Diffuse lesion recognition method, device and equipment and computer readable storage medium

Abstract

The invention discloses a diffuse lesion recognition method, device and equipment and a computer readable storage medium. The method comprises the following steps: acquiring at least two stages of medical images of a target physiological tissue; based on a diffuse lesion area in a target image area in the medical image, determining a target diffuse lesion area of the target physiological tissue; and determining diffuse lesion information according to the target diffuse lesion area. According to the invention, the diffuse lesion identification accuracy can be improved.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a diffuse lesion identification method, device, equipment and computer-readable storage medium. Background Art

[0002] With the development of AI imaging technology, image scanning has become a common auxiliary diagnosis and treatment method in modern medical diagnosis and treatment. Medical imaging can provide doctors with more and more intuitive information about the human body. It speeds up the doctor's diagnosis of the patient's condition. At present, for diffuse lesions such as fatty liver and cirrhosis, it is necessary to track their development through medical imaging, so as to provide a treatment plan that is more in line with the patient's current condition. However, the current medical images for diffuse lesions have problems such as physiological tissue deformation and unclear boundaries, which seriously affect the identification of diffuse lesions. Summary of the invention

[0003] The embodiments of the present invention provide a diffuse lesion recognition method, apparatus, device and computer-readable storage medium, aiming to effectively improve the recognition accuracy of diffuse lesions.

[0004] In a first aspect, an embodiment of the present invention provides a diffuse lesion identification method, wherein the diffuse lesion identification method comprises:

[0005] Obtain at least two phases of medical images of the target physiological tissue;

[0006] Determining a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region within a target image region in the medical image;

[0007] Diffuse lesion information is determined according to the target diffuse lesion area.

[0008] Optionally, determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image includes:

[0009] Performing registration processing on at least two phases of medical images to determine an image region corresponding to at least one key feature organization in the at least two phases of medical images;

[0010] For each key feature organization, selecting a target image area corresponding to the key feature organization from the image area corresponding to the key feature organization;

[0011] Integrate the at least one target image region according to the key feature organization corresponding to the at least one target image region;

[0012] The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

[0013] Optionally, determining the target diffuse lesion region according to the integrated diffuse lesion region in the at least one target image region includes:

[0014] According to the integrated diffuse lesion region in the at least one target image region, the diffuse lesion region in the target image region is spliced ​​to obtain the target diffuse lesion region.

[0015] Optionally, determining the target diffuse lesion region according to the integrated diffuse lesion region in the at least one target image region includes:

[0016] According to the integrated diffuse lesion regions in each of the target image regions, the diffuse lesion region in at least one medical image is optimized to obtain the target diffuse lesion region.

[0017] Optionally, for each key feature organization, selecting a target image area corresponding to the key feature organization from image areas corresponding to the key feature organization includes:

[0018] For each key feature organization, a target image area corresponding to the key feature organization is selected according to image quality data of each image area corresponding to the key feature organization.

[0019] Optionally, before determining the diffuse lesion information according to the target diffuse lesion area, the method further includes:

[0020] For a sub-lesion region in the target diffuse lesion region, obtaining a first diffuse lesion confidence corresponding to the sub-lesion region and a second diffuse lesion confidence corresponding to a neighboring lesion region of the sub-lesion region;

[0021] Performing verification processing on the sub-lesion area according to the first diffuse lesion confidence and the second diffuse lesion confidence;

[0022] The sub-lesion areas in the target diffuse lesion area that have not passed the verification process are screened out.

[0023] Optionally, the verifying the sub-lesion region according to the first diffuse lesion confidence and the second diffuse lesion confidence includes:

[0024] Get pre-configured diffuse lesion area verification strategies;

[0025] The first diffuse lesion confidence and the second diffuse lesion confidence are verified according to the diffuse lesion area verification strategy.

[0026] Optionally, the determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image further includes:

[0027] For each medical image, obtaining the image region that meets the preset image conditions;

[0028] Setting the image region to a corresponding target image region in the medical image;

[0029] integrating the at least one target image region according to a physiological tissue position corresponding to the at least one target image region;

[0030] The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

[0031] Optionally, before determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image, the method further includes:

[0032] For each medical image, obtaining pixel values ​​of each image area in the medical image;

[0033] Determining black areas in the medical image according to pixel values ​​of each image area;

[0034] The black areas in the medical image are removed, and based on the removed medical image, the step of determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image is performed.

[0035] In a second aspect, an embodiment of the present invention provides a diffuse lesion identification device, the diffuse lesion identification device comprising:

[0036] An acquisition module, used for acquiring at least two phases of medical images of a target physiological tissue;

[0037] A first determination module, configured to determine a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region within a target image region in the medical image;

[0038] The second determination module is used to determine diffuse lesion information according to the target diffuse lesion area.

[0039] In a third aspect, an embodiment of the present invention further provides a diffuse lesion identification device, comprising a processor and a memory, wherein the memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of any diffuse lesion identification method provided by the embodiment of the present invention.

[0040] In a fourth aspect, an embodiment of the present invention further provides a computer-readable storage medium, the computer-readable storage medium comprising a computer program. When the computer program is run on an electronic device, the computer program is used to enable the electronic device to execute the steps of any diffuse lesion identification method provided by the embodiment of the present invention.

[0041] The present invention obtains at least two medical images of a target physiological tissue; determines a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region in a target image region in the medical images; and determines diffuse lesion information based on the target diffuse lesion region. A more refined target diffuse lesion region is obtained through the diffuse lesion region in the target image region in multiple medical images, so that more accurate diffuse lesion information can be identified based on the target diffuse lesion region, thereby improving the accuracy of diffuse lesion identification, and the obtained diffuse lesion information can assist medical diagnosis and improve diagnostic efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0043] Figure 1 is a flow chart of an embodiment of a diffuse lesion identification method provided in an embodiment of the present invention;

[0044] Figure 2 is a schematic diagram of a first scene for selecting a target image area provided in an embodiment of the present invention;

[0045] Figure 3 The embodiment of the present invention is based on Figure 3 Schematic diagram of the integration of the target image area;

[0046] Figure 4 is a schematic diagram of the structure of a diffuse lesion identification device provided in an embodiment of the present invention;

[0047] Figure 5 It is a structural schematic diagram of a diffuse lesion identification device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention. At the same time, in the description of the embodiments of the present invention, the terms "first", "second", etc. are only used to distinguish the descriptions and cannot be understood as indicating or implying relative importance. Therefore, the features defined as "first" and "second" may explicitly or implicitly include one or more features. In the description of the embodiments of the present invention, the meaning of "multiple" is two or more, unless otherwise clearly and specifically defined.

[0049] Embodiments of the present invention provide a diffuse lesion identification method, apparatus, device and computer-readable storage medium.

[0050] Specifically, this embodiment will be described from the perspective of a diffuse lesion identification device, which can be specifically integrated in a diffuse lesion identification device, that is, the diffuse lesion identification method of the embodiment of the present invention can be executed by a diffuse lesion identification device, and the diffuse lesion identification device can be a medical device (such as a magnetic resonance imaging device, etc.), a terminal device (such as a computer, a mobile phone, etc.), etc.

[0051] The following is a detailed description in conjunction with the accompanying drawings. In this embodiment, the execution subject is taken as an example of a diffuse lesion identification device. It should be noted that the description order of the following embodiments is not intended to limit the preferred order of the embodiments. Although the logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in an order different from that shown in the accompanying drawings.

[0052] According to the background technology of the present invention, at present, for diffuse lesions such as fatty liver and cirrhosis, it is necessary to track their development through medical imaging, so as to provide a treatment plan that is more in line with the current patient's condition. However, the current medical imaging of diffuse lesions has problems such as physiological tissue deformation and unclear boundaries, which seriously affect the identification of diffuse lesions.

[0053] In order to solve the above problems, the present invention discloses a diffuse lesion identification method, please refer to Figure 1 The specific process of the diffuse lesion identification method can be as follows: Step S10 to Step S30, wherein:

[0054] Step S10, acquiring at least two medical images of the target physiological tissue;

[0055] In this embodiment, for the target physiological tissue to be identified, at least two medical images of the target physiological tissue can be obtained. The target physiological tissue may be a tissue at risk of diffuse lesions, such as the lungs, liver, etc. Diffuse lesions refer to lesions that are widely distributed in a certain organ or tissue, rather than being confined to a specific area. Such lesions may be caused by a variety of reasons, including infection, inflammation, metabolic disorders, autoimmune diseases or tumors. Diffuse lesions can affect different organs, such as the liver, thyroid, lungs, etc., such as fatty liver and cirrhosis are diffuse lesions. It is necessary to collect the corresponding medical images of the target physiological tissue to identify the diffuse lesion information about diffuse lesions in the target physiological tissue. Compared with the boundaries of other types of lesions, the boundaries of diffuse lesions are unclear and widely distributed, so the accuracy of identifying diffuse lesion information is not high.

[0056] Medical images of different phases are medical images acquired at different acquisition times within the same examination cycle for the area where the target physiological tissue is located. The acquisition methods corresponding to medical images of different phases are different. Since medical images of different phases are images acquired based on different acquisition methods, the information contained in the target tissue area of ​​medical images of different phases is different, and may include information such as liver tissue, blood vessels, and lesions. For example, a medical image of a certain phase mainly shows blood vessels, while a medical image of another phase mainly shows lesion information, etc., and this difference in content is also reflected in the different distribution of high image quality areas. For example, in medical image A, the upper left corner area is clearer, while in medical image B, the lower right corner area is clearer.

[0057] Step S20, determining a target diffuse lesion region of the target physiological tissue based on the diffuse lesion region within the target image region in the medical image;

[0058] In this embodiment, in the presence of diffuse lesions, the medical image of the target physiological tissue may capture diffuse lesions in the target physiological tissue, and the diffuse lesion area will be displayed in the medical image. At the same time, the medical image of the target physiological tissue may also capture the tissue area of ​​the target physiological tissue. The tissue area is the substantial area of ​​other target physiological tissues in the medical image, which may include blood vessels, muscles, fat, etc. Segmentation processing is performed on each medical image, and the tissue area and diffuse lesion area corresponding to the target physiological tissue can be determined from the medical image to obtain the tissue area and diffuse lesion area in the target image area. In some embodiments, after determining the target image area, the target image area can be segmented to determine the tissue area and diffuse lesion area in the target image area.

[0059] It should be noted that the target physiological tissue in this embodiment includes diffuse lesions and other tissues except diffuse lesions. In order to distinguish between the two, the diffuse lesion area refers to the image area corresponding to the diffuse lesions, and the tissue area refers to the image area corresponding to other tissues except diffuse lesions. The diffuse lesion area and the tissue area are both substantial areas related to the target physiological tissue.

[0060] Optionally, each period of medical images can be input into a pre-trained diffuse lesion segmentation model to obtain the diffuse lesion area corresponding to each period of medical images, and each period of medical images can also be input into a pre-trained tissue segmentation model to obtain the tissue area corresponding to each period of medical images.

[0061] In this embodiment, it is necessary to optimize the diffuse lesion boundary and the signs of diffuse lesions, such as burrs, color, granularity, etc., according to the diffuse lesion area in the target image area in each period of medical images, so as to obtain a diffuse lesion area that can more accurately characterize the diffuse lesions on the target physiological tissue, which is more conducive to identifying diffuse lesion information.

[0062] The target image region is a partial image region in at least two medical images, and is a region in at least two medical images that meets the recognition requirements, for example, it can be a region with high image quality or a region with higher information content.

[0063] Optionally, the target image area contained in each medical image can be extracted separately, or at least two medical images can be compared together and the target image area can be selected from the two medical images, that is, part of the image area can be selected from each medical image as the target image area, or after comprehensive comparison, the target image area only comes from part of these medical images.

[0064] By using the diffuse lesion area in the target image area in each medical image, the whole or part of the diffuse lesion area of ​​the target physiological tissue can be optimized and spliced ​​to obtain a clearer and more accurate whole or part of the target diffuse lesion area of ​​the target physiological tissue. Before processing, the diffuse lesion area on each medical image is the initially determined diffuse lesion area with low accuracy. It is difficult for each medical image to truly reflect the real diffuse lesion on the target physiological tissue.

[0065] The lesion area reflected by the diffuse lesion area in the target image area in each period of medical imaging is more accurate, and the relevant information of the diffuse lesion contained in the target image area is determined, and the diffuse lesion area in the target image area is used to optimize the initial diffuse lesion area to obtain the target diffuse lesion area of ​​the target physiological tissue, or the target diffuse lesion area of ​​the target physiological tissue is spliced.

[0066] Step S30, determining diffuse lesion information according to the target diffuse lesion area.

[0067] In this embodiment, diffuse lesion information about the target diffuse lesion region can be determined according to the target diffuse lesion region. The target diffuse lesion region can be an image region that can more accurately characterize all or part of the diffuse lesions on the target physiological tissue. In some embodiments, the target diffuse lesion region can be analyzed to achieve local analysis of the diffuse lesions on the target physiological tissue. The target diffuse lesion region can be analyzed to achieve comprehensive analysis of the diffuse lesions on the target physiological tissue. This embodiment does not limit this.

[0068] In some embodiments, according to the target diffuse lesion region, information about the area, size, display characteristics, etc. of the target diffuse lesion region may be determined and output as diffuse lesion information of the target diffuse lesion region.

[0069] In some embodiments, the target diffuse lesion region on the target physiological tissue and the tissue region on the target physiological tissue may be calculated to determine the proportion of the diffuse lesion on the target physiological tissue, and the proportion may be used as the diffuse lesion information of the target physiological tissue.

[0070] In the technical solution disclosed in this embodiment, at least two medical images of the target physiological tissue are obtained; based on the diffuse lesion area in the target image area in the medical images, the target diffuse lesion area of ​​the target physiological tissue is determined; and based on the target diffuse lesion area, the diffuse lesion information is determined. A more refined target diffuse lesion area is obtained through the diffuse lesion area in the target image area in multiple medical images, so that more accurate diffuse lesion information can be identified based on the target diffuse lesion area, thereby improving the accuracy of diffuse lesion identification, and the obtained diffuse lesion information can assist medical diagnosis and improve diagnostic efficiency.

[0071] Furthermore, step S20 “determining a target diffuse lesion region of the target physiological tissue based on the diffuse lesion region within the target image region in the medical image” may include:

[0072] Performing registration processing on at least two phases of medical images to determine an image region corresponding to at least one key feature organization in the at least two phases of medical images;

[0073] For each key feature organization, selecting a target image area corresponding to the key feature organization from the image area corresponding to the key feature organization;

[0074] Integrate the at least one target image region according to the key feature organization corresponding to the at least one target image region;

[0075] The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

[0076] In this embodiment, after segmenting each medical image and determining the diffuse lesion area of ​​the target physiological tissue in each medical image, it is necessary to perform registration processing on at least two medical images so that each tissue in at least two images is registered, that is, for the same tissue, the image area corresponding to the tissue can be found in at least one other medical image, and the positions of the same tissue in at least two medical images are associated. Specifically, when registering at least two medical images, one of the medical images is used as a reference medical image, and the other medical images are registered based on the tissue in the reference medical image, so that each tissue in at least two medical images is registered.

[0077] After the registration process, at least one tissue is selected from these tissues as a key feature tissue, and for each key feature tissue, an image region at a corresponding position of the key feature tissue can be found from at least two medical images of different phases.

[0078] For each key feature tissue, the image regions of the key feature tissue on each medical image are compared, and an optimal image region is selected from the image regions corresponding to the key feature tissue as the target image region corresponding to the key feature tissue. Compared with other target image regions, the target image region can best represent the diffuse lesions on the key feature tissue and can identify more accurate diffuse lesion regions. It can be the image region with the best image quality data, or the image region with the highest amount of information, or the image region with the largest diffuse lesion region, or the image region with richer details of the diffuse lesion region, etc.

[0079] Then, the target image regions are integrated according to the key feature organization of each target image region, the relative position between each target image region and the position of each target image region on the target physiological tissue are determined, and the splicing between each target image region is realized. Figure 2, for the key characteristic organizations A, B, C, D on the target physiological tissue, the corresponding image areas in medical images ①, ②, and ③ are A1, B1, C1, D1, A2, B2, C2, D2, A3, B3, C3, D3, respectively. The corresponding selection of the better target image area distribution is A1, B1, C3, D2. The target image areas are integrated according to the key characteristic organizations to determine the relative positions between the target image areas and the positions of the target image areas on the target physiological tissue. It should be noted that since the medical images of different periods are all taken for the target physiological tissue, it is equivalent to determining the positions of the target image areas and the medical images of each period during the integration, and finally the integrated target image area is obtained. Figure 3 .

[0080] According to the diffuse lesion region in the target image region corresponding to at least one key characteristic tissue, the diffuse lesion region on the corresponding target physiological tissue can be spliced ​​out or optimized to obtain the target diffuse lesion region.

[0081] In this embodiment, since different medical images can see different information, the target image areas for accurately displaying diffuse lesion areas in different medical images are different. For each key feature tissue, the diffuse lesion that best represents the key feature tissue is screened out from each medical image, so as to optimize or splice a more complete and clear target diffuse lesion area based on the diffuse lesion area in the target image area corresponding to at least one key feature tissue, so as to further improve the accuracy of diffuse lesion recognition.

[0082] Furthermore, step S30 of “determining the target diffuse lesion region according to the integrated diffuse lesion region in the at least one target image region” may include:

[0083] According to the integrated target image regions, the diffuse lesion regions in the target image regions are spliced ​​to obtain the target diffuse lesion regions.

[0084] In this embodiment, after registering multiple medical images, the target image area corresponding to at least one key feature tissue can be determined and integrated. After the integration, the relative positions of the target image areas are fixed, and the diffuse lesion areas within the integrated target image areas can be directly spliced ​​according to the integrated positions of the target image areas, and the target tissue areas can be obtained after splicing.

[0085] In an optional implementation scheme of the present embodiment, when the number of key feature tissues exceeds a preset threshold, a scheme of splicing the diffuse lesion area in the target image area according to the integrated target image areas to obtain the target diffuse lesion area can be executed. The more key feature tissues there are, the more corresponding target image area data will be obtained, and then the larger the range of the diffuse lesion area included in all target image areas will be, and thus a more complete target diffuse lesion area can be obtained. Furthermore, after the target image areas corresponding to all key feature tissues are integrated, the complete target physiological tissue can be included, thereby being able to include the diffuse lesions in the complete target physiological tissue, and the most complete target diffuse lesion area can be obtained.

[0086] In this way, after integrating each target image region according to the key feature organization, the diffuse lesion region in the target image region can be spliced ​​to quickly obtain a more accurate target diffuse lesion region related to the target physiological tissue.

[0087] Furthermore, step S30 of “determining the target diffuse lesion region according to the integrated diffuse lesion region in the at least one target image region” may include:

[0088] According to the integrated diffuse lesion regions in each of the target image regions, the diffuse lesion region in at least one medical image is optimized to obtain the target diffuse lesion region.

[0089] In this embodiment, the medical image may contain all or part of the diffuse lesions, but the difference in the accuracy and emphasis of the single-phase medical image may result in the diffuse lesion area contained in the medical image not being able to accurately reflect the diffuse lesions. Therefore, based on the diffuse lesion area in the at least one integrated target image area, the diffuse lesion area in the at least one phase of the medical image may be optimized to obtain the target diffuse lesion area.

[0090] Specifically, for each medical image, since the position of the integrated target image area is fixed relative to the position of the target physiological tissue, the position of the integrated target image area is also fixed relative to the position of the medical image. Therefore, the integrated target image area can be fused or replaced with the image area at the corresponding position of the medical image to optimize the diffuse area and other tissue areas in this medical image. The diffuse lesion area in the integrated target image area can also be fused or replaced with the diffuse lesion area in the image area at the corresponding position of the medical image to optimize the diffuse area in this medical image and obtain the target diffuse lesion area corresponding to the target physiological tissue on the medical image.

[0091] The above method is used to optimize the diffuse lesion area in at least one medical image to obtain the target diffuse lesion area after optimization of at least one medical image. Then, using the target diffuse lesion area after optimization of at least one medical image, the diffuse lesion area corresponding to each optimized medical image can be processed, thereby improving the diffuse lesion recognition rate of a single medical image.

[0092] Further, for each key feature organization, selecting a target image area corresponding to the key feature organization from the image area corresponding to the key feature organization includes:

[0093] For each key feature organization, a target image area corresponding to the key feature organization is selected according to image quality data of each image area corresponding to the key feature organization.

[0094] In this embodiment, there may be differences in image quality between medical images of different periods and between different image regions on the same medical image. Therefore, there are also differences in image quality between image regions corresponding to each key feature tissue in multiple medical images. It can be understood that the better the image quality, the more accurately the image region can characterize diffuse lesions. Therefore, for each key feature tissue, the image quality data of the image region corresponding to the key feature tissue in multiple medical images is obtained. The image quality quantity is a parameter used to evaluate the image quality, including image noise, image color, artifacts, sharpness, etc. The image quality data of each image region corresponding to the key feature tissue is compared, and then the image region with the best image quality can be selected as the target image region corresponding to the key feature tissue.

[0095] In this way, the image quality data of each image area corresponding to the key feature organization is used to select the target image area with better image quality corresponding to the key feature organization, which is used to improve the image quality of the target image area, thereby further improving the accuracy of identifying diffuse lesions.

[0096] Further, based on the diffuse lesion area in the target image area in the medical image, determining the target diffuse lesion area of ​​the target physiological tissue also includes:

[0097] For each medical image, obtaining the image region that meets the preset image conditions;

[0098] Setting the image region to a corresponding target image region in the medical image;

[0099] Integrate the at least one target image region according to the key feature organization corresponding to the at least one target image region;

[0100] The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

[0101] In this embodiment, the target image region may also be an image region in a medical image that meets a preset image condition. The preset image condition may be a condition threshold corresponding to image parameters such as image clarity, contrast, and brightness. When the preset image condition is met, it indicates that the image in the region is more suitable for identification, and it is easier to identify the physiological tissues such as diffuse lesions to be included in the region. From each medical image, an image region in the medical image that meets the preset image condition is extracted as the target image region in the medical image. Each target image region is identified, and the physiological tissues corresponding to different target image regions can be determined. Then, at least one target image region is integrated according to the position of the physiological tissue, so that the target diffuse lesion region can be determined according to the diffuse lesion region in the integrated at least one target image region. The specific method for determining the target diffuse lesion region can refer to the above method, that is, according to the diffuse lesion region in the integrated at least one target image region, the diffuse lesion region in the target image region is spliced ​​to obtain the target diffuse lesion region, or, according to the diffuse lesion region in each target image region after integration, the diffuse lesion region in at least one medical image is optimized to obtain the target diffuse lesion region.

[0102] In this way, by identifying all image areas in the medical image that meet the preset image conditions as the target image area, the information in the medical image can be fully utilized to further improve the accuracy of identifying diffuse lesions.

[0103] Furthermore, before determining the diffuse lesion information according to the target diffuse lesion area, the method further includes:

[0104] For a sub-lesion region in the target diffuse lesion region, obtaining a first diffuse lesion confidence corresponding to the sub-lesion region and a second diffuse lesion confidence corresponding to a neighboring lesion region of the sub-lesion region;

[0105] Performing verification processing on the sub-lesion area according to the first diffuse lesion confidence and the second diffuse lesion confidence;

[0106] The sub-lesion areas in the target diffuse lesion area that have not passed the verification process are screened out.

[0107] In this embodiment, before determining the diffuse lesion information based on the target diffuse lesion area, the target diffuse lesion area can also be verified to filter out the non-diffuse lesion area again and remove the false positive area. The target diffuse lesion area is divided into multiple sub-lesion areas, and at least one of the sub-lesion areas has appeared in at least one medical image. Prior to this, diffuse lesions have been identified in medical images, and the diffuse lesion confidence of the sub-lesion area has been obtained by using a segmentation model and other methods to determine whether the sub-lesion area is a diffuse lesion area. The first diffuse lesion confidence can characterize the possibility that the sub-lesion area is a diffuse lesion. The higher the first diffuse lesion confidence, the more accurate the judgment that the sub-lesion area is a diffuse lesion area.

[0108] It should be noted that the first diffuse lesion confidence corresponding to the sub-lesion area includes the diffuse lesion confidence obtained when the sub-lesion area is identified as a diffuse lesion in at least one medical image, that is, the first diffuse lesion confidence may be the diffuse lesion confidence including the sub-lesion area in at least one medical image.

[0109] The diffuse lesion area is a diffuse lesion and a relatively concentrated lesion area. Based on this, the sub-lesion area can be verified using the neighborhood lesion area of ​​the sub-lesion area. The neighborhood lesion area of ​​the sub-lesion area is the diffuse lesion area adjacent to the sub-lesion area on the target diffuse lesion area. For the sub-lesion area, it is also necessary to obtain the second diffuse lesion confidence of the neighborhood lesion area of ​​the sub-lesion area.

[0110] The neighborhood lesion area has also appeared in at least one medical image. When identifying diffuse lesions in medical images, a segmentation model and other methods are used to calculate the second diffuse lesion confidence of the neighborhood lesion area. The second diffuse lesion confidence can represent the possibility that the neighborhood lesion area is a diffuse lesion. The higher the confidence, the higher the possibility that the neighborhood lesion area is a diffuse lesion area.

[0111] It should be noted that the second diffuse lesion confidence corresponding to the neighborhood lesion area includes the diffuse lesion confidence obtained when the neighborhood lesion area is identified as a diffuse lesion in at least one medical image, that is, the second diffuse lesion confidence may be the diffuse lesion confidence including the neighborhood lesion area in at least one medical image.

[0112] The first diffuse lesion confidence corresponding to the sub-lesion region can characterize the possibility that the sub-lesion region is a diffuse lesion, and the second diffuse lesion confidence of the neighborhood lesion region can characterize the possibility that the neighborhood lesion region is a diffuse lesion. When the possibility that the neighborhood lesion region of the sub-lesion region is a diffuse lesion region is high, and the possibility that it itself is a diffuse lesion region is high, the accuracy of judging that it is a diffuse lesion region is higher. Based on this principle, the sub-lesion region is verified using the first diffuse lesion confidence and the second diffuse lesion confidence to accurately verify whether the sub-lesion region really has a diffuse lesion. The sub-lesion region that passes the verification process is a diffuse lesion region and should be retained. The sub-lesion region that fails the verification process is not a diffuse lesion region and should be screened out. Determining diffuse lesion information based on the screened target diffuse lesion region can improve the accuracy of the diffuse lesion information.

[0113] In this way, the sub-lesion area is verified by the first diffuse lesion confidence corresponding to the sub-lesion area and the second diffuse lesion confidence of the neighboring lesion area of ​​the sub-lesion area, so as to screen out the sub-lesion areas in the target diffuse lesion area that have not passed the verification process, making the target diffuse lesion area more accurate, and further improving the accuracy of diffuse lesions.

[0114] Further, the sub-lesion area is verified according to the first diffuse lesion confidence and the second diffuse lesion confidence, and further includes:

[0115] Get pre-configured diffuse lesion area verification strategies;

[0116] The first diffuse lesion confidence and the second diffuse lesion confidence are verified according to the diffuse lesion area verification strategy.

[0117] In this embodiment, the first diffuse lesion confidence may be the diffuse lesion confidence including the neighborhood lesion region in at least one medical image, and the second diffuse lesion confidence may be the diffuse lesion confidence including the neighborhood lesion region in at least one medical image. That is, there are one or more confidences for the first diffuse lesion confidence, and there are one or more confidences for the second diffuse lesion confidence. Therefore, a pre-configured diffuse lesion region verification strategy may be adopted to verify the first diffuse lesion confidence and the second diffuse lesion confidence. The diffuse lesion region verification strategy includes a confidence processing algorithm and a verification pass condition. The first diffuse lesion confidence and the second diffuse lesion confidence are processed according to the processing algorithm to obtain a processing result, and it is checked whether the processing result meets the verification pass condition to determine whether the sub-lesion region passes the verification process.

[0118] For example, the first average value of each confidence level in the first diffuse lesion confidence level can be calculated, and the second average value of each confidence level in the second diffuse lesion confidence level can be calculated to complete the processing. If both the first average value and the second average value meet the condition of being greater than the preset confidence level, it can be indicated that the sub-lesion area and its neighboring lesion area have a high probability of being diffuse lesions, and then it can be determined that the sub-lesion area has passed the verification process.

[0119] In this way, the first diffuse lesion confidence and the second diffuse lesion confidence are verified according to the pre-configured diffuse lesion area verification strategy, which can quickly complete the verification of the sub-lesion area and improve the efficiency of diffuse lesion recognition.

[0120] Furthermore, before step S20 "determining a target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image", it also includes:

[0121] For each medical image, obtaining pixel values ​​of each image area in the medical image;

[0122] Determining black areas in the medical image according to pixel values ​​of each image area;

[0123] The black areas in the medical image are removed, and based on the removed medical image, the step of determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image is performed.

[0124] In this embodiment, for each medical image, the pixel value of each image area in the medical image is obtained. If the pixel value of the image area is lower than a preset threshold, it is determined that the image area belongs to a black area, thereby determining the black area in the medical image, and then the black area in the medical image is eliminated. Based on the elimination processing, that is, the medical image after the black area elimination processing, the step of determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area within the target image area in the medical image is executed.

[0125] In this way, the pixel values ​​of different image areas in the medical image are used to distinguish the tissue areas of the target physiological tissue in the medical image (such as blood vessels, muscles, fat, etc.), diffuse lesion areas, and other areas related to the target physiological tissue, thereby identifying the black areas in the medical image, and eliminating other areas in the medical image except for the target physiological tissue-related areas, thereby reducing interference and making subsequent processing more accurate.

[0126] This embodiment also provides a diffuse lesion identification device, which can be integrated into a diffuse lesion identification device. Figure 4 As shown, the diffuse lesion identification device may include:

[0127] An acquisition module 1001 is used to acquire at least two medical images of a target physiological tissue;

[0128] A first determination module 1002 is used to determine a target diffuse lesion area of ​​the target physiological tissue based on a diffuse lesion area in a target image area in the medical image;

[0129] The second determination module 1003 is used to determine diffuse lesion information according to the target diffuse lesion area.

[0130] Optionally, the first determining module 1002 is further configured to:

[0131] Performing registration processing on at least two phases of medical images to determine an image region corresponding to at least one key feature organization in the medical images of each phase;

[0132] For each key feature organization, selecting a target image area corresponding to the key feature organization from the image area corresponding to the key feature organization;

[0133] Integrate the at least one target image region according to the key feature organization corresponding to the at least one target image region;

[0134] The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

[0135] Optionally, the first determining module 1002 is further configured to:

[0136] According to the integrated diffuse lesion region in the at least one target image region, the diffuse lesion region in the target image region is spliced ​​to obtain the target diffuse lesion region.

[0137] Optionally, the first determining module 1002 is further configured to:

[0138] According to the integrated diffuse lesion regions in each of the target image regions, the diffuse lesion region in at least one medical image is optimized to obtain the target diffuse lesion region.

[0139] Optionally, the first determining module 1002 is further configured to:

[0140] For each key feature organization, a target image area corresponding to the key feature organization is selected according to image quality data of each image area corresponding to the key feature organization.

[0141] Optionally, the first determining module 1002 is further configured to:

[0142] For each medical image, obtaining the image region that meets the preset image conditions;

[0143] Setting the image region to a corresponding target image region in the medical image;

[0144] integrating the at least one target image region according to a physiological tissue position corresponding to the at least one target image region;

[0145] The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

[0146] Optionally, the acquisition module 1001 is further used for:

[0147] For each medical image, obtaining pixel values ​​of each image area in the medical image;

[0148] Determining black areas in the medical image according to pixel values ​​of each image area;

[0149] The black areas in the medical image are removed, and based on the removed medical image, the step of determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image is performed.

[0150] In this embodiment, at least two medical images of the target physiological tissue are obtained; based on the diffuse lesion area in the target image area in the medical images, the target diffuse lesion area of ​​the target physiological tissue is determined; and based on the target diffuse lesion area, the diffuse lesion information is determined. A more refined target diffuse lesion area is obtained through the diffuse lesion area in the target image area in multiple medical images, so that more accurate diffuse lesion information can be identified based on the target diffuse lesion area, thereby improving the accuracy of diffuse lesion identification, and the obtained diffuse lesion information can assist medical diagnosis and improve diagnostic efficiency.

[0151] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0152] like Figure 5 As shown, Figure 5A schematic diagram of the structure of a diffuse lesion identification device provided in an embodiment of the present invention. The diffuse lesion identification device 1100 includes a processor 1101 having one or more processing cores, a memory 1102 having one or more computer-readable storage media, and a computer program stored in the memory 1102 and executable on the processor. The processor 1101 is electrically connected to the memory 1102. Those skilled in the art will appreciate that the diffuse lesion identification device structure shown in the figure does not constitute a limitation on the diffuse lesion identification device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0153] The processor 1101 is the control center of the diffuse lesion identification device 1100. It uses various interfaces and lines to connect various parts of the entire diffuse lesion identification device 1100, and executes various functions and processes data of the diffuse lesion identification device 1100 by running or loading software programs and / or units stored in the memory 1102, and calling data stored in the memory 1102, so as to monitor the diffuse lesion identification device 1100 as a whole. The processor 1101 can be a processor CPU, a graphics processor GPU, a network processor (Network Processor, NP), etc., and can implement or execute various methods, steps and logic block diagrams disclosed in the embodiments of the present invention.

[0154] In the embodiment of the present invention, the processor 1101 in the diffuse lesion identification device 1100 will load instructions corresponding to the processes of one or more application programs into the memory 1102 according to the following steps, and the processor 1101 will run the application programs stored in the memory 1102, thereby realizing various functions, such as:

[0155] Obtain at least two phases of medical images of the target physiological tissue;

[0156] Determining a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region within a target image region in the medical image;

[0157] Diffuse lesion information is determined according to the target diffuse lesion area.

[0158] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0159] Optional, such as Figure 5As shown, the diffuse lesion identification device 1100 also includes: a touch screen 1103, a radio frequency circuit 1104, an audio circuit 1105, an input unit 1106 and a power supply 1107. Among them, the processor 1101 is electrically connected to the touch screen 1103, the radio frequency circuit 1104, the audio circuit 1105, the input unit 1106 and the power supply 1107 respectively. Those skilled in the art can understand that Figure 5 The structure of the diffuse lesion identification device shown in the figure does not constitute a limitation on the diffuse lesion identification device, and may include more or fewer components than shown in the figure, or combine certain components, or arrange the components differently.

[0160] The touch display screen 1103 can be used to display a graphical user interface and receive operation instructions generated by the user acting on the graphical user interface. The touch display screen 1103 may include a display panel and a touch panel. Among them, the display panel can be used to display information input by the user or information provided to the user and various graphical user interfaces of the diffuse lesion identification device, which can be composed of graphics, text, icons, videos and any combination thereof. Optionally, the display panel can be configured in the form of a liquid crystal display (LCD, Liquid Crystal Display), an organic light-emitting diode (OLED, Organic Light-EmittingDiode), etc. The touch panel can be used to collect the user's touch operation on or near it (such as the user uses any suitable object or accessory such as a finger, a stylus, etc. on the touch panel or near the touch panel), and generate corresponding operation instructions, and the operation instructions execute the corresponding program. Optionally, the touch panel may include two parts: a touch detection device and a touch controller. Among them, the touch detection device detects the user's touch orientation, detects the signal brought by the touch operation, and transmits the signal to the touch controller; the touch controller receives the touch information from the touch detection device, converts it into the touch point coordinates, and then sends it to the processor 1101, and can receive the command sent by the processor 1101 and execute it. The touch panel can cover the display panel. When the touch panel detects a touch operation on or near it, it is transmitted to the processor 1101 to determine the type of touch event, and then the processor 1101 provides corresponding visual output on the display panel according to the type of touch event. In an embodiment of the present invention, the touch panel and the display panel can be integrated into the touch display screen 1103 to realize the input and output functions. However, in some embodiments, the touch panel and the touch panel can be used as two independent components to realize the input and output functions. That is, the touch display screen 1103 can also be used as a part of the input unit 1106 to realize the input function.

[0161] The radio frequency circuit 1104 can be used to send and receive radio frequency signals to establish wireless communication with a network device or other diffuse lesion identification devices through wireless communication, and to send and receive signals between the network device or other diffuse lesion identification devices.

[0162] The audio circuit 1105 can be used to provide an audio interface between the user and the diffuse lesion identification device through a speaker and a microphone. The audio circuit 1105 can transmit the electrical signal converted from the received audio data to the speaker, which is converted into a sound signal for output; on the other hand, the microphone converts the collected sound signal into an electrical signal, which is received by the audio circuit 1105 and converted into audio data, and then the audio data is output to the processor 1101 for processing, and then sent to another diffuse lesion identification device through the radio frequency circuit 1104, or the audio data is output to the memory 1102 for further processing. The audio circuit 1105 may also include an earplug jack to provide communication between an external earphone and the diffuse lesion identification device.

[0163] The input unit 1106 may be used to receive input numbers, character information or user feature information (such as fingerprint, iris, facial information, etc.), and to generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function control.

[0164] The power supply 1107 is used to supply power to various components of the diffuse lesion identification device 1100. Optionally, the power supply 1107 can be logically connected to the processor 1101 through a power management system, so as to manage charging, discharging, and power consumption through the power management system. The power supply 1107 can also include one or more DC or AC power supplies, recharging systems, power failure detection circuits, power converters or inverters, power status indicators, and other arbitrary components.

[0165] although Figure 5 Not shown, the diffuse lesion identification device 1100 may also include a camera, a sensor, a wireless fidelity module, a Bluetooth module, etc., which will not be repeated here.

[0166] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0167] A person of ordinary skill in the art will appreciate that all or part of the steps in the various methods of the above embodiments may be completed by instructions, or by controlling related hardware through instructions. The instructions may be stored in a computer-readable storage medium and loaded and executed by a processor.

[0168] To this end, an embodiment of the present invention provides a computer-readable storage medium, in which a plurality of computer programs are stored, and the computer program can be loaded by a processor to execute any diffuse lesion identification method provided in an embodiment of the present invention. The computer program can execute the following steps of the diffuse lesion identification method:

[0169] Obtain at least two phases of medical images of the target physiological tissue;

[0170] Determining a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region within a target image region in the medical image;

[0171] Diffuse lesion information is determined according to the target diffuse lesion area.

[0172] The specific implementation of the above operations can be found in the previous embodiments, which will not be described in detail here.

[0173] The computer-readable storage medium may include: a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk, etc.

[0174] Since the computer program stored in the computer-readable storage medium can execute any diffuse lesion identification method provided in the embodiments of the present invention, the beneficial effects that can be achieved by any diffuse lesion identification method provided in the embodiments of the present invention can be achieved. Please refer to the previous embodiments for details and will not be repeated here.

[0175] In the above-mentioned diffuse lesion identification device, computer-readable storage medium, diffuse lesion identification equipment, and computer program product embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments. Those skilled in the art can clearly understand that for the convenience and simplicity of description, the specific working process and beneficial effects of the above-described diffuse lesion identification device, computer-readable storage medium, computer program product, diffuse lesion identification equipment, and corresponding units can refer to the description of the diffuse lesion identification method in the above embodiment, and will not be repeated here.

[0176] The above is a detailed introduction to a diffuse lesion identification method, a diffuse lesion identification device, a diffuse lesion identification equipment, a computer-readable storage medium and a computer program product provided in an embodiment of the present invention. Specific examples are used herein to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea. At the same time, for those skilled in the art, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A diffuse lesion identification method, characterized in that: The diffuse lesion identification method comprises: Obtain at least two phases of medical images of the target physiological tissue; Determining a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region within a target image region in the medical image; Diffuse lesion information is determined according to the target diffuse lesion area.

2. The diffuse lesion identification method according to claim 1, characterized in that: The determining of the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image includes: Performing registration processing on at least two phases of medical images to determine an image region corresponding to at least one key feature organization in the at least two phases of medical images; For each key feature organization, selecting a target image area corresponding to the key feature organization from the image area corresponding to the key feature organization; Integrate the at least one target image region according to the key feature organization corresponding to the at least one target image region; The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

3. The diffuse lesion identification method according to claim 2, characterized in that: Determining the target diffuse lesion area according to the integrated diffuse lesion area in the at least one target image area includes: According to the integrated diffuse lesion region in the at least one target image region, the diffuse lesion region in the target image region is spliced ​​to obtain the target diffuse lesion region.

4. The diffuse lesion identification method according to claim 2, characterized in that: Determining the target diffuse lesion area according to the integrated diffuse lesion area in the at least one target image area includes: According to the integrated diffuse lesion regions in each of the target image regions, the diffuse lesion region in at least one medical image is optimized to obtain the target diffuse lesion region.

5. The diffuse lesion identification method according to claim 2, characterized in that: For each key feature organization, selecting a target image area corresponding to the key feature organization from the image area corresponding to the key feature organization includes: For each key feature organization, a target image area corresponding to the key feature organization is selected according to image quality data of each image area corresponding to the key feature organization.

6. The diffuse lesion identification method according to claim 1, characterized in that: Before determining the diffuse lesion information according to the target diffuse lesion area, the method further includes: For a sub-lesion region in the target diffuse lesion region, obtaining a first diffuse lesion confidence corresponding to the sub-lesion region and a second diffuse lesion confidence corresponding to a neighboring lesion region of the sub-lesion region; Performing verification processing on the sub-lesion area according to the first diffuse lesion confidence and the second diffuse lesion confidence; The sub-lesion areas in the target diffuse lesion area that have not passed the verification process are screened out.

7. The diffuse lesion identification method according to claim 6, characterized in that: The verifying process of the sub-lesion region according to the first diffuse lesion confidence and the second diffuse lesion confidence includes: Get pre-configured diffuse lesion area verification strategies; The first diffuse lesion confidence and the second diffuse lesion confidence are verified according to the diffuse lesion area verification strategy.

8. The diffuse lesion identification method according to claim 1, characterized in that: The step of determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image further includes: For each medical image, obtaining the image region that meets the preset image conditions; Setting the image region to a corresponding target image region in the medical image; integrating the at least one target image region according to a physiological tissue position corresponding to the at least one target image region; The target diffuse lesion region is determined according to the integrated diffuse lesion region within the at least one target image region.

9. The diffuse lesion identification method according to any one of claims 1 to 8, characterized in that: Before determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image, the method further includes: For each medical image, obtaining pixel values ​​of each image area in the medical image; Determining black areas in the medical image according to pixel values ​​of each image area; The black areas in the medical image are removed, and based on the removed medical image, the step of determining the target diffuse lesion area of ​​the target physiological tissue based on the diffuse lesion area in the target image area in the medical image is performed.

10. A diffuse lesion identification device, characterized in that: The diffuse lesion identification device comprises: An acquisition module, used for acquiring at least two phases of medical images of a target physiological tissue; A first determination module, configured to determine a target diffuse lesion region of the target physiological tissue based on a diffuse lesion region within a target image region in the medical image; The second determination module is used to determine diffuse lesion information according to the target diffuse lesion area.

11. A diffuse lesion identification device, characterized in that: The method comprises a processor and a memory, wherein the memory stores a computer program. When the computer program is executed by the processor, the processor executes the steps of the diffuse lesion identification method according to any one of claims 1 to 9.

12. A computer-readable storage medium, characterized in that: The computer-readable storage medium comprises a computer program. When the computer program is run on an electronic device, the computer program is used to enable the electronic device to execute the steps of the diffuse lesion identification method according to any one of claims 1 to 9.

Images