Organ lesion detection method and device, processor and program product

Abstract

The invention discloses an organ lesion detection method and device, a processor and a program product. The method comprises the following steps: acquiring an organ image of a to-be-diagnosed organ in a target biological object; identifying the organ image to obtain an identification result; in response to the identification result that the organ image is a target image, based on the organ image, determining a density region in an abnormal state on the to-be-diagnosed organ and the shape of the density region, the organ in the target image being a diseased organ; based on the density area and the shape, a lesion result of the organ to be diagnosed is determined, and the lesion result is used for representing the lesion degree of the organ to be diagnosed. The technical problem of low image recognition efficiency is solved.

Description

Technical Field

[0001] The present application relates to the field of image analysis, and in particular, to a method, device, processor and program product for detecting organ lesions. Background Art

[0002] At present, clinical identification of lesions and assessment of their severity usually rely on clinicians to visually judge and roughly evaluate organ images. The accuracy of the assessment depends entirely on the doctor's personal experience. In addition, the organ images of organs to be diagnosed have shown an explosive growth, which has greatly increased the doctor's workload. Therefore, the above method has the technical problem of low efficiency in recognizing images.

[0003] To address the above-mentioned problems, no effective solution has been proposed yet. Summary of the invention

[0004] The embodiments of the present application provide a method, device, processor and program product for detecting organ lesions, so as to at least solve the technical problem of low image recognition efficiency.

[0005] According to one aspect of an embodiment of the present application, a method for detecting organ lesions is provided, which may include: acquiring an organ image of an organ to be diagnosed in a target biological object; identifying the organ image to obtain a recognition result; in response to the recognition result that the organ image is a target image, determining, based on the organ image, a density area in an abnormal state on the organ to be diagnosed, and a shape of the density area, wherein the organ in the target image is a diseased organ; based on the density area and the shape, determining a lesion result of the organ to be diagnosed, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0006] Optionally, identifying the organ image to obtain the identification result includes: calling an image recognition model to identify the organ image to obtain the identification result, wherein the image recognition model is trained based on historical organ images of multiple sizes.

[0007] Optionally, the method may also include: inputting the historical organ image into the initial image recognition model for feature extraction to obtain shadow pixel features; calibrating the lesion area of ​​the diseased organ in the historical organ image; and adjusting the model parameters of the initial image recognition model using the shadow pixel features and the calibrated historical organ image to obtain an image recognition model.

[0008] Optionally, the model parameters of the initial image recognition model are adjusted using shadow pixel features and calibrated historical organ images to obtain an image recognition model, including: constructing a multidimensional training loss function, wherein the multidimensional training loss function is used to adjust the shadow pixel features output by the initial image recognition model; using the multidimensional training loss function, shadow pixel features and calibrated historical organ images to adjust the model parameters of the initial image recognition model to obtain an image recognition model.

[0009] Optionally, determining the lesion result of the organ to be diagnosed based on the density area and shape includes: classifying the organ image based on the density area and shape to obtain a classification result; and determining the classification result as the lesion result of the organ to be diagnosed.

[0010] Optionally, the organ image is classified based on the density area and shape to obtain a classification result, including: obtaining the lesion density area corresponding to the lesion organ, and the lesion shape corresponding to the lesion organ; based on the lesion density area and the lesion shape, the density area and the shape are classified to obtain a classification result.

[0011] Optionally, before identifying the organ image, the method may further include: performing grayscale correction on the organ image to obtain a corrected image; performing artifact removal on the corrected image to obtain a standard image; and performing segmentation processing on the standard image to obtain the organ image.

[0012] According to another aspect of an embodiment of the present application, a device for detecting organ lesions is also provided, which may include: an acquisition unit, used to acquire an organ image of an organ to be diagnosed in a target biological object; an identification unit, used to identify the organ image and obtain an identification result; a first determination unit, used to determine, in response to the identification result that the organ image is a target image, a density area in an abnormal state on the organ to be diagnosed, and a shape of the density area based on the organ image, wherein the organ in the target image is a diseased organ; a second determination unit, used to determine the lesion result of the organ to be diagnosed based on the density area and the shape, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0013] According to another aspect of an embodiment of the present application, a computer-readable storage medium is also provided, which includes a stored program, wherein when the program is running, the device where the computer-readable storage medium is located is controlled to execute the organ lesion detection method of the embodiment of the present application.

[0014] According to another aspect of an embodiment of the present application, a processor is further provided, the processor being used to run a program, wherein when the program is run by the processor, the organ lesion detection method of the embodiment of the present application is executed.

[0015] According to another aspect of an embodiment of the present application, a computer program product is further provided. The computer program product includes computer instructions, wherein the computer instructions, when executed by a processor, implement the organ lesion detection method of the embodiment of the present application.

[0016] In an embodiment of the present application, an organ image of an organ to be diagnosed in a target biological object is obtained; the organ image is identified to obtain an identification result; in response to the identification result that the organ image is a target image, a density area in an abnormal state on the organ to be diagnosed and the shape of the density area are determined based on the organ image, wherein the organ in the target image is a diseased organ; based on the density area and the shape, a lesion result of the organ to be diagnosed is determined, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed. That is, in an embodiment of the present application, an organ image to be diagnosed is obtained, and the organ image is identified. If the organ image is a target image with lesions, the density area in an abnormal state on the organ to be diagnosed (that is, an abnormal density area) and the shape of the abnormal density area can be determined based on the organ image. By identifying the abnormal density area and shape, the lesion result of the organ to be diagnosed can be determined, and the lesion result can be used to determine the degree of lesion in the organ to be diagnosed, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The drawings described herein are used to provide a further understanding of the present application and constitute a part of the present application. The illustrative embodiments of the present application and their descriptions are used to explain the present application and do not constitute an improper limitation on the present application. In the drawings:

[0018] Figure 1 is a flow chart of a method for detecting organ lesions according to an embodiment of the present application;

[0019] Figure 2 is a schematic diagram of an analysis system for pneumoconiosis images according to an embodiment of the present application;

[0020] Figure 3 is a schematic diagram of a lesion detection device for an organ according to an embodiment of the present application;

[0021] Figure 4 is a structural block diagram of a computer terminal according to an embodiment of the present application;

[0022] Figure 5 It is a block diagram of an electronic device for an organ lesion detection method according to an embodiment of the present application. DETAILED DESCRIPTION

[0023] 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 should fall within the scope of protection of the present application.

[0024] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0025] According to an embodiment of the present application, an embodiment of a method for detecting lesions of an organ is provided. The steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0026] In this embodiment, a method for detecting organ lesions is proposed. The method can obtain an organ image to be diagnosed and identify the organ image. If the organ image is a target image with lesions, the density area (that is, abnormal density area) in an abnormal state and the shape of the abnormal density area on the organ to be diagnosed can be determined based on the organ image. By identifying the abnormal density area and shape, the lesion result of the organ to be diagnosed can be determined, and the lesion result can be used to determine the degree of lesions in the organ to be diagnosed, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency.

[0027] Figure 1 FIG. 1 is a flow chart of a method for detecting organ lesions according to an embodiment of the present application. Figure 1 As shown, the method may include the following steps:

[0028] Step S102, acquiring an organ image of the organ to be diagnosed in the target biological object.

[0029] In the technical solution provided in the above step S102 of the present application, the above target biological object can be an object with organs such as animals and humans. The above organ to be diagnosed can be an organ of a biological object, such as a lung, liver, heart, etc., for example, it can be a tissue organ in the human body. The above organ image can be an image obtained by radiographic imaging technology, for example, it can be a chest X-ray image. It should be noted that this is only an example, and there is no specific limitation on the type of target biological object, the type of organ to be diagnosed, and the type of organ image.

[0030] Optionally, the body part corresponding to the organ to be diagnosed of the target biological object is detected by radiographic imaging technology to obtain an organ image of the organ to be diagnosed.

[0031] Step S104, identifying the organ image to obtain a recognition result.

[0032] In the technical solution provided in the above step S104 of the present application, the acquired organ image can be identified to obtain an identification result, wherein the above identification result can be used to determine whether there is a diseased organ in the organ image.

[0033] Optionally, this embodiment can identify the organ image through a neural network model to obtain a recognition result, and can also obtain the recognition result by directly comparing the target image containing the diseased organ with the organ image, or can identify the organ image through an edge detection algorithm (such as Canny edge detection, Sobel operator, Prewitt operator, etc.), feature matching method, etc. to obtain the recognition result. It should be noted that this is only an example, and there is no specific limitation on the organ image recognition method.

[0034] Step S106, in response to the recognition result that the organ image is a target image, based on the organ image, determining a density area in an abnormal state on the organ to be diagnosed and a shape of the density area, wherein the organ in the target image is a diseased organ.

[0035] In the technical solution provided in the above step S106 of the present application, if the recognition result is that the organ image is the target image, it can be determined that the organ to be diagnosed in the organ image may be an organ with a lesion, and the lesion of the organ to be diagnosed can be further determined. The density area in an abnormal state on the organ to be diagnosed and the shape of the density area can be determined through the organ image, and the density area and shape can be further judged. Among them, the above target image can be a pre-acquired pneumoconiosis image, which can include a lesion organ with pneumoconiosis. It should be noted that this is only an example, and there is no specific limitation on the method of acquiring the target image and the type of the target image.

[0036] Optionally, this embodiment can use the target image to identify the organ image to obtain a recognition result. If the organ image and the target image are similar, the recognition result can be that the organ image is the target image. If the organ image is the target image, the organ to be diagnosed can be further identified based on the organ image to determine the density area in an abnormal state on the organ to be diagnosed, and the shape of the density area. The above-mentioned organ image may include the organ to be diagnosed. The above-mentioned density area in an abnormal state can be a shadow in the pixel density of the pneumoconiosis fiber shadow, can be a shadow on the organ to be diagnosed, can be an abnormal density area, for example, can be a pneumoconiosis shadow. The above-mentioned shape can be a shadow shape.

[0037] Optionally, the "shadow" in the pneumoconiosis fiber shadow pixel density may refer to an abnormal density area observed in an organ image (e.g., an X-ray chest film or a CT image), wherein the density of the abnormal density area is higher than the density of normal lung tissue. The density area may be used to characterize information such as the size, shape, distribution, and density of the shadow, and may be used to determine the degree of pathological changes in the organ to be diagnosed.

[0038] Step S108, determining the lesion result of the organ to be diagnosed based on the density area and the shape, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0039] In the technical solution provided in the above step S108 of the present application, the lesion result of the organ to be diagnosed can be determined based on the density area and shape in an abnormal state. The lesion result can be the lesion grade of the organ to be diagnosed, which can be used to indicate the degree of lesions in the organ to be diagnosed. For example, it can be used to reflect the severity and type of pneumoconiosis, and can be used to stage or grade pneumoconiosis. It should be noted that this is only an example, and there is no specific limitation on the type of lesion result.

[0040] In this embodiment, based on the above-mentioned organ lesion detection method, an analysis system for pneumoconiosis images is also disclosed. The analysis system can be used to execute the above-mentioned organ lesion detection method. Figure 2 is a schematic diagram of an analysis system for pneumoconiosis images according to an embodiment of the present application, such as Figure 2 As shown, the pneumoconiosis image analysis system may include: an image acquisition module 201, an image processing module 202, an image analysis module 203 and a grade classification module 204. The image processing module 202 may include a first image processing unit 2021 and a second image processing unit 2022. The image analysis module 203 may include a model building unit 2031, a model training unit 2032 and an image analysis unit 2033. The grade classification module 204 may include a grade classification unit 2041.

[0041] Optionally, the image acquisition module 201, the image processing module 202, the image analysis module 203 and the grade classification module 204 are connected in sequence.

[0042] Optionally, the image acquisition module 201 can be used to acquire an organ image of an organ to be diagnosed of a target biological object, such as a chest X-ray image. The image processing module 202 can be used to process the chest X-ray image. The image analysis module 203 can be used to construct a pneumoconiosis image analysis model, and input the processed chest X-ray image into the pneumoconiosis image analysis model to complete the recognition of the chest X-ray image and obtain a recognition result, which can be used to determine whether the chest X-ray image is a pneumoconiosis image. The grade classification module 204 can be used to grade the chest X-ray image when the chest X-ray image is a pneumoconiosis image and obtain a classification result (i.e., a lesion result). It should be noted that the module types included in the above system are for illustrative purposes only and are not specifically limited here.

[0043] Through the above steps S102 and S108 of the present application, an organ image of the organ to be diagnosed in the target biological object is obtained; the organ image is identified to obtain an identification result; in response to the identification result that the organ image is a target image, based on the organ image, a density area in an abnormal state on the organ to be diagnosed and the shape of the density area are determined, wherein the organ in the target image is a diseased organ; based on the density area and the shape, the lesion result of the organ to be diagnosed is determined, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed. That is, in an embodiment of the present application, an organ image to be diagnosed is obtained, and the organ image is identified. If the organ image is a target image with lesions, the density area in an abnormal state on the organ to be diagnosed (that is, an abnormal density area) and the shape of the abnormal density area can be determined based on the organ image. By identifying the abnormal density area and shape, the lesion result of the organ to be diagnosed can be determined, and the lesion result can be used to determine the degree of lesion in the organ to be diagnosed, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency.

[0044] The above method of this embodiment is further introduced below.

[0045] As an optional implementation, step S104, identifying the organ image to obtain a recognition result, includes: calling an image recognition model to identify the organ image to obtain a recognition result, wherein the image recognition model is trained based on historical organ images of multiple sizes.

[0046] In this embodiment, an image recognition model can be obtained by training based on historical organ images of different sizes. If the historical organ image is a pneumoconiosis image corresponding to pneumoconiosis, the image recognition model can be a pneumoconiosis image analysis model. Further, the image recognition model is called to recognize the organ image to obtain a recognition result. The historical organ image can be a historical image containing a diseased organ.

[0047] Optionally, if the constructed image recognition model is a pneumoconiosis image analysis model, the pneumoconiosis image analysis model can be obtained by training based on pneumoconiosis images. If the constructed image recognition model is a cardiac image recognition model, the cardiac image recognition model can be obtained by training based on diseased cardiac images. It should be noted that this is only an example, and there is no specific restriction on the type of image recognition model obtained by training. According to actual needs, the required image recognition model can be trained and the corresponding image recognition model can be called to recognize organ images.

[0048] The following is an example of a pneumoconiosis image analysis model. It should be noted that the following numbers are only for illustration and can be adjusted according to actual conditions.

[0049] As an optional implementation, the method may also include: inputting the historical organ image into the initial image recognition model for feature extraction to obtain shadow pixel features; calibrating the lesion area of ​​the diseased organ in the historical organ image; and adjusting the model parameters of the initial image recognition model using the shadow pixel features and the calibrated historical organ image to obtain an image recognition model.

[0050] In this embodiment, the initial image recognition model may be an initial pneumoconiosis image analysis model. The shadow pixel feature may be a pneumoconiosis shadow pixel feature, which may be used to determine the shadow area. The historical organ image may be a historical pneumoconiosis image.

[0051] Optionally, assuming that the constructed image recognition model is a pneumoconiosis image analysis model, the following steps can be performed to train the historical pneumoconiosis images to obtain the pneumoconiosis image analysis model: the historical pneumoconiosis images can be input into the initial pneumoconiosis image analysis model to obtain pneumoconiosis shadow pixel features; the pneumoconiosis shadow pixels in the lung area of ​​the historical pneumoconiosis images are calibrated to obtain the calibrated historical pneumoconiosis images; according to the pneumoconiosis shadow pixel features and the calibrated historical pneumoconiosis images, the model function of the pneumoconiosis image analysis model is trained by a multidimensional training loss function to obtain the image recognition model.

[0052] Optionally, the historical organ image is input into the initial image recognition model for feature extraction to obtain shadow pixel features, and the lesion area of ​​the lesion organ in the historical organ image can be calibrated. The initial image recognition model is trained using the shadow pixel features and the calibrated historical organ image to obtain an image recognition model. For example, the recognition result obtained after the initial image recognition model recognizes the calibrated historical organ image can be obtained, and the model parameters of the initial image recognition model are adjusted using the multi-dimensional training loss function, the recognition result and the shadow pixel features to obtain a trained image recognition model. The obtained image recognition model can be used to identify the organ image to determine whether there is a lesion organ in the organ image. It should be noted that this is only an example, and there is no specific limitation on the process of training the initial image recognition model.

[0053] For example, the initial pneumoconiosis image analysis model may include: 8 convolutional layers connected in sequence, each convolutional layer may use a 1×1 convolution kernel to extract corresponding features; the image pixels of the 6 convolutional layers may be 256×256, 128×128, 64×64, 32×32, and 16×16, respectively. According to the above content, the initial pneumoconiosis image analysis model is constructed. After the initial pneumoconiosis image is constructed, historical pneumoconiosis images with pneumoconiosis shadow pixels of different sizes may be obtained. The initial pneumoconiosis image analysis model may be trained through the historical pneumoconiosis images to obtain a pneumoconiosis image analysis model and obtain a trained pneumoconiosis image analysis model. The trained pneumoconiosis image analysis model is used to identify the chest X-ray image to determine whether the chest X-ray image is a pneumoconiosis image.

[0054] Optionally, the processed chest X-ray image is input into a pneumoconiosis image analysis model, and the pneumoconiosis image analysis model is used to determine whether the chest X-ray image is a pneumoconiosis image. When the chest X-ray image is a pneumoconiosis image, the shadow shape and corresponding coordinates of the pneumoconiosis fiber shadow pixels in the chest X-ray image can be obtained. When the chest X-ray image is not a pneumoconiosis image, no processing is performed.

[0055] As an optional implementation, the model parameters of the initial image recognition model are adjusted using shadow pixel features and calibrated historical organ images to obtain an image recognition model, including: constructing a multidimensional training loss function, wherein the multidimensional training loss function is used to adjust the shadow pixel features output by the initial image recognition model; using the multidimensional training loss function, shadow pixel features and calibrated historical organ images to adjust the model parameters of the initial image recognition model to obtain an image recognition model.

[0056] In this embodiment, the multi-dimensional training loss function may include: a target classification error function and a coordinate error function.

[0057] Optionally, the target classification error function may be:

[0058]

[0059] Wherein, W may be a model parameter of a convolutional neural network model. S = {S1, S2, ..., SN} may be a training sample set, which may include multiple historical organ images. N may be the number of convolutional layers. n The error function l can be n The weight of . Xi can be the labeled sample image block. Yi=(yi,bi) can be the labeled sample label, yi∈{0,1,...,K} can be the real target frame coordinates, and (x, y, w, h) can be represented as the center point coordinates and width and length values ​​of the target frame respectively.

[0060] The above coordinate error function can be:

[0061]

[0062] Where m is the number of convolution layers where the predicted target box is located during detection. i is the initial target box coordinates for detection. j is the real target box coordinates. The coordinate error function can use a smooth loss function (smoothL1()) to solve the coordinate error between the predicted target box and the real target box. (x, y, w, h) can be used to represent the center point coordinates and width and length of the target box, respectively. It can be used to predict the center, length and width coordinates of the target box. It can be the coordinate difference between the predicted target box and the real target box.

[0063] Optionally, the target classification error function and the coordinate error function are jointly trained and learned to fuse the target feature vector output by each convolution layer. That is, the multi-dimensional training loss function can be used to adjust the features output by each convolution layer in the initial image recognition model to obtain accurate shadow pixel features.

[0064] In this embodiment, the model parameters of the initial image recognition model are adjusted using a multi-dimensional training loss function, shadow pixel features, and calibrated historical organ images to obtain an image recognition model.

[0065] For example, the model building unit in the image analysis module can be used to build an initial pneumoconiosis image analysis model. The model training unit in the image analysis module is used to obtain historical pneumoconiosis images with pneumoconiosis shadow pixels of different sizes, and the historical pneumoconiosis images are input into the initial pneumoconiosis image analysis model to obtain pneumoconiosis shadow pixel features; the pneumoconiosis shadow pixels in the lung area of ​​the historical pneumoconiosis images are calibrated to obtain the calibrated historical pneumoconiosis images; according to the pneumoconiosis shadow pixel features and the calibrated historical pneumoconiosis images, the model function of the initial pneumoconiosis image analysis model is trained through a multidimensional training loss function, thereby obtaining a pneumoconiosis image analysis model.

[0066] Optionally, the image analysis unit in the image analysis module may include a trained pneumoconiosis image analysis model, and the processed chest X-ray image may be input into the trained pneumoconiosis image analysis model. Using the pneumoconiosis image analysis model, it may be determined whether the chest X-ray image is a pneumoconiosis image. If the chest X-ray image is determined to be a pneumoconiosis image, the shadow shape and corresponding coordinates of the pneumoconiosis fiber shadow pixels in the chest X-ray image may be obtained to further determine the degree of the lesion of the organ to be diagnosed in the organ image. When the chest X-ray image is a non-pneumoconiosis image, no processing is performed.

[0067] Optionally, the above-mentioned model building unit, model training unit and image analysis unit are connected in sequence.

[0068] In this embodiment, a pneumoconiosis image analysis model is constructed and trained, and the pneumoconiosis image analysis model is used to determine whether a chest X-ray image is a pneumoconiosis image, and whether the chest X-ray image is a pneumoconiosis image can be automatically detected without increasing the workload of doctors, thereby avoiding misdiagnosis and missed diagnosis. If the chest X-ray image is determined to be a pneumoconiosis image, the pathological result of the organ to be diagnosed can be determined based on the organ image, thereby achieving the purpose of improving data processing efficiency.

[0069] As an optional implementation, determining the lesion result of the organ to be diagnosed based on the density area and shape includes: classifying the organ image based on the density area and shape to obtain a classification result; and determining the classification result as the lesion result of the organ to be diagnosed.

[0070] In this embodiment, the density area and shape are obtained, and the organ image can be classified based on the density area and shape to obtain a classification result, and the lesion result of the organ to be diagnosed can be determined based on the classification result. The above classification result can be used to determine which lesion level the organ to be diagnosed belongs to, for example, it can be level one, level two, level three, or can be mid-term, early stage, late stage, etc. It should be noted that this is only an example, and the type of classification result is not specifically limited.

[0071] For example, the pneumoconiosis fiber shadow density (i.e., density area) can be obtained based on the shadow shape and corresponding coordinates of the pneumoconiosis fiber shadow pixels in the chest X-ray image. A shadow density judgment standard can be constructed in advance, and the pneumoconiosis fiber shadow pixel density and shadow shape can be graded according to the shadow density judgment standard to obtain a classification result, which can be determined as a lesion result. For example, based on the density area and shape, it is determined that the lesion level of the organ to be diagnosed corresponds to the pneumoconiosis stage, then the classification result can be determined as the pneumoconiosis stage, and further, the lesion result can be determined as the pneumoconiosis stage.

[0072] As an optional implementation, organ images are classified based on density areas and shapes to obtain classification results, including: obtaining the lesion density area corresponding to the lesion organ, and the lesion shape corresponding to the lesion organ; based on the lesion density area and the lesion shape, the density area and the shape are classified to obtain classification results.

[0073] In this embodiment, the lesion density area and lesion shape corresponding to the lesion organ in different classification results can be obtained. The lesion density area and lesion shape can be used as standard data to judge the density area and shape of the organ image.

[0074] Optionally, the density regions and shapes are classified using the lesion density regions and lesion shapes corresponding to different classification results. If the similarity between the lesion density regions and the density regions is greater than a similarity threshold, and the similarity between the lesion shapes and the shapes is greater than a similarity threshold, the classification result of the organ to be diagnosed can be determined to be the classification result corresponding to the lesion density region. It should be noted that this is only an example, and the method for determining the classification result is not specifically limited.

[0075] For example, the retrieved pneumoconiosis patients can be pre-screened for those who have been confirmed by at least three qualified physicians, have digital chest radiographs, are 18 years old or older, and have no other basic lung diseases. The pneumoconiosis report of the pneumoconiosis patient includes the pneumoconiosis stage, shadow morphology, and shadow density. Obtain and collect the information of the retrieved and screened pneumoconiosis patients, including name, gender, age, occupational history, pneumoconiosis stage, shadow morphology, and shadow density; wherein the pneumoconiosis stage includes: stage 0, stage 1, stage 2, and stage 3. Classify the patients according to their pneumoconiosis stage, small shadow morphology, and small shadow density, set threshold ranges according to different shadow morphologies and shadow densities, and divide them into pneumoconiosis stages, thereby establishing a shadow density judgment standard. The shadow density judgment standard may include the corresponding lesion density area and lesion shape under different classification results. Based on the above standard data, the density area and shape in the chest X-ray image are judged to determine the classification result corresponding to the chest X-ray image. It should be noted that the above method for obtaining the lesion density area and lesion shape is only for illustration and is not specifically limited here.

[0076] In this embodiment, the grading unit in the grading module can be used to obtain the pneumoconiosis fiber shadow density according to the shadow shape and corresponding coordinates of the pneumoconiosis fiber shadow pixels in the chest X-ray image. Further, the pneumoconiosis fiber shadow pixel density and shadow shape can be graded based on the pre-constructed shadow density judgment standard to obtain a classification result.

[0077] In this embodiment, a shadow density judgment standard is pre-constructed, and the pneumoconiosis fiber shadow pixel density and shadow shape are graded according to the shadow density judgment standard, so that the pneumoconiosis grade can be judged quickly, accurately and efficiently, providing a strong basis for subsequent treatment.

[0078] As an optional implementation, before identifying the organ image, the method further includes: performing grayscale correction on the organ image to obtain a corrected image; removing artifacts from the corrected image to obtain a standard image; and performing segmentation processing on the standard image to obtain the organ image.

[0079] In this embodiment, before the organ image is identified, the organ image may be processed to improve the accuracy of organ image identification. Optionally, the processing of the organ image may include the following: grayscale correction may be performed on the organ image to obtain a corrected image. Artifacts may be removed from the corrected image to obtain a standard image. The standard image may be segmented to obtain an organ image to be identified. Further, the processed organ image may be identified using an image recognition model to obtain a recognition result.

[0080] For example, the organ image is a chest X-ray image, and processing the chest X-ray image may include: performing grayscale unevenness correction processing, image denoising processing, and artifact removal processing on the chest X-ray image to obtain a standard chest X-ray image; performing lung parenchyma segmentation, lung nodule segmentation, and lung nodule false positive target removal on the standard chest X-ray image to obtain a processed chest X-ray image.

[0081] In this embodiment, the first image processing unit in the image processing module can be used to perform grayscale non-uniformity correction, image denoising and artifact removal on the chest X-ray image to obtain a standard chest X-ray image (i.e., a standard image). Further, the second image processing unit is used to perform lung parenchyma segmentation, lung nodule segmentation and lung nodule false positive target removal on the standard chest X-ray image to obtain a processed chest X-ray image; wherein the first image processing unit and the second image processing unit are connected.

[0082] Optionally, this embodiment performs grayscale non-uniformity correction, image denoising and artifact removal on the chest X-ray image to obtain a standard chest X-ray image and performs lung parenchyma segmentation, lung nodule segmentation and false positive target removal of lung nodules on the standard chest X-ray image, thereby effectively improving the accuracy of pneumoconiosis diagnosis.

[0083] It should be noted that, in this embodiment, only the process of identifying pneumoconiosis is used as an example to illustrate, but the organ to be diagnosed may be the lung, heart and other organs. Therefore, as long as the organ image of the organ to be diagnosed in the target biological object is obtained; the organ image is identified to obtain a recognition result; in response to the recognition result that the organ image is a target image, based on the organ image, a density area in an abnormal state on the organ to be diagnosed, and the shape of the density area, wherein the organ in the target image is a diseased organ; the method of determining the pathological result of the organ to be diagnosed based on the density area and shape should be within the scope of protection of the present application and no specific restrictions are made here.

[0084] At present, pneumoconiosis is a general term for a group of occupational lung diseases characterized by diffuse fibrosis of lung tissue caused by long-term inhalation of different pathogenic industrial dusts and their retention in the lungs during professional activities. There are many types of dusts that cause pneumoconiosis, including silicosis, coal worker pneumoconiosis, graphite pneumoconiosis, carbon black pneumoconiosis, asbestosis, talc pneumoconiosis, cement pneumoconiosis, mica pneumoconiosis, potter's pneumoconiosis, aluminum pneumoconiosis, welder's pneumoconiosis, foundry worker's pneumoconiosis, etc.

[0085] Exposure time is the most important factor in the occurrence and development of pneumoconiosis. In addition, the nature of dust, dust concentration, and dust particle size are closely related to the progression of pneumoconiosis. There are no specific symptoms in the early stages of pneumoconiosis, but as the disease progresses, clinical symptoms such as coughing, sputum, hemoptysis, chest pain, and dyspnea will appear, and there may be systemic symptoms of varying degrees. As the disease progresses, serious complications of all body systems will occur. Therefore, early detection, diagnosis, and treatment of pneumoconiosis are of great significance to the development of the disease.

[0086] At present, clinical identification of pulmonary fibrosis lesions and assessment of their severity mainly rely on clinicians to visually judge and roughly evaluate chest X-ray images, and the accuracy of the assessment depends entirely on the doctor's personal experience. In addition, the explosive growth of chest X-ray image data has greatly increased the workload of doctors, which is prone to misdiagnosis and missed diagnosis during disease detection. Therefore, the above method has the technical problem of low image recognition efficiency.

[0087] To solve the above problems, in this embodiment, the present application discloses a method and system for analyzing pneumoconiosis images, which may include: obtaining a chest X-ray image (that is, an organ image) of an object to be analyzed (that is, a target biological object), and processing the chest X-ray image; constructing a pneumoconiosis image analysis model (that is, an image recognition model), inputting the processed chest X-ray image into the pneumoconiosis image analysis model, and determining whether the chest X-ray image is a pneumoconiosis image (that is, a target image); when the chest X-ray image is a pneumoconiosis image, the chest X-ray image may be graded and a classification result (that is, a lesion result) may be obtained, and based on the classification result, the degree of lesion of the organ to be diagnosed may be accurately determined.

[0088] In this embodiment, by constructing a pneumoconiosis image analysis model to determine whether a chest X-ray image is a pneumoconiosis image, it is possible to automatically detect whether the chest X-ray image is a pneumoconiosis image without increasing the workload of doctors, thereby avoiding misdiagnosis and missed diagnosis. By constructing a shadow density judgment standard, the pneumoconiosis fiber shadow pixel density and shadow shape are graded according to the shadow density judgment standard, and the pneumoconiosis grade can be judged quickly, accurately and efficiently, providing a strong basis for subsequent treatment, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency.

[0089] According to an embodiment of the present application, a device for detecting organ lesions is also provided. It should be noted that the device for detecting organ lesions of this embodiment can be used to execute the method for detecting organ lesions of embodiment 1 of the present application.

[0090] Figure 3 FIG. 1 is a schematic diagram of an organ lesion detection device according to an embodiment of the present application. Figure 3As shown, the organ lesion detection device 30 may include: an acquisition unit 302 , an identification unit 304 , a first determination unit 306 and a second determination unit 308 .

[0091] The acquisition unit 302 is used to acquire an organ image of the organ to be diagnosed in the target biological object.

[0092] The recognition unit 304 is used to recognize the organ image and obtain a recognition result.

[0093] The first determination unit 306 is used to determine, in response to the recognition result that the organ image is a target image, a density area in an abnormal state on the organ to be diagnosed and a shape of the density area based on the organ image, wherein the organ in the target image is a diseased organ.

[0094] The second determining unit 308 is used to determine the lesion result of the organ to be diagnosed based on the density area and the shape, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0095] The organ lesion detection device of this embodiment acquires the organ image of the organ to be diagnosed in the target biological object through the acquisition unit; recognizes the organ image through the recognition unit to obtain a recognition result; determines the density area in an abnormal state on the organ to be diagnosed and the shape of the density area based on the organ image in response to the recognition result that the organ image is a target image, wherein the organ in the target image is a lesion organ; determines the lesion result of the organ to be diagnosed based on the density area and the shape through the second determination unit, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency.

[0096] The embodiment of the present application may provide a computer terminal, which may be any computer terminal device in a computer terminal group. Optionally, in this embodiment, the computer terminal may also be replaced by a terminal device such as a mobile terminal.

[0097] Optionally, in this embodiment, the computer terminal may be located in at least one network device among a plurality of network devices of the computer network.

[0098] In this embodiment, the above-mentioned computer terminal can execute the program code of the following steps in the organ lesion detection method: obtaining an organ image of the organ to be diagnosed in the target biological object; identifying the organ image to obtain a recognition result; in response to the recognition result that the organ image is a target image, based on the organ image, determining the density area in an abnormal state on the organ to be diagnosed, and the shape of the density area, wherein the organ in the target image is a diseased organ; based on the density area and the shape, determining the lesion result of the organ to be diagnosed, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0099] Optionally, Figure 4 is a structural block diagram of a computer terminal according to an embodiment of the present application, such as Figure 4 As shown, the computer terminal 408 may include: one or more (only one is shown in the figure) processors 402 , a memory 404 and a transmission device 406 .

[0100] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the organ lesion detection method and device in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned organ lesion detection method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, the memory may further include a memory remotely arranged relative to the processor, and these remote memories can be connected to the computer terminal 408 via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0101] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain an organ image of the organ to be diagnosed in the target biological object; identify the organ image to obtain an identification result; in response to the identification result that the organ image is a target image, determine the density area in an abnormal state on the organ to be diagnosed and the shape of the density area based on the organ image, wherein the organ in the target image is a diseased organ; based on the density area and the shape, determine the pathological result of the organ to be diagnosed, wherein the pathological result is used to indicate the degree of the pathological condition of the organ to be diagnosed.

[0102] It can be understood by those skilled in the art that Figure 4 The structure shown is for illustration only, and the computer terminal 408 may also be a smart phone (such as an Android phone, an iOS phone, etc.), a tablet computer, a PDA, a mobile Internet device (MID for short), a PAD, or other terminal devices. Figure 4The structure of the computer terminal 408 is not limited. For example, the computer terminal 408 may also include Figure 4 More or fewer components (such as network interfaces, display devices, etc.) shown in, or having Figure 4 Different configurations are shown.

[0103] A person of ordinary skill in the art can understand that all or part of the steps in the various methods of the above embodiments can be completed by instructing the hardware related to the terminal device through a program, and the program can be stored in a computer-readable storage medium, and the storage medium may include: a flash drive, a read-only memory (ROM), a random access memory (RAM), a disk or an optical disk, etc.

[0104] According to an embodiment of the present application, a computer-readable storage medium is also provided, which includes a stored program, wherein the program executes the organ lesion detection method in the above embodiment.

[0105] Optionally, in this embodiment, the computer-readable storage medium may be located in any one of the computer terminals in a computer terminal group in a computer network, or in any one of the mobile terminals in a mobile terminal group.

[0106] Optionally, in this embodiment, the computer-readable storage medium is configured to store program codes for executing the following steps: acquiring an organ image of an organ to be diagnosed in a target biological object; identifying the organ image to obtain an identification result; in response to the identification result that the organ image is a target image, determining, based on the organ image, a density area in an abnormal state on the organ to be diagnosed, and a shape of the density area, wherein the organ in the target image is a diseased organ; based on the density area and the shape, determining a lesion result of the organ to be diagnosed, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0107] Optionally, the computer-readable storage medium may also execute program code of the following steps: calling an image recognition model to recognize an organ image to obtain a recognition result, wherein the image recognition model is trained based on historical organ images of multiple sizes.

[0108] Optionally, the computer-readable storage medium can also execute the program code of the following steps: inputting the historical organ image into the initial image recognition model for feature extraction to obtain shadow pixel features; calibrating the lesion area of ​​the diseased organ in the historical organ image; and adjusting the model parameters of the initial image recognition model using the shadow pixel features and the calibrated historical organ image to obtain an image recognition model.

[0109] Optionally, the computer-readable storage medium can also execute the program code of the following steps: constructing a multidimensional training loss function, wherein the multidimensional training loss function is used to adjust the shadow pixel features output by the initial image recognition model; using the multidimensional training loss function, shadow pixel features and calibrated historical organ images, adjusting the model parameters of the initial image recognition model to obtain an image recognition model.

[0110] Optionally, the computer-readable storage medium may also execute program codes for the following steps: classifying organ images based on density regions and shapes to obtain classification results; and determining the classification results as pathological results of the organ to be diagnosed.

[0111] Optionally, the computer-readable storage medium may also execute program code for the following steps: obtaining a lesion density area corresponding to the lesion organ, and a lesion shape corresponding to the lesion organ; and classifying the density area and the shape based on the lesion density area and the lesion shape to obtain a classification result.

[0112] Optionally, the computer-readable storage medium may also execute program codes for the following steps: performing grayscale correction on the organ image to obtain a corrected image; performing artifact removal on the corrected image to obtain a standard image; and performing segmentation processing on the standard image to obtain an organ image.

[0113] In this embodiment, an image of an organ to be diagnosed is obtained and the organ image is identified. If the organ image is a target image with a lesion, the density area in an abnormal state (that is, an abnormal density area) and the shape of the abnormal density area on the organ to be diagnosed can be determined based on the organ image. By identifying the abnormal density area and shape, the lesion result of the organ to be diagnosed can be determined, and the lesion result can be used to determine the degree of lesions in the organ to be diagnosed, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency.

[0114] According to an embodiment of the present application, a processor is also provided, which is used to run a program, wherein when the program is run by the processor, the organ lesion detection method in the above embodiment is executed.

[0115] Optionally, in this embodiment, the computer terminal may be located in at least one network device among a plurality of network devices of the computer network.

[0116] In this embodiment, the above-mentioned computer terminal can execute the program code of the following steps in the organ lesion detection method: obtaining an organ image of the organ to be diagnosed in the target biological object; identifying the organ image to obtain a recognition result; in response to the recognition result that the organ image is a target image, based on the organ image, determining the density area in an abnormal state on the organ to be diagnosed, and the shape of the density area, wherein the organ in the target image is a diseased organ; based on the density area and the shape, determining the lesion result of the organ to be diagnosed, wherein the lesion result is used to indicate the degree of lesion in the organ to be diagnosed.

[0117] Among them, the memory can be used to store software programs and modules, such as the program instructions / modules corresponding to the organ lesion detection method and device in the embodiment of the present application. The processor executes various functional applications and data processing by running the software programs and modules stored in the memory, that is, realizing the above-mentioned organ lesion detection method. The memory may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory may further include a memory remotely arranged relative to the processor, and these remote memories may be connected to the computer terminal via a network. Examples of the above-mentioned network include, but are not limited to, the Internet, an intranet, a local area network, a mobile communication network, and combinations thereof.

[0118] The processor can call the information and application stored in the memory through the transmission device to perform the following steps: obtain an organ image of the organ to be diagnosed in the target biological object; identify the organ image to obtain an identification result; in response to the identification result that the organ image is a target image, determine the density area in an abnormal state on the organ to be diagnosed and the shape of the density area based on the organ image, wherein the organ in the target image is a diseased organ; based on the density area and the shape, determine the pathological result of the organ to be diagnosed, wherein the pathological result is used to indicate the degree of the pathological condition of the organ to be diagnosed.

[0119] Optionally, the processor may also execute the program code of the following steps: calling an image recognition model to recognize the organ image to obtain a recognition result, wherein the image recognition model is trained based on historical organ images of multiple sizes.

[0120] Optionally, the processor may also execute the program code of the following steps: inputting the historical organ image into the initial image recognition model for feature extraction to obtain shadow pixel features; calibrating the lesion area of ​​the diseased organ in the historical organ image; and adjusting the model parameters of the initial image recognition model using the shadow pixel features and the calibrated historical organ image to obtain an image recognition model.

[0121] Optionally, the processor may also execute the program code of the following steps: constructing a multidimensional training loss function, wherein the multidimensional training loss function is used to adjust the shadow pixel features output by the initial image recognition model; adjusting the model parameters of the initial image recognition model using the multidimensional training loss function, shadow pixel features and calibrated historical organ images to obtain an image recognition model.

[0122] Optionally, the processor may also execute program codes of the following steps: classifying the organ image based on the density area and shape to obtain a classification result; and determining the classification result as a pathological result of the organ to be diagnosed.

[0123] Optionally, the processor may also execute program codes of the following steps: obtaining a lesion density area corresponding to the lesion organ, and a lesion shape corresponding to the lesion organ; and classifying the density area and the shape based on the lesion density area and the lesion shape to obtain a classification result.

[0124] Optionally, the processor may also execute program codes of the following steps: grayscale correction of the organ image to obtain a corrected image; artifact removal of the corrected image to obtain a standard image; segmentation processing of the standard image to obtain an organ image.

[0125] By adopting the embodiment of the present application, an image of an organ to be diagnosed is obtained, and the organ image is identified. If the organ image is a target image with a lesion, the density area in an abnormal state (that is, an abnormal density area) and the shape of the abnormal density area on the organ to be diagnosed can be determined based on the organ image. By identifying the abnormal density area and shape, the lesion result of the organ to be diagnosed can be determined, and the lesion result can be used to determine the degree of lesions in the organ to be diagnosed, thereby solving the technical problem of low image recognition efficiency and achieving the technical effect of improving image recognition efficiency.

[0126] According to an embodiment of the present application, a computer program product is also provided. The computer program product includes computer instructions, wherein when the computer instructions are executed by a processor, the organ lesion detection method in the above embodiment is implemented.

[0127] An embodiment of the present application may provide an electronic device, which may include a memory and a processor.

[0128] Figure 5It is a block diagram of an electronic device according to a method for detecting lesions of an organ according to an embodiment of the present application. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workbenches, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices, and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present application described and / or required herein.

[0129] like Figure 5 As shown, the device 500 includes a computing unit 501, which can perform various appropriate actions and processes according to a computer program stored in a read-only memory (ROM) 502 or a computer program loaded from a storage unit 508 to a random access memory (RAM) 503. In the RAM 503, various programs and data required for the operation of the device 500 can also be stored. The computing unit 501, the ROM 502, and the RAM 503 are connected to each other via a bus 504. An input / output (I / O) interface 505 is also connected to the bus 504.

[0130] A number of components in the device 500 are connected to the I / O interface 505, including: an input unit 506, such as a keyboard, a mouse, etc.; an output unit 507, such as various types of displays, speakers, etc.; a storage unit 508, such as a disk, an optical disk, etc.; and a communication unit 509, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 509 allows the device 500 to exchange information / data with other devices through a computer network such as the Internet and / or various telecommunication networks.

[0131] The computing unit 501 may be a variety of general and / or special processing components with processing and computing capabilities. Some examples of the computing unit 501 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various dedicated artificial intelligence (AI) computing chips, various computing units running machine learning model algorithms, a digital signal processor (DSP), and any appropriate processor, controller, microcontroller, etc. The computing unit 501 performs the various methods and processes described above, such as a verification method for data. For example, in some embodiments, the verification method for data may be implemented as a computer software program, which is tangibly included in a machine-readable medium, such as a storage unit 508. In some embodiments, part or all of the computer program may be loaded and / or installed on the device 500 via the ROM 502 and / or the communication unit 509. When the computer program is loaded into the RAM 503 and executed by the computing unit 501, one or more steps of the verification method for data described above may be executed. Alternatively, in other embodiments, the computing unit 501 may be configured to execute the data verification method in any other appropriate manner (for example, by means of firmware).

[0132] According to an embodiment of the present application, a method for detecting lesions of an organ is provided. It should be noted that the steps shown in the flowchart of the accompanying drawings can be executed in a computer system such as a set of computer executable instructions, and although a logical order is shown in the flowchart, in some cases, the steps shown or described can be executed in an order different from that shown here.

[0133] Various embodiments of the systems and techniques described above herein may be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), systems on a chip (SOCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include: being implemented in one or more computer programs that are executable and / or interpreted on a programmable system including at least one programmable processor that may be a dedicated or general purpose programmable processor that may receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0134] The program code for implementing the method of the present application can be written in any combination of one or more programming languages. These program codes can be provided to a processor or controller of a general-purpose computer, a special-purpose computer or other programmable data processing device, so that the program code, when executed by the processor or controller, enables the functions / operations specified in the flow chart and / or block diagram to be implemented. The program code can be executed entirely on the machine, partially on the machine, partially on the machine and partially on a remote machine as a stand-alone software package, or entirely on a remote machine or server.

[0135] In the context of the present application, a machine-readable medium may be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, device, or equipment. A machine-readable medium may be a machine-readable signal medium or a machine-readable storage medium. A machine-readable medium may include, but is not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, device, or equipment, or any suitable combination of the foregoing. A more specific example of a machine-readable storage medium may include an electrical connection based on one or more lines, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

[0136] To provide interaction with a user, the systems and techniques described herein may be implemented on a computer having: a display device (e.g., a CRT (cathode ray tube) or an LCD (liquid crystal display, monitor) for displaying information to the user; and a keyboard and a pointing device (e.g., a mouse or a path ball) through which the user can provide input to the computer. Other types of devices may also be used to provide interaction with the user; for example, the feedback provided to the user may be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user may be received in any form (including acoustic input, voice input, or tactile input).

[0137] The systems and techniques described herein may be implemented in a computing system that includes back-end components (e.g., as a data server), or a computing system that includes middleware components (e.g., an application server), or a computing system that includes front-end components (e.g., a user computer with a graphical user interface or a web browser through which a user can interact with implementations of the systems and techniques described herein), or a computing system that includes any combination of such back-end components, middleware components, or front-end components. The components of the system may be interconnected by any form or medium of digital data communication (e.g., a communication network). Examples of communication networks include: a Local Area Network (LAN), a Wide Area Network (WAN), and the Internet.

[0138] A computer system may include a client and a server. The client and the server are generally remote from each other and usually interact through a communication network. The relationship of client and server is generated by computer programs running on respective computers and having a client-server relationship with each other. The server may be a cloud server, a server of a distributed system, or a server combined with a blockchain.

[0139] The serial numbers of the above-mentioned embodiments of the present application are for description only and do not represent the advantages or disadvantages of the embodiments.

[0140] In the above embodiments of the present application, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, please refer to the relevant description of other embodiments.

[0141] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. Among them, the device embodiments described above are only schematic. For example, the division of units can be a logical function division. There may be other division methods in actual implementation. For example, 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 units or modules, which can be electrical or other forms.

[0142] 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 over multiple units. Some or all of the units may be selected according to actual needs to achieve the purpose of the present embodiment.

[0143] 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. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0144] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art or all or part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a personal computer, server or network device, etc.) to perform all or part of the steps of each embodiment method of the present application. The aforementioned storage medium includes: U disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), mobile hard disk, disk or optical disk, etc. Various media that can store program codes.

[0145] The above are only preferred implementations of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for detecting organ lesions, characterized in that: include: Acquiring an organ image of an organ to be diagnosed in a target biological object; Recognizing the organ image to obtain a recognition result; In response to the recognition result that the organ image is a target image, based on the organ image, determining a density region in an abnormal state on the organ to be diagnosed and a shape of the density region, wherein the organ in the target image is a diseased organ; Based on the density area and the shape, the lesion result of the organ to be diagnosed is determined, wherein the lesion result is used to indicate the extent of the lesion in the organ to be diagnosed.

2. The method according to claim 1, characterized in that The step of identifying the organ image to obtain an identification result includes: An image recognition model is called to recognize the organ image to obtain the recognition result, wherein the image recognition model is trained based on historical organ images of multiple sizes.

3. The method according to claim 2, characterized in that The method further comprises: Inputting the historical organ image into an initial image recognition model to extract features and obtain shadow pixel features; calibrating the lesion area of ​​the lesion organ in the historical organ image; The model parameters of the initial image recognition model are adjusted using the shadow pixel features and the calibrated historical organ image to obtain the image recognition model.

4. The method according to claim 3, characterized in that The method of adjusting the model parameters of the initial image recognition model by using the shadow pixel features and the calibrated historical organ image to obtain the image recognition model includes: Constructing a multidimensional training loss function, wherein the multidimensional training loss function is used to adjust the shadow pixel features output by the initial image recognition model; The multidimensional training loss function, the shadow pixel features and the calibrated historical organ image are used to adjust the model parameters of the initial image recognition model to obtain the image recognition model.

5. The method according to claim 1, characterized in that The step of determining the lesion result of the organ to be diagnosed based on the density area and the shape includes: Based on the density region and the shape, the organ image is classified to obtain a classification result; The classification result is determined as the pathological result of the organ to be diagnosed.

6. The method according to claim 5, characterized in that The step of classifying the organ image based on the density region and the shape to obtain a classification result includes: Acquire a lesion density area corresponding to the lesion organ and a lesion shape corresponding to the lesion organ; Based on the lesion density area and the lesion shape, the density area and the shape are classified to obtain the classification result.

7. The method according to any one of claims 1 to 6, characterized in that: Before identifying the organ image, the method further includes: Performing grayscale correction on the organ image to obtain a corrected image; Performing artifact removal on the corrected image to obtain a standard image; The standard image is segmented to obtain the organ image.

8. A device for detecting organ lesions, characterized in that: include: an acquisition unit, used for acquiring an organ image of an organ to be diagnosed in a target biological object; A recognition unit, used to recognize the organ image and obtain a recognition result; A first determination unit is configured to determine, in response to the recognition result that the organ image is a target image, a density region in an abnormal state on the organ to be diagnosed and a shape of the density region based on the organ image, wherein the organ in the target image is a diseased organ; The second determination unit is used to determine the lesion result of the organ to be diagnosed based on the density area and the shape, wherein the lesion result is used to indicate the extent of the lesion in the organ to be diagnosed.

9. A processor, characterized in that: The processor is used to run a program, wherein the program executes the method according to any one of claims 1 to 7 when run by the processor.

10. A computer program product, characterized in that The method comprises computer instructions, which, when executed by a processor, implement the method according to any one of claims 1 to 7.

Images