig.10 A schematic diagram of the hardware structure of an electronic device provided in an embodiment of the present application. The electronic device 2 includes a processor 21 and a memory 22. Optionally, the electronic device 2 also includes an input device 23 and an output device 24. The processor 21, the memory 22, the input device 23 and the output device 24 are coupled via a connector, and the connector includes various interfaces, transmission lines or buses, etc., which are not limited in the embodiments of the present application. It should be understood that in each embodiment of the present application, coupling refers to mutual connection in a specific manner, including direct connection or indirect connection through other devices, for example, it can be connected through various interfaces, transmission lines, buses, etc.

[0158] The processor 21 may include one or more processors, for example, one or more central processing units (CPUs). When the processor is a CPU, the CPU may be a single-core CPU or a multi-core CPU. Optionally, the processor 21 may be a processor group consisting of multiple CPUs, and the multiple processors are coupled to each other through one or more buses. Optionally, the processor may also be other types of processors, etc., which are not limited in the embodiments of the present application.

[0159] The memory 22 can be used to store computer program instructions and various computer program codes including the program code for executing the program code of the present application. Optionally, the memory includes but is not limited to random access memory (RAM), read-only memory (ROM), erasable programmable read only memory (EPROM), or portable read only memory (CD-ROM), which is used for related instructions and data.

[0160] The input device 23 is used to input data and / or signals, and the output device 24 is used to output data and / or signals. The input device 23 and the output device 24 can be independent devices or an integrated device.

[0161] It can be understood that in the embodiment of the present application, the memory 22 can be used not only to store relevant instructions, but also to store relevant data. For example, the memory 22 can be used to store the target ultrasound image, the first foreground image and the background image obtained through the input device 23, or the memory 22 can also be used to store the first target image obtained by the processor 21, etc. The embodiment of the present application does not limit the specific data stored in the memory.

[0162] Understandably, Fig.10 Only a simplified design of an electronic device is shown. In practical applications, the electronic device may also include other necessary components, including but not limited to any number of input / output devices, processors, memories, etc., and all electronic devices that can implement the embodiments of the present application are within the protection scope of the present application.

[0163] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of this application.

[0164] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments, and will not be repeated here. Those skilled in the art can also clearly understand that the descriptions of the various embodiments of the present application have different focuses. For the convenience and brevity of description, the same or similar parts may not be repeated in different embodiments. Therefore, for parts not described or not described in detail in a certain embodiment, refer to the records of other embodiments.

[0165] In the several embodiments provided in the present application, it should be understood that the disclosed systems, devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0166] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0167] In addition, each functional unit in each embodiment of the present application may be integrated into one processing unit, or each unit may exist physically separately, or two or more units may be integrated into one unit.

[0168] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When implemented by software, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function described in the embodiment of the present application is generated in whole or in part. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted through the computer-readable storage medium. The computer instructions may be transmitted from a website site, computer, server or data center to another website site, computer, server or data center by wired (e.g., coaxial cable, optical fiber, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can access or a data storage device such as a server or data center that includes one or more available media integrated. The available medium may be a magnetic medium (eg, a floppy disk, a hard disk, a magnetic tape), an optical medium (eg, a digital versatile disc (DVD)), or a semiconductor medium (eg, a solid state disk (SSD)).

[0169] A person skilled in the art can understand that to implement all or part of the processes in the above-mentioned embodiments, the processes can be completed by a computer program to instruct the relevant hardware, and the program can be stored in a computer-readable storage medium. When the program is executed, it can include the processes of the above-mentioned method embodiments. The aforementioned storage medium includes: a read-only memory (ROM) or a random access memory (RAM), a magnetic disk or an optical disk, and other media that can store program codes.

Claims

1. An image segmentation method, characterized in that: The method comprises: Acquire a target ultrasound image of a target tissue, a first foreground image, and a background image, wherein the target ultrasound image is an image obtained by performing ultrasound scanning on the target tissue, the first foreground image includes an image of the target tissue in the target ultrasound image, and the background image includes an image of an image area other than the target tissue in the target ultrasound image; Determine background tissue pixels whose grayscale values ​​of pixels in the background image are less than a first threshold; Determine foreground tissue pixels whose grayscale values ​​of pixels in the first foreground image are greater than a second threshold; Determining a first position of the background tissue pixel in the background image; Determine a pixel at a first position in the first foreground image to obtain a candidate pixel; Determine pixels among the foreground tissue pixels and pixels among the candidate pixels that are different from the foreground tissue pixels to obtain a first reference pixel; Determine a pixel different from the first reference pixel within a preset distance from the first reference pixel in the first foreground image to obtain a second reference pixel; Taking the first reference pixel and the second reference pixel as target tissue pixels; Obtaining target positioning information according to the position information of the target tissue pixel in the first foreground image; Based on the target positioning information, an image region of the target tissue in the target ultrasound image is determined to obtain a first target image.

2. The method according to claim 1, characterized in that: The target positioning information includes a position of a first feature point in the target ultrasound image, wherein the first feature point includes a point at a boundary of the target tissue; The step of determining an image area of ​​the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image includes: Determining the first feature point from the target ultrasound image based on a position of the first feature point in the target ultrasound image; Determine an image area surrounded by the first feature points in the target ultrasound image to obtain a first area; The first target image is obtained based on the image of the first area in the target ultrasound image.

3. The method according to claim 2, characterized in that The target positioning information includes a position of a second feature point of an image of a diseased tissue in the target tissue in the target ultrasound image, wherein the second feature point is a point at a boundary of the image of the diseased tissue in the target ultrasound image; Before obtaining the first target image based on the image of the first area in the target ultrasound image, the method further includes: Determining the second feature point from the target ultrasound image based on a position of the second feature point in the target ultrasound image; The step of obtaining the first target image based on the image of the first area in the target ultrasound image includes: Determine an image area enclosed by the second feature points in the first target image to obtain a second area; An image region in the first target image corresponding to the second region is marked to obtain a second target image.

4. The method according to claim 3, characterized in that The step of marking the image area in the first target image corresponding to the second area to obtain the second target image includes: The image area corresponding to the second area in the first target image is deleted to obtain the second target image.

5. An image segmentation device, characterized in that: The image segmentation device comprises: an acquisition unit, configured to acquire a target ultrasound image of a target tissue, a first foreground image, and a background image, wherein the target ultrasound image is an image obtained by performing ultrasound scanning on the target tissue, the first foreground image includes an image of the target tissue in the target ultrasound image, and the background image includes an image of an image region other than the target tissue in the target ultrasound image; A determining unit, configured to determine background tissue pixels whose grayscale values ​​of pixels in the background image are less than a first threshold; The determining unit is further used to determine foreground tissue pixels whose grayscale values ​​of pixels in the first foreground image are greater than a second threshold; The determining unit is further used to determine a first position of the background tissue pixel in the background image; The determining unit is further configured to determine a pixel at a first position in the first foreground image to obtain a candidate pixel; The determining unit is further used to determine pixels in the foreground tissue pixels and pixels in the candidate pixels that are different from the foreground tissue pixels to obtain a first reference pixel; The determining unit is further configured to determine a pixel different from the first reference pixel within a preset distance from the first reference pixel in the first foreground image to obtain a second reference pixel; The determining unit is further configured to use the first reference pixel and the second reference pixel as target tissue pixels; The determining unit is further configured to obtain target positioning information according to the position information of the target tissue pixel in the first foreground image; The determination unit is further configured to determine an image region of the target tissue in the target ultrasound image based on the target positioning information to obtain a first target image.

6. An electronic device, characterized in that: include: A processor and a memory, wherein the memory is used to store computer program codes, wherein the computer program codes include computer instructions, and when the processor executes the computer instructions, the electronic device executes the method according to any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, wherein the computer program includes program instructions. When the program instructions are executed by a processor, the processor is enabled to execute the method according to any one of claims 1 to 4.

8. A computer program product, characterized in that The computer program product comprises a computer program or instructions; when the computer program or instructions are run on a computer, the computer is enabled to execute the method according to any one of claims 1 to 4.

Images