Registration method and device of contrast image and CTA image, storage medium and electronic equipment

Abstract

The embodiment of the invention provides a contrastographic image and CTA image registration method and device, a storage medium and electronic equipment. The method comprises the steps of obtaining a first center line of a target blood vessel segment in a key frame angiography image; obtaining a second center line corresponding to the target blood vessel segment in the CTA image, wherein the projection angle of the second center line is consistent with the projection angle of the key frame contrast image; and performing registration processing on the first center line and the second center line by constructing a displacement field to obtain a registration result of the key frame angiography image and the CTA image. According to the embodiment of the invention, by constructing the displacement field and registering the second center line and the first center line, the registration of the cross-modal two-dimensional angiography image and the three-dimensional CTA image is realized, the registration result of the key frame angiography image and the CTA image is obtained, the registration difficulty is reduced, and the registration precision is improved.

Description

Technical Field

[0001] The present disclosure relates to the technical field of image processing, and in particular to a registration method, device, storage medium, and electronic device for angiography and CTA images. Background Art

[0002] Cardiovascular disease is the disease with the highest morbidity and mortality in my country. Currently, percutaneous coronary intervention (PCI) is a very important means of treating coronary heart disease.

[0003] During PCI, coronary angiography (CAG) is routinely used to visualize vascular structure. However, CAG has limitations, making it difficult to visualize plaque structure. To address these issues, CTA (CT Angiography) images can be introduced and coronary angiography images can be registered with them to overcome these limitations.

[0004] In the process of realizing the present invention, the inventors discovered that there are at least the following technical problems in the prior art: the cross-modality registration task of two-dimensional coronary angiography and three-dimensional CTA images is difficult and has low registration accuracy. Summary of the Invention

[0005] The present disclosure provides a registration method, device, storage medium and electronic device for angiography images and CTA images, so as to improve the registration accuracy of angiography images and CTA images.

[0006] According to one aspect of the present disclosure, a method for registering an angiography image with a CTA image is provided, comprising:

[0007] Acquiring a first centerline of a target blood vessel segment in a key-frame angiography image;

[0008] Acquiring a second centerline corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second centerline is consistent with a projection angle of the key frame angiography image;

[0009] By constructing a displacement field, the first center line and the second center line are registered to obtain a registration result of the key frame angiography image and the CTA image.

[0010] Optionally, the first center line and the second center line are registered by constructing a displacement field, including: constructing a first displacement field based on the matching relationship between the key frame angiography image and the key points in the CTA image; deforming the second center line based on the first displacement field to obtain a first registration center line; generating a second displacement field based on the first registration center line and the first center line, and deforming the first registration center line based on the second displacement field to obtain a second registration center line, until the registration end condition is met and the registration process is completed.

[0011] Optionally, the method for determining the matching relationship between key points in the key-frame angiography image and the CTA image includes: identifying a first key point in the key-frame angiography image, and identifying a second key point in the CTA image; wherein, the first key point and the second key point are at least one respectively; extracting point features of the first key point and point features of the second key point; wherein, the point features include at least one of position features and image features; and establishing a matching relationship between the first key point and the second key point based on the point features of the first key point and the point features of the second key point.

[0012] Optionally, a matching relationship between the first key point and the second key point is established based on the point features of the first key point and the point features of the second key point, including: extracting the interaction features corresponding to the first key point and the second key point respectively through the attention mechanism; determining a first key point group and a second key point group having a matching relationship, the second key point group including at least one second key point; the first key point group including at least one first key point; based on the interaction features corresponding to the first key point and the second key point respectively, determining the first key point and the second key point having a matching relationship in the first key point group and the second key point group having a matching relationship.

[0013] Optionally, constructing a first displacement field includes: determining a first displacement vector between key points having a matching relationship in the key frame angiography image and the CTA image; and determining a second displacement vector of other vascular points based on a positional relationship between the key points and other vascular points in the second centerline; and forming the first displacement field based on the first displacement vector and the second displacement vector.

[0014] Optionally, generating a second displacement field based on the first registration centerline and the first centerline includes: using the matching relationship between the key points in the key frame angiography image and the CTA image as a regularization condition, and / or using the similarity function between the first registration centerline and the first centerline as an objective function, determining a third displacement vector for each blood vessel point in the first registration centerline, and forming the second displacement field based on the third displacement vector.

[0015] Optionally, the registration end condition includes at least one of the following: the number of registrations reaches a preset number; the registration metric value of the registration centerline obtained by each registration and the first centerline reaches a convergence state.

[0016] Optionally, the method further includes: obtaining a target registration centerline obtained when the registration end condition is met; superimposing and displaying the target registration centerline on the key frame angiography image; or,

[0017] Based on the displacement fields corresponding to the multiple iterative registration processes, the analysis image corresponding to the CTA image is deformed to obtain a deformed analysis image, and the deformed analysis image is superimposed and displayed on the key frame angiography image;

[0018] The analysis image corresponding to the CTA image includes at least one of the following: a MIP image, a volume reconstruction image, and a vascular abnormality display image reconstructed based on the CTA image.

[0019] According to another aspect of the present disclosure, a device for registering angiographic images and CTA images is provided, characterized by comprising:

[0020] A first centerline acquisition module, configured to acquire a first centerline of a target blood vessel segment in a key frame angiography image;

[0021] a second centerline acquisition module, configured to acquire a second centerline corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second centerline is consistent with a projection angle of the key frame angiography image;

[0022] The registration module performs registration processing on the first center line and the second center line by constructing a displacement field to obtain a registration result of the key frame angiography image and the CTA image.

[0023] According to another aspect of the present disclosure, an electronic device is provided, comprising:

[0024] at least one processor; and

[0025] a memory communicatively connected to the at least one processor; wherein,

[0026] The memory stores a computer program that can be executed by the at least one processor, and the computer program is executed by the at least one processor so that the at least one processor can perform the method for aligning angiography images and CTA images described in any embodiment of the present disclosure.

[0027] According to another aspect of the present disclosure, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the method for registering angiography images and CTA images described in any embodiment of the present disclosure when executed.

[0028] The technical solution of the disclosed embodiment extracts key frame angiography images from angiography image sequences to provide high-quality angiography images for the registration process. By obtaining the first centerline of the target vascular segment in the key angiography image and the second centerline of the corresponding target vascular segment in the CTA image, with the projection angle of the second centerline being consistent with the projection angle of the key angiography image, unnecessary interference is eliminated through the consistency of the projection angles. A two-dimensional second centerline is further obtained through projection. By constructing a displacement field, the second centerline is registered with the first centerline, achieving cross-modal registration of the two-dimensional angiography image with the three-dimensional CTA image, and obtaining the registration result of the key frame angiography image and the CTA image, the registration difficulty is reduced and the registration accuracy is improved.

[0029] It should be understood that the contents described in this section are not intended to identify the key or important features of the embodiments of the present disclosure, nor are they intended to limit the scope of the present disclosure. Other features of the present disclosure will become readily understood through the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present disclosure, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present disclosure. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0031] Figure 1 This is a flow chart of a registration method for angiography images and CTA images provided by an embodiment of the present disclosure;

[0032] Figure 2 This is a flow chart of a registration method for angiography images and CTA images provided by an embodiment of the present disclosure;

[0033] Figure 3 Schematic diagram of a process for determining a matching relationship between key points in a key-frame angiography image and a CTA image provided by an embodiment of the present disclosure;

[0034] Figure 4is a schematic diagram of an image display provided by an embodiment of the present disclosure;

[0035] Figure 5 1 is a schematic structural diagram of a device for registering angiographic images and CTA images provided by an embodiment of the present disclosure;

[0036] Figure 6 It is a structural diagram of an electronic device provided by an embodiment of the present disclosure. DETAILED DESCRIPTION

[0037] In order to enable those skilled in the art to better understand the solutions of the present disclosure, the technical solutions in the embodiments of the present disclosure will be clearly and completely described below in conjunction with the drawings in the embodiments of the present disclosure. Obviously, the embodiments described are only part of the embodiments of the present disclosure, not all of the embodiments. Based on the embodiments of the present disclosure, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present disclosure.

[0038] It should be noted that the terms "first", "second", etc. in the specification and claims of the present disclosure and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that the numbers used in this way can be interchanged where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0039] Figure 1 This is a flow chart of a registration method for angiography images and CTA images provided by an embodiment of the present disclosure. This embodiment is applicable to situations where angiography images and CTA images are efficiently and accurately registered. The method can be executed by an angiography image and CTA image registration device. The angiography image and CTA image registration device can be implemented in the form of hardware and / or software. The angiography image and CTA image registration device can be configured in one or more electronic devices such as a terminal device, a computer device, a server, an image acquisition device equipped with a processor, and a surgical robot. The terminal device can include but is not limited to a mobile phone and a tablet computer. Figure 1 As shown, the method includes:

[0040] S110 , obtaining a first centerline of a target blood vessel segment in a key-frame angiography image.

[0041] S120 , obtaining a second center line corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second center line is consistent with a projection angle of the key frame angiography image.

[0042] S130 , performing registration processing on the first center line and the second center line by constructing a displacement field to obtain a registration result of the key frame angiography image and the CTA image.

[0043] In the disclosed embodiments, keyframe angiographic images and CTA images of a target object are acquired. The target object can be a human or an animal, without limitation. The angiographic image sequence and CTA image sequence of the target object can be acquired separately using corresponding acquisition devices, read from a database based on identification information corresponding to the target object, or imported from an external storage device. The method for acquiring the keyframe angiographic images and CTA images is not limited herein.

[0044] The key-frame angiography image and the CTA image include the same target blood vessel segment. In some embodiments of the present disclosure, the target blood vessel segment may be a coronary artery, or a local blood vessel segment in the coronary artery.

[0045] A keyframe angiographic image is a frame of angiographic image in an angiographic image sequence. As can be understood, taking the target vascular segment as a coronary artery or a localized segment thereof as an example, the acquisition process for an angiographic image sequence is as follows: when a contrast agent is injected into the target subject, the contrast agent rapidly fills the coronary artery along with the blood flow, causing it to be visualized under X-rays. The image acquisition system of the cardiovascular angiography machine is activated, and angiographic images of the coronary arteries are acquired at different projection angles and time phases according to preset procedures and parameters. In other words, through the above acquisition method, an angiographic image sequence can be acquired, which includes multiple frames of angiographic images corresponding to multiple time points in the time series.

[0046] Optionally, the key-frame angiography image can be the one with the highest vascular clarity in the angiography image sequence. For example, an image quality assessment model can be used to evaluate the clarity of the target vascular segment in each angiography image frame in the angiography image sequence to obtain a clarity index for the target vascular segment corresponding to each angiography image frame. The clarity index corresponding to each angiography image frame is then compared. The clarity index can be a numerical value between 0 and 1, with the magnitude of the numerical value representing clarity. For example, a larger numerical value indicates a clearer target vascular segment in the image. The angiography image frame with the highest clarity index is then selected as the key-frame angiography image. For example, multiple angiography image frames within the angiography image sequence can be displayed. In response to a user selecting any angiography image, the selected angiography image is designated as the key-frame angiography image. Selecting the key-frame angiography image with the highest target vascular segment clarity facilitates accurate registration of the key-frame angiography image with the CTA image.

[0047] Optionally, the key-frame angiography image can be the angiography image frame in the angiography image sequence that has the highest similarity to the CTA image. To simplify the similarity matching process, the first centerline of the target vascular segment in each angiography image frame in the angiography image sequence is extracted, and the second centerline of the corresponding target vascular segment in the CTA image is extracted. A similarity metric is determined between the first and second centerlines of the target vascular segment in each angiography image frame, and the angiography image frame corresponding to the maximum similarity metric is identified as the key-frame angiography image. It will be appreciated that CTA images are three-dimensional images, and accordingly, the vascular centerline of the CTA image is a three-dimensional centerline. The second centerline is obtained by projecting the three-dimensional centerline of the CTA image based on the projection angle corresponding to the angiography image sequence. Alternatively, the CTA image can be projected based on the projection angle corresponding to the angiography image sequence to obtain a projected image, and the vascular centerline is extracted from the projected image to obtain the second centerline of the target vascular segment. The projection angle of the second centerline is consistent with the projection angle of the key-frame angiography image, facilitating accurate registration of the key-frame angiography image and the CTA image.

[0048] In the above embodiments, the first centerline, the second centerline of any angiography image or the three-dimensional centerline of a CTA image can be obtained based on a vascular centerline extraction algorithm, which includes but is not limited to a machine learning algorithm.

[0049] In this embodiment, the keyframe angiography image and the CTA image are registered by performing registration processing on the first centerline and the second centerline, thereby achieving registration of the keyframe angiography image and the CTA image, and obtaining a registration result of the keyframe angiography image and the CTA image. It is understandable that the three-dimensional centerline of the CTA image includes multiple vascular points, each of which is provided with an identifier. The second centerline of the target vascular segment corresponding to the CTA image includes multiple vascular points, and the multiple vascular points on the second centerline are projected vascular points of the vascular points on the three-dimensional centerline. Accordingly, the identifiers of the vascular points at the same position corresponding to the three-dimensional centerline and the second centerline are consistent. When the first centerline and the second centerline of the keyframe angiography image are registered, the registration result of the keyframe angiography image and the CTA image can be obtained based on the mapping relationship between the second centerline and the identifiers of the vascular points in the CTA image.

[0050] In some embodiments of the present disclosure, the first center line and the second center line are aligned by constructing a displacement field, wherein the alignment process can be a multiple-iteration configuration process to achieve a precise alignment effect of the first center line and the second center line.

[0051] The process of registering the first centerline and the second centerline includes: constructing a first displacement field, deforming the second centerline using the first displacement field to obtain a first registration centerline; generating a second displacement field based on the first registration centerline and the first centerline; deforming the first registration centerline based on the second displacement field to obtain a second registration centerline; and generating a displacement field for the next iteration based on the second registration centerline and the first centerline, until a registration termination condition is met, thereby completing the registration process.

[0052] The first displacement field may be a uniform zero field, that is, the initial displacements of all points are zero; or the first displacement field may be a random displacement field, that is, the initial displacements of all points are random values within a preset range.

[0053] An objective function is set to measure the registration effect of each iterative registration process. For example, the objective function can be a similarity function between the registration centerline obtained in each registration process and the first centerline. The minimization of the similarity function is used as the optimization goal to implement iterative registration processing and obtain the registration result.

[0054] Specifically, the second centerline is deformed using the first displacement field to obtain a first registration centerline. A first similarity function is determined between the first registration centerline and the first centerline. The first displacement field is updated using the first similarity function to obtain a second displacement field. Similarly, the first registration centerline is deformed based on the second displacement field to obtain a second registration centerline. A second similarity function is determined between the second registration centerline and the first centerline. The second displacement field is updated using the second similarity function to obtain a third displacement field. This process is repeated until the registration end condition is met and the target registration centerline is obtained.

[0055] The first similarity function and the second similarity function are the objective functions of the first and second iterations respectively. The updating method of the displacement field can be based on the gradient descent method, and the displacement field is updated along the negative gradient direction of the objective function. Specifically, where θ k is the displacement field parameter in the k-th iteration displacement field, α is the learning rate, is the objective function J in θ k The gradient at .

[0056] The above method is iterated continuously until a preset number of iterations is met or the objective function reaches convergence. The registration termination condition is then determined to be satisfied, and the registration is completed, resulting in the registration result of the keyframe angiography image and the CTA image. Compared to regular topological registration methods, which have difficulty handling missing or variable vascular branches, this embodiment achieves registration of the keyframe angiography image and the CTA image by constructing a displacement field and optimizing it during multiple iterative registration processes. This registration process is applicable to non-rigidly varying vessels and can achieve high-precision registration in the presence of missing or variable vascular branches.

[0057] The technical solution provided in this embodiment provides high-quality angiographic images for the registration process by extracting keyframe angiographic images from angiographic image sequences. By obtaining the first centerline of the target vascular segment in the key angiographic image and the second centerline of the corresponding target vascular segment in the CTA image, with the projection angle of the second centerline consistent with the projection angle of the key angiographic image, unnecessary interference is eliminated through the consistency of the projection angles. A two-dimensional second centerline is further obtained through projection. By constructing a displacement field, the second centerline is registered with the first centerline, achieving cross-modal registration of the two-dimensional angiographic image with the three-dimensional CTA image. The registration results of the keyframe angiographic image and the CTA image are obtained, reducing the registration difficulty and improving the registration accuracy.

[0058] Figure 2 This is a flowchart of a method for registering angiographic images with CTA images, provided in an embodiment of the present disclosure. This method is optimized based on the above-mentioned embodiment, using the matching relationship between key points in keyframe angiographic images and CTA images as a priori information to construct a first displacement field and improve registration efficiency. For detailed implementation details, please refer to the technical solution of this embodiment. Technical terms that are identical or similar to those in the above-mentioned embodiments are not repeated here.

[0059] S210 , obtaining a first centerline of a target blood vessel segment in a key-frame angiography image.

[0060] S220 , obtaining a second center line corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second center line is consistent with a projection angle of the key frame angiography image.

[0061] S230: Construct a first displacement field based on the matching relationship between the key frame angiography image and the key points in the CTA image.

[0062] S240: Deform the second center line based on the first displacement field to obtain a first registration center line.

[0063] S250. Generate a second displacement field based on the first registration centerline and the first centerline, deform the first registration centerline based on the second displacement field to obtain a second registration centerline, and complete the registration process until the registration end condition is met to obtain the registration result of the key frame angiography image and the CTA image.

[0064] The key points in the key-frame angiography image and the CTA image can be understood as points on the center of the blood vessel with clear anatomical significance or characteristics, such as but not limited to branch points and stenosis points. The matching relationship between the key points in the key-frame angiography image and the CTA image can be understood as the position correspondence or identification correspondence between the same key point on the target blood vessel segment in the key-frame angiography image and the CTA image respectively. The first displacement field constructed by the matching relationship between the key points in the key-frame angiography image and the CTA image is a displacement field with prior information corresponding to the above matching relationship. Compared with the zero displacement field or the random displacement field, it has high accuracy, realizes effective initial registration, simplifies the iterative registration process, and improves the registration efficiency.

[0065] In some embodiments of the present disclosure, the matching relationship between key points in the key frame angiography image and the CTA image can be achieved based on the following steps: Figure 3 It is a schematic diagram of the process of determining the matching relationship between key points in the key frame angiography image and the CTA image provided by the embodiment of the present disclosure.

[0066] S231. Identify a first key point in the key frame angiography image, and identify a second key point in the CTA image; wherein the number of the first key point and the number of the second key point are at least one.

[0067] S232: Extract point features of the first key point and point features of the second key point.

[0068] S233: Establish a matching relationship between the first key point and the second key point based on the point feature of the first key point and the point feature of the second key point.

[0069] The first key point in the keyframe angiography image can be understood as the first key point on the target vascular segment in the keyframe angiography image, and the second key point in the CTA image can be understood as the second key point on the target vascular segment in the CTA image. Specifically, the key point recognition model can be used to identify key points in the keyframe angiography image to obtain the first key point, and the key point recognition model can be used to identify the CTA image to obtain the second key point. The number of the first key points and the number of the second key points can be the same or different.

[0070] The point features of any key point include at least one of a position feature and an image feature. The position feature represents the spatial location of the key point, for example, its coordinates within the image. The image feature can be a feature vector that represents the global semantic characteristics of the key point within the image. The point features of the first key point and the second key point can be respectively implemented based on feature extraction models. In this embodiment, by extracting the position features and image features of each key point, the registration process is adapted to non-rigidly varying blood vessels, thereby improving the accuracy of blood vessel registration.

[0071] In some embodiments of the present disclosure, a pre-trained key point recognition network model is obtained. The key point recognition network model can be understood as a deep learning network model for detecting and describing key points in an image. The first key point and the point features of the first key point are identified by the key point recognition network model, and the second key point and the point features of the second key point are identified by the key point recognition network model. The key point recognition network model may include a feature extraction module, a key point detection branch module, and a descriptor generation branch module. The key point detection branch module and the descriptor generation branch module are respectively connected to the feature extraction module. Among them, the feature extraction module can be a convolutional neural network structure, which is used to extract a feature map from a key frame angiography image or a CTA image, and convert the key frame angiography image or a CTA image into a feature representation with rich semantic information. The key point detection branch module processes the feature map and outputs a key point response map. Each pixel value in the key point response map represents the probability of whether the corresponding pixel point is a key point. By setting an appropriate threshold, the key points and the location information of the key points are extracted from the response map. The descriptor generation branch processes the feature map to generate a descriptor for each keypoint. A descriptor is a fixed-length vector that describes the keypoint's global semantic characteristics within the image, known as the image feature. This keypoint recognition network model enables end-to-end processing, improving processing efficiency. Furthermore, a robust cross-modal feature descriptor based on deep learning methods is constructed to overcome interference from differences in vascular topology and local deformation.

[0072] In some embodiments of the present disclosure, the matching relationship between the first key point and the second key point can be determined by the similarity between the point feature of the first key point and the point feature of the second key point. Specifically, for any first key point, the second key points are traversed to determine the similarity data between the first key point and each second key point. The similarity data can be calculated by a cosine similarity algorithm or other similarity algorithms, which are not limited here. A matching relationship is established between the second key point corresponding to the maximum similarity data and the first key point. Based on the above method, the matching relationship between each first key point and the second key point is determined respectively.

[0073] In some embodiments of the present disclosure, a matching relationship between the first key point and the second key point is established based on the point features of the first key point and the point features of the second key point, including: extracting the interaction features corresponding to the first key point and the second key point respectively through an attention mechanism; determining a first key point group and a second key point group having a matching relationship, the second key point group including at least one second key point; the first key point group including at least one first key point; based on the interaction features corresponding to the first key point and the second key point respectively, determining the first key point and the second key point group having a matching relationship in the first key point group and the second key point group having a matching relationship.

[0074] Optionally, the attention mechanism can be a cross-attention mechanism, which extracts the interaction features corresponding to the first key point and the second key point, respectively. Specifically, the point features of the first key point and the second key point are input into the cross-attention mechanism model to obtain the interaction features corresponding to the first key point and the second key point, respectively. Specifically, the point features of the first key point are used as the query (Query), and the point features of the second key point are used as the key (Key) and the value (Value). An attention weight is calculated between the point features of the first key point and the point features of the second key point. This weight represents the correlation between each feature element in the point features of the first key point and each feature element in the point features of the second key point. The point features of the second key point are weighted and summed according to the attention weights to obtain a feature representation of the second key point after interacting with the point features of the first key point, i.e., the interaction features of the second key point. Similarly, the point features of the second key point are used as the query (Query), and the point features of the first key point are used as the key (Key) and the value (Value). A feature representation of the first key point after interacting with the point features of the second key point, i.e., the interaction features of the first key point, can be obtained.

[0075] By extracting the interactive features of key points in key-frame angiography images and CTA images, feature enhancement and cross-modal feature alignment are achieved. The different information about blood vessels provided by different cross-modal medical images is fused together through feature interaction, enriching the features and thus performing image registration more accurately. Furthermore, feature interaction can better overcome the impact of modality differences, accurately match and align blood vessel segments in images of different modalities, and improve the accuracy of image registration.

[0076] Optionally, the matching relationship between the first key point and the second key point can be determined by the similarity between the interaction feature of the first key point and the interaction feature of the second key point, which will not be described in detail here.

[0077] Optionally, a first key point group and a second key point group having a matching relationship are determined, the second key point group includes a local number of second key points, the first key point group includes a local number of second key points, and the number of key points in the first key point group and the second key point group may be the same or different.

[0078] Clustering is performed based on the interaction features of the second key points to obtain at least one second key point group, and clustering is performed based on the interaction features of the first key points to obtain at least one first key point group. The clustering algorithm is not limited here, and may include, for example, but not limited to, the k-means clustering algorithm. For each first key point group, the center point features of the first key point group are determined based on the interaction features of a local number of first key points in the first key point group. The center point features of the first key point group may be the mean features of the interaction features of a local number of first key points in the first key point group. Similarly, the center point features of each second key point group are determined based on the interaction features of the local number of first key points in the first key point group.

[0079] The first keypoint group and the second keypoint group having a matching relationship are determined based on the similarity between the center point feature of each second keypoint group and the center point feature of each first keypoint group. For example, the first keypoint group and the second keypoint group may be determined to have a matching relationship when the similarity between the center point feature of any second keypoint group and the center point feature of the first keypoint group is greater than a similarity threshold.

[0080] In a first key point group and a second key point group having a matching relationship, the first key points and the second key points therein are traversed. For any first key point in the first key point group, key point matching is performed based on the interaction feature of the first key point and the interaction features with each second key point in the second key point group, and a second key point having a matching relationship with the first key point is determined. The first key point and the second key point having a matching relationship are determined based on the above method.

[0081] By first determining the first key point group and the second key point group with a matching relationship, narrowing the matching range of the key points, and determining the first key point and the second key point with a matching relationship in the first key point group and the second key point group with a matching relationship, the amount of calculation in the process of determining the matching relationship of the key points is reduced and the processing efficiency is improved.

[0082] On the basis of the above embodiment, prior information is provided for constructing a displacement field based on the first key point and the second key point having a matching relationship, thereby improving the accuracy of the displacement field and providing an effective displacement field for the registration process.

[0083] Optionally, constructing a first displacement field includes: determining a first displacement vector between key points having a matching relationship in the key frame angiography image and the CTA image; and determining a second displacement vector of other vascular points based on a positional relationship between the key points and other vascular points in the second centerline; and forming the first displacement field based on the first displacement vector and the second displacement vector.

[0084] It can be understood that for the first key point A and the second key point B in the matching relationship, in the case of centerline alignment, the first key point A and the second key point B are aligned to the same position or a similar position, and accordingly, the displacement field is used to displace the second key point B to the position of the first key point A.

[0085] A first displacement vector is determined based on the position information corresponding to key points with matching relationships in the keyframe angiography image and the CTA image. Specifically, for a first key point and a second key point with matching relationships, the first position information of the first key point and the second position information of the second key point are obtained, and a first displacement vector is determined between the second position information and the first position information. The first displacement vector can be used to move the second key point to the position of the first key point. The first displacement vector can be determined based on the horizontal position difference and the vertical position difference corresponding to the second position information and the first position information, respectively.

[0086] For other vascular points outside of the matched keypoints, second displacement vectors are determined based on the positional relationship between the other vascular points and the second keypoint. The closer the other vascular point is to the second keypoint, the closer it is to the first displacement vector of the second keypoint. Optionally, the second displacement vectors of each other vascular point are determined based on the first displacement of at least one second keypoint using interpolation. On the second centerline, the first displacement vector of the second keypoint and the second displacement vectors of the other vascular points form a first displacement field.

[0087] Optionally, determining the first displacement field may further include obtaining position information corresponding to a first key point and a second key point that have a matching relationship, as well as position information of other vascular points, and inputting the position information corresponding to the first key point and the second key point, as well as the position information of other vascular points, into a pre-trained neural network model to obtain the first displacement field. The neural network model is a pre-trained machine learning model that has a displacement field generation function.

[0088] Based on the above method, a first displacement field is determined, and the second centerline is deformed based on the first displacement field. Specifically, a displacement vector corresponding to each blood vessel point in the second centerline in the first displacement field is obtained. Here, each blood vessel point in the second centerline includes the second key point and the other blood vessel points described above. Based on the displacement vector read from the first displacement field, the corresponding blood vessel point is deformed to obtain deformed position information of the blood vessel point. The deformed position information of the multiple blood vessel points forms the first registration centerline.

[0089] In order to improve the registration accuracy, the first registration centerline is optimized through multiple iterative registration processes. Specifically, based on the first registration centerline and the first centerline optimized displacement field, a second displacement field is obtained, and a second registration process is performed.

[0090] In some embodiments of the present disclosure, a second displacement field is generated based on the first registration centerline and the first centerline, including: using the matching relationship between the key points in the key frame angiography image and the CTA image as a regularization condition, and / or using the similarity function between the first registration centerline and the first centerline as an objective function, determining a third displacement vector for each blood vessel point in the first registration centerline, and forming the second displacement field based on the third displacement vector.

[0091] To prevent overfitting, the matching relationship between the key frame angiography image and the key points in the CTA image is used as a regularization condition to optimize the displacement field and improve the rationality and stability of the registration result.

[0092] In some embodiments of the present disclosure, the displacement field is optimized using the matching relationship between the key frame angiography image and the key points in the CTA image as a regularization condition, and the similarity function between the first registration center line and the first center line as an objective function to obtain a second displacement field.

[0093] In some embodiments of the present disclosure, a similarity function between the first registration center line and the first center line is used as an objective function to optimize the displacement field to obtain a second displacement field.

[0094] The first displacement field is updated based on the objective function, a third displacement vector is determined for each blood vessel point in the first registration centerline, and the second displacement field is formed based on the third displacement vector. The method for determining the third displacement vector for each blood vessel point is not further described here.

[0095] The first registration center line is deformed based on the second displacement field to obtain a second registration center line. Specifically, the displacement vector corresponding to each blood vessel point in the first registration center line in the second displacement field is obtained. Based on the displacement vector read from the second displacement field, the corresponding blood vessel point is deformed to obtain the deformed position information of the blood vessel point. The deformed position information of multiple blood vessel points forms the second registration center line.

[0096] It is understood that iterative optimization continues based on the second registration centerline, a third displacement field is determined based on the second registration centerline and the first centerline, and registration processing continues on the second registration centerline based on the third displacement field, and so on, until the registration end condition is met and the target registration centerline is obtained. During each iteration, the matching relationship between the key points in the keyframe angiography image and the CTA image is used as a regularization condition, and / or the similarity function between the registration centerline obtained from the previous registration process and the first centerline is used as the objective function to obtain the displacement field corresponding to the current registration process, thereby updating the displacement field and further implementing the registration process of the centerline.

[0097] The registration end condition includes at least one of the following: the number of registrations reaches a preset number; the registration metric value of the registration center line obtained by each registration reaches a convergence state with the first center line.

[0098] During the registration process, the registration times field is set. When each iteration is completed, the value in the registration times field is increased by 1. After each iteration is completed, the number of times in the registration times field is read. When the number reaches the preset number, the registration end condition is met.

[0099] After each iteration, the obtained registration centerline is measured to obtain a registration metric value to measure the registration effect of the iterative registration process. The registration metric value can be a similarity metric value, specifically, the similarity data between the registration centerline obtained in each iteration and the first centerline can be used as the above-mentioned registration metric value.

[0100] A registration metric sequence is formed based on the registration metric values corresponding to multiple iterations. Based on the registration metric sequence, it is determined whether the registration metric value has reached convergence. If so, it is determined that the registration end condition is met. If not, the next iteration is continued until the registration end condition is met.

[0101] On the basis of the above embodiment, when the registration process is completed, a correspondence is established between the first center line and the blood vessel points whose positions are consistent or within the error range in the target registration center line to obtain a registration result.

[0102] The technical solution of this embodiment obtains the matching relationship between key points in the keyframe angiography image and the CTA image, and uses this matching relationship as prior information to construct a first displacement field, thereby improving the effectiveness and accuracy of the first displacement field. Based on the first displacement field, the second centerline is registered, and the displacement field is further optimized. The first registered centerline is then registered again until a registration result is obtained, thereby improving the accuracy of the registration result.

[0103] In some embodiments of the present disclosure, a target registration centerline is obtained when a registration termination condition is satisfied; the target registration centerline is superimposed and displayed on a keyframe angiographic image. Specifically, two layers may be provided, wherein a first layer displays the keyframe angiographic image, and a second layer displays the target registration centerline, with the second layer being positioned above the first layer.

[0104] For example, see Figure 4 , Figure 4 It is a schematic diagram of an image display provided by an embodiment of the present disclosure. Figure 4 The first layer shows the key frame angiography image, and the second layer shows the target registration centerline. Figure 4 The green line in the image represents the target registration centerline. By superimposing the target registration centerline on the display image, the vascular centerline of the display image is intuitively displayed, improving the presentation quality of the keyframe angiography image.

[0105] In some embodiments of the present disclosure, the analysis image corresponding to the CTA image is deformed based on the displacement fields corresponding to the multiple iterative registration processes to obtain a deformed analysis image, and the deformed analysis image is superimposed and displayed on the key frame angiography image; wherein the analysis image corresponding to the CTA image includes at least one of the following: an MIP image, a volume reconstruction image, and a vascular abnormality display image reconstructed based on the CTA image.

[0106] The MIP (Maximum Intensity Projection) image can be understood as projecting the maximum CT value (i.e., the pixel value with the highest density) along the line of sight from a specific perspective to form a two-dimensional image. Volume rendering images (VR) can be understood as processing the entire volume data to display tissues and structures of different densities with different transparencies and colors, thereby achieving a three-dimensional effect. The vascular abnormality display image can be understood as an image obtained by performing vascular abnormality analysis on the CTA image, and the vascular abnormality display image includes marked vascular abnormality points. Among them, the vascular abnormality display image includes but is not limited to plaque images and stenosis images. There is no limitation on the type and acquisition method of the above-mentioned analysis images.

[0107] For any of the aforementioned analysis images, the analysis image corresponding to the CTA image is deformed based on the displacement fields corresponding to the respective multiple iterative registration processes to obtain a deformed analysis image. It will be appreciated that the analysis image corresponding to the CTA image is obtained based on the CTA image, and that deformation of the analysis image using the displacement fields during the registration process of the keyframe angiography image and the CTA image can similarly achieve registration of the keyframe angiography image and the analysis image, and accordingly, the keyframe angiography image matches the deformed analysis image.

[0108] The deformed analysis image is superimposed on the keyframe angiography image using two layers: the first layer displays the keyframe angiography image, and the second layer displays the deformed analysis image, with the second layer positioned above the first. Furthermore, the transparency of the deformed analysis image displayed in the second layer can be set, enabling the superimposition of vascular information in different dimensions, increasing the diversity and comprehensiveness of the displayed information.

[0109] In some embodiments of the present disclosure, the registration of key-frame angiography images and CTA images can be achieved by means of a machine learning model, which can be, for example, a neural network model. Specifically, the registration result of the key-frame angiography image and the CTA image is obtained by inputting the first center line and the second center line into the above-mentioned machine learning model, wherein the machine learning model realizes the registration process by constructing a displacement field. Optionally, the machine learning model realizes the registration of the key-frame angiography image and the CTA image by identifying key points in the key-frame angiography image and the CTA image, extracting point features and interaction features of the key points, determining the matching relationship between the key points in the key-frame angiography image and the CTA image, and constructing a displacement field based on the matching relationship between the key points in the key-frame angiography image and the CTA image, and obtaining the registration result of the key-frame angiography image and the CTA image.

[0110] The machine learning model is trained using a sample dataset, which includes multiple sample data. Each sample data set includes a first centerline of a target vessel segment in a keyframe angiography image and a second centerline of the corresponding target vessel segment in a CTA image of the same subject. The label information corresponding to the sample data is the registration result of the keyframe angiography image and the CTA image.

[0111] In this embodiment, a keyframe angiography image and a CTA image of the same sample subject are acquired, and a registration result of the keyframe angiography image and the CTA image is obtained by marking. The CTA image is projected at a projection angle corresponding to the keyframe angiography image to obtain a projected image. This projected image is then image-enhanced to obtain a first extended image. Similarly, the keyframe angiography image can be image-enhanced to obtain a second extended image. Image enhancement methods include, but are not limited to, one or more of rotation, mirroring, cropping, scaling, and translation. Centerlines are identified on the projected image and the first extended image to obtain multiple second centerlines corresponding to the target vascular segment in the CTA image. Centerlines are identified on the keyframe angiography image and the second extended image to obtain multiple first centerlines of the target vascular segment. Any first centerline and any second centerline are used to form sample data. These sample data all correspond to the keyframe angiography image and CTA image of the same sample subject, i.e., these sample data all correspond to a unified registration result.

[0112] Compared with the traditional model's supervised learning method's reliance on labeled data, which leads to difficulties in labeling and high consumption costs, the sample data used in the training process of the machine learning model in this embodiment reduces the demand for sample labeling and reduces the difficulty of collecting sample data.

[0113] Figure 5 FIG. 1 is a schematic diagram of a device for registering angiographic images and CTA images provided by an embodiment of the present disclosure. Figure 5 As shown, the device includes:

[0114] A first centerline acquisition module 310 is configured to acquire a first centerline of a target blood vessel segment in a key frame angiography image;

[0115] A second centerline acquisition module 320 is configured to acquire a second centerline corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second centerline is consistent with a projection angle of the key frame angiography image;

[0116] The registration module 330 performs registration processing on the first center line and the second center line by constructing a displacement field, thereby obtaining a registration result of the key frame angiography image and the CTA image.

[0117] The technical solution of this embodiment provides high-quality angiographic images for the registration process by extracting keyframe angiographic images from the angiographic image sequence. By obtaining the first centerline of the target vascular segment in the key angiographic image and the second centerline of the corresponding target vascular segment in the CTA image, with the projection angle of the second centerline consistent with the projection angle of the key angiographic image, unnecessary interference is eliminated through the consistency of the projection angles. A two-dimensional second centerline is further obtained through projection. By constructing a displacement field, the second centerline is registered with the first centerline, achieving cross-modal registration of the two-dimensional angiographic image and the three-dimensional CTA image. The registration results of the keyframe angiographic image and the CTA image are obtained, reducing the registration difficulty and improving the registration accuracy.

[0118] Based on the above embodiment, optionally, the registration module 330 is used to: construct a first displacement field based on the matching relationship between the key frame angiography image and the key points in the CTA image; deform the second center line based on the first displacement field to obtain a first registration center line; generate a second displacement field based on the first registration center line and the first center line, and deform the first registration center line based on the second displacement field to obtain a second registration center line, until the registration end condition is met and the registration process is completed.

[0119] Optionally, the registration module 330 is also used to: identify a first key point in the key frame angiography image, and identify a second key point in the CTA image; wherein the first key point and the second key point are at least one each; extract point features of the first key point and the point features of the second key point; wherein the point features include at least one of position features and image features; establish a matching relationship between the first key point and the second key point based on the point features of the first key point and the point features of the second key point.

[0120] Optionally, the registration module 330 is also used to: extract the interaction features corresponding to the first key point and the second key point respectively through the attention mechanism; determine a first key point group and a second key point group with a matching relationship, wherein the second key point group includes at least one second key point; based on the interaction features corresponding to the first key point and the second key point respectively, determine the first key point and the second key point group with a matching relationship in the first key point group and the second key point group with a matching relationship.

[0121] Optionally, the registration module 330 is also used to: determine a first displacement vector between key points having a matching relationship in the key frame angiography image and the CTA image; and, based on the positional relationship between other vascular points in the second centerline and the key points, determine a second displacement vector of the other vascular points; and form the first displacement field based on the first displacement vector and the second displacement vector.

[0122] Optionally, the registration module 330 is further used to: determine the third displacement vector of each blood vessel point in the first registration centerline based on the matching relationship between the key frame angiography image and the key points in the CTA image as a regularization condition, and form the second displacement field based on the third displacement vector.

[0123] Optionally, the registration end condition includes at least one of the following: the number of registrations reaches a preset number; the registration metric value of the registration centerline obtained by each registration and the first centerline reaches a convergence state.

[0124] Based on the above embodiment, the device also includes: a display module, which is used to obtain the target registration center line obtained when the registration end condition is met; superimpose the target registration center line on the display image, and the display image includes at least one of the following: the key frame angiography image, the maximum intensity projection image reconstructed based on the CTA image, the virtual reality image and the vascular abnormality display image.

[0125] The angiography image and CTA image registration device provided in the embodiment of the present disclosure can execute the angiography image and CTA image registration method provided in any embodiment of the present disclosure, and has the corresponding functional modules and beneficial effects of the execution method.

[0126] Figure 6 1 is a schematic diagram of the structure of an electronic device provided by an embodiment of the present disclosure. The electronic device 10 is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processing, cellular phones, smart phones, wearable devices (such as helmets, glasses, watches, etc.) and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely examples and are not intended to limit the implementation of the present disclosure described and / or required herein.

[0127] like Figure 6As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12, a random access memory (RAM) 13, etc., which is communicatively connected to the at least one processor 11. The memory stores a computer program that can be executed by the at least one processor, and the processor 11 can perform various appropriate actions and processes according to the computer program stored in the read-only memory (ROM) 12 or the computer program loaded from the storage unit 18 into the random access memory (RAM) 13. Various programs and data required for the operation of the electronic device 10 can also be stored in the random access memory (RAM) 13. The processor 11, the read-only memory (ROM) 12, and the random access memory (RAM) 13 are connected to each other via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0128] Multiple components in the electronic device 10 are connected to the I / O interface 15, including an input unit 16, such as a keyboard, a mouse, etc.; an output unit 17, such as various types of displays, speakers, etc.; a storage unit 18, such as a magnetic disk, an optical disk, etc.; and a communication unit 19, such as a network card, a modem, a wireless communication transceiver, etc. The communication unit 19 allows the electronic device 10 to exchange information / data with other devices via a computer network such as the Internet and / or various telecommunication networks.

[0129] The processor 11 can be any general-purpose and / or specialized processing component with processing and computing capabilities. Some examples of the processor 11 include, but are not limited to, a central processing unit (CPU), a graphics processing unit (GPU), various specialized artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, a digital signal processor (DSP), and any other suitable processor, controller, microcontroller, etc. The processor 11 executes the various methods and processes described above, such as the registration method for angiographic images and CTA images.

[0130] In some embodiments, the method for registering angiographic images with CTA images can be implemented as a computer program tangibly embodied in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program can be loaded and / or installed on electronic device 10 via read-only memory (ROM) 12 and / or communication unit 19. When the computer program is loaded into random access memory (RAM) 13 and executed by processor 11, one or more steps of the method for registering angiographic images with CTA images described above can be performed. Alternatively, in other embodiments, processor 11 can be configured to perform the method for registering angiographic images with CTA images in any other suitable manner (e.g., via firmware).

[0131] Various embodiments of the systems and techniques described herein can be implemented in digital electronic circuit systems, integrated circuit systems, field programmable gate arrays (FPGAs), application specific integrated circuits (ASICs), application specific standard products (ASSPs), system-on-chip systems (SOCs), programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments can include being implemented in one or more computer programs that are executable and / or interpreted on a programmable system that includes at least one programmable processor, which can be a special purpose or general purpose programmable processor that can receive data and instructions from a storage system, at least one input device, and at least one output device, and transmit data and instructions to the storage system, the at least one input device, and the at least one output device.

[0132] Computer programs for implementing the disclosed methods for registering angiographic images with CTA images can be written in any combination of one or more programming languages. These computer programs can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the functions / operations specified in the flowcharts and / or block diagrams are implemented. The computer programs can be executed entirely on the machine, partially on the machine, as a stand-alone software package, partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0133] The present disclosure also provides a computer-readable storage medium storing computer instructions for causing a processor to execute a registration method for angiographic images and CTA images, the method comprising:

[0134] A first centerline of a target vascular segment in a keyframe angiography image is obtained; a second centerline corresponding to the target vascular segment in a CTA image is obtained, wherein a projection angle of the second centerline is consistent with a projection angle of the keyframe angiography image; and the first centerline and the second centerline are registered by constructing a displacement field to obtain a registration result of the keyframe angiography image and the CTA image.

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

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

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

[0138] A computing system may include clients and servers. The clients and servers are typically remote from each other and typically interact via a communication network. This client-server relationship arises through computer programs running on the respective computers, creating a client-server relationship. The server may be a cloud server, also known as a cloud computing server or cloud host. This server is a hosting product within the cloud computing service ecosystem that addresses the management difficulties and limited scalability of traditional physical hosting and VPS services.

[0139] It should be understood that the various forms of processes shown above can be used to reorder, add, or delete steps. For example, the steps described in this disclosure can be performed in parallel, sequentially, or in a different order, as long as the desired results of the technical solutions of this disclosure can be achieved, and this document is not limited here.

[0140] The above specific embodiments do not constitute a limitation on the scope of protection of this disclosure. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may be made based on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this disclosure shall be included within the scope of protection of this disclosure.

Claims

1. A registration method for angiography images and CTA images, characterized in that: include: Acquiring a first centerline of a target blood vessel segment in a key-frame angiography image; Acquiring a second centerline corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second centerline is consistent with a projection angle of the key frame angiography image; By constructing a displacement field, the first center line and the second center line are registered to obtain a registration result of the key frame angiography image and the CTA image.

2. The method according to claim 1, characterized in that Performing registration processing on the first center line and the second center line by constructing a displacement field includes: constructing a first displacement field based on a matching relationship between key points in the key frame angiography image and the CTA image; deforming the second center line based on the first displacement field to obtain a first registration center line; A second displacement field is generated based on the first registration centerline and the first centerline, and the first registration centerline is deformed based on the second displacement field to obtain a second registration centerline, until a registration end condition is met, thereby completing the registration process.

3. The method according to claim 2, characterized in that The method for determining the matching relationship between the key frame angiography image and the key points in the CTA image includes: Identifying a first key point in the key frame angiography image, and identifying a second key point in the CTA image; wherein the first key point and the second key point are each at least one; Extracting point features of the first key point and point features of the second key point; wherein the point features include at least one of a position feature and an image feature; A matching relationship between the first key point and the second key point is established based on the point feature of the first key point and the point feature of the second key point.

4. The method according to claim 3, characterized in that Establishing a matching relationship between the first key point and the second key point based on the point feature of the first key point and the point feature of the second key point includes: Extracting the interaction features corresponding to the first key point and the second key point respectively through the attention mechanism; Determining a first key point group and a second key point group having a matching relationship, wherein the second key point group includes at least one second key point; and the first key point group includes at least one first key point; Based on the interaction features respectively corresponding to the first key point and the second key point, a first key point and a second key point having a matching relationship are determined in the first key point group and the second key point group having a matching relationship.

5. The method according to claim 3, characterized in that Construct the first displacement field, including: Determining a first displacement vector between key points having a matching relationship in the key frame angiography image and the CTA image; and determining a second displacement vector of the other blood vessel points based on a positional relationship between the other blood vessel points in the second centerline and the key points; The first displacement field is formed based on the first displacement vector and the second displacement vector.

6. The method according to claim 3, characterized in that Generating a second displacement field based on the first registration centerline and the first centerline includes: Using the matching relationship between the key points in the key-frame angiography image and the CTA image as a regularization condition, and / or using the similarity function between the first registration centerline and the first centerline as an objective function, determine the third displacement vector of each blood vessel point in the first registration centerline, and form the second displacement field based on the third displacement vector.

7. The method according to claim 2, characterized in that The registration end condition includes at least one of the following: The number of registrations reaches the preset number; The registration metric value of the registration center line obtained by each registration reaches a convergence state with the registration metric value of the first center line.

8. The method according to claim 2, characterized in that The method further comprises: Obtaining a target registration centerline obtained when the registration end condition is met; superimposing and displaying the target registration centerline on the key frame angiography image; or, Based on the displacement fields corresponding to the multiple iterative registration processes, the analysis image corresponding to the CTA image is deformed to obtain a deformed analysis image, and the deformed analysis image is superimposed and displayed on the key frame angiography image; The analysis image corresponding to the CTA image includes at least one of the following: a MIP image, a volume reconstruction image, and a vascular abnormality display image reconstructed based on the CTA image.

9. A device for registering angiographic images and CTA images, characterized in that: include: A first centerline acquisition module, configured to acquire a first centerline of a target blood vessel segment in a key frame angiography image; A second centerline acquisition module is configured to acquire a second centerline corresponding to the target blood vessel segment in the CTA image, wherein a projection angle of the second centerline is consistent with a projection angle of the key frame angiography image; The registration module performs registration processing on the first center line and the second center line by constructing a displacement field to obtain a registration result of the key frame angiography image and the CTA image.

10. An electronic device, characterized in that: The electronic device comprises: at least one processor; and a memory communicatively connected to the at least one processor; wherein, The memory stores a computer program executable by the at least one processor, and the computer program is executed by the at least one processor to enable the at least one processor to perform the registration method for angiography images and CTA images according to any one of claims 1 to 8.

11. A computer-readable storage medium, characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a processor to implement the registration method of angiography images and CTA images according to any one of claims 1 to 8 when executed.

Images