Interventional navigation method and apparatus based on image registration, and device and medium

Through the image registration-based method, the problem of low accuracy of interventional structure positioning in the three-dimensional vascular model images is solved, and high-precision positioning of the interventional structure in the three-dimensional environment is achieved, which improves the accuracy of interventional navigation.

WO2024234458A1PCT designated stage expired Publication Date: 2024-11-21SHENZHEN INST OF ADVANCED BIOMEDICAL ROBOT CO LTD

Patent Information

Application Number
PCT/CN2023/105203
Authority / Receiving Office
WO · WO
Patent Type
Applications
Current Assignee / Owner
Priority Date
2023-05-15
Filing Date
2023-06-30
Publication Date
2024-11-21

AI Technical Summary

Technical Problem

The prior art has low accuracy in positioning the catheter or guidewire in three-dimensional vascular model images, which makes it difficult to judge the precise position of the interventional structure in a three-dimensional environment.

Method used

Using an image registration-based method, by acquiring the original DSA image and the three-dimensional vascular model image, the loss function is constructed to obtain the differential homoembryonic deformation field, and the smooth reversible registration of the labeled DSA image to the three-dimensional vascular model image is achieved, so as to accurately Position the location of the interventional structure.

Benefits of technology

It improves the accuracy of interventional structure positioning in the three-dimensional vascular model images, can display the location of the interventional structure in real time, and enhances the accuracy of interventional navigation.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN2023105203_21112024_PF_FP_ABST
    Figure CN2023105203_21112024_PF_FP_ABST
Patent Text Reader

Abstract

The present application belongs to the technical field of medical image processing. Disclosed are an interventional navigation method and apparatus based on image registration, and a device and a medium. The method comprises: extracting a first head end position of an intervention structure in an original DSA image in real time, and marking the first head end position, so as to obtain a marked DSA image; extracting a three-dimensional vascular model image, and registering the marked DSA image to the three-dimensional vascular model image, so as to obtain a registered image; searching the registered image for a second head end position corresponding to the intervention structure in the marked DSA image; and sending the second head end position to a display end, so as to display the second head end position. A mapping process of the registration method has the advantage of smoothness and reversibility, and a marked DSA image can be registered to a three-dimensional vascular model image with relatively high precision, such that a second head end position of an intervention structure is precisely displayed in real time.
Need to check novelty before this filing date? Find Prior Art

Description

Interventional navigation method, device, equipment and medium based on image registration

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on May 15, 2023, with application number 2023105475843, and invention name “Interventional navigation method, device, equipment and medium based on image registration”, the entire contents of which are incorporated by reference into this application. Technical Field

[0002] The present application relates to the technical field of medical image processing, for example, to an interventional navigation method, apparatus, device and medium based on image registration. Background Art

[0003] Interventional technology has been increasingly widely used in the medical field. Existing interventional technology involves doctors determining the real-time position of interventional structures, such as catheters or guidewires, within the three-dimensional human body based on DSA images. Different doctors use different judgment criteria, and there is a lack of a unified position determination standard. Existing technologies, such as CN111681254A, input a video sequence containing a catheter region into a trained deep learning-based encoding and decoding structure to generate a binary mask sequence corresponding to the catheter. Based on the mask sequence, the position of the catheter in the video sequence is determined. This existing technology performs pixel-level target prediction and image segmentation on the catheter in the video sequence and can only locate the catheter in an X-ray transmission sequence. However, three-dimensional vascular model images are more complex than X-ray images and contain more information. Using the aforementioned encoding and decoding structure to perform pixel-level target prediction and image segmentation on the catheter or guidewire in the three-dimensional vascular model image is ineffective, resulting in low accuracy in locating the catheter or guidewire in the three-dimensional vascular model image.

[0004] In summary, the existing technology has the problem of low accuracy in positioning interventional structures such as catheters or guidewires in three-dimensional vascular model images. Technical issues

[0005] The purpose of this application is to provide an interventional navigation method, device, equipment and medium based on image registration, which can solve the problem of low accuracy in positioning interventional structures such as catheters or guidewires in three-dimensional vascular model images in the prior art. Technical Solutions

[0006] The purpose of this application is to provide an interventional navigation method, device, equipment and medium based on image registration, which can solve the problem of low accuracy in positioning interventional structures such as catheters or guidewires in three-dimensional vascular model images in the prior art.

[0007] To achieve the above objectives, the present application provides an interventional navigation method based on image registration, which is applied to a data processing end of an interventional navigation system, wherein the data processing end is connected to a display end. The method includes the following steps performed by the data processing end:

[0008] Obtaining an original digital subtraction angiography (DSA) image, where the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel;

[0009] extracting in real time the first head end position of the interventional structure in the original DSA image, marking the first head end position, and obtaining a marked DSA image;

[0010] extracting a three-dimensional vascular model image, and registering the labeled DSA image to the three-dimensional vascular model image to obtain a registered image;

[0011] searching for a second tip position corresponding to the interventional structure in the marked DSA image from the registered image;

[0012] The second head end position is sent to the display end to display the second head end position.

[0013] Preferably, registering the marked DSA image to the three-dimensional vascular model image to obtain a registered image includes:

[0014] Constructing a first loss function for transforming the labeled DSA image into the three-dimensional blood vessel model image;

[0015] constructing a second loss function for transforming the three-dimensional blood vessel model image into the labeled DSA image;

[0016] Adding the first loss function and the second loss function to obtain a final loss function;

[0017] Optimizing the final loss function to obtain a final differential homeomorphic deformation field;

[0018] An interpolation operation is performed on the labeled DSA image according to the final diffeomorphic deformation field to obtain the registered image.

[0019] Preferably, constructing a first loss function for transforming the labeled DSA image into the three-dimensional blood vessel model image comprises:

[0020] The first loss function is constructed by the following formula: in, is the first loss function, is the noise parameter, is the spatial uncertainty parameter, is a constraint parameter, Sim is a similarity metric function, F is the three-dimensional vascular model image, M is the labeled DSA image, s is the differential homeomorphic deformation field, u is the updated deformation field, Reg represents the constraint function, ||u||2 represents the result of the two-norm operation of the updated deformation field, It indicates that the marked DSA image is registered according to the sum of the diffeomorphic deformation field and the updated deformation field.

[0021] Preferably, constructing a second loss function for transforming from the three-dimensional blood vessel model image to the labeled DSA image comprises:

[0022] The second loss function is constructed by the following formula: ;in, is the inverse matrix of the diffeomorphic deformation field, is the second loss function.

[0023] Preferably, the optimizing the final loss function to obtain a final differential homeomorphic deformation field comprises:

[0024] Optimizing the final loss function using Newton's method to obtain the minimum function value of the final loss function;

[0025] The differential homeomorphic deformation field corresponding to the minimum function value is used as the final differential homeomorphic deformation field.

[0026] Preferably, searching for the second tip position corresponding to the interventional structure in the marked DSA image from the registered image comprises:

[0027] Detecting whether each pixel in the registered image has a mark, and if so, taking the corresponding pixel as a target pixel;

[0028] The second head end position is determined based on all the target pixel points.

[0029] Preferably, the real-time extraction of the first tip position of the interventional structure in the original DSA image comprises:

[0030] The first head end position is extracted using a target recognition method based on a neural network model, or the first head end position is extracted using a target recognition method based on an image threshold.

[0031] The present application provides an interventional navigation device based on image registration, which is applied to a data processing end of an interventional navigation system. The data processing end is connected to a display end. The device includes:

[0032] An image acquisition module is used to acquire an original digital subtraction angiography (DSA) image, where the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel;

[0033] a tip end position marking module, configured to extract in real time the first tip end position of the interventional structure in the original DSA image, mark the first tip end position, and obtain a marked DSA image;

[0034] a registration module, configured to extract a three-dimensional vascular model image and register the labeled DSA image to the three-dimensional vascular model image to obtain a registered image;

[0035] a tip end position search module, configured to search the registered image for a second tip end position corresponding to the interventional structure in the marked DSA image;

[0036] The head end position display module is used to send the second head end position to the display end to display the second head end position.

[0037] The present application also provides a computer device, comprising a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, it implements any one of the above-mentioned interventional navigation methods based on image registration and / or any one of the above-mentioned steps of the interventional navigation method based on image registration.

[0038] The present application also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements any one of the above-mentioned interventional navigation methods based on image registration and / or any one of the above-mentioned steps of the interventional navigation method based on image registration. Beneficial effects

[0039] The present application discloses an interventional navigation method based on image registration, which is applied to the data processing end of the interventional navigation system. The data processing end is connected to the display end. The method comprises the following steps performed by the data processing end: obtaining an original digital subtraction angiography image DSA, which is an image reflecting the position of the interventional structure in the human blood vessel. Extracting the first head end position of the interventional structure in the original DSA image in real time, marking the first head end position, and obtaining a marked DSA image. Extracting a three-dimensional vascular model image, registering the marked DSA image to the three-dimensional vascular model image, and obtaining a registered image. Searching for a second head end position corresponding to the interventional structure in the marked DSA image from the registered image. Sending the second head end position to the display end to display the second head end position. The marked DSA image is a two-dimensional image containing the first tip position of an interventional structure such as a catheter or guidewire. The mapping process of the above-mentioned registration method has the advantages of being smooth and reversible, and can register the marked DSA image to the three-dimensional vascular model image with high accuracy, thereby accurately displaying the second tip position of the interventional structure in real time in the three-dimensional registered image. BRIEF DESCRIPTION OF THE DRAWINGS

[0040] FIG1 is a schematic flow chart of an interventional navigation method based on image registration according to an embodiment;

[0041] FIG2 is a schematic diagram of a process for registering a marked DSA image according to an embodiment;

[0042] FIG3 is a schematic diagram of a process for calculating a final diffeomorphic deformation field according to an embodiment;

[0043] FIG4 is a schematic diagram of a process of searching for the position of a second head end according to an embodiment;

[0044] FIG5 is a schematic block diagram of the structure of an interventional navigation device based on image registration according to an embodiment;

[0045] FIG6 is a schematic block diagram of the structure of a computer device according to an embodiment.

[0046] The realization of the objectives, functional features and advantages of this application will be further explained in conjunction with embodiments and with reference to the accompanying drawings. Best Mode for Carrying Out the Invention

[0047] In order to make the purpose, technical solutions and advantages of this application more clear, the following further describes this application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain this application and are not intended to limit this application.

[0048] It will be understood by those skilled in the art that, unless expressly stated otherwise, the singular forms "a", "an", "above", and "the" used herein may also include plural forms. It should be further understood that the term "comprising" used in the specification of the present invention refers to the presence of features, integers, steps, operations, elements, modules, and / or components, but does not exclude the presence or addition of one or more other features, integers, steps, operations, elements, modules, components, and / or groups thereof. It should be understood that when we refer to an element as being "connected" or "coupled" to another element, it may be directly connected or coupled to the other element, or there may be intermediate elements. In addition, "connected" or "coupled" as used herein may include wireless connections or wireless couplings. The term "and / or" used herein includes all or any module and all combinations of one or more associated listed items.

[0049] Those skilled in the art will understand that, unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by those skilled in the art in the art to which this invention belongs. It should also be understood that terms such as those defined in common dictionaries should be understood to have meanings consistent with their meanings in the context of the prior art and, unless specifically defined as such, will not be interpreted in an idealized or overly formal sense.

[0050] In one embodiment, referring to FIG1 , which is a flow chart of an interventional navigation method based on image registration disclosed in the present application, the method is applied to a data processing end of an interventional navigation system, the data processing end being connected to a display end, and the method includes the following steps performed by the data processing end:

[0051] S1: Acquire an original digital subtraction angiography (DSA) image. The original DSA image is an image reflecting the position of the interventional structure in the human blood vessel.

[0052] DSA (Digital Subtraction Angiography) is a digital subtraction technique based on sequential images. By subtracting the first and second frames of the same body part, the difference is obtained, eliminating bone and soft tissue structures. This allows the contrast-filled blood vessels in the second frame to appear in the subtracted image, enhancing contrast. Raw DSA images are unprocessed DSA images.

[0053] The intervention structure may be an intervention catheter or an intervention guide wire, which is not limited here.

[0054] S2: extracting the first head end position of the interventional structure in the original DSA image in real time, marking the first head end position, and obtaining a marked DSA image.

[0055] The first head end position is extracted using a target recognition method based on a neural network model, or the first head end position is extracted using a target recognition method based on an image threshold.

[0056] As an example, the YoloV5 neural network model can be used to extract the first head end position, and the Fast RCNN neural network model can also be used to extract the first head end position.

[0057] As an example, the PCA method with adaptive image threshold may be used to extract the first head end position, or the histogram bimodal method may be used to extract the first head end position.

[0058] The outline of the first head end position in the DSA image may be set to a preset color, or the coordinate set of the first head end position may be recorded and a corresponding label may be set for the first head end position.

[0059] S3: Extracting a three-dimensional blood vessel model image, and registering the labeled DSA image to the three-dimensional blood vessel model image to obtain a registered image.

[0060] Constructing a first loss function for transforming the labeled DSA image into the three-dimensional blood vessel model image;

[0061] constructing a second loss function for transforming the three-dimensional blood vessel model image into the labeled DSA image;

[0062] Adding the first loss function and the second loss function to obtain a final loss function;

[0063] Optimizing the final loss function to obtain a final differential homeomorphic deformation field;

[0064] An interpolation operation is performed on the labeled DSA image according to the final diffeomorphic deformation field to obtain the registered image.

[0065] A three-dimensional vascular model image is extracted based on the vascular CTA (computed tomographic angiography) image.

[0066] Differentiomorphism is a concept applicable to the category of differential manifolds. Both the labeled DSA image and the three-dimensional vascular model image can be regarded as differential manifolds. The symmetric differentiomorphism registration method can be used to smoothly and reversibly map the labeled DSA image to the coordinate system of the three-dimensional vascular model image.

[0067] S4: Searching for a second tip position corresponding to the interventional structure in the marked DSA image from the registered image.

[0068] Detecting whether each pixel in the registered image has a mark, and if so, taking the corresponding pixel as a target pixel;

[0069] The second head end position is determined based on all the target pixel points.

[0070] If the outline of the first tip end position in the DSA image is set to a preset color, the color of this position after mapping will be different from the colors of other positions in the registered image. The second tip end position in the registered image can be searched based on the color.

[0071] If the coordinate set of the first head end position is recorded, the coordinate set of the first head end position is mapped using the symmetric differential homeomorphism registration method to obtain another coordinate set, and the second head end position corresponding to the interventional structure in the registered image can be determined based on the coordinate set.

[0072] S5: Send the second head end position to the display end to display the second head end position.

[0073] The display end may be a liquid crystal display screen or other display devices, which is not limited here.

[0074] The image to be sent can be obtained by taking the second head end position as the center and including the background area near the second head end position, and the image to be sent can be sent to the display end, with the second head end position highlighted. Alternatively, the entire registered image including the second head end position can be sent to the display end, with the second head end position displayed.

[0075] An interventional navigation method based on image registration according to an embodiment of the present application is applied to a data processing end of an interventional navigation system, the data processing end being connected to a display end, and the method comprises the following steps performed by the data processing end: obtaining an original digital subtraction angiography image DSA, the original DSA image being an image reflecting the position of the interventional structure in the human blood vessel. Extracting the first head end position of the interventional structure in the original DSA image in real time, marking the first head end position, and obtaining a marked DSA image. Extracting a three-dimensional vascular model image, registering the marked DSA image to the three-dimensional vascular model image, and obtaining a registered image. Searching for a second head end position corresponding to the interventional structure in the marked DSA image from the registered image. Sending the second head end position to the display end to display the second head end position. The marked DSA image is a two-dimensional image containing the first tip position of the interventional structure such as a catheter or guidewire. The mapping process of the symmetric differential homeomorphism registration method has the advantages of being smooth and reversible, and can align the marked DSA image to the three-dimensional vascular model image with high accuracy, thereby accurately displaying the second tip position of the interventional structure in real time in the three-dimensional registered image.

[0076] In one embodiment, referring to FIG. 2 , registering the labeled DSA image to the three-dimensional vascular model image to obtain a registered image includes:

[0077] S32: Constructing a first loss function for transforming the labeled DSA image into the three-dimensional blood vessel model image.

[0078] Before step S32, the method further includes step S31: extracting a three-dimensional blood vessel model image.

[0079] The first loss function is constructed by the following formula: ;in, is the first loss function, is the noise parameter, is the spatial uncertainty parameter, is a constraint parameter, Sim is a similarity metric function, F is the three-dimensional vascular model image, M is the labeled DSA image, s is the differential homeomorphic deformation field, u is the updated deformation field, Reg represents the constraint function, ||u||2 represents the result of the two-norm operation of the updated deformation field, It indicates that the marked DSA image is registered according to the sum of the diffeomorphic deformation field and the updated deformation field.

[0080] Used to measure the difference between the registration result of the marked DSA image and the 3D vascular model image. Used to constrain the updated deformation field, Used to impose manifold constraints on diffeomorphic deformation fields.

[0081] The first loss function is the loss function for differential homeomorphism registration of the labeled DSA image to the three-dimensional vascular model image. The smaller the first loss function, the higher the accuracy of transforming the labeled DSA image to the coordinate system of the three-dimensional vascular model image, and the higher the similarity between the registered image and the three-dimensional vascular model image.

[0082] S33: Constructing a second loss function for transforming the three-dimensional blood vessel model image into the labeled DSA image.

[0083] The second loss function is constructed by the following formula: ;in, is the inverse matrix of the diffeomorphic deformation field, is the second loss function.

[0084] Used to measure the difference between the registration result of the 3D vascular model image and the labeled DSA image. Used to constrain the updated deformation field, Used to impose manifold constraints on the inverse transformation of a diffeomorphic deformation field.

[0085] The second loss function is the loss function for aligning the three-dimensional vascular model image to the labeled DSA image. The smaller the second loss function, the higher the accuracy of transforming the three-dimensional vascular model image to the coordinate system of the labeled DSA image, and the higher the similarity between the registered vascular model image and the labeled DSA image.

[0086] S34: Add the first loss function and the second loss function to obtain a final loss function.

[0087] ;

[0088] in, is the final loss function, is the first loss function, is the second loss function.

[0089] The smaller the final loss function is, the better the results of both the transformation of the labeled DSA image to the three-dimensional vascular model image and the transformation of the three-dimensional vascular model image to the labeled DSA image can be achieved.

[0090] S35: Optimizing the final loss function to obtain a final differential homeomorphic deformation field.

[0091] 3 , step S35 includes steps S351 - S352 .

[0092] S351: Optimize the final loss function using Newton's method to obtain the minimum function value of the final loss function.

[0093] S352: Taking the differential homeomorphic deformation field corresponding to the minimum function value as the final differential homeomorphic deformation field.

[0094] During the optimization of the final loss function, the value of the final loss function decreases. When the value of the final loss function is less than the final loss function threshold, or when the number of optimizations of the final loss function exceeds the optimization number threshold, the optimization of the final loss function is stopped, and the minimum function value of the final loss function is obtained. The differential homeomorphic deformation field corresponding to the minimum function value is used as the final differential homeomorphic deformation field.

[0095] After the optimization is complete, the final loss function value is small, achieving good registration results in both forward and backward registration. When both forward and backward registration are performed, the reversible transformation from the labeled DSA image to the 3D vascular model image can be further guaranteed. Forward registration is the transformation from the labeled DSA image to the 3D vascular model image, while backward registration is the transformation from the 3D vascular model image to the labeled DSA image.

[0096] S36: performing an interpolation operation on the labeled DSA image according to the final diffeomorphic deformation field to obtain the registered image.

[0097] The coordinates of each pixel point of the marked DSA image are added to the coordinates of the corresponding point in the final differential homeomorphic deformation field to obtain the registered coordinates. The registered pixel values ​​corresponding to the registered coordinates are calculated using the interpolation method based on the registered coordinates. The registered coordinates and the registered pixel values ​​together constitute the registered image.

[0098] The interpolation method can be the nearest neighbor interpolation method, bilinear interpolation method or bicubic interpolation method. The nearest neighbor interpolation method has the fastest operation speed but the worst interpolation effect; the bicubic interpolation method has the slowest operation speed but the best interpolation effect; the bilinear interpolation method has an operation speed between the nearest neighbor interpolation method and the bicubic interpolation method, and the interpolation effect of the bilinear interpolation method is between the nearest neighbor interpolation method and the bicubic interpolation method. The embodiment of the present application takes the bilinear interpolation method as an example.

[0099] The registered image is a three-dimensional image, and the position of the second tip of the interventional structure can be accurately displayed in the registered image in real time.

[0100] As mentioned above, by continuously optimizing the final loss function, the first loss function and the second loss function are both small, ensuring that both forward registration and reverse registration have good effects, and further ensuring that the process of transforming the labeled DSA image to the three-dimensional vascular model image is reversible.

[0101] In one embodiment, referring to FIG. 4 , searching for a second tip position corresponding to the interventional structure in the marked DSA image from the registered image includes:

[0102] S41: Detect whether each pixel in the registered image has a mark, and if so, use the corresponding pixel as a target pixel.

[0103] Detect whether each pixel in the registered image has a color mark or label mark, for example, detect whether each pixel has a red mark or a label corresponding to the first head end position. If so, the pixel is used as the target pixel.

[0104] S42: Determine the second head end position according to all the target pixel points.

[0105] Based on all target pixels, the outline of the intervening structure in the registered image can be determined, and the center coordinates of this outline can be used as the second tip position. Once the second tip position is determined, it can be displayed in real time in the registered image to better illustrate changes in the second tip position of the intervening structure.

[0106] As described above, searching for the second head end position corresponding to the interventional structure in the marked DSA image from the registered image includes detecting whether each pixel in the registered image has a mark, and if so, using the corresponding pixel as a target pixel. The second head end position is determined based on all the target pixels. After the second head end position is determined, the second head end position can be displayed in real time in the registered image to better show the changes in the second head end position of the interventional structure.

[0107] 5 is a schematic block diagram of the structure of an interventional navigation device based on image registration disclosed in the present application. The interventional navigation device based on image registration is applied to the data processing end of the interventional navigation system, and the data processing end is connected to the display end. The device includes:

[0108] An image acquisition module 10 is used to acquire an original digital subtraction angiography (DSA) image, where the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel;

[0109] a tip end position marking module 20 for extracting in real time the first tip end position of the interventional structure in the original DSA image, marking the first tip end position, and obtaining a marked DSA image;

[0110] A registration module 30 is configured to extract a three-dimensional vascular model image and register the labeled DSA image to the three-dimensional vascular model image to obtain a registered image;

[0111] a tip end position search module 40 for searching, from the registered image, a second tip end position corresponding to the interventional structure in the marked DSA image;

[0112] The head end position display module 50 is used to send the second head end position to the display end to display the second head end position.

[0113] As described above, the interventional navigation device based on image registration can implement the interventional navigation method based on image registration.

[0114] In one embodiment, the registration module 30 further includes:

[0115] a first loss function construction unit, configured to construct a first loss function for transforming the labeled DSA image into the three-dimensional blood vessel model image;

[0116] a second loss function construction unit, configured to construct a second loss function for transforming the three-dimensional blood vessel model image into the labeled DSA image;

[0117] a final loss function calculation unit, configured to add the first loss function and the second loss function to obtain a final loss function;

[0118] A final loss function optimization unit, configured to optimize the final loss function to obtain a final differential homeomorphic deformation field;

[0119] An interpolation operation unit is used to perform an interpolation operation on the marked DSA image according to the final diffeomorphic deformation field to obtain the registered image.

[0120] In one embodiment, the first loss function construction unit further includes:

[0121] The first loss function construction subunit is used to construct the first loss function using the following formula: ;in, is the first loss function, is the noise parameter, is the spatial uncertainty parameter, is a constraint parameter, Sim is a similarity metric function, F is the three-dimensional vascular model image, M is the labeled DSA image, s is the differential homeomorphic deformation field, u is the updated deformation field, Reg represents the constraint function, ||u||2 represents the result of the two-norm operation of the updated deformation field, It indicates that the marked DSA image is registered according to the sum of the diffeomorphic deformation field and the updated deformation field.

[0122] In one embodiment, the second loss function construction unit further includes:

[0123] The second loss function construction subunit is used to construct the second loss function using the following formula: ;in, is the inverse matrix of the diffeomorphic deformation field, is the second loss function.

[0124] In one embodiment, the final loss function optimization unit further includes:

[0125] A final loss function optimization subunit, configured to optimize the final loss function using Newton's method to obtain a minimum function value of the final loss function;

[0126] The final differential homeomorphic deformation field definition subunit is used to use the differential homeomorphic deformation field corresponding to the minimum function value as the final differential homeomorphic deformation field.

[0127] In one embodiment, the head end location search module 40 further includes:

[0128] a target pixel definition unit, configured to detect whether each pixel in the registered image has a mark, and if so, to use the corresponding pixel as a target pixel;

[0129] The second head end position determining unit is configured to determine the second head end position according to all the target pixel points.

[0130] In one embodiment, the head end position marking module 20 further includes:

[0131] The first head end position extraction unit is used to extract the first head end position by using a target recognition method based on a neural network model, or to extract the first head end position by using a target recognition method based on an image threshold.

[0132] 6 , a computer device is also provided in an embodiment of the present application, and the internal structure of the computer device can be shown in FIG6 . The computer device includes a processor, a memory, a network interface and a database connected via a system bus. The processor designed for the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating device, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store the first head end position and the second head end position, etc. The network interface of the computer device is used to communicate with an external terminal via a network connection. Furthermore, the above-mentioned computer device can also be provided with an input device and a display screen, etc. When the computer program is executed by the processor, it implements an interventional navigation method based on image registration. The method is applied to the data processing end of the interventional navigation system, and the data processing end is connected to the display end. The method includes the following steps performed by the data processing end: obtaining an original digital subtraction angiography image DSA, the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel; extracting the first head end position of the interventional structure in the original DSA image in real time, marking the first head end position, and obtaining a marked DSA image; extracting a three-dimensional blood vessel model image, and registering the marked DSA image to the three-dimensional blood vessel model image to obtain a registered image; searching for a second head end position corresponding to the interventional structure in the marked DSA image from the registered image; and sending the second head end position to the display end to display the second head end position. Those skilled in the art will understand that the structure shown in Figure 6 is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation on the computer device to which the present application scheme is applied.

[0133] An embodiment of the present application also provides a computer-readable storage medium having a computer program stored thereon, which implements an interventional navigation method based on image registration when the computer program is executed by a processor, the method being applied to a data processing end of an interventional navigation system, the data processing end being connected to a display end, the method comprising the following steps performed by the data processing end: acquiring an original digital subtraction angiography image DSA, the original DSA image being an image reflecting the position of the interventional structure in a human blood vessel; extracting a first head end position of the interventional structure in the original DSA image in real time, marking the first head end position to obtain a marked DSA image; extracting a three-dimensional vascular model image, registering the marked DSA image to the three-dimensional vascular model image to obtain a registered image; searching for a second head end position corresponding to the interventional structure in the marked DSA image from the registered image; and sending the second head end position to the display end to display the second head end position.

[0134] It can be understood that the computer-readable storage medium in this embodiment can be a volatile readable storage medium or a non-volatile readable storage medium.

[0135] Those skilled in the art will understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus direct RAM (RDRAM), direct RAM bus dynamic RAM (DRDRAM), and RAM bus dynamic RAM (RDRAM).

[0136] It should be noted that, in this document, the terms "comprises," "includes," or any other variations thereof are intended to encompass non-exclusive inclusion, such that a process, apparatus, article, or method comprising a series of elements includes not only those elements but also other elements not explicitly listed, or elements inherent to such process, apparatus, article, or method. In the absence of further limitations, an element defined by the phrase "comprising a ..." does not exclude the presence of other identical elements in the process, apparatus, article, or method comprising the element.

[0137] The above description is only a preferred embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made by using the contents of the present invention description and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. An interventional navigation method based on image registration, wherein: A data processing end applied to an interventional navigation system, the data processing end being connected to a display end, the method comprising the following steps performed by the data processing end: Acquire an original digital subtraction angiography (DSA) image, where the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel; extracting in real time the first head end position of the intervention structure in the original DSA image, marking the first head end position, and obtaining a marked DSA image; Extracting a three-dimensional vascular model image, and registering the labeled DSA image to the three-dimensional vascular model image to obtain a registered image; Searching for a second tip position corresponding to the intervention structure in the marked DSA image from the registered image; The second head end position is sent to the display end to display the second head end position.

2. The interventional navigation method based on image registration according to claim 1, wherein: The step of registering the marked DSA image to the three-dimensional blood vessel model image to obtain a registered image includes: Constructing a first loss function for transforming the labeled DSA image to the three-dimensional blood vessel model image; Constructing a second loss function for transforming the three-dimensional blood vessel model image to the labeled DSA image; Adding the first loss function and the second loss function to obtain a final loss function; Optimizing the final loss function to obtain a final differential homeomorphic deformation field; An interpolation operation is performed on the labeled DSA image according to the final diffeomorphic deformation field to obtain the registered image.

3. The interventional navigation method based on image registration according to claim 2, wherein: The interpolation operation is performed on the marked DSA image according to the final differential homeomorphic deformation field to obtain the registered image, including: Adding the coordinates of each pixel point of the marked DSA image to the coordinates of the corresponding point in the final diffeomorphic deformation field to obtain the registered coordinates; The registered pixel values ​​corresponding to the registered coordinates are calculated using an interpolation method according to the registered coordinates, and the registered coordinates and the registered pixel values ​​together constitute a registered image.

4. The interventional navigation method based on image registration according to claim 2, wherein: The constructing a first loss function for transforming the labeled DSA image to the three-dimensional blood vessel model image includes: The first loss function is constructed by the following formula: ;in, is the first loss function, is the noise parameter, is the spatial uncertainty parameter, is a constraint parameter, Sim is a similarity measurement function, F is the three-dimensional vascular model image, M is the labeled DSA image, s is the differential homeomorphic deformation field, u is the updated deformation field, Reg represents a constraint function, ||u||2 represents the result of the two-norm operation of the updated deformation field, It indicates that the marked DSA image is registered according to the sum of the differential homeomorphic deformation field and the updated deformation field.

5. The interventional navigation method based on image registration according to claim 4, wherein: The constructing a second loss function for transforming the three-dimensional blood vessel model image to the labeled DSA image includes: The second loss function is constructed by the following formula: ;in, is the inverse matrix of the diffeomorphic deformation field, is the second loss function.

6. The interventional navigation method based on image registration according to claim 2, wherein: The final loss function is optimized to obtain a final differential homeomorphic deformation field, including: Using Newton's method to optimize the final loss function to obtain the minimum function value of the final loss function; The differential homeomorphic deformation field corresponding to the minimum function value is used as the final differential homeomorphic deformation field.

7. The interventional navigation method based on image registration according to claim 6, wherein: The using Newton's method to optimize the final loss function to obtain the minimum function value of the final loss function includes: When the value of the final loss function is less than the final loss function threshold, or when the number of optimizations of the final loss function is greater than the optimization number threshold, stop optimizing the final loss function; The minimum function value of the final loss function is obtained.

8. The interventional navigation method based on image registration according to claim 1, wherein: The step of searching the registered image for a second tip position corresponding to the intervention structure in the marked DSA image comprises: Detecting whether each pixel in the registered image has a mark, and if so, taking the corresponding pixel as a target pixel; The second head end position is determined according to all the target pixel points.

9. The interventional navigation method based on image registration according to claim 8, wherein: After the step of determining the position of the second head end according to all the target pixel points, the method further comprises: The second head end position is displayed in real time in the registered image to better show the change of the second head end position of the interventional structure.

10. The interventional navigation method based on image registration according to claim 1, wherein: The real-time extraction of the first head end position of the intervention structure in the original DSA image comprises: The first head end position is extracted using a target recognition method based on a neural network model, or the first head end position is extracted using a target recognition method based on an image threshold.

11. An interventional navigation device based on image registration, wherein: A data processing end applied to an interventional navigation system, the data processing end is connected to a display end, and the device comprises: An image acquisition module is used to acquire an original digital subtraction angiography (DSA) image, where the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel; A head end position marking module, used for extracting the first head end position of the intervention structure in the original DSA image in real time, marking the first head end position, and obtaining a marked DSA image; A registration module, used for extracting a three-dimensional vascular model image, registering the marked DSA image to the three-dimensional vascular model image, and obtaining a registered image; A tip end position search module, used to search for a second tip end position corresponding to the intervention structure in the marked DSA image from the registered image; The head end position display module is used to send the second head end position to the display end to display the second head end position.

12. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: When the processor executes the computer program, an intervention navigation method based on image registration is implemented; The interventional navigation method based on image registration is applied to a data processing end of an interventional navigation system, the data processing end is connected to a display end, and the method includes the following steps performed by the data processing end: Acquire an original digital subtraction angiography (DSA) image, where the original DSA image is an image reflecting the position of the interventional structure in the human blood vessel; extracting in real time the first head end position of the intervention structure in the original DSA image, marking the first head end position, and obtaining a marked DSA image; Extracting a three-dimensional vascular model image, and registering the labeled DSA image to the three-dimensional vascular model image to obtain a registered image; Searching for a second tip position corresponding to the intervention structure in the marked DSA image from the registered image; The second head end position is sent to the display end to display the second head end position.

13. A computer device according to claim 12, wherein registering the marked DSA image to the three-dimensional vascular model image to obtain a registered image comprises: Constructing a first loss function for transforming the labeled DSA image to the three-dimensional blood vessel model image; Constructing a second loss function for transforming the three-dimensional blood vessel model image to the labeled DSA image; Adding the first loss function and the second loss function to obtain a final loss function; Optimizing the final loss function to obtain a final differential homeomorphic deformation field; An interpolation operation is performed on the labeled DSA image according to the final diffeomorphic deformation field to obtain the registered image.

14. A computer device according to claim 13, wherein the step of performing an interpolation operation on the labeled DSA image according to the final diffeomorphic deformation field to obtain the registered image comprises: Adding the coordinates of each pixel point of the marked DSA image to the coordinates of the corresponding point in the final diffeomorphic deformation field to obtain the registered coordinates; The registered pixel values ​​corresponding to the registered coordinates are calculated using an interpolation method according to the registered coordinates, and the registered coordinates and the registered pixel values ​​together constitute a registered image.

15. The computer device according to claim 13, wherein the step of constructing a first loss function for transforming the labeled DSA image to the three-dimensional blood vessel model image comprises: The first loss function is constructed by the following formula: ;in, is the first loss function, is the noise parameter, is the spatial uncertainty parameter, is a constraint parameter, Sim is a similarity measurement function, F is the three-dimensional vascular model image, M is the labeled DSA image, s is the differential homeomorphic deformation field, u is the updated deformation field, Reg represents a constraint function, ||u||2 represents the result of the two-norm operation of the updated deformation field, It indicates that the marked DSA image is registered according to the sum of the differential homeomorphic deformation field and the updated deformation field.

16. A computer device according to claim 15, wherein the step of constructing a second loss function for transforming the three-dimensional blood vessel model image to the labeled DSA image comprises: The second loss function is constructed by the following formula: ;in, is the inverse matrix of the diffeomorphic deformation field, is the second loss function.

17. A computer device according to claim 15, wherein the step of optimizing the final loss function to obtain a final differential homeomorphic deformation field comprises: Using Newton's method to optimize the final loss function to obtain the minimum function value of the final loss function; The differential homeomorphic deformation field corresponding to the minimum function value is used as the final differential homeomorphic deformation field.

18. A computer device according to claim 12, wherein searching the registered image for a second tip position corresponding to the interventional structure in the marked DSA image comprises: Detecting whether each pixel in the registered image has a mark, and if so, taking the corresponding pixel as a target pixel; The second head end position is determined according to all the target pixel points.

19. The computer device according to claim 12, wherein the step of extracting the first tip position of the interventional structure in the original DSA image in real time comprises: The first head end position is extracted using a target recognition method based on a neural network model, or the first head end position is extracted using a target recognition method based on an image threshold.

20. A computer-readable storage medium having a computer program stored thereon, wherein: When the computer program is executed by a processor, the steps of the intervention navigation method based on image registration are implemented as claimed in any one of claims 1 to 10.

Citation Information

Patent Citations

  • System and method for motion-adjusted device guidance using vascular roadmaps

    CN114025658A

  • Intraoperative navigation system based on DSA imaging device, computing device and program product

    CN115035001A

  • DSA image-based surgical navigation method and system, medium and electronic equipment

    CN115737123A

Cited By

  • Intelligent planning and intraoperative navigation system for spinal surgery

    CN121421674A