Method and device for acquiring degree of artifact in medical image, computer device and medium

Abstract

Embodiments of the present application provide a method and device for obtaining the degree of artifacts in a medical image, a computer device and a medium, and relate to the technical field of images. In the method, the computer device obtains a medical image of a target part, wherein the medical image comprises a plurality of continuous slice images; obtains an artifact intensity value of artifacts in each slice image, and obtains the degree of artifacts in the medical image based on the artifact intensity values of the artifacts in the plurality of slice images. Thus, the severity of the artifacts can be directly quantified. Meanwhile, the computer device can comprehensively evaluate the degree of artifacts in the medical image based on the artifact intensity values of the artifacts in the plurality of slice images, thereby ensuring more comprehensive evaluation of the degree of artifacts in the medical image and improving the accuracy of the obtained degree of artifacts.

Description

Technical Field

[0001] This application belongs to the field of image technology, specifically relating to a method, apparatus, computer equipment, and medium for obtaining the degree of artifacts in medical images. Background Technology

[0002] With the widespread application of medical imaging technology, medical images such as computed tomography (CT) and magnetic resonance imaging (MRI) play an important role in clinical diagnosis. However, these medical images are often affected by various artifacts, such as metal artifacts and motion artifacts. These artifacts may obscure or distort important pathological information, resulting in lower image quality and consequently misdiagnosis or missed diagnosis.

[0003] In related technologies, computer equipment can only detect whether there are artifacts in medical images, but there is little analysis of the degree of artifacts in medical images. Summary of the Invention

[0004] This application provides a method, apparatus, computer device, and medium for obtaining the degree of artifacts in medical images, in order to solve the problem that there is little analysis of the degree of artifacts in medical images in the prior art.

[0005] The first aspect of this application provides a method for obtaining the degree of artifacts in medical images, the method comprising: Acquire medical images of the target area, which include multiple consecutive slice images; Obtain the artifact intensity value of the artifact in each slice image; The degree of artifacts in a medical image is obtained by measuring the artifact intensity values ​​of artifacts in multiple slice images.

[0006] Optionally, the degree of artifacts in a medical image can be obtained based on the artifact intensity values ​​of artifacts in multiple slice images, including: For each slice image, the artifact intensity value of the artifact in the slice image is corrected to obtain the artifact intensity correction value of the artifact in the slice image. The degree of artifacts in a medical image is obtained by using artifact intensity correction values ​​from multiple slice images.

[0007] Optionally, the artifact intensity values ​​of the artifacts in the slice image are corrected to obtain artifact intensity correction values ​​for the artifacts in the slice image, including: Obtain the attenuation weight of the slice image, and the attenuation weight of at least one target slice image adjacent to the slice image; Based on the attenuation weight and artifact intensity value of the slice image, and the attenuation weight and artifact intensity value of at least one target slice image, the artifact intensity correction value of the artifact in the slice image is obtained.

[0008] Optionally, the attenuation weight of the sliced ​​image is a preset value; When there is at least one target slice image, the attenuation weight of the target slice image is determined based on the corresponding layer difference; When there are multiple target slice images, the attenuation weights of the multiple target slice images are attenuated sequentially in order of increasing difference in the corresponding layer number. The difference in the number of layers corresponding to each target slice image is the number of images between the target slice image and the slice image.

[0009] Optionally, the attenuation weight for each target slice image is an exponential function with a natural constant as the base and a linear attenuation factor as the exponent, wherein the linear attenuation factor is determined based on the negative of the first target product, and the first target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0010] Optionally, the attenuation weight for each target slice image is an exponential function with a natural constant as the base and a quadratic attenuation factor as the exponent, wherein the quadratic attenuation factor is determined based on the negative of the second target product, and the second target product is determined based on the product of the attenuation coefficient and the square of the corresponding layer difference.

[0011] Optionally, the attenuation weight for each target slice image is determined based on the reciprocal of the target sum value, which is determined based on the sum of the third target product and the target parameter, and the third target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0012] Optionally, the decay rate of the decay weight is positively correlated with the decay coefficient.

[0013] Optionally, based on the attenuation weights and artifact intensity values ​​of the slice image, and the attenuation weights and artifact intensity values ​​of at least one target slice image, an artifact intensity correction value for the artifact in the slice image is obtained, including: Based on the attenuation weight of the slice image and the attenuation weight of at least one target slice image, the artifact intensity value of the artifact in the slice image and the artifact intensity value of the artifact in at least one target slice image are weighted to obtain the artifact intensity correction value of the artifact in the slice image.

[0014] Optionally, the degree of artifacts in a medical image is obtained based on artifact intensity correction values ​​from multiple slice images, including: Obtain the target proportion of artifacts in each slice image; For each slice image, the artifact intensity correction value and the target proportion in the slice image are weighted to obtain the artifact degree; The artifact level of a medical image is obtained by analyzing the artifact levels of multiple slice images.

[0015] Optionally, the method also includes: The level of artifact severity is determined based on the range of severity levels within which the artifacts are located.

[0016] A second aspect of this application provides an apparatus for obtaining the degree of artifacts in medical images, the apparatus comprising: The acquisition module is used to acquire medical images of the target area, which include multiple consecutive slice images; The artifact intensity recognition module is used to obtain the artifact intensity value of artifacts in each slice image; The evaluation module is used to obtain the degree of artifacts in medical images based on the artifact intensity values ​​of artifacts in multiple slice images.

[0017] Optionally, the evaluation module is used to correct the artifact intensity value of the artifact in each slice image to obtain the artifact intensity correction value in the slice image; based on the artifact intensity correction values ​​of the artifact in multiple slice images, the degree of artifact in the medical image is obtained.

[0018] Optionally, the evaluation module is used to obtain the attenuation weight of the slice image and the attenuation weight of at least one target slice image adjacent to the slice image; based on the attenuation weight of the slice image and the artifact intensity value of the artifact, and the attenuation weight of at least one target slice image and the artifact intensity value of the artifact, the artifact intensity correction value of the artifact in the slice image is obtained.

[0019] Optionally, the attenuation weight of the slice image is a preset value; when there is at least one target slice image, the attenuation weight of the target slice image is determined based on the corresponding layer difference; when there are multiple target slice images, the attenuation weights of the multiple target slice images are attenuated sequentially in ascending order of the corresponding multiple layer differences; wherein, the layer difference corresponding to each target slice image is the number of images between the target slice image and the slice image.

[0020] Optionally, the attenuation weight for each target slice image is an exponential function with a natural constant as the base and a linear attenuation factor as the exponent, wherein the linear attenuation factor is determined based on the negative of the first target product, and the first target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0021] Optionally, the attenuation weight for each target slice image is an exponential function with a natural constant as the base and a quadratic attenuation factor as the exponent, wherein the quadratic attenuation factor is determined based on the negative of the second target product, and the second target product is determined based on the product of the attenuation coefficient and the square of the corresponding layer difference.

[0022] Optionally, the attenuation weight for each target slice image is determined based on the reciprocal of the target sum value, which is determined based on the sum of the third target product and the target parameter, and the third target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0023] Optionally, the decay rate of the decay weight is positively correlated with the decay coefficient.

[0024] Optionally, the evaluation module is used to perform weighted processing on the artifact intensity values ​​of artifacts in the slice image and the artifact intensity values ​​of artifacts in at least one target slice image based on the attenuation weight of the slice image and the attenuation weight of at least one target slice image, to obtain the artifact intensity correction value of artifacts in the slice image.

[0025] Optionally, the evaluation module is used to obtain the target proportion of artifacts in each slice image; for each slice image, the artifact intensity correction value and the target proportion of artifacts in the slice image are weighted to obtain the artifact degree; based on the artifact degree of multiple slice images, the artifact degree of the medical image is obtained.

[0026] Optionally, the evaluation module is also used to obtain the level of artifact severity based on the severity range of the artifact severity.

[0027] A third aspect of this application provides a computer device, including: a processor, the processor being connected to a memory, the memory storing a computer program or instructions executable on the processor, the computer program or instructions being executed by the processor to implement a method for obtaining the degree of artifacts in a medical image as described in any of the first aspects.

[0028] A fourth aspect of this application provides a computer-readable storage medium storing a computer program or instructions, which, when executed by a processor, implements a method for obtaining the degree of artifacts in a medical image as described in any of the first aspects.

[0029] The fifth aspect of this application provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement a method for obtaining the degree of artifacts in a medical image as described in any of the first aspects.

[0030] The method, apparatus, computer device, and medium for obtaining the degree of artifacts in medical images provided in this application embodiment, after acquiring a medical image of a target area using a computer device (the medical image includes multiple consecutive slice images), acquires the artifact intensity value of artifacts in each slice image, and obtains the degree of artifacts in the medical image based on the artifact intensity values ​​of artifacts in multiple slice images. This allows for direct quantification of the severity of artifacts. Furthermore, since the computer device can comprehensively evaluate the degree of artifacts in the medical image based on the artifact intensity values ​​of artifacts in multiple slice images, a more comprehensive evaluation of the degree of artifacts in the medical image can be ensured, improving the accuracy of the obtained artifact degree. Attached Figure Description

[0031] Figure 1 A flowchart illustrating a method for obtaining the degree of artifacts in a medical image, provided in an embodiment of this application; Figure 2 A flowchart illustrating another method for obtaining the degree of artifacts in a medical image, provided in an embodiment of this application; Figure 3 A block diagram of a device for acquiring the degree of artifacts in a medical image, provided in an embodiment of this application; Figure 4 This is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0032] The technical solutions of the embodiments of this application will be clearly described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application are within the scope of protection of this application.

[0033] The terms "first," "second," etc., used in this application are used to distinguish similar objects and not to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of this application can be implemented in orders other than those illustrated or described herein, and the objects distinguished by "first" and "second" are generally of the same class, without limiting the number of objects; for example, the first object can be one or more. Furthermore, "or" in this application indicates at least one of the connected objects. For example, "A or B" covers three scenarios: Scenario 1: including A but not B; Scenario 2: including B but not A; Scenario 3: including both A and B. The character " / " generally indicates that the preceding and following objects are in an "or" relationship.

[0034] The term "instruction" in this application can be either a direct instruction (or explicit instruction) or an indirect instruction (or implicit instruction). A direct instruction can be understood as one in which the sender explicitly informs the receiver of specific information, the operation to be performed, or the requested result, etc.; an indirect instruction can be understood as one in which the receiver determines the corresponding information based on the instruction sent by the sender, or makes a judgment and determines the operation to be performed or the requested result, etc., based on the judgment result.

[0035] With the widespread application of medical imaging technology, CT and MRI images have become indispensable key evidence for clinical diagnosis. However, the quality of acquired medical images is often low due to factors such as equipment performance, scanning parameter settings, and individual differences between patients and physicians. For example, common factors affecting image quality in medical images include, but are not limited to, inaccurate scanning range, artifact interference, noise interference, and non-standard patient positioning. Inaccurate scanning range may lead to missed lesions or incomplete display of anatomical structures in medical images, directly affecting the comprehensiveness and accuracy of diagnosis. Artifacts (such as motion artifacts and metal artifacts) can distort tissue morphology in medical images, increasing the risk of misdiagnosis or missed diagnosis. Noise interference reduces the signal-to-noise ratio of medical images, making it difficult to identify small lesions or low-contrast structures. Non-standard patient positioning may cause occlusion between tissue structures in medical images, thus affecting the physician's judgment of structural morphology. These problems may not only delay patient treatment and increase the risk of radiation exposure from repeated examinations, but also lead to unnecessary consumption of medical resources, including prolonged equipment occupancy and increased examination costs.

[0036] In related technologies, computer equipment can only detect whether there are artifacts in medical images, but the degree of artifacts in medical images is rarely analyzed.

[0037] This application proposes a method for obtaining the degree of artifacts in medical images. In this method, after acquiring a medical image of a target area (including multiple consecutive slice images), a computer device obtains the artifact intensity value of each slice image and, based on these values, determines the degree of artifacts in the medical image. This allows for direct quantification of the severity of artifacts. Furthermore, since the computer device can comprehensively evaluate the degree of artifacts in the medical image based on the artifact intensity values ​​across multiple slice images, a more comprehensive assessment of the artifact degree is ensured, improving the accuracy of the obtained artifact degree.

[0038] The technical solution of the method for obtaining the degree of artifacts in medical images according to this application is described below using several specific embodiments as examples: Figure 1A flowchart illustrating a method for obtaining the degree of artifacts in medical images, provided in this application embodiment, is available. This method can be applied to computer devices, such as... Figure 1 As shown, the flow of the method in this embodiment is as follows: S11: Acquire medical images of the target area, which include multiple consecutive slice images.

[0039] Optionally, the medical image of the target area can be extracted from the original medical image, which can be three-dimensional, and correspondingly, each slice image is two-dimensional.

[0040] Medical images include, but are not limited to, CT images, MRI images, and ultrasound images. The target site refers to the part of the original medical image to be diagnosed and analyzed. For example, if the original medical image includes the lung parenchyma, trachea, cardiovascular system, chest wall, or muscles, and the lung parenchyma is to be diagnosed and analyzed, the target site is the lung parenchyma. Another example is that if the original medical image includes the liver, pancreas, gastrointestinal tract, abdominal cavity, and bladder, and the gastrointestinal tract is to be diagnosed and analyzed, the target site is the gastrointestinal tract.

[0041] As an alternative implementation, the computer device can acquire the original medical image from the scanning device, segment the original medical image along one direction to obtain multiple consecutive original slice images, and determine the slice image of the target region based on the original slice images containing the target region among the multiple original slice images. For example, if the medical image includes CT images, the scanning device can be a CT device.

[0042] Computer equipment can use an original slice image containing the target region as the slice image of the target region, meaning the resolution of the slice image of the target region can be the same as the resolution of the corresponding original slice image. Alternatively, it can extract a sub-image containing the target region from the original slice image containing the target region and use this sub-image as the slice image of the target region, where the resolution of the sub-image is lower than the resolution of the corresponding original slice image.

[0043] As another alternative implementation, the computer device can input the identifier of the target area and the original medical image into the area recognition model to obtain the recognition result output by the area recognition model. If the original medical image includes the target area, the recognition result can be a target medical image containing the target area. Then, the target medical image is divided along one direction to obtain multiple continuous slice images.

[0044] The target medical image is a three-dimensional image. The site recognition model can be an image classification model, a coordinate prediction model, or a site segmentation model. The image classification model (or coordinate prediction model) can output a bounding box image of the target site. If the site recognition model is a site segmentation model, the site segmentation model can output a three-dimensional segmented image of the target site.

[0045] In this embodiment, the image classification model can be a convolutional neural network model such as ResNet-50. The computer device can acquire multiple first sample data and train the image classification model on the multiple first sample data. Each first sample data may include a sample first original medical image, a sample site identifier, and a sample target medical image of the sample site.

[0046] Each target medical image is a bounding box image containing the sample site. The computer device can segment the first original medical image of the sample from three directions, obtaining multiple original sample slice images in each direction. From these multiple original sample slice images in each direction, multiple target sample original slice images containing the sample site are obtained, and then a target medical image is generated based on these multiple target sample original slice images in the three directions. Each target sample original slice image is a bounding box image containing the sample site.

[0047] In this embodiment of the application, the coordinate prediction model can be a 3D convolutional neural network model such as ResNet. The computer device can acquire multiple second sample data and train the coordinate prediction model on the multiple second sample data. Each second sample data can include a sample second original medical image, a sample part identifier, and a sample target medical image of the sample part.

[0048] Specifically, for each sample second original medical image, the computer device can annotate the bounding box of the sample region in the sample second original medical image, thereby obtaining a sample target medical image of the sample region corresponding to the sample second original medical image. For example, the bounding box of the sample region can be the area enclosed by three sets of diagonal vertices of a cuboid.

[0049] In this embodiment of the application, the part segmentation model can be a convolutional neural network model such as U-Net. The computer device can acquire multiple third sample data and train the part segmentation model on the multiple third sample data. Each third sample data can include the original third medical image of the sample, the sample part identifier, and the sample target medical image of the sample part.

[0050] Specifically, for each sample third original medical image, the computer device can label the pixel values ​​of the region where the sample site is located in the sample third original medical image as a first value, and label the pixel values ​​of other tissue regions besides the region where the sample site is located as a second value, thereby obtaining a sample target medical image of the sample site corresponding to the sample third original medical image. For example, the first value can be 1, and the second value can be 0.

[0051] S12: Obtain the artifact intensity value of each slice image.

[0052] Artifacts present in slice images include, but are not limited to, motion artifacts and metal artifacts. During medical image scanning, if the patient carries high-density metallic materials, such as dental fillings, artificial joints, nails, or vascular clamps, these materials will absorb more X-rays, resulting in bright radial streaks in the medical image—a metal artifact. If the patient exhibits voluntary movements (such as swallowing and moving) and involuntary movements (such as heartbeat, breathing, and intestinal peristalsis), it will cause striped ghosting or blurring in the medical image—a motion artifact.

[0053] Artifact intensity refers to the degree to which an artifact causes a change in the pixel value of an image, and can be used to characterize the brightness of the artifact. A higher artifact intensity value indicates a brighter artifact, and correspondingly, a more severe distortion of the tissue structure near the artifact, which has a greater impact on the doctor's observation and diagnosis of medical images. Conversely, a lower artifact intensity value indicates a lower brightness artifact, and correspondingly, a less impact on the doctor's observation and diagnosis of medical images.

[0054] Computer equipment can input individual slice images into an artifact intensity prediction model to obtain the artifact intensity values ​​for each slice image. This artifact intensity prediction model can be a deep network model, such as a regression network model.

[0055] The artifact intensity prediction model can be trained on fourth sample data, each of which can include a sample slice image and sample artifact intensity values.

[0056] For example, sample slice images are manually scored to obtain corresponding sample artifact intensity values. For instance, a sample containing a large brightness artifact is labeled with an artifact intensity value of 1, a sample containing a small brightness artifact with an artifact intensity value of 0, and a sample containing a medium brightness artifact with an artifact intensity value of 0.5. Then, a computer device can input the slice image into an artifact intensity prediction model, which outputs a predicted artifact intensity value between 0 and 1. If the artifact intensity value is close to 1, it indicates the presence of a certain high-intensity artifact in the medical image; if the artifact intensity value is close to 0, it indicates the absence of that artifact in the medical image, or that the existing artifact is negligible.

[0057] S13: Obtain the degree of artifacts in a medical image based on the artifact intensity values ​​of artifacts in multiple slice images.

[0058] In this embodiment, a computer device acquires a medical image of a target area, which includes multiple consecutive slice images. The device then acquires the artifact intensity value of each slice image. Based on these artifact intensity values, the degree of artifacts in the medical image can be directly quantified. Furthermore, because the computer device comprehensively evaluates the degree of artifacts in the medical image based on the artifact intensity values ​​of multiple slice images, a more comprehensive assessment of the artifact degree is ensured, thus improving the accuracy of the acquired artifact level.

[0059] Figure 2 This is a flowchart illustrating another method for obtaining the degree of artifacts in medical images provided in an embodiment of this application. Figure 2 Is Figure 1 Based on the illustrated embodiment, a further description of a possible implementation of S13 is provided, in which the method can be applied to computer devices, such as... Figure 2 As shown, the flow of the method in this embodiment is as follows: S131: For each slice image, the artifact intensity value of the artifact in the slice image is corrected to obtain the artifact intensity correction value of the artifact in the slice image.

[0060] Because artifacts in multiple slice images differ in morphology, intensity, coverage area, and distribution, determining the artifact intensity value of each slice image is challenging. This leads to discrepancies between the artifact intensity values ​​obtained in step S12 and the actual artifact intensity values ​​in those slice images, resulting in lower accuracy of the obtained artifact intensity values. Furthermore, due to the spatial continuity and content correlation of multiple consecutive slice images, artifacts are continuously distributed across them, and the actual artifact intensity values ​​in consecutive reference slice images are relatively close. Therefore, when using step S12 to determine the artifact intensity value of a slice image, if the artifact intensity value of one reference slice image is 0 while the artifact intensity values ​​of other reference slice images are all 1, it can be determined that the artifact intensity value of that particular reference slice image is inaccurate.

[0061] In this embodiment, for each slice image, the computer device can correct the artifact intensity value of the artifact in the slice image based on the artifact intensity value of at least one target slice image adjacent to the slice image. Since at least one target slice image is adjacent to the slice image, the actual artifact intensity value of the artifact in at least one target slice image is relatively close to the actual artifact intensity value of the slice image. Therefore, if the accuracy of the artifact intensity value of the slice image determined in step S12 is low, correcting the artifact intensity value of the slice image by using at least one target slice image adjacent to the slice image can correct the artifact intensity value of the slice image to a value close to the artifact intensity value of at least one target slice image. This can improve the accuracy of the artifact degree of the subsequently determined slice image, and thus improve the accuracy of the artifact degree of the subsequently acquired medical images.

[0062] Optionally, the process of correcting the artifact intensity values ​​of artifacts in the sliced ​​image using computer equipment to obtain corrected artifact intensity values ​​may include the following steps A1 and A2: Step A1: Obtain the attenuation weight of the slice image and the attenuation weight of at least one target slice image adjacent to the slice image.

[0063] In the case where the slice image is the first slice image among multiple slice images, at least one target slice image adjacent to the slice image is located after the slice image.

[0064] When the slice image is located between the first and last slice images in a plurality of slice images, at least one target slice image adjacent to the slice image may all be located after the slice image, or all be located before the slice image, or some of the target slice images may be located after the slice image and others may be located before the slice image.

[0065] In the case where the slice image is the last slice image among multiple slice images, at least one target slice image adjacent to the slice image is located before the slice image.

[0066] Optionally, the attenuation weight of the sliced ​​image is a preset value, which can be pre-stored in the computer device. For example, the preset value can be 1.

[0067] When there is only one target slice image, the attenuation weight of the target slice image is determined based on the corresponding layer number difference. When there are multiple target slice images, the attenuation weights of the multiple target slice images are attenuated sequentially in ascending order of the corresponding layer number differences.

[0068] The difference in the number of layers corresponding to each target slice image is the number of images between the target slice image and the slice image.

[0069] As an optional implementation, the attenuation weight of each target slice image can be an exponential function with the natural constant as the base and the linear attenuation factor as the exponent, wherein the linear attenuation factor is determined based on the negative of the first target product, and the first target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0070] The computer device can pre-store attenuation coefficients; for example, the attenuation coefficient could be 0.2. The first target product can be the product of the attenuation coefficient and the corresponding layer difference. The linear attenuation factor can be the negative of the first target product, or it can be the negative of the first target value. This first target value can be determined based on the first product and the first preset coefficient. Optionally, the first target value can be determined based on the product of the first product and the first preset coefficient. For example, the first target value can be the product of the first product and the first preset coefficient.

[0071] Optionally, the decay rate of the decay weight is positively correlated with the decay coefficient.

[0072] Here, the attenuation weight of the target slice image is an exponential function with a natural constant as the base and a linear attenuation factor as the exponent, where the linear attenuation factor is the negative of the first target product, and the first target product is the product of the attenuation coefficient and the corresponding layer difference. This is an example of the attenuation weight of the slice image and each target slice image. It can be expressed by the following formula:

[0073] in, The product of the first objective, Represents the linear decay factor. This represents the sliced ​​image, and the difference in the number of layers corresponding to each target slice image. This represents the attenuation coefficient. Since there are no images between slice images, the layer difference between the corresponding slice images is 0, therefore the attenuation weight of the slice image is 1.

[0074] As an alternative implementation, the attenuation weight for each target slice image can be an exponential function with the natural constant as the base and a quadratic attenuation factor as the exponent, where the quadratic attenuation factor is determined based on the negative of the second target product, and the second target product is determined based on the product of the attenuation coefficient and the square of the corresponding layer difference.

[0075] The second target product can be the product of the attenuation coefficient and the corresponding layer difference. The secondary attenuation factor can be the negative of the second target product or the negative of the second target value. The second target value can be determined based on the second product and the second preset coefficient. Optionally, the second target value can be determined based on the product of the second product and the second preset coefficient. For example, the second target value can be the product of the second product and the second preset coefficient.

[0076] Taking the attenuation weight of the target slice image as an exponential function with a natural constant as the base and a quadratic attenuation factor as the exponent, where the quadratic attenuation factor is the negative of the second target product, and the second target product is the product of the attenuation coefficient and the square of the corresponding layer difference, as an example, the attenuation weight of this slice image and each target slice image. It can be expressed by the following formula:

[0077] in, For the product of the second objective, This represents the second-order attenuation factor. This represents the sliced ​​image, and the difference in the number of layers corresponding to each target slice image. This represents the attenuation coefficient. Since there are no images between slice images, the layer difference between the corresponding slice images is 0, therefore the attenuation weight of the slice image is 1.

[0078] As another alternative implementation, the attenuation weight of each target slice image can be determined based on the reciprocal of the target sum value, which is determined based on the sum of the third target product and the target parameter, and the third target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0079] The computer equipment can pre-store target parameters to avoid division by zero errors; for example, the target parameter could be 1. The third target product can be the product of the attenuation coefficient and the corresponding layer difference. The target sum can be the sum of the third target product and the target parameter, or it can be the sum of the third target value and the target parameter. This third target value can be determined based on the third product and a third preset coefficient. Optionally, the third target value can be determined based on the product of the third product and the third preset coefficient. For example, the third target value can be the product of the third product and the third preset coefficient.

[0080] Taking the attenuation weight of the target slice image as the reciprocal of the target sum value, where the target sum value is the sum of the third target product and the target parameter, and the third target product is the product of the attenuation coefficient and the corresponding layer difference, with the target parameter being 1, as an example, the attenuation weight of this slice image and each target slice image is... It can be expressed by the following formula:

[0081] in, For the target and value, The product of the third objective. This represents the sliced ​​image, and the difference in the number of layers corresponding to each target slice image. This represents the attenuation coefficient. Since there are no images between slice images, the layer difference between the corresponding slice images is 0, therefore the attenuation weight of the slice image is 1.

[0082] It is understandable that the fewer the number of images between a target slice image and that slice image, the greater the correlation between them. As can be seen from the formulas above, the greater the attenuation weight of the target slice image. Conversely, the more the number of images between a target slice image and that slice image, the smaller the correlation between them. As can be seen from the formulas above, the smaller the attenuation weight of the target slice image. Therefore, the attenuation weight of multiple target slice images... According to the corresponding multiple layer differences The decay occurs sequentially from smallest to largest; decay coefficient The larger the value, the lower the decay weight. The faster it decays.

[0083] Step A2: Based on the attenuation weight and artifact intensity value of the slice image, and the attenuation weight and artifact intensity value of at least one target slice image, obtain the artifact intensity correction value of the artifact in the slice image.

[0084] The computer device can perform weighted processing on the artifact intensity values ​​of artifacts in the slice image and the artifact intensity values ​​of artifacts in at least one target slice image based on the attenuation weight of the slice image and the attenuation weight of at least one target slice image to obtain the artifact intensity correction value of the artifacts in the slice image.

[0085] Optionally, this weighting process can be a weighted average. Assume that in multiple slice images, the th... There are multiple target slice images corresponding to the nth slice image, and half of the target slice images are located in the nth slice image. Before the j-th slice image, the other half of the target slice image is located after the j-th slice image. The artifact intensity value of the artifact in each slice image is Then the first Artifact intensity correction value for artifacts in a slice image It can be expressed by the following formula:

[0086] in, Indicates the total number of slice images; The index value represents the slice image. less than or equal to Integers; This represents half the number of multiple target slice images. Indicates the index value of the target slice image. Indicates in and Take the minimum value from the middle. Indicates in 1 and Take the maximum value from the middle. Represents at least one target slice image and the first The total number of slice images, Indicates the first Attenuation weights for each target slice image Indicates the first The artifact intensity value of artifacts in a target slice image. The corresponding attenuation weight and artifact intensity value, i.e., the attenuation weight and artifact intensity value of the slice image.

[0087] Furthermore, when the actual artifact intensity value in at least one target slice image is close to the actual artifact intensity value in that slice image, for each slice image, the fewer the number of images between the target slice image and the slice image in at least one adjacent target slice image, the closer the artifact intensity value in the target slice image is to the actual artifact intensity value in that slice image.

[0088] Since the attenuation weight of at least one target slice image is attenuated sequentially in ascending order according to the corresponding multiple layer differences, the artifact intensity correction value of the artifact in the slice image determined based on the above method is closer to the artifact intensity value of the artifact in the target slice image that is close to the slice image. This ensures the accuracy of the determined artifact intensity correction value of the slice image.

[0089] S132: Obtain the degree of artifacts in a medical image based on the artifact intensity correction values ​​of artifacts in multiple slice images.

[0090] In some embodiments, the computer device may obtain the artifact level of a medical image based on the average of artifact intensity correction values ​​for artifacts in multiple slice images. For example, the computer device may use this average value as the artifact level of the medical image.

[0091] In other embodiments, the computer device may also weight the artifact intensity correction values ​​of artifacts in multiple slice images to obtain the artifact level of the medical image. For example, the weighting process may be a weighted average calculation.

[0092] In some other embodiments, the computer device may also obtain the target proportion of artifacts in each slice image. For each slice image, the artifact intensity correction value and the target proportion of artifacts in the slice image are weighted to obtain the artifact degree of the slice image. Then, based on the artifact degree of multiple slice images, the artifact degree of the medical image is obtained.

[0093] The target proportion is the area proportion, which is the ratio of the area of ​​artifacts in each slice image to the total area of ​​the corresponding slice image. The computer device can use an artifact segmentation model to obtain the artifact regions in the slice image, and then calculate the ratio of the area of ​​the artifact regions in the slice image to the total area of ​​the slice image. The ratio ranges from 0 to 1. The artifact segmentation model can be trained on fifth sample data, and each fifth sample data can include a sample image and sample artifact regions.

[0094] Computer equipment can take the average of the artifact levels from multiple slice images as the artifact level of a medical image. Alternatively, it can take the sum of the artifact levels from multiple slice images as the artifact level of a medical image.

[0095] Furthermore, the artifact severity of the slice image is calculated by weighting the artifact intensity correction value and the target proportion in the slice image. The artifact severity of multiple slice images is then averaged to obtain the artifact severity of the medical image. For example, the artifact severity of a medical image is calculated as follows: It can be expressed by the following formula:

[0096] in, This represents the weight of the artifact intensity correction value; for example, this weight could be set to 0.5. Indicates the first The proportion of the target area of ​​artifacts in a slice image.

[0097] Furthermore, the computer device can set the weight of the artifact intensity correction value according to the type of artifact. When the impact of metal artifacts on the image is more important, the weight of the artifact intensity correction value can be set to be greater than the weight of the target proportion; when the impact of motion artifacts on the image is more important, the weight of the artifact intensity correction value can be set to be less than the weight of the target proportion.

[0098] For example, when a patient undergoes a medical image scan, if the patient carries a lot of metal in their body, there may be more metal artifacts. In this case, the weight of the artifact intensity correction value can be set to be greater than the weight of the target proportion. If the patient has a lot of voluntary and involuntary movements, there may be more motion artifacts. In this case, the weight of the artifact intensity correction value can be set to be less than the weight of the target proportion.

[0099] In some embodiments, the computer device may also perform a weighted summation calculation on the artifact intensity correction value and the target proportion of artifacts in the slice image to obtain the artifact degree of the slice image; and then obtain the artifact degree of the medical image based on the artifact degree of multiple slice images.

[0100] In the above embodiments, the computer device can also obtain the level of artifact severity based on the range of severity of the artifact.

[0101] However, since there is a lack of unified quantitative standards for assessing the quality of medical images based on the degree of artifacts, it involves a certain degree of subjectivity. Therefore, a first-degree threshold can be preset. Second degree threshold The computer equipment determines the level of artifact severity based on a first-level threshold and a second-level threshold. The first-level threshold is lower than the second-level threshold.

[0102] The computer device can determine three degree ranges, from the first degree range to the third degree range, based on the first degree threshold and the second degree threshold. The upper limit of the first degree range is the first degree threshold, the lower limit of the second degree range is greater than the first degree threshold, and the upper limit of the second degree range is less than the second degree threshold. The lower limit of the third degree range is equal to the second degree threshold.

[0103] Furthermore, the artifact severity levels corresponding to the first severity range, the second severity range, and the third severity range increase sequentially.

[0104] Optionally, the computer device can normalize the artifact severity of the medical image based on a first severity threshold and a second severity threshold to obtain a normalized artifact severity level, and then determine the artifact severity level based on the normalized artifact severity level. The level of artifact severity is positively correlated with the magnitude of the normalized artifact severity level.

[0105] The greater the degree of artifacts after normalization, the higher the level of artifact severity; the smaller the degree of artifacts after normalization, the lower the level of artifact severity.

[0106] If the artifact severity is less than or equal to the first severity threshold, the computer device can use the third value as the normalized artifact severity.

[0107] When the artifact severity is greater than a first severity threshold and less than a second severity threshold, the computer device can obtain a first difference between the artifact severity and the first severity threshold, and a second difference between the second severity threshold and the first severity threshold, and determine the normalized artifact severity based on the ratio of the first difference and the second difference.

[0108] Optionally, the computer device may determine the ratio of the first difference to the second difference as the normalized degree of artifact.

[0109] If the artifact severity is greater than or equal to the second severity threshold, the computer device can use the fourth value as the normalized artifact severity.

[0110] When the third value equals 0, the fourth value equals 1, and the normalized artifact severity is the ratio of the first difference to the second difference, the normalized artifact severity... It can be expressed by the following formula:

[0111] Specifically, when the normalized artifact severity is the first value, the artifact severity level can be determined as severe; when the normalized artifact severity is the second value, the artifact severity level can be determined as mild; and when the normalized artifact severity is... At that time, the level of artifact severity can be determined as moderate.

[0112] Furthermore, the sensitivity of determining the artifact severity level can be adjusted by changing the first and second severity thresholds according to the specific needs of the application scenario, so as to quantitatively assess the artifact severity level of medical images.

[0113] Furthermore, computer equipment can also assess the quality of medical images based on the degree of artifacts. The quality of a medical image is negatively correlated with the degree of artifacts; that is, a higher degree of artifacts results in lower image quality, and vice versa.

[0114] In this embodiment of the application, the computer device displays the level of artifact severity in numerical form, thereby providing doctors with a more intuitive understanding of the artifact severity of medical images and improving the efficiency of doctors in screening medical images.

[0115] Artifacts (such as motion artifacts and metallic artifacts) can distort tissue morphology in medical images, leading to risks of misdiagnosis or missed diagnosis. Therefore, effective medical image quality assessment helps improve image quality and is of great practical significance for ensuring the reliability of clinical decisions, protecting patients' health rights, and achieving efficient use of medical resources.

[0116] In current techniques, only artifacts are typically detected in medical images. Images with artifacts are generally considered to be of poor quality and unusable for diagnosis, requiring rescanning. However, some medical images may contain artifacts that are minute or do not affect critical tissue structure information. These images are of acceptable quality and do not require rescanning. Using such techniques can lead to a high misjudgment rate of medical image quality, resulting in inaccurate medical image quality assessment and consequently affecting the accuracy and efficiency of doctors' diagnoses.

[0117] Therefore, in this embodiment of the application, by obtaining the degree of artifacts in a medical image, the quality of the medical image can be quantified more accurately than simply determining whether artifacts exist in the image. The computer device corrects the artifact intensity value of artifacts in the slice image based on the attenuation weight of the slice image and the attenuation weight of at least one target slice image adjacent to it. The determined corrected artifact intensity value is closer to the artifact intensity value of the target slice image, improving the accuracy of the obtained degree of artifacts in the slice image, thereby improving the accuracy of the obtained degree of artifacts in the medical image and ultimately improving the accuracy of medical image quality assessment. Simultaneously, the computer device assesses the degree of artifacts based on both the area ratio and the artifact intensity value, improving the objectivity and comprehensiveness of the obtained degree of artifacts in the slice image, further improving the accuracy of the obtained degree of artifacts in the medical image and thus improving the accuracy of medical image quality assessment. Furthermore, the computer device obtains the level of artifact degree based on the range of artifact degree, further clarifying the quality of the medical image corresponding to the artifact degree, thereby improving the efficiency of medical image quality assessment.

[0118] In other embodiments, a third-degree threshold may be preset, and the computer device may qualitatively assess whether there are artifacts in the medical image based on the normalized artifact level and the third-degree threshold.

[0119] Specifically, if the normalized artifact level is greater than the third-level threshold, the medical image can be considered to have either acceptable or unacceptable artifacts. If the normalized artifact level is less than the third-level threshold, the medical image can be considered to have either no artifacts or acceptable artifacts.

[0120] In some embodiments of this application, the computer device can calculate the average value of artifact intensity values ​​in multiple slice images to obtain the artifact level of the medical image. In other embodiments of this application, the computer device can also perform a weighted average calculation of artifact intensity values ​​in multiple slice images to obtain the artifact level of the medical image. That is, there is no need to correct the artifact intensity values.

[0121] In some embodiments of this application, the computer device may also perform weighted processing on the artifact intensity value and target proportion of each slice image to obtain the artifact degree.

[0122] Figure 3 A block diagram of a device for acquiring the degree of artifacts in medical images provided in this application embodiment, as shown below. Figure 3 As shown, the device includes: an acquisition module 301, an artifact intensity recognition module 302, and an evaluation module 303, wherein, The acquisition module 301 is used to acquire medical images of the target area, which include multiple consecutive slice images; The artifact intensity recognition module 302 is used to obtain the artifact intensity value of artifacts in each slice image; Evaluation module 303 is used to obtain the degree of artifacts in a medical image based on the artifact intensity values ​​of artifacts in multiple slice images.

[0123] Optionally, the evaluation module 303 is used to correct the artifact intensity value of the artifact in each slice image to obtain the artifact intensity correction value of the artifact in the slice image; and to obtain the artifact degree of the medical image based on the artifact intensity correction values ​​of the artifact in multiple slice images.

[0124] Optionally, the evaluation module 303 is used to obtain the attenuation weight of the slice image and the attenuation weight of at least one target slice image adjacent to the slice image; based on the attenuation weight of the slice image and the artifact intensity value of the artifact, and the attenuation weight of at least one target slice image and the artifact intensity value of the artifact, to obtain the artifact intensity correction value of the artifact in the slice image.

[0125] Optionally, the attenuation weight of the slice image is a preset value; when there is at least one target slice image, the attenuation weight of the target slice image is determined based on the corresponding layer difference; when there are multiple target slice images, the attenuation weights of the multiple target slice images are attenuated sequentially in ascending order of the corresponding multiple layer differences; wherein, the layer difference corresponding to each target slice image is the number of images between the target slice image and the slice image.

[0126] Optionally, the attenuation weight for each target slice image is an exponential function with a natural constant as the base and a linear attenuation factor as the exponent, wherein the linear attenuation factor is determined based on the negative of the first target product, and the first target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0127] Optionally, the attenuation weight for each target slice image is an exponential function with a natural constant as the base and a quadratic attenuation factor as the exponent, wherein the quadratic attenuation factor is determined based on the negative of the second target product, and the second target product is determined based on the product of the attenuation coefficient and the square of the corresponding layer difference.

[0128] Optionally, the attenuation weight for each target slice image is determined based on the reciprocal of the target sum value, which is determined based on the sum of the third target product and the target parameter, and the third target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

[0129] Optionally, the decay rate of the decay weight is positively correlated with the decay coefficient.

[0130] Optionally, the evaluation module 303 is used to perform weighted processing on the artifact intensity value of the artifact in the slice image and the artifact intensity value of the artifact in the at least one target slice image based on the attenuation weight of the slice image and the attenuation weight of at least one target slice image, to obtain the artifact intensity correction value of the artifact in the slice image.

[0131] Optionally, the evaluation module 303 is used to obtain the target proportion of artifacts in each slice image; for each slice image, the artifact intensity correction value and the target proportion of artifacts in the slice image are weighted to obtain the artifact degree; based on the artifact degree of multiple slice images, the artifact degree of the medical image is obtained.

[0132] Optionally, the evaluation module 303 is also used to obtain the level of artifact severity based on the severity range of the artifact severity.

[0133] The apparatus in this embodiment can be used to execute the solutions of the above method embodiments. Its implementation principle and technical effect are similar, and will not be described again here.

[0134] Figure 4 This application provides a schematic diagram of the structure of a computer device, as shown in the embodiment of the present application. Figure 4 As shown, the computer device 40 includes a memory 401, a processor 402, and a computer program stored in the memory 401 and executable on the processor 402. When the processor 402 executes the computer program, it implements the method for obtaining the degree of artifacts in medical images according to the above embodiment.

[0135] The processor 402 may be a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof.

[0136] This application also provides a computer-readable storage medium storing a computer program or instructions, which, when executed by a processor, implements the method for obtaining the degree of artifacts in a medical image as described above.

[0137] This application also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implement the method for obtaining the degree of artifacts in a medical image as described above.

[0138] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes that element. Furthermore, it should be noted that the scope of the methods and apparatuses in the embodiments of this application is not limited to performing functions in the order shown or discussed, but may also include performing functions substantially simultaneously or in the reverse order, depending on the functions involved. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Additionally, features described with reference to certain examples may be combined in other examples.

[0139] From the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of computer software products plus necessary general-purpose hardware platforms, and of course, they can also be implemented by hardware. The computer software product is stored in a storage medium (such as ROM, RAM, magnetic disk, optical disk, etc.) and includes several instructions to cause the terminal or network-side device to execute the methods described in the various embodiments of this application.

[0140] The embodiments of this application have been described above with reference to the accompanying drawings. However, this application is not limited to the specific embodiments described above. The specific embodiments described above are merely illustrative and not restrictive. Those skilled in the art can make many other implementations under the guidance of this application without departing from the spirit and scope of the claims. All of these implementations are within the protection scope of this application.

Claims

1. A method for obtaining the degree of artifacts in medical images, characterized in that, The method includes: Acquire medical images of the target area, the medical images including multiple consecutive slice images; Obtain the artifact intensity value of the artifact in each of the slice images; The degree of artifacts in the medical image is obtained based on the artifact intensity values ​​of the artifacts in the multiple slice images. The degree of artifact in the medical image is obtained based on the artifact intensity values ​​of artifacts in multiple slice images, including: For each slice image, the artifact intensity value of the artifact in the slice image is corrected to obtain the artifact intensity correction value of the artifact in the slice image; The degree of artifacts in the medical image is obtained based on the artifact intensity correction values ​​of the artifacts in the multiple slice images. The artifact intensity values ​​of the artifacts in the sliced ​​image are corrected to obtain the artifact intensity correction values ​​of the artifacts in the sliced ​​image, including: Obtain the attenuation weight of the slice image, and the attenuation weight of at least one target slice image adjacent to the slice image; Based on the attenuation weight and artifact intensity value of the slice image, and the attenuation weight and artifact intensity value of at least one target slice image, the artifact intensity correction value of the artifact in the slice image is obtained. The attenuation weight of the sliced ​​image is a preset value; When there is at least one target slice image, the attenuation weight of the target slice image is determined based on the corresponding layer difference; When there are multiple target slice images, the attenuation weights of the multiple target slice images are attenuated sequentially in order of increasing difference in the corresponding layer number; Wherein, the layer difference corresponding to each target slice image is the number of images between the target slice image and the slice image.

2. The method according to claim 1, characterized in that, The attenuation weight for each target slice image is an exponential function with a natural constant as the base and a linear attenuation factor as the exponent, wherein the linear attenuation factor is determined based on the negative of a first target product, which is determined based on the product of the attenuation coefficient and the corresponding layer difference.

3. The method according to claim 1, characterized in that, The attenuation weight for each target slice image is an exponential function with a natural constant as the base and a quadratic attenuation factor as the exponent, wherein the quadratic attenuation factor is determined based on the negative of the second target product, which is determined based on the product of the attenuation coefficient and the square of the corresponding layer difference.

4. The method according to claim 1, characterized in that, The attenuation weight for each target slice image is determined based on the reciprocal of the target sum value, which is determined based on the sum of a third target product and a target parameter, and the third target product is determined based on the product of the attenuation coefficient and the corresponding layer difference.

5. The method according to any one of claims 2 to 4, characterized in that, The decay rate of the decay weight is positively correlated with the decay coefficient.

6. The method according to any one of claims 1 to 4, characterized in that, Based on the attenuation weight and artifact intensity value of the sliced ​​image, and the attenuation weight and artifact intensity value of at least one target sliced ​​image, an artifact intensity correction value for the artifact in the sliced ​​image is obtained, including: Based on the attenuation weight of the slice image and the attenuation weight of at least one target slice image, the artifact intensity value of the artifact in the slice image and the artifact intensity value of the artifact in at least one target slice image are weighted to obtain the artifact intensity correction value of the artifact in the slice image.

7. The method according to any one of claims 1 to 4, characterized in that, The artifact severity of the medical image is obtained based on artifact intensity correction values ​​in multiple slice images, including: Obtain the target proportion of artifacts in each of the sliced ​​images; For each slice image, the artifact intensity correction value and the target proportion of the artifact in the slice image are weighted to obtain the artifact degree; The artifact level of the medical image is obtained based on the artifact level of multiple slice images.

8. The method according to any one of claims 1 to 4, characterized in that, The method further includes: Based on the range of the artifact severity, the level of the artifact severity is obtained.

9. An apparatus for acquiring the degree of artifacts in medical images, suitable for control using the method for acquiring the degree of artifacts in medical images according to any one of claims 1-8, characterized in that, The device includes: An acquisition module is used to acquire medical images of a target area, the medical images including multiple consecutive slice images; The artifact intensity recognition module is used to obtain the artifact intensity value of the artifacts in each slice image; An evaluation module is used to obtain the degree of artifacts in the medical image based on the artifact intensity values ​​of artifacts in multiple slice images.

10. A computer device, characterized in that, include: A processor for connection to a memory storing a computer program or instructions executable on the processor, the computer program or instructions being executed by the processor to implement the method for obtaining the degree of artifacts in a medical image as described in any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program or instructions, which, when executed by a processor, implement the method for obtaining the degree of artifacts in a medical image as described in any one of claims 1 to 8.

Images