An ablation zone localization system
By acquiring and registering static and dynamic images of the ablation target, and combining them with a digital model of the ablation catheter, the three-dimensional spatial coordinates of the ablation target are generated and marked. This solves the problem of inaccurate target localization and achieves precise localization of the ablation area and completeness and accuracy of the operation.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- NINGBO FIRST HOSPITAL
- Filing Date
- 2026-03-11
- Publication Date
- 2026-07-10
AI Technical Summary
In existing technologies, the localization of ablation targets is easily affected by environmental and physiological factors, leading to spatial misalignment between the ablation area and the actual target tissue, which affects the integrity and accuracy of ablation.
By acquiring static baseline images and real-time dynamic images of the ablation target, image registration is performed using a registration algorithm. Combined with the digital model of the ablation catheter, the three-dimensional spatial coordinates of the ablation target are generated, and a visual marker is generated on the superimposed and fused images to achieve precise positioning of the ablation area.
It improves the positioning accuracy of the ablation area, ensures precise coverage of the ablation area and the actual target tissue, avoids missed ablation or repeated ablation, and enhances the integrity and accuracy of the ablation operation.
Smart Images
Figure CN122368166A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of image processing technology, and more specifically, to an ablation region localization system. Background Technology
[0002] The prerequisite for the effective operation of an ablation system is the accurate positioning of ablation targets to ensure that the ablation area formed by these targets can completely cover the target tissue, while avoiding missed ablation or repeated ablation.
[0003] In existing technologies, the localization of ablation targets typically relies on real-time image guidance of the target tissue (such as digital subtraction angiography, DSA). However, such images are susceptible to interference from various environmental factors (such as equipment vibration and patient movement) and physiological factors (such as respiration, heartbeat, and vascular pulsation) during acquisition, causing the target tissue to exhibit continuous dynamic displacement in the images. Because the image content is constantly changing, the system struggles to stably and accurately identify and lock onto the ablation target, resulting in spatial misalignment between the ablation area and the actual target tissue. This localization deviation directly affects the integrity and accuracy of the ablation. Summary of the Invention
[0004] The problem addressed by this invention is how to improve the positioning accuracy of the ablation zone.
[0005] To address the above problems, this invention provides an ablation area positioning system.
[0006] In a first aspect, the present invention provides an ablation region positioning system, comprising: The data acquisition unit is used to acquire a static reference image of the ablation target and to acquire a dynamic real-time image of the ablation target in real time. The static reference image contains the ablation target location of the ablation target. The registration unit is used to register the static reference image and the dynamic real-time image using a registration algorithm to obtain a superimposed and fused image of the static reference image and the dynamic real-time image. The mapping unit is used to map the preset digital model of the ablation catheter to the ablation target location of the superimposed fused image to obtain the three-dimensional spatial coordinates of the ablation target. A marking unit is used to generate a visual marker matching the ablation catheter on the superimposed fusion image based on the three-dimensional spatial coordinates of the ablation target and in response to an ablation operation command. The positioning unit is used to generate an ablation area distribution map of the ablation target based on the visualization markers.
[0007] Optionally, the data acquisition unit is specifically used for: Obtain angiographic data of the ablation target and use the angiographic data as the static reference image; The static reference image is processed using an artificial intelligence segmentation algorithm to obtain the segmentation result of the ablation target; Based on the segmentation results, a three-dimensional model of the ablation target is constructed, and the location of the ablation target point is determined from the three-dimensional model.
[0008] Optionally, the data acquisition unit is further configured to: Real-time acquisition of digital subtraction angiography images of the ablation target, and using the digital subtraction angiography images as the initial images; The initial image is subjected to fully automatic contour segmentation to obtain the real-time dynamic contour information of the ablation target; Based on the real-time dynamic contour information and combined with the initial image, the dynamic real-time image of the ablation target is generated.
[0009] Optionally, the registration unit is specifically used for: By using a feature point registration algorithm, feature extraction is performed on the static reference image and the dynamic real-time image respectively to obtain the three-dimensional anatomical feature points of the static reference image and the two-dimensional feature points of the dynamic real-time image. By using the iterative nearest point algorithm, the three-dimensional anatomical feature points are projected onto a two-dimensional plane and matched with the two-dimensional feature points in the dynamic real-time image to obtain the initial registration transformation matrix. Dynamic elastic deformation compensation is performed on the initial registration transformation matrix to obtain real-time registration parameters; Based on the real-time registration parameters, the 3D model and the dynamic real-time image are spatially superimposed to generate the superimposed and fused image.
[0010] Optionally, the mapping unit is specifically used for: Construct a digital model of the ablation catheter that matches the ablation catheter; Real-time acquisition of the positional change information of the ablation catheter in the dynamic real-time image; Based on the location change information, the digital model of the ablation catheter is dynamically mapped to the corresponding position in the overlaid fusion image; The corresponding position is used as the ablation target point to obtain the three-dimensional spatial coordinates of the ablation target point.
[0011] Optionally, the ablation operation instruction includes multiple single ablation operations; The marking unit is specifically used for: In response to the single ablation operation command, a marker image matching the unfolded shape of the ablation catheter is generated at the spatial position of the digital model of the ablation catheter in the superimposed fused image; For each of the marked images corresponding to a single ablation operation, color encoding and timestamp marking are performed to generate a visual marker for the single ablation operation instruction.
[0012] Optionally, the positioning unit is specifically used for: Based on the visual markers of each single ablation operation command, the ablation sub-region corresponding to the single ablation operation command is obtained; The ablation region distribution map is obtained by spatially superimposing the ablation sub-regions corresponding to all the single ablation operations in the ablation operation instruction.
[0013] Optionally, the positioning unit is further configured to: Based on all the ablation sub-regions in the ablation region distribution map and the preset target region of the ablation target, the ablation coverage ratio after executing the ablation operation command is determined; Based on morphological algorithms, gap regions and uncovered regions within the ablation regions are obtained according to the distribution of all the ablation sub-regions; Based on the gap area and the uncovered area, and in conjunction with the ablation coverage ratio, an ablation assessment report for the ablation area is generated.
[0014] Optionally, it further includes: a supplementary unit, the supplementary unit being used for: Based on the spatial distribution of the gap regions and the uncovered regions, determine the geometric features of each of the gap regions and the uncovered regions; Based on the geometric features of the gap region and the uncovered region, the location of the supplementary ablation target and the movement path of the ablation catheter are determined.
[0015] Optionally, the supplementary unit is specifically used for: Based on the spatial distribution of the gap region and the uncovered region, determine the area, perimeter, and minimum distance from the ablated region corresponding to the gap region and the uncovered region, respectively; The gap regions and the uncovered regions are sorted to determine the ablation order; The movement path is generated based on the ablation order and the location of the supplementary ablation target using a path planning algorithm.
[0016] The ablation area localization system of this invention acquires static reference images containing the location of the ablation target and real-time dynamic images through a data acquisition unit, providing complete and real-time image data support for localization and avoiding localization deviations caused by incomplete information from a single image. A registration unit registers the static reference image and the real-time dynamic image using a registration algorithm to obtain a superimposed and fused image, effectively combining static anatomical information and real-time dynamic information, solving the problem of their relative independence and incompatibility, ensuring accurate target location even in dynamic scenarios. A mapping unit maps a preset digital model of the ablation catheter to the ablation target location in the superimposed and fused image to obtain three-dimensional spatial coordinates. With the accurate mapping of the digital model, a spatial association between the ablation catheter and the target location is established, avoiding subjective judgment errors regarding the target location. A marking unit generates visual markers matching the ablation catheter based on the three-dimensional spatial coordinates in response to ablation operation commands, providing clear and accurate identification of the ablation target location and reducing ambiguity in target identification. A localization unit generates an ablation area distribution map based on the visual markers, achieving a systematic presentation of the ablation target location and making target localization more comprehensive and accurate. In summary, through the synergistic effect of the aforementioned units, this invention gradually improves the positioning accuracy of the ablation region from data acquisition, image fusion, spatial mapping, precise marking to region presentation. Attached Figure Description
[0017] Figure 1 This is a schematic diagram of the ablation area positioning system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of image fusion and ablation marking in an embodiment of the present invention. Detailed Implementation
[0018] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Although some embodiments of the present invention are shown in the drawings, it should be understood that the present invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the present invention. It should be understood that the accompanying drawings and embodiments of the present invention are for illustrative purposes only and are not intended to limit the scope of protection of the present invention.
[0019] It should be understood that the various steps described in the method embodiments of the present invention may be performed in different orders and / or in parallel. Furthermore, the method embodiments may include additional steps and / or omit the steps shown. The scope of the present invention is not limited in this respect.
[0020] The term "comprising" and its variations as used herein are open-ended, meaning "including but not limited to"; the term "based on" means "at least partially based on"; the term "one embodiment" means "at least one embodiment"; the term "another embodiment" means "at least one additional embodiment"; the term "some embodiments" means "at least some embodiments"; and the term "optionally" means "optional embodiments". Definitions of other terms will be given in the following description. It should be noted that the concepts of "first," "second," etc., mentioned in this invention are used only to distinguish different devices, modules, or units, and are not intended to limit the order of functions performed by these devices, modules, or units or their interdependencies.
[0021] It should be noted that the terms "a" and "a plurality of" used in this invention are illustrative rather than restrictive. Those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".
[0022] Combination Figure 1 As shown, an ablation area localization system provided in this embodiment of the invention includes: The data acquisition unit is used to acquire a static reference image of the ablation target and to acquire a dynamic real-time image of the ablation target in real time. The static reference image contains the ablation target location of the ablation target.
[0023] Specifically, the static reference images acquired by the data acquisition unit can be cardiac computed tomography angiography (CTA) images or magnetic resonance imaging (MRI) images. These images can clearly present key anatomical structures, such as the left atrium, right atrium, superior vena cava, and atrial septum, thus accurately containing the location of the ablation target and providing a solid preoperative anatomical foundation for localization. The dynamic real-time images acquired in real time can be digital subtraction angiography (DSA) images during the ablation process, which can capture the dynamic changes of the ablation target in real time. This makes up for the shortcomings of the existing technology, where static images are disconnected from the dynamic scene, and provides comprehensive and real-time image data support for subsequent accurate localization, ensuring that the localization process can take into account both the accuracy of ablation planning and the real-time nature of the scene during ablation.
[0024] The registration unit is used to register the static reference image and the dynamic real-time image using a registration algorithm to obtain a superimposed and fused image of the static reference image and the dynamic real-time image.
[0025] Specifically, the registration unit registers the static reference image with the dynamic real-time image through a registration algorithm and generates an overlay fused image. This breaks the limitation of the relative independence of the static reference image and the dynamic real-time image in the existing technology, and realizes the fusion of the two types of images. This allows the ablation target to still correspond to the precise anatomical position determined before the operation during the dynamic ablation process, eliminating the target positioning offset problem caused by image fragmentation. It provides a unified and accurate image carrier for subsequent target coordinate calculation and marking, and improves the basic accuracy of positioning.
[0026] The mapping unit is used to map the preset digital model of the ablation catheter to the ablation target location of the superimposed fused image, so as to obtain the three-dimensional spatial coordinates of the ablation target.
[0027] Specifically, based on the registration of the superimposed and fused images, the digital model is accurately mapped to the corresponding ablation target according to the position of the ablation target in the fused image. By reading the spatial coordinate parameters of the digital model in the fused image, the three-dimensional spatial coordinates of the ablation target are obtained.
[0028] A marking unit is used to generate a visual marker matching the ablation catheter on the superimposed fusion image based on the three-dimensional spatial coordinates of the ablation target and in response to an ablation operation command.
[0029] Specifically, after obtaining the three-dimensional spatial coordinates of the ablation target, the marking unit performs a marking operation on the superimposed fusion image based on the three-dimensional spatial coordinates when it receives the ablation operation command. Combined with the structural morphology of the ablation catheter, it generates a visual marker that matches the unfolded shape of the ablation catheter, ensuring that the marking position corresponds precisely to the ablation target.
[0030] The positioning unit is used to generate an ablation area distribution map of the ablation target based on the visualization markers.
[0031] Specifically, the localization unit records the visual markers generated after each ablation operation, summarizes the marker information of all single ablation targets, integrates and sorts them according to the spatial relationship of each marker, integrates the scattered individual target markers into a systematic regional distribution presentation, and finally constructs an ablation area distribution map that can completely present the distribution of all ablation targets.
[0032] The ablation area localization system of this invention acquires static reference images containing the location of the ablation target and real-time dynamic images through a data acquisition unit, providing complete and real-time image data support for localization and avoiding localization deviations caused by incomplete information from a single image. A registration unit registers the static reference image and the real-time dynamic image using a registration algorithm to obtain a superimposed and fused image, effectively combining static anatomical information and real-time dynamic information, solving the problem of their relative independence and incompatibility, ensuring accurate target location even in dynamic scenarios. A mapping unit maps a preset digital model of the ablation catheter to the ablation target location in the superimposed and fused image to obtain three-dimensional spatial coordinates. With the accurate mapping of the digital model, a spatial association between the ablation catheter and the target location is established, avoiding subjective judgment errors regarding the target location. A marking unit generates visual markers matching the ablation catheter based on the three-dimensional spatial coordinates in response to ablation operation commands, providing clear and accurate identification of the ablation target location and reducing ambiguity in target identification. A localization unit generates an ablation area distribution map based on the visual markers, achieving a systematic presentation of the ablation target location and making target localization more comprehensive and accurate. In summary, through the synergistic effect of the aforementioned units, this invention gradually improves the positioning accuracy of the ablation region from data acquisition, image fusion, spatial mapping, precise marking to region presentation.
[0033] Optionally, the data acquisition unit is specifically used for: Obtain angiographic data of the ablation target and use the angiographic data as the static reference image; The static reference image is processed using an artificial intelligence segmentation algorithm to obtain the segmentation result of the ablation target; Based on the segmentation results, a three-dimensional model of the ablation target is constructed, and the location of the ablation target point is determined from the three-dimensional model.
[0034] Specifically, the process begins by acquiring angiographic data through computed tomography (CTA) and exporting DICOM-compliant image data as a static baseline image. Then, an AI-powered automatic segmentation algorithm is invoked to precisely process the static baseline image, separating key anatomical structures of the target tissue to obtain the segmentation result of the ablation target. Based on the segmented anatomical data, a three-dimensional model reconstruction of the ablation target is completed using an image fusion system. Combined with lesion region characteristics, such as pulmonary vein vestibular lesions related to atrial fibrillation, the location of the ablation target is accurately determined from the three-dimensional model. The entire process relies on CTA image acquisition technology, AI automatic segmentation algorithms, and three-dimensional reconstruction technology to achieve its core functions. In a preferred embodiment of the present invention, target ablation is performed on the heart. The data acquisition unit first performs a cardiac CTA examination, collects and exports DICOM format angiography data as a static reference image. Then, an automatic segmentation algorithm is used to segment key anatomical structures such as the left atrium, pulmonary veins, and atrial septum in the static reference image, obtaining independent image data for each structure. Based on the segmentation results, the image fusion system quickly reconstructs a three-dimensional model of the heart. Through this model, the connection morphology and lesion extent between the pulmonary veins and the left atrium are observed. Finally, the abnormal electrical activity area around the pulmonary vein vestibule is determined in the three-dimensional model as the ablation target location. In particular, the artificial intelligence automatic segmentation algorithm usually relies on a pre-trained deep learning model (such as U-Net, 3DU-Net, and other network architectures suitable for medical image segmentation). The training dataset of this model contains a large number of CTA image samples annotated with key cardiac anatomical structures. The annotated structures cover key target tissue structures such as the left atrium, right atrium, superior vena cava, atrial septum, fossa ovalis, pulmonary veins, left ventricle, right ventricle, and aorta. Through multiple rounds of iterative training, the model learns the grayscale features, morphological features, and spatial relationship rules of each anatomical structure.
[0035] In this embodiment of the invention, by using CTA images as static reference images, high-resolution anatomical information of the ablation target can be obtained. The application of AI automatic segmentation algorithm realizes efficient and accurate separation of key anatomical structures, avoids subjective errors of manual segmentation, and improves the consistency and accuracy of segmentation results. Furthermore, the three-dimensional model constructed based on the segmentation results can intuitively and comprehensively present the three-dimensional anatomical relationship of the ablation target, making the determination of the ablation target location more scientific and accurate, and effectively solving the positioning deviation problem caused by relying on two-dimensional images to judge the target in traditional technology.
[0036] Optionally, the data acquisition unit is further configured to: Real-time acquisition of digital subtraction angiography images of the ablation target, and using the digital subtraction angiography images as the initial images; The initial image is subjected to fully automatic contour segmentation to obtain the real-time dynamic contour information of the ablation target; Based on the real-time dynamic contour information and combined with the initial image, the dynamic real-time image of the ablation target is generated.
[0037] Specifically, the system first uses a video splitter to split the real-time image signal output from the digital subtraction angiography (DSA) device during the ablation process into two. One signal is connected to the video acquisition card of the image fusion system host to achieve real-time acquisition of DSA images, which are then used as the initial image. Subsequently, an artificial intelligence automatic segmentation algorithm is called to process the initial image, quickly and accurately segmenting the two-dimensional anatomical contour of the heart to obtain the real-time dynamic contour information of the ablation target. Specifically, especially for target area ablation of the heart, the system starts a pre-trained lightweight AI segmentation model (adapted to real-time processing scenarios during ablation, such as the optimized U-Net and MobileNet combined with the segmentation head architecture). This model has been trained through a large number of DSA image samples during ablation and has learned the two-dimensional anatomical contour of the target tissue (such as the contour features of core structures such as the atria, ventricles, and main blood vessels), the gray-scale distribution pattern under contrast agent filling, and the dynamic change pattern, which can improve the computing speed while ensuring segmentation accuracy. The model performs frame-by-frame computation on each pre-processed DSA initial image. The encoder rapidly extracts edge features, grayscale gradient features, and regional connectivity features from the image, focusing on the image region corresponding to the anatomical structure. The decoder reconstructs these features, and combined with a dynamic threshold adjustment mechanism, accurately distinguishes the anatomical contour from background areas (such as perivascular tissue and areas not covered by contrast agent), automatically segmenting the two-dimensional anatomical contour of the target tissue. Notably, to address dynamic changes during ablation, such as cardiac pulsation and catheter movement, the algorithm compares the contour information of each frame after segmentation with the contour information of the previous frame. Optical flow is used to track the dynamic displacement of the contour, achieving continuous real-time updates of the dynamic contour. The segmented real-time dynamic contour information is then combined with the original acquired DSA initial image, and through integrated processing by an image fusion system, a dynamic real-time image clearly showing the dynamic changes of the ablation target is generated.
[0038] In this embodiment of the invention, a video splitter enables real-time acquisition and splitting of DSA images, which does not affect the regular observation of DSA images and provides high-quality initial data for the generation of dynamic real-time images. The application of AI automatic segmentation algorithm realizes fully automatic contour segmentation of the initial image without manual intervention, which not only improves the efficiency of contour extraction, but also ensures the accuracy and consistency of real-time dynamic contour information, avoiding subjective errors and time delays caused by manual segmentation. Combining dynamic contour information with the initial image to generate dynamic real-time images can intuitively present the dynamic changes of the ablation target, ensuring that the registration process can adapt to the dynamic changes of the ablation target, thereby improving the accuracy of the overall ablation area positioning.
[0039] Optionally, the registration unit is specifically used for: By using a feature point registration algorithm, feature extraction is performed on the static reference image and the dynamic real-time image respectively to obtain the three-dimensional anatomical feature points of the static reference image and the two-dimensional feature points of the dynamic real-time image. By using the iterative nearest point algorithm, the three-dimensional anatomical feature points are projected onto a two-dimensional plane and matched with the two-dimensional feature points in the dynamic real-time image to obtain the initial registration transformation matrix. Dynamic elastic deformation compensation is performed on the initial registration transformation matrix to obtain real-time registration parameters; Based on the real-time registration parameters, the 3D model and the dynamic real-time image are spatially superimposed to generate the superimposed and fused image.
[0040] Specifically, for target ablation of the heart, a feature point registration algorithm is first used to extract three-dimensional anatomical feature points such as the left atrium, atrial septum, and pulmonary vein openings from the three-dimensional cardiac model corresponding to static reference images (such as CTA images), while two-dimensional feature points such as the cardiac edge and vascular branches are extracted from dynamic real-time images (such as DSA images). Then, the extracted three-dimensional anatomical feature points are projected onto a two-dimensional plane using the Iterative Closest Point (ICP) algorithm and matched point by point with the two-dimensional feature points of the dynamic real-time images to obtain an initial registration transformation matrix. Next, dynamic elastic deformation compensation technology is used to correct the initial registration transformation matrix in real time to obtain accurate real-time registration parameters for the anatomical structure displacement caused by cardiac pulsation and respiratory movements during ablation. Finally, based on the real-time registration parameters, the three-dimensional cardiac model constructed before the operation is spatially aligned and superimposed with the dynamic real-time images during the ablation process using an image fusion system to generate a superimposed fused image that combines three-dimensional anatomical information and real-time dynamic information. Specifically, the coordinates of three-dimensional anatomical feature points extracted from static reference images (CTA images), such as the center of the pulmonary vein opening, the geometric center of the atrial septum, and the coordinates of the superior vena cava opening, are first standardized and uniformly converted into three-dimensional coordinate data (X1, Y1, Z1) in the world coordinate system. Simultaneously, the coordinates of two-dimensional feature points extracted from dynamic real-time images (DSA images), such as atrial contour inflection points and vascular branch intersections, are calibrated and converted into two-dimensional coordinate data (X2, Y2) in the image pixel coordinate system. Imaging parameters of the dynamic real-time images (such as focal length and imaging angle) are also recorded. Based on the imaging geometry model of the DSA images (such as a perspective projection model), the standardized three-dimensional anatomical feature points (X1, Y1, Z1) are projected onto a two-dimensional plane, and the corresponding two-dimensional projection coordinates (X1', Y1') of each three-dimensional feature point are calculated. During the projection process, the real-time imaging angle of the DSA device during ablation is considered to ensure that the projection result is consistent with the imaging perspective of the dynamic real-time images, simulating the presentation effect of the three-dimensional structure in the two-dimensional image. Using the Euclidean distance metric, the distance between the two-dimensional projected coordinates (X1', Y1') of each three-dimensional feature point and all two-dimensional feature points (X2, Y2) in the dynamic real-time image is calculated. The nearest two-dimensional feature point is matched for each projection point, and an initial feature point correspondence set {(X1', Y1'), (X2, Y2)} is constructed to initially establish the association between three-dimensional features and two-dimensional features.The ICP algorithm iterative process is then initiated. First, based on the initial set of corresponding point pairs, the initial transformation matrix (including translation and rotation parameters) is solved using the least squares method to minimize the overall error between the projected coordinates of the 3D feature points and the matched 2D feature points. Next, the error value of each corresponding point pair is calculated, and abnormal corresponding point pairs exceeding a preset error threshold (set based on clinical anatomical accuracy requirements) are removed (such as mismatched points caused by image noise). Based on the filtered valid corresponding point pairs, the transformation matrix is re-solved until the change in transformation matrix parameters between two consecutive iterations is less than the set threshold, at which point the iteration terminates. When the iteration converges, the final transformation matrix is the initial registration transformation matrix. This matrix contains the spatial alignment parameters between the 3D model and the dynamic real-time image, enabling preliminary coordinate mapping between the 3D anatomical structure and the 2D dynamic image.
[0041] In this embodiment of the invention, key feature points of two types of images are accurately extracted by the feature point registration algorithm, providing a reliable matching basis for registration and ensuring accurate correspondence at the feature level; the application of the iterative nearest point algorithm realizes efficient and accurate matching of three-dimensional feature points and two-dimensional feature points, improving the accuracy of the initial registration transformation matrix.
[0042] Optionally, the mapping unit is specifically used for: Construct a digital model of the ablation catheter that matches the ablation catheter; Real-time acquisition of the positional change information of the ablation catheter in the dynamic real-time image; Based on the location change information, the digital model of the ablation catheter is dynamically mapped to the corresponding position in the overlaid fusion image; The corresponding position is used as the ablation target point to obtain the three-dimensional spatial coordinates of the ablation target point.
[0043] Specifically, firstly, the digital model of the ablation catheter, which is matched with the ablation catheter, is usually petal-shaped. Therefore, based on the actual structural parameters of the petal-shaped ablation catheter (such as the number of petals, unfolding angle, length, and diameter), a digital model that perfectly matches the physical ablation catheter needs to be constructed using 3D modeling technology. Subsequently, with the help of real-time data acquisition capabilities of dynamic real-time imaging (such as DSA imaging and ICE ultrasound imaging), combined with image recognition technology, the spatial position change information of the ablation catheter during the ablation process (including translation, rotation, and other displacement parameters) is captured in real time. Based on the captured position change data, the digital model of the ablation catheter is dynamically superimposed onto the corresponding physical position in the registered superimposed fusion image using a coordinate mapping algorithm, ensuring that the spatial attitude of the digital model and the physical catheter is synchronized in real time. Finally, by reading the spatial coordinate parameters of the digital model in the 3D model associated with the superimposed fusion image, the corresponding position is directly determined as the ablation target, thereby accurately obtaining the 3D spatial coordinates of the ablation target.
[0044] In this embodiment of the invention, a digital model that perfectly matches the physical ablation catheter is constructed using 3D modeling technology, providing a precise virtual reference carrier for target localization and ensuring structural consistency in the mapping process. The combination of image recognition technology and coordinate mapping algorithm enables real-time synchronization of the position changes between the digital model and the physical catheter, allowing the target position to be dynamically updated as the catheter moves, avoiding localization lag caused by static marking. The spatial coordinates of the digital model are directly used as the 3D spatial coordinates of the ablation target, accurately associating the catheter position with the target coordinates, abandoning the method of relying on subjective judgment to determine the target, and significantly improving the accuracy of obtaining the 3D spatial coordinates of the ablation target.
[0045] Optionally, the ablation operation instruction includes multiple single ablation operations; The marking unit is specifically used for: In response to the single ablation operation command, a marker image matching the unfolded shape of the ablation catheter is generated at the spatial position of the digital model of the ablation catheter in the superimposed fused image; For each of the marked images corresponding to a single ablation operation, color encoding and timestamp marking are performed to generate a visual marker for the single ablation operation instruction.
[0046] Specifically, the system first pre-sets the ablation operation command to include the execution logic of multiple single ablation operations. When a single ablation operation command is received, based on the real-time spatial position of the digital model of the ablation catheter in the superimposed fusion image, a marker image is created using graphics generation technology to match the unfolding shape of the pulsed electric field ablation (PFA) petal-shaped catheter at a 1:1 ratio, ensuring that the shape of the marker image is consistent with the actual effective range of the ablation catheter. Subsequently, color coding technology is used to assign a unique color identifier to the marker image corresponding to different single ablation operations. At the same time, time stamp recording technology is used to add operation time information to each marker image, embedding the color identifier and time stamp information into the marker image, ultimately generating a visual marker for single ablation operations that combines morphological matching, distinguishability, and traceability.
[0047] In summary, combining Figure 2As shown, preoperative image data of the patient was first acquired through cardiac computed tomography angiography (CTA). After being exported in a DICOM-compliant format, the data was imported into an image fusion processing system. An artificial intelligence (AI) automatic segmentation algorithm was used to accurately separate key anatomical structures such as the left atrium and atrial septum. Based on the segmentation results, a CTA-3D model was reconstructed and optimized using a multi-color labeling method. During the ablation process, real-time digital subtraction angiography (DSA) images were acquired through a video splitter. Simultaneously, an AI algorithm was used to segment the two-dimensional anatomical contour of the heart in real time. Then, a registration algorithm was used to accurately register the preoperative CTA-3D model with the DSA images during the ablation process, generating a superimposed fused image of the DSA and 3D models. During the ablation operation, a pre-set digital model of the ablation catheter was dynamically mapped to the target location of the fused image. Based on the three-dimensional spatial coordinates of the target point, petal-shaped ablation markers matching the shape of the ablation catheter were generated. Through the accumulation and integration of multiple target markers, the distribution of the ablation area was fully presented.
[0048] In this embodiment of the invention, a marker image matching the unfolded shape of the ablation catheter is generated using graphic generation technology, enabling the visual marker to accurately reflect the actual effective range of a single ablation, thus avoiding the problem of the marker being out of sync with the actual ablation area. Color coding technology enables rapid differentiation of different single ablation operation markers, and timestamp recording technology provides a clear operational timeline for each marker. The combination of the two can intuitively distinguish the location, range, and timeline relationship of different ablation targets, effectively solving the problem of difficulty in remembering ablation areas in traditional technologies. At the same time, it provides clear and traceable marker data for subsequent ablation area distribution organization, improving the standardization and traceability of ablation operations.
[0049] Optionally, the positioning unit is specifically used for: Based on the visual markers of each single ablation operation command, the ablation sub-region corresponding to the single ablation operation command is obtained; The ablation region distribution map is obtained by spatially superimposing the ablation sub-regions corresponding to all the single ablation operations in the ablation operation instruction.
[0050] Specifically, firstly, region extraction technology is used to accurately extract the spatial range covered by each marker from the superimposed fusion image based on the visual markers corresponding to each single ablation operation (including marker images matching the deployment morphology of the ablation catheter, color codes, and timestamps). This range is determined as the ablation sub-region corresponding to each single ablation operation, ensuring that the ablation sub-region is consistent with the actual ablation range. Then, a spatial overlay algorithm is used to overlay and integrate all the ablation sub-regions corresponding to each single ablation operation according to their three-dimensional spatial coordinates. By calculating the spatial intersection and union of each sub-region, the duplicate overlay parts are removed and the complete coverage is retained, finally generating an ablation region distribution map that can completely present the spatial distribution of all ablation regions.
[0051] In this embodiment of the invention, the ablation sub-region corresponding to each single ablation operation is accurately extracted by region extraction technology, ensuring the accuracy of the individual ablation range and avoiding the deviation between the sub-region and the actual ablation range. The application of the spatial overlay algorithm realizes the systematic integration of all ablation sub-regions, transforming the scattered individual ablation ranges into an overall ablation region distribution, effectively solving the problem of difficult memory of ablation regions in traditional technologies. At the same time, it provides a clear visual basis for judging whether there is missed ablation or repeated ablation, ensuring the integrity and accuracy of the ablation operation.
[0052] Optionally, the positioning unit is further configured to: Based on all the ablation sub-regions in the ablation region distribution map and the preset target region of the ablation target, the ablation coverage ratio after executing the ablation operation command is determined; Based on morphological algorithms, gap regions and uncovered regions within the ablation regions are obtained according to the distribution of all the ablation sub-regions; Based on the gap area and the uncovered area, and in conjunction with the ablation coverage ratio, an ablation assessment report for the ablation area is generated.
[0053] Specifically, firstly, the total spatial range of all ablation sub-regions in the ablation area distribution map is extracted using spatial proportion calculation technology. This is then combined with the spatial range of the pre-defined ablation target area (such as the pulmonary vein vestibular lesion area in atrial fibrillation patients) based on preoperative static baseline images. The ablation coverage ratio is determined by calculating the spatial proportion of these two areas. Subsequently, morphological algorithms (such as erosion expansion and connected component analysis) are used to process the distribution of all ablation sub-regions, identifying non-overlapping gaps between ablation sub-regions and uncovered areas not covered by any ablation sub-regions. The spatial coordinates and range information of these two types of areas are obtained. Specifically, connected component analysis algorithms are used to perform spatial clustering on all ablation sub-regions. Sub-regions that are adjacent, overlapping, or within a preset threshold (based on the ablation catheter's effective range, such as 2 mm) are identified as connected regions and merged to obtain several independent fusion ablation regions. At the same time, the three-dimensional contour, volume, and spatial coordinate range of each fusion ablation region are recorded to identify the core areas that have been covered. Morphological expansion is performed on the merged ablation area. Structured elements (such as circles or petals) matching the unfolded shape of the ablation catheter are selected, and the area boundary is expanded at a preset expansion step size (such as 0.5 mm) to fill the tiny gaps between the ablation sub-regions (such as small gaps caused by marking errors), resulting in a fully expanded coverage area. This ensures the accuracy of subsequent gap identification and avoids misjudging small errors as gap areas. The three-dimensional contour data of the pre-defined ablation target area, determined preoperatively based on static reference images (CTA images), is extracted. Spatial difference operations are performed between the expanded fully covered area and the pre-defined target area to obtain two types of candidate areas: areas within the pre-defined target area not included in the fully covered area (preliminary uncovered areas), and blank areas within the fully covered area not covered by the original ablation sub-regions (preliminary gap areas). Morphological erosion is performed on the fully covered area, with the erosion step size consistent with the expansion step size, to restore the true boundary of the fused ablation area. Simultaneously, the initial gap area and the initial uncovered area are corrected by erosion. The corrected candidate areas are then evaluated for features. If an area is located within a preset target area and is surrounded by at least two fused ablation areas but not covered by any original ablation sub-regions, it is identified as a gap area. If an area is located within a preset target area and is not covered by any fused ablation area or any original ablation sub-region, it is identified as an uncovered area. Three-dimensional spatial coordinates (center point coordinates, boundary vertex coordinates) and extent information (volume, surface area, two-dimensional projected area) of the two types of areas are obtained through geometric calculations, ultimately outputting structured area data. Finally, the relevant data on ablation coverage ratio, gap areas, and uncovered areas are integrated and compiled into an ablation assessment report containing coverage information, details of missing areas, and quantitative indicators according to a preset format using report generation technology.
[0054] In this embodiment of the invention, by accurately identifying gap areas and uncovered areas, the problem of omissions in manual judgment is avoided, providing clear targets for supplementary ablation; the report generation technology systematically organizes various evaluation data into an ablation evaluation report, clearly presenting quantitative indicators of ablation effect and specific defects, effectively improving the accuracy and completeness of ablation surgery and reducing the risk of residual lesions.
[0055] Optionally, it further includes: a supplementary unit, the supplementary unit being used for: Based on the spatial distribution of the gap regions and the uncovered regions, determine the geometric features of each of the gap regions and the uncovered regions; Based on the geometric features of the gap region and the uncovered region, the location of the supplementary ablation target and the movement path of the ablation catheter are determined.
[0056] Specifically, firstly, geometric feature extraction technology is used to analyze the spatial distribution data of the gap areas and uncovered areas clearly identified in the ablation assessment report, and calculate and obtain the geometric features of each area, including area, perimeter, three-dimensional coordinates of the center point, boundary contour parameters, and minimum distance to the ablated sub-area. Then, a target optimization algorithm is used to determine the precise location of the supplementary ablation target by combining these geometric features (such as prioritizing areas with larger areas and close connections to the ablated areas). At the same time, a path planning algorithm (such as based on the shortest path principle) is used to plan a non-redundant and efficient catheter movement path according to the spatial distribution order of the supplementary ablation target and the current position of the ablation catheter.
[0057] In this embodiment of the invention, key missed areas are preferentially selected as supplementary targets based on regional geometric features to ensure the targeting and necessity of supplementary ablation; it reduces the ineffective movement of the ablation catheter in the body, reduces surgical trauma and operation time, and the complete evaluation report formed by combining the ablation coverage ratio makes the supplementary ablation plan more scientific and operable.
[0058] Optionally, the supplementary unit is specifically used for: Based on the spatial distribution of the gap region and the uncovered region, determine the area, perimeter, and minimum distance from the ablated region corresponding to the gap region and the uncovered region, respectively; The gap regions and the uncovered regions are sorted to determine the ablation order; The movement path is generated based on the ablation order and the location of the supplementary ablation target using a path planning algorithm.
[0059] Specifically, firstly, spatial geometric calculation technology is used to calculate the area (converted by three-dimensional spatial integration or two-dimensional projected area), perimeter (length accumulation after extracting the boundary contour of the area), and minimum distance to the ablated area (by calculating the shortest spatial distance between the boundary of the area and the boundary of the ablated sub-area) of each region based on the spatial distribution data of the gap region and the uncovered area. Then, a priority sorting algorithm is used, combined with the ablation requirements, such as prioritizing the processing of areas with larger areas and close connections to the ablated areas, to sort the gap region and uncovered area according to their importance and determine the optimal ablation order. Finally, a path planning algorithm is used to plan a catheter movement path that is non-redundant, efficient, and conforms to the surgical operation logic, based on the sorted ablation order and the three-dimensional spatial coordinates of the supplementary ablation target points and the current position of the ablation catheter. Specifically, in a preferred embodiment of the present invention, for target area ablation of the heart, a shortest path algorithm (such as Dijkstra's algorithm) is used as the basic path planning framework. Starting from the current position of the catheter, the straight-line distance and spatial path between adjacent target points are calculated sequentially according to the sorted ablation target point order to generate preliminary candidate paths. At the same time, considering the maneuverability of the catheter in the body, based on the anatomical orientation of the cardiac blood vessels, 2-3 alternative paths (such as detour paths that bypass vascular branches) are generated to avoid a single path being unexecutable due to anatomical obstacles. Each candidate path is evaluated in multiple dimensions: first, a spatial collision detection algorithm is used to verify whether the path has a collision risk with important anatomical structures, and paths that violate the safety distance constraints are eliminated; second, the total length of each path and the time required for catheter movement are calculated (based on the average movement speed of the catheter), and the efficient path with the shortest length and least time consumption is selected. The selected optimal candidate paths are refined by adding navigation nodes at the starting point, each supplementary ablation target, and key turning points along the path. The spatial coordinates, orientation (e.g., probe direction), and movement speed of the ablation catheter at each node are clearly defined. For potentially narrow passages or complex anatomical areas within the path, intermediate transition nodes are added to optimize the catheter's trajectory, ensuring a path free of redundancy and obstruction. The optimized path nodes are then integrated sequentially to generate complete catheter movement path data, including the three-dimensional coordinates of each node and the direction and distance of movement between nodes.
[0060] In this embodiment of the invention, the optimal ablation sequence is determined based on clinical needs through a priority ranking algorithm, ensuring that important missed areas are given priority for supplementary ablation, thereby improving the targeting and effectiveness of supplementary ablation. The efficient movement path planned by the path planning algorithm reduces the ineffective movement of the ablation catheter in the body, reduces surgical trauma and operation time, and avoids unnecessary interference with surrounding normal tissues during catheter movement, further improving the accuracy, safety and efficiency of the ablation operation.
[0061] While the present invention has been disclosed above, its scope of protection is not limited thereto. Those skilled in the art can make various changes and modifications without departing from the spirit and scope of the present invention, and all such changes and modifications will fall within the scope of protection of the present invention.
Claims
1. An ablation area positioning system, characterized in that, include: The data acquisition unit is used to acquire a static reference image of the ablation target and to acquire a dynamic real-time image of the ablation target in real time. The static reference image contains the ablation target location of the ablation target. The registration unit is used to register the static reference image and the dynamic real-time image using a registration algorithm to obtain a superimposed and fused image of the static reference image and the dynamic real-time image. The mapping unit is used to map the preset digital model of the ablation catheter to the ablation target location of the superimposed fused image to obtain the three-dimensional spatial coordinates of the ablation target. A marking unit is used to generate a visual marker matching the ablation catheter on the superimposed fusion image based on the three-dimensional spatial coordinates of the ablation target and in response to an ablation operation command. The positioning unit is used to generate an ablation area distribution map of the ablation target based on the visualization markers.
2. The ablation area positioning system according to claim 1, characterized in that, The data acquisition unit is specifically used for: Obtain angiographic data of the ablation target and use the angiographic data as the static reference image; The static reference image is processed using an artificial intelligence segmentation algorithm to obtain the segmentation result of the ablation target; Based on the segmentation results, a three-dimensional model of the ablation target is constructed, and the location of the ablation target point is determined from the three-dimensional model.
3. The ablation area positioning system according to claim 1, characterized in that, The data acquisition unit is further configured to: Real-time acquisition of digital subtraction angiography images of the ablation target, and using the digital subtraction angiography images as the initial images; The initial image is subjected to fully automatic contour segmentation to obtain the real-time dynamic contour information of the ablation target; Based on the real-time dynamic contour information and combined with the initial image, the dynamic real-time image of the ablation target is generated.
4. The ablation area positioning system according to claim 2, characterized in that, The registration unit is specifically used for: By using a feature point registration algorithm, feature extraction is performed on the static reference image and the dynamic real-time image respectively to obtain the three-dimensional anatomical feature points of the static reference image and the two-dimensional feature points of the dynamic real-time image. By using the iterative nearest point algorithm, the three-dimensional anatomical feature points are projected onto a two-dimensional plane and matched with the two-dimensional feature points in the dynamic real-time image to obtain the initial registration transformation matrix. Dynamic elastic deformation compensation is performed on the initial registration transformation matrix to obtain real-time registration parameters; Based on the real-time registration parameters, the 3D model and the dynamic real-time image are spatially superimposed to generate the superimposed and fused image.
5. The ablation area positioning system according to claim 1, characterized in that, The mapping unit is specifically used for: Construct a digital model of the ablation catheter that matches the ablation catheter; Real-time acquisition of the positional change information of the ablation catheter in the dynamic real-time image; Based on the location change information, the digital model of the ablation catheter is dynamically mapped to the corresponding position in the overlaid fusion image; The corresponding position is used as the ablation target point to obtain the three-dimensional spatial coordinates of the ablation target point.
6. The ablation area positioning system according to claim 1, characterized in that, The ablation operation instructions include multiple single ablation operations; The marking unit is specifically used for: In response to the single ablation operation command, a marker image matching the unfolded shape of the ablation catheter is generated at the spatial position of the digital model of the ablation catheter in the superimposed fused image; For each of the marked images corresponding to a single ablation operation, color encoding and timestamp marking are performed to generate a visual marker for the single ablation operation instruction.
7. The ablation area positioning system according to claim 6, characterized in that, The positioning unit is specifically used for: Based on the visual markers of each single ablation operation command, the ablation sub-region corresponding to the single ablation operation command is obtained; The ablation region distribution map is obtained by spatially superimposing the ablation sub-regions corresponding to all the single ablation operations in the ablation operation instruction.
8. The ablation area positioning system according to claim 7, characterized in that, The positioning unit is further configured to: Based on all the ablation sub-regions in the ablation region distribution map and the preset target region of the ablation target, the ablation coverage ratio after executing the ablation operation command is determined; Based on morphological algorithms, gap regions and uncovered regions within the ablation regions are obtained according to the distribution of all the ablation sub-regions; Based on the gap area and the uncovered area, and in conjunction with the ablation coverage ratio, an ablation assessment report for the ablation area is generated.
9. The ablation area positioning system according to claim 8, characterized in that, Also includes: Supplementary unit, the supplementary unit being used for: Based on the spatial distribution of the gap regions and the uncovered regions, determine the geometric features of each of the gap regions and the uncovered regions; Based on the geometric features of the gap region and the uncovered region, the location of the supplementary ablation target and the movement path of the ablation catheter are determined.
10. The ablation area positioning system according to claim 9, characterized in that, The supplementary unit is specifically used for: Based on the spatial distribution of the gap region and the uncovered region, determine the area, perimeter, and minimum distance from the ablated region corresponding to the gap region and the uncovered region, respectively; The gap regions and the uncovered regions are sorted to determine the ablation order; The movement path is generated based on the ablation order and the location of the supplementary ablation target using a path planning algorithm.