Image registration method, image registration device, electronic equipment and storage medium

Abstract

The invention discloses an image registration method, an image registration device, electronic equipment and a storage medium, and the image registration method comprises the steps: obtaining a preoperative image and an intraoperative image; performing rigid registration on the preoperative image and the intraoperative image to obtain a corresponding initial narrow field angle rigid matrix; performing transformation processing on the preoperative image according to the initial narrow field angle rigid matrix to obtain a corresponding rigid registration preoperative image; according to the rigid registration preoperative image and the intraoperative image, obtaining a corresponding initial narrow field angle elastic deformation field; and according to the initial narrow field angle rigid matrix and the initial narrow field angle elastic deformation field, performing transformation processing on the preoperative image to obtain a target registration image. According to the method, rigid registration and elastic registration based on the preoperative image and the intraoperative image are realized, the preoperative image is registered to the intraoperative image, and the registration accuracy of image registration is improved.

Description

Technical Field

[0001] The present application belongs to the field of image processing technology, and in particular relates to an image registration method, an image registration device, an electronic device and a storage medium. Background Art

[0002] In robotic interventional puncture surgery, the registration of preoperative computed tomography (CT) images with intraoperative real-time images is a crucial technology. Preoperative images provide comprehensive anatomical structure and lesion information. Doctors usually locate and analyze the size, shape, location of the patient's lesions and the location of surrounding tissues and organs on the preoperative image sequence to plan interventional puncture surgery. Intraoperative real-time images provide real-time navigation and monitoring for the surgery.

[0003] In order to maintain real-time performance, the field of view (FOV) of intraoperative real-time images is usually very narrow. In contrast, preoperative images have a wider FOV. In puncture surgery, it is often necessary to register the preoperative images with the intraoperative real-time images.

[0004] At present, the registration method of preoperative images and intraoperative real-time images mainly relies on manual operation. However, manual operation requires doctors to make manual adjustments based on their experience, which is time-consuming and depends on the doctor's experience, and is prone to human errors, resulting in low registration accuracy of image registration. Summary of the invention

[0005] In view of this, embodiments of the present application provide an image registration method, an image registration device, an electronic device, and a storage medium to overcome or at least partially solve the above problems of the prior art.

[0006] In a first aspect, an embodiment of the present application provides an image registration method, which acquires a preoperative image and an intraoperative image; rigidly registers the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix; transforms the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image; obtains a corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative image and the intraoperative image; transforms the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registration image.

[0007] In the second aspect, the embodiment of the present application provides an image registration device, including an image acquisition module, a registration module, a sampling module, an acquisition module and a processing module. The image acquisition module is used to acquire preoperative images and intraoperative images; the registration module is used to rigidly register the preoperative images with the intraoperative images to obtain the corresponding initial narrow field of view rigid matrix; the first processing module is used to transform the preoperative images according to the initial narrow field of view rigid matrix to obtain the corresponding rigidly registered preoperative images; the acquisition module is used to obtain the corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative images and intraoperative images; the second processing module is used to transform the preoperative images according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain the target registration image.

[0008] In a third aspect, an embodiment of the present application provides an electronic device, comprising a memory; one or more processors connected to the memory; and one or more applications, wherein the one or more applications are stored in the memory and configured to be executed by the one or more processors, and the one or more applications are configured to execute the image registration method provided in the first aspect above.

[0009] In a fourth aspect, an embodiment of the present application provides a computer-readable storage medium, in which a program code is stored. The program code can be called by a processor to execute the image registration method provided in the first aspect above.

[0010] In a fifth aspect, an embodiment of the present application provides a computer program product. When the computer program product is run on a computer device, the computer device executes the image registration method provided in the first aspect above.

[0011] The solution provided by the present application obtains a preoperative image and an intraoperative image, and rigidly aligns the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix, and transforms the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image, and obtains a corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative image and the intraoperative image, and transforms the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registered image, thereby realizing rigid registration and elastic registration based on the preoperative image and the intraoperative image, registering the preoperative image to the intraoperative image, and improving the registration accuracy of image registration. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0013] Figure 1 A schematic diagram of a scenario of an image registration system provided in an embodiment of the present application is shown.

[0014] Figure 2 A flow chart of an image registration method provided in an embodiment of the present application is shown.

[0015] Figure 3a A schematic diagram of a scene of a preoperative image in the image registration method provided in an embodiment of the present application is shown.

[0016] Figure 3b A schematic diagram of a scene of an intraoperative image in the image registration method provided in an embodiment of the present application is shown.

[0017] Figure 3c A schematic diagram of a scene of a target registration image in the image registration method provided in an embodiment of the present application is shown.

[0018] Figure 4 Another flow chart of the image registration method provided in the embodiment of the present application is shown.

[0019] Figure 5a A schematic diagram of a scene of a wide FOV elastic deformation field in the image registration method provided in an embodiment of the present application is shown.

[0020] Figure 5b Another schematic diagram of a scene of a wide FOV elastic deformation field in the image registration method provided in an embodiment of the present application is shown.

[0021] Figure 5c Another schematic diagram of a wide FOV elastic deformation field in the image registration method provided in an embodiment of the present application is shown.

[0022] Figure 5d Another schematic diagram of a scene of a wide FOV elastic deformation field in the image registration method provided in an embodiment of the present application is shown.

[0023] Figure 6 A structural block diagram of an image registration device provided in an embodiment of the present application is shown.

[0024] Figure 7 A functional block diagram of an electronic device provided in an embodiment of the present application is shown.

[0025] Figure 8 A computer-readable storage medium provided in an embodiment of the present application and used for storing or carrying a program code for implementing an image registration method provided in an embodiment of the present application is shown.

[0026] Fig. 9 A computer program product provided in an embodiment of the present application and used for storing or carrying program codes for implementing the image registration method provided in an embodiment of the present application is shown. DETAILED DESCRIPTION

[0027] In order to make the purpose, features, and advantages of the invention of this application more obvious and easy to understand, the technical solutions in the embodiments of this application will be clearly and completely described below in conjunction with the drawings in the embodiments of this application. Obviously, the embodiments described below are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0028] It should be understood that when used in this specification and the appended claims, the term "comprising" indicates the presence of described features, integers, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or groups thereof.

[0029] It should also be understood that the terms used in this application specification are only for the purpose of describing specific embodiments and are not intended to limit the application. As used in this application specification and the appended claims, the singular forms "a", "an" and "the" are intended to include plural forms unless the context clearly indicates other working conditions.

[0030] It should be further understood that the term “and / or” used in the specification and appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0031] In addition, in the description of the present application, the terms "first", "second", "third", etc. are only used to distinguish the description and cannot be understood as indicating or implying relative importance.

[0032] In robotic interventional puncture surgery, the registration of preoperative computed tomography (CT) images with intraoperative real-time images is a crucial technology. Preoperative images provide comprehensive anatomical structure and lesion information. Doctors usually locate and analyze the size, shape, location of the patient's lesions and the location of surrounding tissues and organs on the preoperative image sequence to plan interventional puncture surgery. Intraoperative real-time images provide real-time navigation and monitoring for the surgery.

[0033] In order to maintain real-time performance, the field of view (FOV) of intraoperative real-time images is usually very narrow. In contrast, preoperative images have a wider FOV. In puncture surgery, it is often necessary to register the preoperative images with the intraoperative real-time images.

[0034] At present, the registration method of preoperative images and intraoperative real-time images mainly relies on manual operation. However, manual operation requires doctors to make manual adjustments based on their experience, which is time-consuming and depends on the doctor's experience, and is prone to human errors, resulting in low registration accuracy of image registration.

[0035] In response to the above problems, the image registration method, image registration device, electronic device and storage medium provided in the embodiments of the present application obtain preoperative images and intraoperative images, and rigidly register the preoperative images with the intraoperative images to obtain the corresponding initial narrow field of view rigid matrix, and transform the preoperative images according to the initial narrow field of view rigid matrix to obtain the corresponding rigidly registered preoperative images, and obtain the corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative images and intraoperative images, and transform the preoperative images according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain the target registered image, thereby realizing rigid registration and elastic registration based on the preoperative images and the intraoperative images, registering the preoperative images to the intraoperative images, and improving the registration accuracy of image registration.

[0036] The technical solutions in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application.

[0037] See also Figure 1 , which shows a schematic diagram of an application scenario of the image registration system provided in an embodiment of the present application, which may include an image acquisition device 100 and a processing device 200, wherein the image acquisition device 100 is communicatively connected to the processing device 200 and performs data exchange with the processing device 200.

[0038] Among them, the image acquisition device 100 can be used to acquire images of a target object (for example, a person or an animal, etc.). The image acquisition device 100 can be any one of a computed tomography (CT) device, a positron emission tomography (PETCT) device, or a magnetic resonance imaging (MRI) device, etc. The type of the image acquisition device 100 is not limited here, and can be set according to actual needs.

[0039] The processing device 200 may be a terminal device or a server, etc. The type of the processing device 200 is not limited here and may be specifically configured according to actual needs.

[0040] The terminal device can be a mobile terminal device (for example, any one of a mobile phone, a personal digital assistant (PDA), a tablet personal computer (Tablet PC), a laptop computer, a smart watch, a smart bracelet or a wearable device, etc.), or a fixed terminal device (desktop computer, smart panel, etc.), etc.

[0041] The server can be an independent physical server, a server cluster or distributed system composed of multiple physical servers, or any one of the cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDN), big data or artificial intelligence platforms, etc.

[0042] In some embodiments, the image acquisition device 100 can be integrated with the processing device 200 to form an integrated device. For example, the processing device 200 can be embedded in the image acquisition device 100 and form an integrated device with the image acquisition device 100. The image acquisition device 100 can also be separately and independently set from the processing device 200. The image acquisition device 100 is connected to the processing device 200 via a cable or a wireless network, and data is exchanged with the processing device 200 via a cable or a wireless network.

[0043] Among them, the wireless network can be any one of a ZigBee network, a Bluetooth (BT) network, a Wireless Fidelity (Wi-Fi) network, a Long Range Radio (LoRa) network, a Low-Power Wide-Area Network (LPWAN), an infrared network or a Wireless Personal Area Network (WPAN), etc. The type of wireless network is not limited here and can be set according to actual needs.

[0044] See also Figure 2 , which shows a flow chart of an image registration method provided by an embodiment of the present application. In a specific embodiment, the image registration method can be applied to Figure 1 The processing device 200 in the image registration system shown in FIG. Figure 2 The process shown is described in detail, and the image registration method may include the following steps S110 to S150.

[0045] Step S110: Acquire preoperative images and intraoperative images.

[0046] In the embodiment of the present application, when the user needs to perform image registration, a registration instruction can be sent to the processing device, and the processing device receives and responds to the registration instruction to obtain preoperative images and intraoperative images. Among them, the preoperative images and intraoperative images are acquired by the imaging device on the same machine and bed, and the preoperative images and intraoperative images are composed of multiple slices in a three-axis (x-axis, y-axis and z-axis) coordinate system, the x-axis is the front to back direction axis when the target object is facing up and back down, the y-axis is the right to left direction axis when the target object is facing up and back down, and the z-axis is the direction axis from feet to head when the target object is facing up and back down.

[0047] In some embodiments, the imaging device pre-stores preoperative imaging images, which are imaging images captured by the imaging device on the target object before the operation.

[0048] When the user needs to perform image registration, a registration instruction can be sent to the processing device, the processing device receives and responds to the registration instruction, sends an image acquisition instruction to the imaging device, the imaging device receives and responds to the image acquisition instruction, performs intraoperative image acquisition on the target object during surgery, obtains intraoperative image images, and sends preoperative image images and intraoperative image images to the processing device, and the processing device receives the preoperative image images and intraoperative image images sent by the imaging device.

[0049] In some embodiments, when a user needs to perform image registration, a registration instruction can be sent to a processing device. The processing device receives and responds to the registration instruction and generates corresponding image upload prompt information. The image upload prompt information is used to prompt the user to upload preoperative images and intraoperative images to the processing device. The processing device receives the preoperative images and intraoperative images uploaded by the user.

[0050] Among them, the image upload prompt information can be at least any one of text image upload prompt information, sound image upload prompt information or light image upload prompt information, etc. The type of image upload prompt information is not limited here and can be set according to actual needs.

[0051] In some embodiments, the processing device may include an input panel. When a user needs to perform image registration, the user may input a registration instruction on the input panel of the electronic device, and the processing device receives the registration instruction through the input panel.

[0052] In some embodiments, the processing device may include a speech recognition module. When a user requires image registration, a speech message carrying a registration instruction may be sent within the speech collection range of the speech recognition module. The processing device collects the speech information through the speech recognition module, and performs speech recognition on the collected speech information to obtain the registration instruction.

[0053] In some implementations, the image registration system may further include a client, which is connected to the processing device via a wireless network and performs data exchange with the client via the wireless network.

[0054] When the user needs to perform image registration, a registration instruction can be sent to the client. The client receives and responds to the registration instruction and forwards the registration instruction to the processing device via the wireless network. The processing device receives the registration instruction forwarded by the client.

[0055] Among them, the client can be any one of a mobile client (for example, a mobile phone client, a personal digital assistant (PDA) client, a tablet computer (Tablet Personal Computer, Tablet PC) client, a laptop computer client, a smart watch client, a smart bracelet client or a wearable client, etc.) or a fixed client (for example, a desktop computer client, a smart panel client, etc.). The type of client is not limited here and can be set according to actual needs.

[0056] Step S120: rigidly register the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix.

[0057] In an embodiment of the present application, after acquiring the preoperative image and the intraoperative image, the processing device may rigidly register the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view (FOV) rigid matrix.

[0058] Specifically, after acquiring the preoperative image and the intraoperative image, the processing device can preprocess the preoperative image to obtain the corresponding local preoperative image, and perform local rigid registration between the local preoperative image and the intraoperative image to obtain the corresponding local narrow FOV rigid matrix, and according to the first corresponding position of the local preoperative image in the preoperative image, restore the local narrow FOV rigid matrix to the first corresponding position corresponding to the global narrow FOV rigid matrix corresponding to the preoperative image to obtain the initial narrow FOV rigid matrix.

[0059] In some embodiments, the processing device can obtain the alignment matrix of the preoperative image and the intraoperative image, and transform the preoperative image based on the alignment matrix and the interpolation calculation method to obtain the corresponding transformed preoperative image, and crop the transformed preoperative image according to the preset rules to obtain the corresponding local preoperative image, thereby realizing FOV cropping of the preoperative image, reducing the impact of non-target areas in the preoperative image on the registration accuracy of the rigid registration of the preoperative image and the intraoperative image, and improving the registration accuracy of the rigid registration of the preoperative image and the intraoperative image.

[0060] Among them, the preset rules can be used to characterize the selection rules of image slices. For example, the preset rule can be a rule that takes a certain amount of redundant slices in the z-axis direction and below the image based on the initial alignment slices of the transformed preoperative image; the preset rule can also be a rule that takes a certain amount of redundant slices in the x-axis direction and below the image based on the initial alignment slices of the transformed preoperative image; the preset rule can also be a rule that takes a certain amount of redundant slices in the y-axis direction and below the image based on the initial alignment slices of the transformed preoperative image, etc.; the initial alignment slice is the image slice corresponding to the intraoperative image in the transformed preoperative image. The type of preset rule is not limited here, and can be set according to actual needs.

[0061] In some embodiments, the preoperative image and the intraoperative image have the same reference coordinate system (Frame of Reference UID), so the coordinate position information recorded during scanning by the imaging device, such as the image position (Image Position) information and image orientation (Image Orientation) information recorded in Digital Imaging and Communications in Medicine (DICOM), can be used to obtain the alignment matrix of the preoperative image and the intraoperative image.

[0062] The processing device can obtain the first coordinate of the preoperative image and the second coordinate of the intraoperative image, and obtain the alignment matrix of the preoperative image and the intraoperative image according to the first coordinate and the second coordinate. The first coordinate and the second coordinate can be the starting scanning point coordinate in the preoperative image and the starting scanning point coordinate in the intraoperative image, respectively. The first coordinate and the second coordinate can also be the coordinates of the same feature point in the preoperative image and the coordinates of the same feature point in the intraoperative image, etc.

[0063] In some implementations, the processing device may perform local rigid registration of the local preoperative image with the intraoperative image based on an image similarity algorithm to obtain a corresponding local narrow FOV rigid matrix.

[0064] Step S130: transforming the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image.

[0065] In an embodiment of the present application, after the processing device rigidly aligns the preoperative image with the intraoperative image to obtain the corresponding initial narrow FOV rigid matrix, it can transform the preoperative image according to the initial narrow FOV rigid matrix to obtain the corresponding rigidly registered preoperative image.

[0066] Step S140: Obtaining a corresponding initial narrow field-of-view elastic deformation field according to the rigid registration of the preoperative image and the intraoperative image.

[0067] In an embodiment of the present application, after the processing device samples the preoperative image according to the initial narrow FOV rigid matrix and obtains the corresponding rigidly registered preoperative image, it can obtain the corresponding initial narrow FOV elastic deformation field according to the rigidly registered preoperative image and the intraoperative image.

[0068] Specifically, after the processing device transforms the preoperative image according to the initial narrow FOV rigid matrix to obtain the corresponding rigidly registered preoperative image, it can elastically register the rigidly registered preoperative image with the intraoperative image to obtain the corresponding initial global narrow FOV elastic deformation field, determine the registration region of interest, and based on the registration region of interest, crop the rigidly registered preoperative image to obtain a local rigidly registered preoperative image, crop the intraoperative image to obtain a local intraoperative image, and elastically register the local rigidly registered preoperative image with the local intraoperative image to obtain the corresponding initial local narrow FOV elastic deformation field, and obtain the corresponding initial narrow FOV elastic deformation field based on the initial global narrow FOV elastic deformation field and the initial local narrow FOV elastic deformation field.

[0069] In some embodiments, the processing device can input the rigidly registered preoperative image and the intraoperative image into the target image registration network model, the target image registration network model receives and responds to the rigidly registered preoperative image and the intraoperative image, elastically registers the rigidly registered preoperative image and the intraoperative image, obtains and outputs the corresponding initial global narrow FOV elastic deformation field to the processing device, and the processing device receives the initial global narrow field of view elastic deformation field output by the target image registration network model, realizes elastic registration of the rigidly registered preoperative image and the intraoperative image based on the target image registration network model, and improves the elastic registration speed of the rigidly registered preoperative image and the intraoperative image. Among them, the target image registration network model is obtained based on the lightweight network model.

[0070] In some embodiments, before inputting rigidly registered preoperative images and intraoperative images into the target image registration network model, the processing device may input historical preoperative images and historical intraoperative images into a lightweight network model for training to obtain an initial image registration network model, and perform inference acceleration processing on the initial image registration network model based on a graphics processing unit (Graphics Processing Unit, GPU) to obtain a target image registration network model, thereby achieving inference acceleration of the target image registration network model and improving the registration speed of the target image registration network model.

[0071] Among them, the historical preoperative images are reference images, and the historical intraoperative images are moving images. During the training process of the lightweight network model, the output is a three-channel registration deformation field of the moving image registered to the reference image. In the feature extraction process of the lightweight network model, considering that the historical preoperative images and historical intraoperative images have a narrow FOV in the z-axis direction and fewer slices, the convolution downsampling only performs downsampling operations in the x-axis and y-axis directions, and the z-axis direction keeps the original size unchanged, and the features of different layers are fused. The lightweight network model uses image-based similarity measurement parameters (mutual information, cross-correlation, etc.) and regularized constraint loss to train the lightweight network model until convergence.

[0072] In some embodiments, the processing device can obtain the puncture length of the puncture needle, obtain a puncture circular area with the needle tip of the puncture needle as the center and the puncture length as the radius, and use the minimum positive circumscribed rectangle of the puncture circular area as the registration interest region.

[0073] The processing device may calculate the puncture length of the current puncture needle entering the target object based on the located needle tip position of the puncture needle, and the region of interest is continuously updated and enlarged as the depth of the needle tip entering the target object.

[0074] In some embodiments, the processing device can restore the initial local narrow FOV elastic deformation field to the second position of the global narrow FOV elastic deformation field corresponding to the rigidly registered preoperative image image according to the second corresponding position of the local rigidly registered preoperative image in the rigidly registered preoperative image, obtain the current local narrow FOV elastic deformation field, and fuse the current local narrow FOV elastic deformation field and the initial global narrow FOV elastic deformation field to obtain the corresponding initial narrow FOV elastic deformation field.

[0075] In some embodiments, the processing device can input the local rigid registration preoperative image and the local intraoperative image into the target image registration network model, the target image registration network model receives and responds to the local rigid registration preoperative image and the local intraoperative image, elastically registers the local rigid registration preoperative image with the local intraoperative image, obtains and outputs the corresponding initial local narrow FOV elastic deformation field to the processing device, and receives the initial local narrow FOV elastic deformation field output by the target image registration network model, thereby realizing elastic registration of the local rigid registration preoperative image and the local intraoperative image based on the target image registration network model, and improving the elastic registration speed of the local rigid registration preoperative image and the local intraoperative image.

[0076] Step S150: transforming the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registration image.

[0077] In an embodiment of the present application, after the processing device obtains the corresponding initial narrow FOV elastic deformation field based on the rigid registration of the preoperative image and the intraoperative image, it can transform the preoperative image according to the initial narrow FOV rigid matrix and the initial narrow FOV elastic deformation field to obtain the target registration image, thereby realizing rigid registration and elastic registration based on the preoperative image and the intraoperative image, registering the preoperative image to the intraoperative image, and improving the registration accuracy of the image registration.

[0078] Specifically, after the processing device obtains the corresponding initial narrow FOV elastic deformation field based on the rigid registration of the preoperative image and the intraoperative image, it can perform global processing on the initial narrow FOV rigid matrix to obtain the corresponding initial wide FOV rigid matrix, and perform global processing on the initial narrow FOV elastic deformation field to obtain the corresponding initial wide FOV elastic deformation field, and based on the initial wide FOV rigid matrix and the initial wide FOV elastic deformation field, perform transformation processing on the preoperative image to obtain the target registration image.

[0079] The processing device may process the translation amount in the initial narrow FOV rigid matrix according to the alignment matrix of the preoperative image and the intraoperative image to obtain the corresponding initial wide FOV rigid matrix.

[0080] The processing device can restore the initial narrow FOV elastic deformation field to the position corresponding to the wide FOV elastic deformation field of the preoperative image according to the alignment matrix of the preoperative image and the intraoperative image, and obtain the corresponding initial wide FOV elastic deformation field.

[0081] In the initial wide FOV elastic deformation field, only the area corresponding to the initial narrow FOV elastic deformation field has a value, and the values ​​of other areas are 0; the narrow FOV elastic deformation field corresponds to the intraoperative image, and the wide FOV elastic deformation field corresponds to the preoperative image.

[0082] In one application scenario, the preoperative imaging image is a wide FOV real-time CT image, such as Figure 3a As shown in Figure 2, the intraoperative imaging images are narrow FOV real-time CT images. Figure 3b As shown in FIG. 1 , based on the initial wide FOV rigid matrix and the initial wide FOV elastic deformation field corresponding to the wide FOV real-time CT image and the narrow FOV real-time CT image, the wide FOV real-time CT image is transformed and processed, and the target registration image obtained is as shown in FIG. Figure 3c As shown in FIG. 1 , the region images corresponding to the initial wide FOV rigid matrix and the initial wide FOV elastic deformation field in the target registration image are from the narrow FOV real-time CT image, and the other region images are from the wide FOV real-time CT image.

[0083] The solution provided in this embodiment obtains a preoperative image and an intraoperative image, and rigidly aligns the preoperative image with the intraoperative image to obtain a corresponding initial narrow FOV rigid matrix, and transforms the preoperative image according to the initial narrow FOV rigid matrix to obtain a corresponding rigidly registered preoperative image, and obtains a corresponding initial narrow FOV elastic deformation field according to the rigidly registered preoperative image and the intraoperative image, and transforms the preoperative image according to the initial narrow FOV rigid matrix and the initial narrow FOV elastic deformation field to obtain a target registered image, thereby realizing rigid registration and elastic registration based on the preoperative image and the intraoperative image, registering the preoperative image to the intraoperative image, and improving the registration accuracy of image registration.

[0084] See also Figure 4 , which shows a flow chart of an image registration method provided by another embodiment of the present application. In a specific embodiment, the image registration method can be applied to Figure 1 The processing device 200 in the image registration system shown in FIG. Figure 4 The process shown is described in detail, and the image registration method may include the following steps S210 to S290.

[0085] Step S210: Acquire preoperative images and intraoperative images.

[0086] Step S220: rigidly register the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix.

[0087] Step S230: transforming the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image.

[0088] Step S240: Obtaining a corresponding initial narrow field-of-view elastic deformation field according to the rigid registration of the preoperative image and the intraoperative image.

[0089] Step S250: performing global processing on the initial narrow field angle rigid matrix to obtain a corresponding initial wide field angle rigid matrix.

[0090] Step S260: performing global processing on the initial narrow field angle elastic deformation field to obtain a corresponding initial wide field angle elastic deformation field.

[0091] In this embodiment, step S210 , step S220 , step S230 , step S240 , step S250 , and step S260 may refer to the contents of the corresponding steps in the aforementioned embodiments, and will not be described in detail here.

[0092] Step S270: performing edge expansion on a region corresponding to the initial narrow field angle elastic deformation field in the initial wide field angle elastic deformation field to obtain an initial expanded wide field angle elastic deformation field.

[0093] In this embodiment, since the initial wide FOV elastic deformation field is discontinuous in the FOV connecting part, that is, there is a smoothness loss, the processing device can perform edge expansion on the area corresponding to the initial narrow FOV elastic deformation field in the initial wide FOV elastic deformation field to obtain the initial expanded wide FOV elastic deformation field, thereby ensuring that the initial expanded wide FOV elastic deformation field is continuous in the FOV connecting part.

[0094] The edge expansion is performed on the region corresponding to the initial narrow FOV elastic deformation field in the initial wide FOV elastic deformation field, that is, the edge expansion is performed on the region having a value in the initial wide FOV elastic deformation field.

[0095] Step S280: fusing the weight map generated based on the Gaussian smoothing function and the initial expanded wide-field-angle elastic deformation field to obtain a target wide-field-angle elastic deformation field.

[0096] In this embodiment, after the processing device performs edge expansion on the initial wide FOV elastic deformation field to obtain the initial expanded wide FOV elastic deformation field, it can fuse the weight map generated based on the Gaussian smoothing function and the initial expanded wide FOV elastic deformation field to obtain the target wide FOV elastic deformation field, thereby reducing the smoothness loss of the target wide FOV elastic deformation field.

[0097] Among them, the area data corresponding to the original narrow FOV of the target wide FOV elastic deformation field remains unchanged, and the expanded FOV area value is 0.

[0098] Step S290: Based on the initial wide field angle rigid matrix and the target wide field angle elastic deformation field, the preoperative image is transformed to obtain a target registration image.

[0099] In this embodiment, after fusing the weight map generated based on the Gaussian weight rule and the initial expanded wide FOV elastic deformation field to obtain the target wide FOV elastic deformation field, the processing device can transform the preoperative image based on the initial wide FOV rigid matrix and the target wide FOV elastic deformation field to obtain the target registration image, thereby realizing the registration of the preoperative image based on the initial wide FOV rigid matrix and the target wide FOV elastic deformation field, reducing the smoothness loss of the target wide FOV elastic deformation field, and further improving the registration accuracy of the image registration.

[0100] In one application scenario, such as Figure 5aAs shown in , the initial wide FOV elastic deformation field is obtained by global processing of the initial narrow FOV elastic deformation field. In the initial wide FOV elastic deformation field, only the area corresponding to the initial narrow FOV elastic deformation field has a value, and the other areas are 0. The initial wide FOV elastic deformation field is discontinuous in the FOV connection part. After the initial wide FOV elastic deformation field is edge expanded, the initial expanded wide FOV elastic deformation field is obtained, as shown in Figure 5b As shown, the weight graph generated by the Gaussian weight rule is as follows Figure 5c As shown in , the weight map and the initial expanded wide FOV elastic deformation field are fused to obtain the target wide FOV elastic deformation field, as shown in Figure 5d shown.

[0101] The solution provided in this embodiment obtains a preoperative image and an intraoperative image, and rigidly aligns the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view angle rigid matrix, and transforms the preoperative image according to the initial narrow field of view angle rigid matrix to obtain a corresponding rigidly registered preoperative image, and obtains a corresponding initial narrow field of view angle elastic deformation field according to the rigidly registered preoperative image and the intraoperative image, and performs global processing on the initial narrow field of view angle rigid matrix to obtain a corresponding initial wide field of view angle rigid matrix, and performs global processing on the initial narrow field of view angle elastic deformation field to obtain a corresponding initial wide field of view angle The method further comprises the steps of: generating an elastic deformation field of the initial wide field of view angle, expanding the edges of the area corresponding to the initial narrow field of view angle in the initial wide field of view angle, obtaining the initial expanded wide field of view angle elastic deformation field, fusing the weight map generated based on the Gaussian smoothing function and the initial expanded wide field of view angle elastic deformation field, obtaining the target wide field of view angle elastic deformation field, and transforming the preoperative image based on the initial wide field of view angle rigid matrix and the target wide field of view angle elastic deformation field to obtain the target registration image, realizing the rigid registration and elastic registration based on the preoperative image and the intraoperative image, registering the preoperative image to the intraoperative image, and improving the registration accuracy of the image registration.

[0102] Furthermore, based on the initial wide FOV rigid matrix and the target wide FOV elastic deformation field, the preoperative imaging images are registered, which reduces the smoothness loss of the target wide FOV elastic deformation field and further improves the registration accuracy of the image registration.

[0103] See also Figure 6 , which shows an image registration device 300 provided by an embodiment of the present application. The image registration device 300 can be applied to Figure 1 The processing device 200 in the image registration system shown in FIG. Figure 6The image registration device 300 shown is described in detail. The image registration device 300 may include an image acquisition module 310 , a registration module 320 , a first processing module 330 , an acquisition module 340 and a second processing module 350 .

[0104] The image acquisition module 310 can be used to acquire preoperative images and intraoperative images; the registration module 320 can be used to rigidly register the preoperative images with the intraoperative images to obtain the corresponding initial narrow field of view rigid matrix; the first processing module 330 can be used to transform the preoperative images according to the initial narrow field of view rigid matrix to obtain the corresponding rigidly registered preoperative images; the acquisition module 340 can be used to obtain the corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative images and intraoperative images; the second processing module 350 can be used to transform the preoperative images according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain the target registered image.

[0105] In some implementations, the registration module 320 may include a pre-processing unit, a first registration unit, and a restoration unit.

[0106] The preprocessing unit can be used to preprocess the preoperative image to obtain the corresponding local preoperative image; the first registration unit can be used to perform local rigid registration on the local preoperative image and the intraoperative image to obtain the corresponding local narrow field of view rigid matrix; the restoration unit can be used to restore the local narrow field of view rigid matrix to the first corresponding position of the global narrow field of view rigid matrix corresponding to the preoperative image according to the first corresponding position of the local preoperative image in the preoperative image, so as to obtain the initial narrow field of view rigid matrix.

[0107] In some implementations, the preprocessing unit may include a first acquisition subunit, a transformation subunit, and a cropping subunit.

[0108] The first acquisition subunit can be used to obtain the alignment matrix of the preoperative image and the intraoperative image; the transformation subunit can be used to transform the preoperative image according to the alignment matrix to obtain the corresponding transformed preoperative image; the cropping subunit can be used to crop the transformed preoperative image according to preset rules to obtain the corresponding local preoperative image.

[0109] In some implementations, the obtaining module 340 may include a second registration unit, a determination unit, a cropping unit, a third registration unit, and a deriving unit.

[0110] The second registration unit can be used to elastically register the rigidly registered preoperative image with the intraoperative image to obtain the corresponding initial global narrow field of view elastic deformation field; the determination unit can be used to determine the registration area of ​​interest; the cropping unit can be used to crop the rigidly registered preoperative image to obtain the local rigidly registered preoperative image, and crop the intraoperative image to obtain the local intraoperative image based on the registration area of ​​interest; the third registration unit can be used to elastically register the local rigidly registered preoperative image with the local intraoperative image to obtain the corresponding initial local narrow field of view elastic deformation field; the obtaining unit can be used to obtain the corresponding initial narrow field of view elastic deformation field based on the initial global narrow field of view elastic deformation field and the initial local narrow field of view elastic deformation field.

[0111] In some embodiments, a obtaining unit may include a reducing subunit and a fusing subunit.

[0112] The restoration subunit can be used to restore the initial local narrow field of view angle elastic deformation field to the second position of the global narrow field of view angle elastic deformation field corresponding to the rigid registration preoperative image image according to the second corresponding position of the local rigid registration preoperative image image in the rigid registration preoperative image image, so as to obtain the current local narrow field of view angle elastic deformation field; the fusion subunit can be used to fuse the current local narrow field of view angle elastic deformation field and the initial global narrow field of view angle elastic deformation field, so as to obtain the corresponding initial narrow field of view angle elastic deformation field.

[0113] In some embodiments, the second registration unit may include a first input subunit and a first receiving subunit.

[0114] The first input subunit can be used to input the rigidly registered preoperative image and intraoperative image into the target image registration network model, so that the target image registration network model elastically registers the rigidly registered preoperative image and the intraoperative image to obtain and output the corresponding initial global narrow field of view elastic deformation field; the first receiving subunit can be used to receive the initial global narrow field of view elastic deformation field output by the target image registration network model.

[0115] In some implementations, the image registration apparatus 300 may further include an input module and an inference acceleration processing module.

[0116] The input module can be used to input historical preoperative images and historical intraoperative images into a lightweight network model for training before the first input unit inputs the rigidly registered preoperative images and intraoperative images into the target image registration network model to obtain an initial image registration network model; the inference acceleration processing module can be used to perform inference acceleration processing on the initial image registration network model to obtain a target image registration network model.

[0117] In some implementations, the determining unit may include a second obtaining subunit, a third obtaining subunit, and an acting subunit.

[0118] The second acquisition subunit can be used to acquire the puncture length of the puncture needle; the third acquisition subunit can be used to acquire the puncture circular area with the needle tip of the puncture needle as the center and the puncture length as the radius; the as subunit can be used to use the minimum positive circumscribed rectangle of the puncture circular area as the registration interest area.

[0119] In some embodiments, the third registration unit may include a second input subunit and a second receiving subunit.

[0120] The second input subunit can be used to input the local rigid registration preoperative image and the local intraoperative image into the target image registration network model, so that the target image registration network model elastically registers the local rigid registration preoperative image with the local intraoperative image to obtain and output the corresponding initial local narrow field of view angle elastic deformation field; the second receiving subunit can be used to receive the initial local narrow field of view angle elastic deformation field output by the target image registration network model.

[0121] In some implementations, the processing module 350 may include a first global processing unit, a second global processing unit, and a processing unit.

[0122] The first global processing unit can be used to perform global processing on the initial narrow field of view rigid matrix to obtain the corresponding initial wide field of view rigid matrix; the second global processing unit can be used to perform global processing on the initial narrow field of view elastic deformation field to obtain the corresponding initial wide field of view elastic deformation field; the processing unit can be used to transform the preoperative image based on the initial wide field of view rigid matrix and the initial wide field of view elastic deformation field to obtain the target registration image.

[0123] In some implementations, the image registration apparatus 300 may further include an edge expansion module and a fusion module.

[0124] The edge expansion module can be used for the processing unit to transform the preoperative image based on the initial wide field of view rigid matrix and the initial wide field of view elastic deformation field, and before obtaining the target registration image, the edge expansion is performed on the area corresponding to the initial narrow field of view elastic deformation field in the initial wide field of view elastic deformation field to obtain the initial expanded wide field of view elastic deformation field; the fusion module can be used to fuse the weight map generated based on the Gaussian smoothing function and the initial expanded wide field of view elastic deformation field to obtain the target wide field of view elastic deformation field.

[0125] In some embodiments, the processing unit may include a first processing sub-unit.

[0126] The first processing subunit can be used to transform the preoperative image based on the initial wide field angle rigid matrix and the target wide field angle elastic deformation field to obtain the target registration image.

[0127] In some implementations, the first global processing unit may include a second processing sub-unit.

[0128] The second processing subunit may be used to process the translation amount in the initial narrow field angle rigid matrix according to the alignment matrix of the preoperative image and the intraoperative image to obtain the corresponding initial wide field angle rigid matrix.

[0129] In some implementations, the second global processing unit may include a third processing sub-unit.

[0130] The third processing subunit can be used to restore the initial narrow field angle elastic deformation field to the position corresponding to the wide field angle elastic deformation field of the preoperative image according to the alignment matrix of the preoperative image and the intraoperative image, and obtain the corresponding initial wide field angle elastic deformation field.

[0131] The solution provided in this embodiment obtains a preoperative image and an intraoperative image, and rigidly aligns the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix, and transforms the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image, and obtains a corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative image and the intraoperative image, and transforms the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registered image, thereby realizing rigid registration and elastic registration based on the preoperative image and the intraoperative image, registering the preoperative image to the intraoperative image, and improving the registration accuracy of image registration.

[0132] It should be noted that each embodiment in this specification is described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same and similar parts between the embodiments can be referred to each other. For the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment. Any processing method described in the method embodiment can be implemented by the corresponding processing module in the device embodiment, and will not be repeated one by one in the device embodiment.

[0133] In addition, each functional module in each embodiment of the present application can be integrated into a processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The above integrated modules can be implemented in the form of hardware or software functional modules.

[0134] See also Figure 7 , which shows a functional block diagram of an electronic device 600 provided by an embodiment of the present application, and the electronic device 600 may include one or more of the following components: a memory 610, a processor 620, and one or more applications, wherein the one or more applications may be stored in the memory 610 and configured to be executed by the one or more processors 620, and the one or more applications are configured to execute the method described in the aforementioned method embodiment.

[0135] The memory 610 may include a random access memory (RAM) or a read-only memory (ROM). The memory 610 may be used to store instructions, programs, codes, code sets or instruction sets.The memory 610 may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for implementing at least one function (such as obtaining a preoperative image, obtaining an intraoperative image, rigidly registering a preoperative image with an intraoperative image, obtaining an initial narrow field of view rigid matrix, transforming and processing a preoperative image, obtaining a rigidly registered preoperative image, obtaining an initial narrow field of view elastic deformation field, processing a preoperative image, obtaining a target registration image, preprocessing a preoperative image, obtaining a local preoperative image, locally rigidly registering a local preoperative image with an intraoperative image, obtaining a local narrow field of view rigid matrix, restoring a local narrow field of view rigid matrix, obtaining an initial narrow field of view elastic deformation field, processing a preoperative image, obtaining a target registration image, preprocessing a preoperative image, obtaining a local preoperative image, locally rigidly registering a local preoperative image with an intraoperative image, obtaining a local narrow field of view rigid matrix, restoring a local narrow field of view rigid matrix, obtaining an initial narrow field of view Rigid matrix, obtain alignment matrix, transform and process preoperative image, obtain transformed preoperative image, crop sampled and changed preoperative image, obtain local preoperative image, elastically register rigidly register preoperative image and intraoperative image, obtain initial global narrow field of view elastic deformation field, determine registration area of ​​interest, crop rigidly register preoperative image, obtain local rigidly register preoperative image, crop intraoperative image, obtain local intraoperative image, elastically register local rigidly register preoperative image and local intraoperative image, obtain initial local narrow field of view elastic deformation field, obtain initial narrow field of view elastic deformation field, restore initial local narrow field of view elastic deformation field, obtain current local narrow field of view elastic deformation field, Fusion of the current local narrow field of view elastic deformation field and the initial global narrow field of view elastic deformation field, input of rigid registration preoperative image, input of intraoperative image, obtain of the initial global narrow field of view elastic deformation field, output of the initial global narrow field of view elastic deformation field, reception of the initial global narrow field of view elastic deformation field, input of historical preoperative image, input of historical intraoperative image, training of lightweight network model, obtain of the initial image registration network model, inference acceleration processing of the initial image registration network model, obtain of the target image registration network model, obtain of the puncture length, obtain of the puncture circular area as the registration interest area, input of local rigid registration preoperative image, input of local intraoperative image, obtain of the initial local narrow field of view elastic deformation field, outputting an initial local narrow field of view elastic deformation field, receiving an initial local narrow field of view elastic deformation field, globally processing an initial narrow field of view rigid matrix, obtaining an initial wide field of view rigid matrix, globally processing an initial narrow field of view elastic deformation field, obtaining an initial wide field of view elastic deformation field, edge expanding an initial wide field of view elastic deformation field, obtaining an initial expanded wide field of view elastic deformation field, fusing a weight map and an initial expanded wide field of view elastic deformation field, obtaining a target wide field of view elastic deformation field, processing the translation amount in the initial narrow field of view rigid matrix, obtaining an initial wide field of view rigid matrix, restoring an initial narrow field of view elastic deformation field, and obtaining an initial wide field of view elastic deformation field, etc.), instructions for implementing the following various method embodiments, etc.

[0136] The storage data area may also store data created by the electronic device 600 during use (such as preoperative images, intraoperative images, initial narrow field angle rigid matrix, rigid registration preoperative images, initial narrow field angle elastic deformation field, target registration images, local preoperative images, local narrow field angle rigid matrix, first corresponding position, global narrow field angle rigid matrix, alignment matrix, transformed preoperative images, preset rules, local preoperative images, initial global narrow field angle elastic deformation field, registration interest region, local rigid registration preoperative images, local Intraoperative image, initial local narrow field of view elastic deformation field, second position, global narrow field of view elastic deformation field, current local narrow field of view elastic deformation field, target image registration network model, historical preoperative image, historical intraoperative image, lightweight network model, puncture needle, puncture length, needle tip, center of circle, radius puncture circular area, minimum positive circumscribed rectangle, initial wide field of view rigid matrix, initial wide field of view elastic deformation field, initial expanded wide field of view elastic deformation field, Gaussian smoothing function, weight map, target wide field of view elastic deformation field and translation) etc.

[0137] The processor 620 may include one or more processing cores. The processor 620 uses various interfaces and lines to connect various parts of the entire electronic device 600, and executes various functions and processes data of the electronic device 600 by running or executing instructions, programs, code sets or instruction sets stored in the memory 610, and calling data stored in the memory 610. Optionally, the processor 620 can be implemented in at least one hardware form of digital signal processing (DSP), field-programmable gate array (FPGA), and programmable logic array (PLA). The processor 620 can integrate one or a combination of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes the operating system, user interface, and application programs; the GPU is responsible for rendering and drawing display content; and the modem is used to process wireless communications. It can be understood that the above-mentioned modem may not be integrated into the processor 620, but may be implemented separately through a communication chip.

[0138] Please refer to Figure 8, which shows a structural block diagram of a computer-readable storage medium provided in an embodiment of the present application. The computer-readable storage medium 700 stores a program code 710, which can be called by a processor to execute the method described in the above method embodiment.

[0139] The computer readable storage medium 700 can be an electronic memory such as a flash memory, an EEPROM (electrically erasable programmable read-only memory), an EPROM, a hard disk, or a ROM. Optionally, the computer readable storage medium 700 includes a non-transitory computer-readable storage medium. The computer readable storage medium 700 has storage space for program code 710 that performs any method steps in the above method. These program codes can be read from or written to one or more computer program products. The program code 710 can be compressed, for example, in an appropriate form.

[0140] Please refer to Fig. 9 , which shows a structural block diagram of a computer program product 800 provided in an embodiment of the present application. The computer program product 800 includes a computer program / instruction 810, and the computer program / instruction 810 is stored in a computer-readable storage medium of a computer device. When the computer program product 800 is run on a computer device, a processor of the computer device reads the computer program / instruction 810 from the computer-readable storage medium, and the processor executes the computer program / instruction 810, so that the computer device performs the method described in the above method embodiment.

[0141] The solution provided in this embodiment obtains a preoperative image and an intraoperative image, and rigidly aligns the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix, and transforms the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image, and obtains a corresponding initial narrow field of view elastic deformation field according to the rigidly registered preoperative image and the intraoperative image, and transforms the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registered image, thereby realizing rigid registration and elastic registration based on the preoperative image and the intraoperative image, registering the preoperative image to the intraoperative image, and improving the registration accuracy of image registration.

[0142] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. An image registration method, characterized in that: include: Acquiring a preoperative image and an intraoperative image, wherein the field of view of the preoperative image is wider than the field of view of the intraoperative image; Performing rigid registration of the preoperative image and the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix; According to the initial narrow field of view rigid matrix, the preoperative image is transformed to obtain a corresponding rigidly registered preoperative image; Obtaining a corresponding initial narrow field-of-view elastic deformation field according to the rigidly registered preoperative image and the intraoperative image; According to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field, the preoperative image is transformed to obtain a target registration image.

2. The image registration method according to claim 1, characterized in that: The rigid registration of the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix includes: Preprocessing the preoperative image to obtain a corresponding local preoperative image; Performing local rigid registration of the local preoperative image and the intraoperative image to obtain a corresponding local narrow field of view rigid matrix; According to the first corresponding position of the local preoperative image in the preoperative image, the local narrow field of view angle rigid matrix is ​​restored to the first corresponding position of the global narrow field of view angle rigid matrix corresponding to the preoperative image to obtain an initial narrow field of view angle rigid matrix.

3. The image registration method according to claim 2, characterized in that: The pre-processing of the pre-operative image to obtain a corresponding local pre-operative image includes: Acquire an alignment matrix between the preoperative image and the intraoperative image; According to the alignment matrix, transforming the preoperative image to obtain a corresponding transformed preoperative image; According to preset rules, the transformed preoperative image is cropped to obtain a corresponding local preoperative image.

4. The image registration method according to claim 1, characterized in that: The step of obtaining a corresponding initial narrow field angle elastic deformation field according to the rigid registration preoperative image and the intraoperative image comprises: Elastically registering the rigidly registered pre-operative image with the intra-operative image to obtain a corresponding initial global narrow field-of-view elastic deformation field; Determine the registration region of interest; Based on the registration interest region, cropping the rigid registration pre-operative image to obtain a local rigid registration pre-operative image, and cropping the intra-operative image to obtain a local intra-operative image; Elastically registering the local rigidly registered pre-operative image with the local intra-operative image to obtain a corresponding initial local narrow field of view elastic deformation field; Based on the initial global narrow field angle elastic deformation field and the initial local narrow field angle elastic deformation field, a corresponding initial narrow field angle elastic deformation field is obtained.

5. The image registration method according to claim 4, characterized in that: The obtaining of a corresponding initial narrow field angle elastic deformation field based on the initial global narrow field angle elastic deformation field and the initial local narrow field angle elastic deformation field comprises: According to the second corresponding position of the local rigid registration pre-operative image in the rigid registration pre-operative image, the initial local narrow field of view angle elastic deformation field is restored to the second corresponding position of the global narrow field of view angle elastic deformation field corresponding to the rigid registration pre-operative image, to obtain the current local narrow field of view angle elastic deformation field; The current local narrow field angle elastic deformation field and the initial global narrow field angle elastic deformation field are fused to obtain a corresponding initial narrow field angle elastic deformation field.

6. The image registration method according to claim 4, characterized in that: The elastic registration of the rigidly registered pre-operative image and the intra-operative image to obtain a corresponding initial global narrow field of view elastic deformation field includes: Inputting the rigidly registered preoperative image and the intraoperative image into a target image registration network model, so that the target image registration network model elastically registers the rigidly registered preoperative image with the intraoperative image, and obtains and outputs a corresponding initial global narrow field of view elastic deformation field; Receive the initial global narrow field of view angle elastic deformation field output by the target image registration network model.

7. The image registration method according to claim 6, characterized in that: Before inputting the rigid registration pre-operative image and the intra-operative image into the target image registration network model, the image registration method further includes: Input historical preoperative images and historical intraoperative images into the lightweight network model for training to obtain an initial image registration network model; The initial image registration network model is subjected to inference acceleration processing to obtain a target image registration network model.

8. The image registration method according to claim 4, characterized in that: The determining of the registration region of interest comprises: Obtaining the puncture length of the puncture needle; Obtaining a puncture circular area with the needle tip of the puncture needle as the center and the puncture length as the radius; The minimum positive circumscribed rectangle of the puncture circular area is used as the registration interest region.

9. The image registration method according to claim 4, characterized in that: The elastic registration of the local rigid registration pre-operative image with the local intra-operative image to obtain a corresponding initial local narrow field of view elastic deformation field includes: Inputting the local rigid registration preoperative image and the local intraoperative image into the target image registration network model, so that the target image registration network model elastically registers the local rigid registration preoperative image with the local intraoperative image, and obtains and outputs the corresponding initial local narrow field of view angle elastic deformation field; Receive the initial local narrow field of view angle elastic deformation field output by the target image registration network model.

10. The image registration method according to any one of claims 1 to 9, characterized in that: The step of transforming the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registration image includes: Performing global processing on the initial narrow field angle rigidity matrix to obtain a corresponding initial wide field angle rigidity matrix; Performing global processing on the initial narrow field angle elastic deformation field to obtain a corresponding initial wide field angle elastic deformation field; Based on the initial wide field angle rigid matrix and the initial wide field angle elastic deformation field, the preoperative image is transformed to obtain a target registration image.

11. The image registration method according to claim 10, characterized in that: Before the preoperative image is transformed based on the initial wide field angle rigid matrix and the initial wide field angle elastic deformation field to obtain a target registration image, the image registration method further includes: Performing edge expansion on a region corresponding to the initial narrow field angle elastic deformation field in the initial wide field angle elastic deformation field to obtain an initial expanded wide field angle elastic deformation field; Fusing the weight map generated based on the Gaussian smoothing function and the initial expanded wide-field-angle elastic deformation field to obtain a target wide-field-angle elastic deformation field; The step of transforming the preoperative image based on the initial wide field angle rigid matrix and the initial wide field angle elastic deformation field to obtain a target registration image includes: Based on the initial wide-viewing angle rigidity matrix and the target wide-viewing angle elastic deformation field, the preoperative image is transformed to obtain a target registration image.

12. The image registration method according to claim 10, characterized in that: The globally processing the initial narrow field angle rigidity matrix to obtain the corresponding initial wide field angle rigidity matrix includes: According to the alignment matrix of the preoperative image and the intraoperative image, the translation amount in the initial narrow field angle rigid matrix is ​​processed to obtain the corresponding initial wide field angle rigid matrix.

13. The image registration method according to claim 10, characterized in that: The globally processing the initial narrow field angle elastic deformation field to obtain the corresponding initial wide field angle elastic deformation field includes: According to the alignment matrix of the preoperative image and the intraoperative image, the initial narrow field angle elastic deformation field is restored to a position corresponding to the wide field angle elastic deformation field of the preoperative image to obtain a corresponding initial wide field angle elastic deformation field.

14. An image registration device, characterized in that: include: An image acquisition module, used to acquire preoperative images and intraoperative images; A registration module, used for rigidly registering the preoperative image with the intraoperative image to obtain a corresponding initial narrow field of view rigid matrix; A first processing module, configured to transform the preoperative image according to the initial narrow field of view rigid matrix to obtain a corresponding rigidly registered preoperative image; An acquisition module, used for acquiring a corresponding initial narrow field angle elastic deformation field according to the rigidly registered preoperative image and the intraoperative image; The second processing module is used to transform the preoperative image according to the initial narrow field of view rigid matrix and the initial narrow field of view elastic deformation field to obtain a target registration image.

15. An electronic device, characterized in that: include: Memory; One or more processors connected to the memory; One or more applications, wherein the one or more applications are stored in the memory and are configured to be executed by one or more processors, and the one or more applications are configured to execute the image registration method according to any one of claims 1 to 13.

16. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores program codes, and the program codes can be called by a processor to execute the image registration method according to any one of claims 1 to 13.

Images