Correction algorithm automatic calling method, device, electronic device and storage medium

Abstract

The present invention provides a method, device, electronic device and storage medium for automatically calling a correction algorithm, the method comprising: obtaining a medical image to be corrected, and determining a region of interest in the medical image to be corrected; judging whether the region of interest needs to be corrected based on a preset discriminant model; and when the region of interest needs to be corrected, calling a preset correction algorithm to correct the medical image to be corrected. The present invention saves the time of presenting the medical data to be corrected to a doctor and waiting for the doctor to judge whether correction is needed, thereby improving the efficiency of judging whether correction is needed, and further improving the efficiency of correcting the medical image to be corrected.

Description

Technical Field

[0001] The present invention relates to the technical field of medical image processing, and in particular to a correction algorithm automatic calling method, device, electronic equipment and storage medium. Background Art

[0002] With the continuous development of medical imaging technology, in order to better examine the human body, a variety of technologies are integrated to detect the human body. For example, Positron Emission Tomography / Computed Tomography (PET / CT) uses PET for organ and soft tissue detection and CT for human body layer detection. CT images and PET images are obtained at the same time, and the advantages of the two images complement each other, allowing doctors to understand biological metabolic information while obtaining precise anatomical positioning, so as to make a comprehensive and accurate judgment on the disease.

[0003] However, when patients are currently examined by a PET / CT system, due to the high scanning speed of CT, the CT image corresponds to a single frame or adjacent frames of human respiratory motion imaging. However, PET images usually have a long scanning time, so PET images correspond to average human respiratory imaging. Therefore, when patients are examined, the patient's breathing and movement are large, which will cause a large difference between the CT image and the PET image. When the CT image is used to perform attenuation correction on the PET image, the reconstructed PET image will have respiratory artifacts and / or motion artifacts, thereby affecting the judgment of the disease.

[0004] In order to avoid adverse effects of respiratory artifacts and / or motion artifacts on disease diagnosis results, doctors are usually required to manually read the images, that is, doctors manually check the reconstructed PET images to determine whether there are artifacts. If there are artifacts, the correction algorithm is manually called to correct the PET images to improve the reliability of disease diagnosis. The existing technology has the following problems: relying on doctors to manually read the images, resulting in a long time and low efficiency for correcting PET images. Summary of the invention

[0005] In view of this, it is necessary to provide a correction algorithm automatic calling method, device, electronic device and storage medium to solve the technical problems of long correction time and low efficiency in correcting medical images in the prior art.

[0006] In one aspect, the present invention provides a method for automatically calling a correction algorithm, comprising:

[0007] Acquire a medical image to be corrected, and determine a region of interest in the medical image to be corrected;

[0008] Determining whether the region of interest needs to be corrected based on a preset discriminant model;

[0009] When the region of interest needs to be corrected, a preset correction algorithm is called to correct the medical image to be corrected.

[0010] In some possible implementations, the discriminant model includes a lesion recognition sub-model and an image artifact detection sub-model, and judging whether the region of interest needs to be corrected based on a preset discriminant model includes:

[0011] determining whether the region of interest needs correction based on the lesion recognition sub-model and the image artifact detection sub-model in sequence;

[0012] or,

[0013] determining whether the region of interest needs correction based on the image artifact detection sub-model and the lesion recognition sub-model in sequence;

[0014] or,

[0015] At the same time, whether the region of interest needs to be corrected is determined based on the lesion recognition sub-model and the image artifact detection sub-model.

[0016] In some possible implementations, the determining whether the region of interest needs to be corrected based on the lesion recognition sub-model and the image artifact detection sub-model in sequence includes:

[0017] Determining whether the region of interest is a lesion region by using the lesion recognition sub-model;

[0018] When the region of interest is a lesion region, detecting whether there is an image artifact in the region of interest by using the image artifact detection sub-model;

[0019] When image artifacts exist in the region of interest, the region of interest needs to be corrected.

[0020] In some possible implementations, the determining whether the region of interest needs to be corrected based on the image artifact detection sub-model and the lesion recognition sub-model in sequence includes:

[0021] Detecting whether there are image artifacts in the region of interest by using the image artifact detection sub-model;

[0022] When there are image artifacts in the region of interest, determining whether the region of interest is a lesion region by using the lesion recognition sub-model;

[0023] When the region of interest is a lesion region, the region of interest needs to be corrected.

[0024] In some possible implementations, the determining whether the region of interest needs correction based on both the lesion recognition sub-model and the image artifact detection sub-model includes:

[0025] Determining whether the region of interest is a lesion region by using the lesion recognition sub-model;

[0026] Detecting whether there are image artifacts in the region of interest by using the image artifact detection sub-model;

[0027] When the region of interest is a lesion region and image artifacts exist in the region of interest, the region of interest needs to be corrected.

[0028] In some possible implementations, determining the region of interest in the to-be-corrected medical image includes:

[0029] Segmenting the medical image to be corrected based on an image segmentation algorithm to obtain multiple image regions;

[0030] The multiple image regions are calibrated based on a part recognition algorithm to obtain the region of interest.

[0031] On the other hand, the present invention also provides a correction algorithm automatic calling device, comprising:

[0032] An image acquisition unit, used for acquiring a medical image to be corrected and determining a region of interest in the medical image to be corrected;

[0033] A correction determination unit, used for determining whether the region of interest needs to be corrected based on a preset determination model;

[0034] The correction algorithm calling unit is used to call a preset correction algorithm to correct the medical image to be corrected when the region of interest needs to be corrected.

[0035] On the other hand, the present invention also provides an electronic device, including a memory and a processor, wherein:

[0036] The memory is used to store programs;

[0037] The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the correction algorithm automatic calling method described in any of the above implementations.

[0038] On the other hand, the present invention also provides a computer-readable storage medium for storing a computer-readable program or instruction, which, when executed by a processor, can implement the steps in the correction algorithm automatic calling method described in any of the above implementations.

[0039] The beneficial effect of adopting the above embodiment is that the correction algorithm automatic calling method provided by the present invention determines whether the region of interest needs to be corrected based on a preset discrimination model, and when the region of interest needs to be corrected, calls the preset correction algorithm to correct the medical image to be corrected. There is no need for doctors to manually read the images and determine whether the medical image to be corrected needs to be corrected, which saves the time of presenting the medical data to be corrected to the doctor and waiting for the doctor to determine whether correction is needed, thereby improving the efficiency of determining whether correction is needed, and further improving the efficiency of correcting the medical image to be corrected.

[0040] Furthermore, by judging whether the region of interest needs to be corrected based on a preset discrimination model, the human error rate caused by manual reading of the film can be avoided, thereby improving the accuracy of judging whether the region of interest needs to be corrected. Moreover, when the region of interest needs to be corrected, the preset correction algorithm is called to correct the medical image to be corrected, without waiting for the doctor to manually issue a call instruction to call the correction algorithm, further improving the efficiency of correcting the medical image to be corrected. BRIEF DESCRIPTION OF THE DRAWINGS

[0041] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0042] Figure 1 A schematic flow chart of an embodiment of the method for automatically calling a correction algorithm provided by the present invention;

[0043] Figure 2 For the present invention Figure 1 The first embodiment flow chart of S102;

[0044] Figure 3 For the present invention Figure 1 A second embodiment flow chart of S102;

[0045] Figure 4 For the present invention Figure 1 A schematic flow chart of a third embodiment of S102;

[0046] Figure 5 For the present invention Figure 1 A schematic flow chart of an embodiment of S101;

[0047] Figure 6 A schematic diagram of the structure of an embodiment of the correction algorithm automatic calling device provided by the present invention;

[0048] Figure 7A schematic structural diagram of an embodiment of an electronic device provided by the present invention. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0050] In the description of the embodiments of the present invention, unless otherwise specified, "multiple" means two or more than two. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, "A and / or B" may mean: A exists alone, A and B exist at the same time, and B exists alone.

[0051] Reference to "embodiments" herein means that a particular feature, structure, or characteristic described in conjunction with the embodiments may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor is it an independent or alternative embodiment that is mutually exclusive with other embodiments. It is explicitly and implicitly understood by those skilled in the art that the embodiments described herein may be combined with other embodiments.

[0052] The present invention provides a correction algorithm automatic calling method, device, electronic device and storage medium, which are described below respectively.

[0053] Figure 1 A flow chart of an embodiment of the method for automatically calling the correction algorithm provided by the present invention is shown as follows: Figure 1 As shown, the automatic calling method of the correction algorithm includes:

[0054] S101, obtaining a medical image to be corrected, and determining a region of interest in the medical image to be corrected;

[0055] S102, judging whether the region of interest needs to be corrected based on a preset discriminant model;

[0056] S103: When the region of interest needs to be corrected, a preset correction algorithm is called to correct the medical image to be corrected.

[0057] Compared with the prior art, the automatic correction algorithm calling method provided in the embodiment of the present invention determines whether the region of interest needs to be corrected based on a preset discrimination model. When the region of interest needs to be corrected, the preset correction algorithm is called to correct the medical image to be corrected. There is no need for doctors to manually read the images and determine whether the medical image to be corrected needs to be corrected, which saves the time of presenting the medical data to be corrected to the doctor and waiting for the doctor to determine whether correction is needed, thereby improving the efficiency of determining whether correction is needed, and further improving the efficiency of correcting the medical image to be corrected.

[0058] Furthermore, by judging whether the region of interest needs to be corrected based on a preset discrimination model, the human error rate caused by manual reading of the film can be avoided, thereby improving the accuracy of judging whether the region of interest needs to be corrected. Moreover, when the region of interest needs to be corrected, the preset correction algorithm is called to correct the medical image to be corrected, without waiting for the doctor to manually issue a call instruction to call the correction algorithm, further improving the efficiency of correcting the medical image to be corrected.

[0059] In some embodiments of the present invention, the medical image to be corrected in step S101 may be a medical image obtained by scanning a test object with a medical imaging device. The test object may be biological or non-biological. For example, the test object may be a patient, an artificial object (such as a human model), etc.

[0060] In some embodiments of the present invention, the region of interest may be a specific part, organ and / or tissue of the subject. For example, when the subject is a patient, the region of interest may be any one of the patient's head, neck, abdomen, chest, shoulder, arm, leg, heart, stomach, liver, lung, etc.

[0061] In some specific embodiments, the medical imaging device may be a single-modality medical imaging device or a multi-modality medical imaging device. Specifically, the single-modality medical imaging device may be a PET medical imaging device, and the multi-modality medical imaging device may be a PET / CT medical imaging device.

[0062] In an actual situation, when the medical imaging device is a PET / CT medical imaging device, due to the long PET scanning time, the different respiratory phases of PET and CT will lead to banana artifacts at the liver-lung junction, commonly known as respiratory artifacts.

[0063] In the actual disease diagnosis process, what is of concern is whether there are image artifacts in the lesion area, while the existence of image artifacts in the non-lesion area will not affect the disease diagnosis results. Therefore, in order to reduce the correction workload of the medical image to be corrected, in some embodiments of the present invention, the discriminant model includes a lesion recognition sub-model and an image artifact detection sub-model.

[0064] The embodiment of the present invention sets a discriminant model including a lesion recognition sub-model and an image artifact detection sub-model, so that the medical image to be corrected with image artifacts in the lesion area can be corrected without correcting all the medical images to be corrected with artifacts, thereby reducing the correction workload and improving the correction accuracy of the medical images to be corrected.

[0065] It should be noted that the image detection submodel may include a respiratory artifact detection submodel and / or a motion artifact detection submodel. By setting the image detection submodel to include a respiratory artifact detection submodel and / or a motion artifact detection submodel, respiratory artifacts and / or motion artifacts may be detected, thereby improving the comprehensiveness and reliability of artifact detection.

[0066] Since when the region of interest is not a lesion region, even if there are image artifacts in the region of interest, it will not affect the diagnosis result, that is, there is no need to correct the region of interest that is not a lesion region but has image artifacts; similarly, there is no need to correct the image artifacts for the region of interest that is a lesion region but does not have image artifacts. Therefore, in order to reduce the correction workload, in some embodiments of the present invention, step S102 includes:

[0067] Determine whether the region of interest needs correction based on the lesion recognition sub-model and the image artifact detection sub-model in turn;

[0068] or,

[0069] Whether the region of interest needs correction is determined based on the image artifact detection sub-model and the lesion recognition sub-model.

[0070] That is, the operation mode of the lesion recognition submodel and the image artifact detection submodel is serial, rather than simultaneous. By setting the operation mode of the lesion recognition submodel and the image artifact detection submodel to be serial, only when there is an image artifact in the region of interest and the region of interest is the region of interest determined to need correction, thereby avoiding unnecessary correction work and improving the accuracy and efficiency of correcting the image to be corrected. In addition, the technical problem of excessive memory usage and errors in the operation process of the lesion recognition submodel and the image artifact detection submodel caused by running the lesion recognition submodel and the image artifact detection submodel at the same time can be avoided, thereby improving the safety and reliability of the model operation.

[0071] In order to further improve the discriminant model's discriminant speed, in some embodiments of the present invention, step S102 includes:

[0072] At the same time, based on the lesion recognition sub-model and the image artifact detection sub-model, it is determined whether the region of interest needs to be corrected.

[0073] That is, the operation mode of the lesion recognition submodel and the image artifact detection submodel are parallel. By setting the operation mode of the lesion recognition submodel and the image artifact detection submodel to be parallel, the lesion recognition submodel and the image artifact model can be run simultaneously, thereby improving the efficiency of the discrimination model in judging whether the region of interest needs to be corrected, thereby further improving the efficiency of correcting the medical images to be corrected.

[0074] It should be understood that: in practical applications, it can be selected according to the hardware conditions or actual needs of running the discriminant model whether to use the method of judging whether the region of interest needs to be corrected based on the lesion recognition sub-model and the image artifact detection sub-model in sequence, or based on the image artifact detection sub-model and the lesion recognition sub-model in sequence, or using the method of judging whether the region of interest needs to be corrected based on both the lesion recognition sub-model and the image artifact detection sub-model at the same time. No specific limitation is made here.

[0075] In some embodiments of the present invention, Figure 2 As shown, judging whether the region of interest needs to be corrected is based on the lesion recognition sub-model and the image artifact detection sub-model in turn, including:

[0076] S201, determining whether the region of interest is a lesion region through a lesion recognition sub-model;

[0077] S202, when the region of interest is a lesion region, detecting whether there are image artifacts in the region of interest by using an image artifact detection sub-model;

[0078] S203: When image artifacts exist in the region of interest, the region of interest needs to be corrected.

[0079] In some embodiments of the present invention, Figure 3 As shown, judging whether the region of interest needs to be corrected is based on the image artifact detection sub-model and the lesion recognition sub-model in turn, including:

[0080] S301, detecting whether there are image artifacts in the region of interest by using an image artifact detection sub-model;

[0081] S302, when there are image artifacts in the region of interest, determining whether the region of interest is a lesion region by using a lesion recognition sub-model;

[0082] S303: When the region of interest is a lesion region, the region of interest needs to be corrected.

[0083] The embodiment of the present invention improves the diversity and flexibility of the judgment methods of whether the region of interest needs to be corrected by setting two judgment methods: first determining whether the region of interest is a lesion region, then determining whether there are image artifacts in the lesion region to determine whether the region of interest needs to be corrected; and first determining whether there are image artifacts in the region of interest, then determining whether the region of interest with image artifacts is a lesion region to determine whether the region of interest needs to be corrected in steps S301-S303.

[0084] In some embodiments of the present invention, Figure 4 As shown, whether the region of interest needs to be corrected is determined based on the lesion recognition sub-model and the image artifact detection sub-model, including:

[0085] S401, determining whether the region of interest is a lesion region through a lesion recognition sub-model;

[0086] S402, detecting whether there are image artifacts in the region of interest by using an image artifact detection sub-model;

[0087] S403: When the region of interest is a lesion region and image artifacts exist in the region of interest, the region of interest needs to be corrected.

[0088] The embodiment of the present invention is configured to simultaneously determine whether a region of interest is a lesion region and whether there are image artifacts in the region of interest through a lesion recognition sub-model and an image artifact detection sub-model, without the need to perform detection through the image artifact detection sub-model or in the recognition sub-model only after the lesion recognition sub-model has a recognition result or the image artifact detection sub-model has a detection result, thereby improving the efficiency of determining whether the region of interest needs to be corrected.

[0089] In some embodiments of the present invention, before step S102, the method further includes:

[0090] Historical medical images are obtained, and an image artifact detection sub-model and a lesion recognition sub-model are established based on the historical medical images based on a deep learning algorithm.

[0091] It should be understood that deep learning algorithms can be able to learn certain knowledge and capabilities from existing data (historical medical images) for processing new data, and can be designed to perform various tasks. In an embodiment of the present invention, they are used to determine whether a region of interest is a lesion area and / or whether there are image artifacts in the region of interest.

[0092] In some embodiments of the present invention, examples of deep learning algorithms include, but are not limited to, deep belief networks (DBN), convolutional neural networks (CNN), recurrent neural networks (RNN), and the like.

[0093] In a specific embodiment of the present invention, the network structure of the lesion recognition sub-model can be any one of 2DU-net, 3DU-net, U-net++, U-net3+, and V-net.

[0094] In some embodiments of the present invention, the specific process of establishing an image artifact detection sub-model is: first, an initial model is established according to any of the deep learning algorithms mentioned above, and then the initial model is trained based on historical medical images. After the training is completed, the image artifact detection sub-model can be obtained, and the image artifact detection sub-model can be used to detect whether there are image artifacts in the area of ​​interest.

[0095] It should be understood that historical medical images include medical images with image artifacts and medical images without image artifacts.

[0096] It should also be understood that the specific process of establishing the lesion recognition sub-model is the same as the specific process of establishing the image artifact detection sub-model, and will not be described in detail here.

[0097] In some embodiments of the present invention, Figure 5 As shown, determining the region of interest in the medical image to be corrected in step S101 includes:

[0098] S501, segmenting the medical image to be corrected based on an image segmentation algorithm to obtain multiple image regions;

[0099] S502: Calibrate multiple image regions based on a part recognition algorithm to obtain a region of interest.

[0100] The embodiment of the present invention can improve the speed of obtaining the region of interest by setting an image segmentation algorithm to segment the medical image to be corrected, without the need for doctors to manually segment the medical image with images, thereby further improving the efficiency of judging whether the medical image to be corrected needs to be corrected.

[0101] It should be understood that the image segmentation algorithm in step S501 may be any one of a threshold-based segmentation algorithm, an edge-based segmentation algorithm, a clustering analysis-based segmentation algorithm, a wavelet transform-based segmentation algorithm or a neural network-based segmentation algorithm.

[0102] In some embodiments of the present invention, when the medical image to be corrected is a human body image, the part recognition algorithm in step S502 is used to identify and calibrate which part of the human body the multiple image regions are, and determine the region of interest from the identified multiple image regions.

[0103] For example, when the abdomen, chest, liver, and lungs are identified by the part recognition algorithm, at least one of the abdomen, chest, liver, and lungs can be selected as the region of interest according to needs.

[0104] It should be understood that the part recognition algorithm in step S502 can be any one of the algorithms such as decision tree, random forest algorithm, support vector machine (SVM) and the like.

[0105] It should be noted that: considering the real-time display requirements and non-real-time display requirements of the to-be-corrected medical image corrected by the preset correction algorithm, in some embodiments of the present invention, after step S103, the following is further included:

[0106] storing the medical image to be corrected after being corrected by the correction algorithm;

[0107] In response to the display signal, the stored corrected medical image to be corrected is called and displayed.

[0108] The embodiment of the present invention can meet the real-time display requirements and non-real-time display requirements of the medical image to be corrected after correction by a preset correction algorithm by setting the corrected medical image to be corrected to be stored and calling and displaying it in response to a display signal.

[0109] In order to better implement the correction algorithm automatic calling method in the embodiment of the present invention, based on the correction algorithm automatic calling method, correspondingly, Figure 6 As shown, the embodiment of the present invention further provides a correction algorithm automatic calling device, and the correction algorithm automatic calling device 600 includes:

[0110] An image acquisition unit 601 is used to acquire a medical image to be corrected and determine a region of interest in the medical image to be corrected;

[0111] A correction determination unit 602 is used to determine whether the region of interest needs to be corrected based on a preset determination model;

[0112] The correction algorithm calling unit 603 is used to call a preset correction algorithm to correct the medical image to be corrected when the region of interest needs to be corrected.

[0113] The correction algorithm automatic calling device 600 provided in the above embodiment can implement the technical solution described in the above correction algorithm automatic calling method embodiment. The specific implementation principles of the above modules or units can refer to the corresponding contents in the above correction algorithm automatic calling method embodiment, which will not be repeated here.

[0114] like Figure 7 As shown, the present invention also provides an electronic device 700. The electronic device 700 includes a processor 701, a memory 702 and a display 703. Figure 7 Only some components of the electronic device 700 are shown, but it should be understood that it is not required to implement all of the components shown, and more or fewer components may be implemented instead.

[0115] In some embodiments, the processor 701 may be a central processing unit (CPU), a microprocessor or other data processing chip, used to run program codes or process data stored in the memory 702, such as the automatic calling method of the correction algorithm in the present invention.

[0116] In some embodiments, the processor 701 may be a single server or a server group. The server group may be centralized or distributed. In some embodiments, the processor 701 may be local or remote. In some embodiments, the processor 701 may be implemented in a cloud platform. In one embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an internal cloud, a multi-cloud, etc., or any combination thereof.

[0117] In some embodiments, the memory 702 may be an internal storage unit of the electronic device 700, such as a hard disk or memory of the electronic device 700. In other embodiments, the memory 702 may also be an external storage device of the electronic device 700, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the electronic device 700.

[0118] Furthermore, the memory 702 may include both an internal storage unit of the electronic device 700 and an external storage device. The memory 702 is used to store application software installed in the electronic device 700 and various data.

[0119] In some embodiments, the display 703 may be an LED display, a liquid crystal display, a touch-sensitive liquid crystal display, an OLED (Organic Light-Emitting Diode) touch device, etc. The display 703 is used to display information on the electronic device 700 and to display a visual user interface. The components 701-703 of the electronic device 700 communicate with each other via a system bus.

[0120] In one embodiment, when the processor 701 executes the correction algorithm automatic calling program in the memory 702, the following steps may be implemented:

[0121] Acquire a medical image to be corrected, and determine a region of interest in the medical image to be corrected;

[0122] Determine whether the region of interest needs to be corrected based on a preset discriminant model;

[0123] When the region of interest needs to be corrected, a preset correction algorithm is called to correct the medical image to be corrected.

[0124] It should be understood that: when the processor 701 executes the correction algorithm in the memory 702 to automatically call the program, in addition to the above functions, other functions can also be implemented. For details, please refer to the description of the corresponding method embodiment above.

[0125] Furthermore, the embodiment of the present invention does not specifically limit the type of the electronic device 700 mentioned, and the electronic device 700 may be a portable electronic device such as a mobile phone, a tablet computer, a personal digital assistant (PDA), a wearable device, a laptop computer, etc. Exemplary embodiments of portable electronic devices include but are not limited to portable electronic devices equipped with IOS, Android, Microsoft or other operating systems. The above-mentioned portable electronic device may also be other portable electronic devices, such as a laptop computer with a touch-sensitive surface (e.g., a touch panel). It should also be understood that in some other embodiments of the present invention, the electronic device 700 may not be a portable electronic device, but a desktop computer with a touch-sensitive surface (e.g., a touch panel).

[0126] Accordingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, it can implement the steps or functions in the correction algorithm automatic calling method provided in the above-mentioned method embodiments.

[0127] The automatic correction algorithm calling method provided by the embodiment of the present invention determines whether the region of interest needs to be corrected based on a preset discrimination model, and when the region of interest needs to be corrected, calls the preset correction algorithm to correct the medical image to be corrected. There is no need for doctors to manually read the images and determine whether the medical image to be corrected needs to be corrected, which saves the time of presenting the medical data to be corrected to the doctor and waiting for the doctor to determine whether correction is needed, thereby improving the efficiency of determining whether correction is needed, and further improving the efficiency of correcting the medical image to be corrected.

[0128] Furthermore, by judging whether the region of interest needs to be corrected based on a preset discrimination model, the human error rate caused by manual reading of the film can be avoided, thereby improving the accuracy of judging whether the region of interest needs to be corrected. Moreover, when the region of interest needs to be corrected, the preset correction algorithm is called to correct the medical image to be corrected, without waiting for the doctor to manually issue a call instruction to call the correction algorithm, further improving the efficiency of correcting the medical image to be corrected.

[0129] Furthermore, the present invention can improve the accuracy of the obtained coincidence event counts by performing decay correction on the coincidence event counts in real time, thereby improving the accuracy of the estimated injection dose.

[0130] Those skilled in the art will appreciate that all or part of the processes of the above-mentioned embodiments can be implemented by instructing related hardware (such as a processor, a controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium, wherein the computer-readable storage medium is a disk, an optical disk, a read-only storage memory, or a random access memory, etc.

[0131] The above is a detailed introduction to the correction algorithm automatic calling method, device, electronic device and storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A method for automatically calling a correction algorithm, characterized in that: include: Acquire a medical image to be corrected, and determine a region of interest in the medical image to be corrected; Determining whether the region of interest needs to be corrected based on a preset discriminant model; the discriminant model includes a lesion recognition sub-model and an image artifact detection sub-model; When the region of interest needs to be corrected, a preset correction algorithm is called to correct the medical image to be corrected.

2. The method for automatically calling a correction algorithm according to claim 1, characterized in that: The determining whether the region of interest needs to be corrected based on a preset discriminant model includes: determining whether the region of interest needs correction based on the lesion recognition sub-model and the image artifact detection sub-model in sequence; or, determining whether the region of interest needs correction based on the image artifact detection sub-model and the lesion recognition sub-model in sequence; or, At the same time, whether the region of interest needs to be corrected is determined based on the lesion recognition sub-model and the image artifact detection sub-model.

3. The method for automatically calling a correction algorithm according to claim 2, characterized in that: The determining whether the region of interest needs to be corrected based on the lesion recognition sub-model and the image artifact detection sub-model in sequence includes: Determining whether the region of interest is a lesion region by using the lesion recognition sub-model; When the region of interest is a lesion region, detecting whether there is an image artifact in the region of interest by using the image artifact detection sub-model; When image artifacts exist in the region of interest, the region of interest needs to be corrected.

4. The method for automatically calling a correction algorithm according to claim 2, characterized in that: The determining whether the region of interest needs to be corrected based on the image artifact detection sub-model and the lesion recognition sub-model in sequence includes: Detecting whether there are image artifacts in the region of interest by using the image artifact detection sub-model; When there are image artifacts in the region of interest, determining whether the region of interest is a lesion region by using the lesion recognition sub-model; When the region of interest is a lesion region, the region of interest needs to be corrected.

5. The method for automatically calling a correction algorithm according to claim 2, characterized in that: The step of judging whether the region of interest needs correction based on both the lesion recognition sub-model and the image artifact detection sub-model comprises: Determining whether the region of interest is a lesion region by using the lesion recognition sub-model; Detecting whether there are image artifacts in the region of interest by using the image artifact detection sub-model; When the region of interest is a lesion region and image artifacts exist in the region of interest, the region of interest needs to be corrected.

6. The method for automatically calling a correction algorithm according to any one of claims 2 to 5, characterized in that: Before judging whether the region of interest needs to be corrected based on the preset discrimination model, the method further includes: Historical medical images are acquired, and the image artifact detection sub-model and the lesion recognition sub-model are established according to the historical medical images based on a deep learning algorithm.

7. The method for automatically calling a correction algorithm according to claim 1, characterized in that: The determining of the region of interest in the medical image to be corrected comprises: Segmenting the medical image to be corrected based on an image segmentation algorithm to obtain multiple image regions; The multiple image regions are calibrated based on a part recognition algorithm to obtain the region of interest.

8. A correction algorithm automatic calling device, characterized in that: include: An image acquisition unit, used for acquiring a medical image to be corrected and determining a region of interest in the medical image to be corrected; A correction discrimination unit is used to judge whether the region of interest needs to be corrected based on a preset discrimination model; the discrimination model includes a lesion recognition sub-model and an image artifact detection sub-model, The correction algorithm calling unit is used to call a preset correction algorithm to correct the medical image to be corrected when the region of interest needs to be corrected.

9. An electronic device, characterized in that: comprising a memory and a processor, wherein: The memory is used to store programs; The processor is coupled to the memory and is used to execute the program stored in the memory to implement the steps in the correction algorithm automatic calling method described in any one of claims 1 to 7 above.

10. A computer-readable storage medium, characterized in that: Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the steps in the correction algorithm automatic calling method described in any one of claims 1 to 7 above.

Images