Targeted puncture navigation method, system, and storage medium

Abstract

The application provides a target puncture navigation method, system and storage medium, the method comprising: constructing a three-dimensional model of a target biological organ based on image data of the target biological organ obtained before surgery, and labeling a target region and a safety boundary on the three-dimensional model; performing spatial registration on the three-dimensional model based on a first image collected during surgery to obtain a fusion coordinate system; mapping pose data of a puncture needle in a surgical field space into the fusion coordinate system, and superimposing and displaying a first position of a needle tip of the puncture needle in the three-dimensional model on the first image; dynamically compensating the first position of the needle tip based on displacement estimated based on a plurality of continuous first images to obtain a second position of the needle tip; when the second position of the needle tip continuously locates within the target region for a first time length, and a constraint condition of the safety boundary is met, controlling a pulsed electric field generator to trigger a pulsed electric field for ablation.

Description

Technical Field

[0001] This application relates to the field of ablation technology, and more specifically, to a targeted puncture navigation method, system, and storage medium. Background Technology

[0002] Hypertrophic obstructive cardiomyopathy (HOCM) is a hereditary cardiomyopathy characterized by asymmetric left ventricular hypertrophy, impaired diastolic function, and dynamic obstruction of the left ventricular outflow tract (LVOT). Patients typically experience chest tightness, chest pain, dyspnea, syncope, heart failure, and even sudden cardiac death.

[0003] Currently, minimally invasive treatment of HOCM is gradually shifting from traditional radiofrequency or alcohol ablation to pulsed electric field (PEF) ablation, due to its advantages such as high tissue selectivity and minimal damage to the conduction system and vascular structures.

[0004] However, existing technologies used in the treatment of HOCM via pulsed electric field ablation have problems such as inaccurate target localization due to single ultrasound guidance and interference from cardiac motion on the puncture path. Summary of the Invention

[0005] This application addresses the shortcomings of existing methods by proposing a targeted puncture navigation method, system, and storage medium, aiming to solve the problems of inaccurate target localization and interference of cardiac motion on the puncture path when existing schemes are applied to the treatment of HOCM by pulsed electric field ablation.

[0006] In a first aspect, embodiments of this application provide a targeted puncture navigation system, comprising: a processor, a controller, a pulsed electric field generator, and an image acquisition unit, wherein, The image acquisition device is used to acquire image data of the target biological organ; The processor is used to perform the following operations: Based on the image data of the target biological organ acquired by the image acquisition device before surgery, a three-dimensional model of the target biological organ is constructed, and the target area and safety boundary are marked on the three-dimensional model; The first image acquired during the operation by the image acquisition device is spatially registered with the three-dimensional model to obtain a fused coordinate system. The first image is an ultrasound image of the first region of the target biological organ. The pose data of the puncture needle in the surgical field space is mapped to the fused coordinate system, and the first position of the tip of the puncture needle in the three-dimensional model is superimposed on the first image. Based on the displacement estimated from multiple consecutive frames of the first image, the first position of the needle tip is dynamically compensated to obtain the second position of the needle tip. The controller is configured to control the pulse electric field generator to trigger a pulse electric field for ablation when the second position of the needle tip remains within the target area for a first duration and the constraint conditions of the safety boundary are met.

[0007] Secondly, embodiments of this application provide a targeted puncture navigation method, including: Based on the image data of the target biological organ obtained before surgery, a three-dimensional model of the target biological organ is constructed, and the target area and safety boundary are marked on the three-dimensional model; The first image acquired during the operation is spatially registered with the three-dimensional model to obtain a fused coordinate system. The first image is an ultrasound image of the first region of the target biological organ. The pose data of the puncture needle in the surgical field space is mapped to the fused coordinate system, and the first position of the tip of the puncture needle in the three-dimensional model is superimposed on the first image. Based on the displacement estimated from multiple consecutive frames of the first image, the first position of the needle tip is dynamically compensated to obtain the second position of the needle tip. When the second position of the needle tip remains within the target area for a first duration and the constraint conditions of the safety boundary are met, the control pulse electric field generator triggers a pulse electric field for ablation.

[0008] Thirdly, embodiments of this application also disclose a signal processing system, including the targeted puncture navigation system as described in the first aspect.

[0009] Fourthly, embodiments of this application also disclose a computer-readable storage medium storing a computer program that, when executed by a processor, implements one or more of the targeted puncture navigation methods described in the embodiments of the second aspect of this application.

[0010] Fifthly, embodiments of this application also disclose a computer program product, including a computer program that, when executed by a processor, implements one or more of the targeted puncture navigation methods described in the embodiments of the second aspect of this application.

[0011] The beneficial technical effects of the technical solutions provided in this application include: By precisely registering high-resolution preoperative images (such as MRI / CT) with intraoperative ultrasound images, and performing real-time pose tracking and motion compensation of the puncture needle within the resulting fusion space, a closed-loop navigation mechanism is formed. This solves the problems of inaccurate target area positioning and interference of cardiac motion on the puncture path caused by single ultrasound guidance. At the same time, it reduces operational risks and the probability of repeated punctures.

[0012] Additional aspects and advantages of this application will be set forth in part in the description which follows, and will become apparent from the description or may be learned by practice of this application. Attached Figure Description

[0013] The above and / or additional aspects and advantages of this application will become apparent and readily understood from the following description of the embodiments taken in conjunction with the accompanying drawings, wherein: Figure 1 This is a schematic diagram of the structure of a targeted puncture navigation system provided in an embodiment of this application; Figure 2 A flowchart illustrating a targeted puncture navigation method provided in an embodiment of this application; Figure 3 This is a schematic diagram of the signal processing system provided in an embodiment of this application. Detailed Implementation The embodiments of this application are described below with reference to the accompanying drawings. It should be understood that the embodiments described below with reference to the accompanying drawings are exemplary descriptions for explaining the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions of the embodiments of this application.

[0014] Those skilled in the art will understand that, unless specifically stated otherwise, the terms "described" and "the" as used herein may also include plural forms. It should be further understood that the term "comprising" as used in this application's specification means the presence of the stated features, integers, steps, operations, elements, and / or components, but does not exclude implementations of other features, information, data, steps, operations, elements, components, and / or combinations thereof supported by this art. It should be understood that when we say an element is "connected" or "coupled" to another element, the element may be directly connected or coupled to the other element, or it may mean that the element and the other element are connected through an intermediate element. Furthermore, "connected" or "coupled" as used herein may include wireless connections or wireless coupling. The term "and / or" as used herein refers to at least one of the items defined by the term; for example, "A and / or B" may be implemented as "A," or as "B," or as "A and B."

[0015] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0016] Currently, minimally invasive treatment for hypertrophic obstructive cardiomyopathy (HOCM) is gradually shifting from traditional radiofrequency or alcohol ablation to pulsed electric field (PEF) ablation due to its advantages such as high tissue selectivity and minimal damage to the conduction system and vascular structures. Some existing techniques utilize a percutaneous apical puncture approach, under ultrasound guidance, to deliver a nanosecond pulsed electrode needle to the hypertrophic area of ​​the interventricular septum for ablation, and combine this with impedance feedback to dynamically adjust energy parameters, thus improving the safety and effectiveness of the treatment.

[0017] However, these techniques still largely rely on intraoperative ultrasound for real-time guidance. While ultrasound offers good real-time performance and safety, its spatial resolution is limited, especially in obese patients, those with lung air interference, or those with variations in cardiac position, making it difficult to clearly visualize the deep structures of the interventricular septum and accurately define the boundaries of the hypertrophic target area. Furthermore, the apical puncture path must traverse beating myocardium, subject to interference from both respiratory movements and cardiac pulsation, leading to potential deviations in the puncture trajectory and affecting the needle's precise arrival at the pre-set target. Even with the introduction of 3D ultrasound and AI modeling, or the use of neural networks to predict ablation sites, it still relies on intraoperative images and lacks rigid or non-rigid registration with high-precision preoperative anatomical structures (such as MRI / CT), making it difficult to achieve sub-millimeter-level spatial correspondence.

[0018] On the other hand, existing navigation methods mostly focus on parameter control during the energy release phase, while paying insufficient attention to positional stability and targeting accuracy during the puncture process. Although some existing technologies propose needle insertion path planning based on preoperative CT, they do not solve the problem of path drift caused by organ deformation during the operation; other existing technologies mention multimodal fusion, but are mainly used for arrhythmia ablation and do not integrate real-time puncture navigation functions.

[0019] Therefore, existing technologies lack a closed-loop navigation mechanism that accurately registers high-resolution preoperative images (such as MRI / CT) with intraoperative ultrasound and performs real-time pose tracking and motion compensation of the puncture needle within this fusion space, resulting in limited targeting accuracy and increased operational risks and the probability of repeated punctures.

[0020] The targeted puncture navigation method proposed in this application aims to solve the problems of inaccurate target area localization of the ventricular septum and interference of cardiac motion on the puncture path caused by single ultrasound guidance in the existing pulsed electric field ablation treatment of HOCM. It provides a spatial registration method based on the fusion of preoperative MRI / CT and intraoperative ultrasound images, and combines it with an electromagnetic or optical tracking system to realize real-time position monitoring and motion compensation of the puncture needle in the fused coordinate system, thereby improving the targeting accuracy and operational stability of percutaneous apical puncture.

[0021] The technical solution of this application and how it solves the above-mentioned technical problems are described in detail below with specific embodiments. It should be noted that the following embodiments can be referenced, borrowed, or combined with each other, and the same terms, similar features, and similar implementation steps in different embodiments will not be described again.

[0022] The targeted puncture navigation system and targeted puncture navigation method proposed in the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0023] In some embodiments, a targeted puncture navigation system is provided, such as Figure 1 As shown, the targeted puncture navigation system 10 includes an image acquisition unit 11, a processor 12, a controller 13, and a pulsed electric field generator 14. The image acquisition unit 11, processor 12, controller 13, and pulsed electric field generator 14 are connected sequentially. The pulsed electric field generator 14 is also connected to the puncture needle.

[0024] In some embodiments, the image uploader 11 is used to acquire image data of the target biological organ, and the processor 12 is used to perform the following operations: Based on the image data of the target biological organ acquired by the image acquisition device before surgery, a three-dimensional model of the target biological organ is constructed, and the target area and safety boundary are marked on the three-dimensional model; The first image acquired during the operation by the image acquisition device is spatially registered with the three-dimensional model to obtain a fused coordinate system; The pose data of the puncture needle in the surgical field space is mapped to the fused coordinate system, and the first position of the tip of the puncture needle in the three-dimensional model is superimposed on the first image. Based on the displacement estimated from multiple consecutive frames of the first image, the first position of the needle tip is dynamically compensated to obtain the second position of the needle tip.

[0025] The controller 13 is used to control the pulse electric field generator 14 to trigger the pulse electric field for ablation when the second position of the needle tip is continuously located within the target area for a first duration and the constraint conditions of the safety boundary are met.

[0026] Optionally, the first image is an ultrasound image of a first region of the target biological organ.

[0027] In some embodiments, when the processor 12 spatially registers the first image acquired during surgery with the three-dimensional model to obtain a fused coordinate system, it has the following functions: The first image is obtained by selecting keyframe images from the ultrasound images of the first region acquired in real time during the operation; Based on the three-dimensional model, a simulated ultrasound image corresponding to the ultrasound image is generated; A feature extraction network is used to extract features from the first image and the simulated ultrasound image respectively to obtain multi-scale feature maps; Based on the multi-scale feature map, the deformation field for registering the first image to the coordinate system of the three-dimensional model is calculated, and the deformation field is used to characterize the non-rigid displacement of the first image relative to the three-dimensional model. Based on the deformation field, the first image and the three-dimensional model are elastically registered to obtain the fused coordinate system.

[0028] In some embodiments, when the processor 12 calculates the deformation field for registering the first image to the coordinate system of the three-dimensional model based on the multi-scale feature map, it has the following functions: The multi-scale feature map is input into the registration network, and the initial deformation field is obtained through feature fusion and regression. The initial deformation field and the registration network are iteratively optimized based on the registration loss function to obtain the deformation field. The registration loss function includes: Data fidelity items used to measure the similarity between the first image and the simulated ultrasound image; A regularization term used to constrain the deformation smoothness of the deformation field; and, Temporal continuity constraints are used to limit the difference between the deformation field of the current frame and the deformation field of the previous frame.

[0029] In some embodiments, when the processor 12 performs elastic registration between the first image and the three-dimensional model based on the deformation field to obtain the fused coordinate system, it has the following functions: Based on the deformation field, a non-rigid spatial transformation is performed on the three-dimensional model to generate a dynamic three-dimensional model corresponding to the current first image. A unified coordinate system is established based on the three-dimensional space where the dynamic three-dimensional model is located, and the first image is mapped to the unified coordinate system to achieve flexible registration between the first image and the dynamic three-dimensional model; The unified coordinate system is defined as the fused coordinate system.

[0030] In some embodiments, the processor 12, when mapping the pose data of the puncture needle in the surgical field space to the fused coordinate system, has the following functions: Acquire the pose data of the puncture needle in the surgical field space, and the pose data has a synchronous temporal relationship with the first image; Based on the preset spatial transformation relationship between the external tracking device coordinate system and the ultrasound image coordinate system, the pose data is converted to the ultrasound image space; Based on the inverse transformation of the deformation field, the transformed pose data is mapped onto the fused coordinate system to determine the real-time position of the puncture needle in the fused coordinate system.

[0031] In some embodiments, when the processor 12 dynamically compensates for the first position of the needle tip based on the displacement estimated from multiple consecutive frames of the first image to obtain the second position of the needle tip, it has the following functions: Based on the registration deformation field between multiple consecutive frames of the first image, the motion phase of the target biological organ at the current moment is determined. Using a pre-trained neural network model, the motion trend vector of the target biological organ at the next moment is predicted based on the motion phase and historical displacement information. When it is determined that the needle tip has entered the state of approaching the target area, the first position of the needle tip is compensated for by reverse motion based on the motion trend vector to obtain the second position of the needle tip.

[0032] In some embodiments, the processor 12 is further configured to: if it is detected that the distance between the needle tip and the boundary of the target area is less than a first threshold and the moving speed of the needle tip is less than a second threshold, determine that the needle tip has entered a state of approaching the target area.

[0033] In some embodiments, the processor 12 is further configured to: estimate the overall displacement of the target biological organ based on the positional changes of the diaphragm in multiple consecutive frames of the first image; When the overall displacement exceeds a preset safety threshold, a breathing control prompt message is output, and the current navigation guidance is paused; When the overall displacement of the target biological organ is detected to return to within the safety threshold, the navigation guidance is restored.

[0034] In some embodiments, before the controller 13 controls the pulse electric field generator 14 to trigger pulses for electric field ablation, it is also used to light up the status indicator light to a preset color to indicate that the target is ready; the processor 12 is also used to mark the qualified area that meets the safety distance constraint on the ultrasound image displayed on the operation interface in a way that highlights the graphic overlay.

[0035] In some embodiments, when the controller 13 controls the pulse electric field generator 14 to trigger a pulse for electric field ablation, it is specifically used to: send a control signal to the pulse electric field generator 13; and when the pulse electric field generator 13 receives the control signal and verifies in real time that the second position of the needle tip is within the target area, it outputs a pulse.

[0036] In some embodiments, the processor 13 is further configured to: during the ablation process, if the pose drift of the puncture needle is continuously detected to be greater than a fifth threshold within a second duration, control the pulse electric field generator 14 to stop pulse output.

[0037] In the above embodiments, the constraint conditions of the safety boundary include at least one of the following: The distance between the second position of the needle tip and the aortic valve annulus is greater than or equal to the third threshold; The distance between the second position of the needle tip and the His bundle travel path is greater than or equal to the fourth threshold. The second position of the needle tip avoids the left bundle branch trunk region.

[0038] It should be understood that the aforementioned processor may include a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), a FPGA (Field-Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor may also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0039] The targeted puncture navigation system provided in this application embodiment can achieve the following: Figure 2 The various processes implemented in the method embodiments shown will not be described again here to avoid repetition.

[0040] Based on the same inventive concept, this application provides a targeted puncture navigation method. By constructing a unified spatial reference framework of "preoperative high-precision anatomy, intraoperative real-time imaging, and puncture instrument position", the puncture targeting accuracy is significantly improved, and ineffective ablation or damage to adjacent structures caused by positioning deviation is reduced, providing reliable front-end navigation support for pulsed electric field ablation.

[0041] The following is combined with Figure 2 The targeted puncture navigation method provided in the embodiments of this application will be described in detail. For example... Figure 2 As shown, the method includes: S1. Based on the image data of the target biological organ obtained before surgery, construct a three-dimensional model of the target biological organ, and mark the target area and safety boundary on the three-dimensional model.

[0042] In some embodiments, high-resolution MRI or CT imaging data of the patient's target biological organ can be acquired preoperatively.

[0043] Optionally, if the target biological organ is the heart, then delayed-enhanced MRI (LGE-MRI) can be used to acquire cardiac image data to clearly show the areas of myocardial fibrosis and the degree of ventricular septal hypertrophy.

[0044] In some embodiments, the structural regions of the target biological organ can be segmented in three dimensions using a deep learning segmentation network (such as nnU-Net or Swin-Unet) to generate an anatomical model of the target biological organ. To obtain a more accurate anatomical model, modifications from professionals can be incorporated to generate the anatomical model of the target biological organ.

[0045] Optionally, a deep learning segmentation network can be used to automatically perform three-dimensional semantic segmentation of the left ventricle, right ventricle, interventricular septum, aortic valve annulus, and adjacent areas of His bundle, and combined with interactive correction by physicians to generate an accurate anatomical model.

[0046] In some embodiments, based on the above anatomical model, potential ablation target areas can be automatically recommended according to the wall thickness threshold (usually ≥15mm) and hemodynamic simulation results, and the physician can manually delineate the final target area and its safety boundary on the three-dimensional model.

[0047] Optionally, the safety boundary should avoid the left bundle branch, the course of the His bundle, and the 2mm area below the aortic valve. The course of the His bundle can be determined based on anatomical priors from literature or by DTI fiber tracing.

[0048] It should be noted that all annotation information can be embedded into the preoperative 3D model coordinate system with voxel-level precision, providing structured priors for subsequent registration and navigation.

[0049] S2. Spatial registration is performed between the first image acquired during the operation and the three-dimensional model to obtain a fused coordinate system.

[0050] Optionally, the first image is an ultrasound image of a first region of the target biological organ.

[0051] For example, the first image may be an intraoperative real-time apical four-chamber or five-chamber section ultrasound image, but is not limited thereto.

[0052] This step is used to establish a high-precision spatial mapping relationship between intraoperative ultrasound images and preoperative MRI / CT three-dimensional models.

[0053] In some embodiments, step S2 above may specifically include: S21 (not shown in the figure): Select keyframe images from the ultrasound images of the first region acquired in real time during the operation to obtain the first image.

[0054] In some embodiments, an ultrasound video stream of the apical four-chamber view can be acquired at a preset frame rate, and then keyframes can be selected from the ultrasound video stream at preset time intervals as the first image.

[0055] Optionally, the preset frame rate can be 30 fps, and the preset time interval is one frame selected every 5 frames, but it is not limited to this.

[0056] For example, the first image can be obtained by selecting one keyframe image every five frames from the intraoperative real-time acquisition of the apical four-chamber or five-chamber ultrasound video stream, but it is not limited to this.

[0057] S22 (not shown in the figure): Generate a simulated ultrasound image corresponding to the ultrasound image based on the three-dimensional model.

[0058] In some embodiments, the three-dimensional model can be processed using a differentiable ultrasound simulator (such as DeepUS or UltrasoundSim), and then, based on the ray projection and scattering model, a synthetic ultrasound image at the corresponding viewpoint can be generated as the simulated ultrasound image.

[0059] S23 (not shown in the figure): Feature extraction network is used to extract features from the first image and the simulated ultrasound image respectively to obtain multi-scale feature maps.

[0060] In some embodiments, the feature extraction network may employ F3D (a lightweight 3D visualization viewer) or VoxelMorph (a CNN-based unsupervised medical image registration deep learning framework), but is not limited to these.

[0061] In some embodiments, a feature extraction network is used to extract local texture features and edge features of the first image, as well as the corresponding hierarchical features of the simulated ultrasound image.

[0062] Optionally, a dual-branch feature extraction network is constructed using an improved VoxelMorph architecture (adapted to ultrasound modality), with the two branches processing the first image and the simulated ultrasound image respectively, and outputting multi-scale feature maps.

[0063] S24 (not shown in the figure): Based on the multi-scale feature map, calculate the deformation field that registers the first image to the coordinate system of the three-dimensional model.

[0064] Optionally, the deformation field is used to characterize the non-rigid displacement of the first image relative to the three-dimensional model.

[0065] In some embodiments, the deformation field from the ultrasound space to the preoperative model space can be calculated using a bidirectional consistency-consistent optical flow field estimation or a thin-plate spline (TPS)-based non-rigid transformation model. It should be understood that this deformation field considers not only overall cardiac displacement but also models non-uniform deformation caused by local myocardial contraction and introduces time continuity regularization terms (such as inter-frame deformation smoothing constraints) to suppress noise interference.

[0066] In some embodiments, step S24 above may include: The multi-scale feature map is input into the registration network, and the initial deformation field is obtained through feature fusion and regression. The initial deformation field and the registration network are iteratively optimized based on the registration loss function to obtain the deformation field.

[0067] Optionally, the registration loss function includes: Data fidelity items used to measure the similarity between the first image and the simulated ultrasound image; A regularization term used to constrain the deformation smoothness of the deformation field; and, Temporal continuity constraints are used to limit the difference between the deformation field of the current frame and the deformation field of the previous frame.

[0068] S25 (not shown in the figure): Based on the deformation field, the first image and the three-dimensional model are elastically registered to obtain the fused coordinate system.

[0069] In some embodiments, a dynamically updated 4×4 spatial transformation matrix sequence obtained in step S24 can be used to achieve real-time elastic alignment of ultrasound images and preoperative models with sub-millimeter precision, thereby constructing a unified fusion coordinate system.

[0070] In some embodiments, step S25 may include: Based on the deformation field, a non-rigid spatial transformation is performed on the three-dimensional model to generate a dynamic three-dimensional model corresponding to the current first image. A unified coordinate system is established based on the three-dimensional space where the dynamic three-dimensional model is located, and the first image is mapped to the unified coordinate system to achieve flexible registration between the first image and the dynamic three-dimensional model; The unified coordinate system is defined as the fused coordinate system.

[0071] In some embodiments, after obtaining the deformation field of the current frame, it is applied to the preoperative 3D model to generate a registered dynamic model. A unified coordinate system established based on the 3D space where the dynamic model is located is determined as the fusion coordinate system.

[0072] S3. Map the pose data of the puncture needle in the surgical field space to the fused coordinate system, and overlay the first position of the puncture needle tip in the three-dimensional model on the first image.

[0073] In some embodiments, the pose data of the puncture needle is acquired synchronously, and the pose data is strictly synchronized with the ultrasound image frame through hardware triggering or software timestamp alignment mechanism. After coordinate system transformation, the pose data can be accurately mapped to the aforementioned fused coordinate system, realizing the virtual overlay display of the needle body on the preoperative anatomical model.

[0074] Optionally, pose data may include, but is not limited to, three-dimensional position [x,y,z] and Euler angles [α,β,γ] or orientation represented by quaternions.

[0075] In some embodiments, step S3 may include: Acquire the pose data of the puncture needle in the surgical field space, and the pose data has a synchronous temporal relationship with the first image; Based on the preset spatial transformation relationship between the external tracking device coordinate system and the ultrasound image coordinate system, the pose data is converted to the ultrasound image space; Based on the inverse transformation of the deformation field, the transformed pose data is mapped onto the fused coordinate system to determine the real-time position of the puncture needle in the fused coordinate system.

[0076] In some embodiments, the pose data of the puncture needle can be acquired at a preset frequency (e.g., 100 Hz, but not limited thereto). While ensuring that the position data has the same timestamp as the first image, the pose data is converted to the ultrasound space using a rigid transformation matrix from the electromagnetic coordinate system to the ultrasound coordinate system calibrated before surgery. Then, the data is mapped to the preoperative model space through the inverse transformation of the deformation field to obtain the precise position of the needle tip in the fused coordinate system. It should be noted that, to achieve precise visualization of the puncture instruments in the fusion coordinate system, a miniature electromagnetic sensor or a passive infrared reflective sphere array can be integrated into the proximal end or handle of the dedicated pulsed electric field ablation needle. If electromagnetic tracking is used, an electromagnetic field generator is deployed near the operating table, and a shielding algorithm is used to compensate for magnetic field distortions caused by metal instruments and pacemakers, etc. If optical tracking is used, a high frame rate (≥60Hz) infrared stereo camera system is set up above the surgical field to capture the spatial coordinates of the reflective markers in real time.

[0077] In some embodiments, the real-time pose data of the puncture needle can be deeply fused with the elastic registration result, and the corresponding position of the needle tip in the preoperative 3D model can be overlaid on the ultrasound image, providing a multi-view navigation interface. Optionally, the multi-view navigation interface can display: the original ultrasound image, 3D model slices, target area heatmap, but is not limited to these.

[0078] S4. Based on the displacement estimated from multiple consecutive frames of the first image, the first position of the needle tip is dynamically compensated to obtain the second position of the needle tip.

[0079] In some embodiments, step S4 may include: Based on the registration deformation field between multiple consecutive frames of the first image, the motion phase of the target biological organ at the current moment is determined. Using a pre-trained neural network model, the motion trend vector of the target biological organ at the next moment is predicted based on the motion phase and historical displacement information. When it is determined that the needle tip has entered the state of approaching the target area, the first position of the needle tip is compensated for by reverse motion based on the motion trend vector to obtain the second position of the needle tip.

[0080] The key feature of this application's embodiments lies in the introduction of a dynamic compensation mechanism, such as a dynamic compensation mechanism based on cardiac motion. Specifically, the current cardiac cycle phase (e.g., end-diastolic or mid-systolic) can be estimated based on the registration deformation field between consecutive multiple ultrasound images, and a recurrent neural network (e.g., ConvLSTM) can be used to predict the displacement trend of cardiac tissue at the next moment. When the needle tip is detected approaching the target area, the registration state under the current cardiac phase is automatically locked, and the puncture needle trajectory is compensated for in reverse motion. That is, based on the predicted myocardial displacement vector, the displayed position of the needle tip in the fusion space is dynamically adjusted so that it is "anchored" to the real anatomical target point, rather than the instantaneous ultrasound position affected by heartbeat artifacts.

[0081] In some embodiments, the above method may further include: If the distance between the needle tip and the boundary of the target area is less than a first threshold and the moving speed of the needle tip is less than a second threshold, it is determined that the needle tip has entered the state of approaching the target area.

[0082] In some embodiments, when the distance between the needle tip and the target area boundary is detected to be less than a first threshold and the moving speed of the needle tip is less than a second threshold, the system determines that it has entered the "approaching the target area" state. Optionally, the first threshold can be 5 mm and the second threshold can be 2 mm / s, but it is not limited thereto.

[0083] In some embodiments, the above method may further include: Based on the positional changes of the diaphragm in multiple consecutive frames of the first image, the overall displacement of the target biological organ is estimated. When the overall displacement exceeds a preset safety threshold, a breathing control prompt message is output, and the current navigation guidance is paused; When the overall displacement of the target biological organ is detected to return to within the safety threshold, the navigation guidance is restored.

[0084] In some embodiments, a breathing gating strategy can also be set: when breathing causes the overall displacement of the heart to exceed a preset safety threshold, navigation guidance is paused and the operator is prompted to hold their breath until the displacement returns to within the safety threshold, ensuring puncture stability.

[0085] Optionally, the preset safety threshold can be 2mm, but is not limited to this.

[0086] S5. When the second position of the needle tip remains within the target area for a first duration and the constraint conditions of the safety boundary are met, the pulse electric field generator is controlled to trigger the pulse electric field for ablation.

[0087] In some embodiments, the second position of the needle tip can be determined to be within the target area by the pose variance or standard deviation of the sliding window. Optionally, if the pose standard deviation within the sliding window is less than 0.5 mm, then the second position of the needle tip is within the target area.

[0088] Optionally, the first duration can be greater than or equal to 1 second, for example, 1.2 seconds, but is not limited to this.

[0089] In some embodiments, the constraints of the security boundary include at least one of the following: The distance between the second position of the needle tip and the aortic valve annulus is greater than or equal to the third threshold; The distance between the second position of the needle tip and the His bundle travel path is greater than or equal to the fourth threshold. The second position of the needle tip avoids the left bundle branch trunk region.

[0090] Optionally, the third threshold may be greater than or equal to 2 mm, but is not limited to this.

[0091] Optionally, the fourth threshold may be greater than or equal to 3 mm, but is not limited to this.

[0092] For example, if the needle tip remains within the target area for 1.2 seconds after compensation (position standard deviation within the sliding window < 0.5 mm) and the Euclidean distance from the nearest safety boundary (such as the aortic valve annulus) is greater than 2 mm, then the pulse electric field generator is controlled to trigger the pulse electric field for ablation.

[0093] In some embodiments, before controlling the pulsed electric field generator to trigger pulses for electric field ablation, the above method may further include: Light up the status indicator light to a preset color to indicate that the target is ready; On the ultrasound images displayed on the user interface, areas that meet the safety distance constraints are marked by highlighting and overlaying graphics.

[0094] In some embodiments, the preset color can be green, red, or yellow, and this application embodiment does not limit this.

[0095] For example, if the needle tip remains within the target area for 1.2 seconds after compensation (position standard deviation within the sliding window < 0.5 mm) and the Euclidean distance from the nearest safety boundary (such as the aortic valve annulus) is greater than 2 mm, the green "target ready" indicator light will be illuminated, and the target area will be highlighted in a green semi-transparent circle on the ultrasound image.

[0096] In some embodiments, controlling a pulsed electric field generator to trigger a pulsed electric field for ablation may include: Send a control signal to the pulse electric field generator; When the pulse electric field generator receives the control signal and verifies in real time that the second position of the needle tip is within the target area, it outputs a pulse.

[0097] For example, if the needle tip remains within the target area for 1.2 seconds after compensation (positional standard deviation within the sliding window < 0.5 mm), and the Euclidean distance from the nearest safety boundary (such as the aortic valve annulus) is greater than 2 mm, the green "Target Ready" indicator light will illuminate, and the target area will be highlighted in a green semi-transparent circle on the ultrasound image. At this time, the operator can trigger pulsed electric field ablation with a single button press.

[0098] It should be noted that high-voltage pulses are only permitted to be released when the needle tip is confirmed to be in the effective target area and in a stable position.

[0099] Optionally, the parameters of the high-voltage pulse output by the pulse electric field generator include, but are not limited to, eight pulses with a voltage of 1800 V and a pulse width of 100 μs.

[0100] In some embodiments, the above method may further include: During the ablation process, if the positional drift of the puncture needle is continuously detected to be greater than the fifth threshold within the second time period, the pulse electric field generator is controlled to stop pulse output.

[0101] Optionally, the second duration can be greater than or equal to 3 seconds, but is not limited to this.

[0102] Optionally, the fifth threshold may be greater than or equal to 1 mm, but is not limited to this.

[0103] In some embodiments, if the pose drift of the puncture needle is continuously detected to be greater than a fifth threshold during the ablation process for a second duration, an interruption command can be sent to the pulse electric field generator, which immediately stops pulse output upon receiving the interruption command.

[0104] For example, if the positional drift of the puncture needle is detected to be greater than 1.5 mm for 5 consecutive seconds during the ablation process, the pulse energy output will be immediately interrupted to prevent accidental injury.

[0105] The entire process of this application forms a closed-loop control chain of "location, confirmation, release, and monitoring," which effectively improves the safety margin and clinical reliability of the treatment.

[0106] In some embodiments, the targeted puncture navigation method proposed in this application, when applied to the pulsed electric field ablation treatment of HOCM, may include the following steps: 1. Preoperative high-resolution image acquisition and target area annotation: MRI or CT images of the patient's heart are acquired before the operation to reconstruct a three-dimensional heart model. The physician then accurately annotates the target area of ​​ventricular septal hypertrophy and its safe boundaries (avoiding the conduction bundle and aortic valve annulus) on the model.

[0107] 2. Intraoperative ultrasound-preoperative image elastic registration: Acquire apical four-chamber or five-chamber ultrasound images during the operation. Through feature point matching or deep learning-driven non-rigid registration algorithm, the real-time ultrasound images are spatially aligned with the preoperative MRI / CT three-dimensional model to establish a unified fusion coordinate system.

[0108] 3. Real-time tracking of puncture needle position: A miniature electromagnetic sensor or optical reflective marker is integrated on the puncture needle. Through an external electromagnetic field generator or infrared optical camera system, the six-degree-of-freedom position (position and orientation) of the puncture needle in the surgical field space can be acquired in real time.

[0109] 4. Needle tip visualization and motion compensation in fusion space: The real-time pose data of the puncture needle is mapped to the aforementioned fusion coordinate system, and the corresponding position of the needle tip in the preoperative model is displayed in the ultrasound image overlay; at the same time, based on continuous inter-frame cardiac motion estimation, the needle tip trajectory is dynamically compensated to eliminate pseudo displacement caused by breathing and heartbeat.

[0110] 5. Targeting and confirmation: When the needle tip is stably located within the preset target area in the fusion view and meets the safety distance constraint, the pulse electric field ablation is triggered.

[0111] In summary, the embodiments of this application provide a targeted puncture navigation method, which accurately registers high-resolution preoperative images (such as MRI / CT) with intraoperative ultrasound effects, and performs real-time pose tracking and motion compensation of the puncture needle within the obtained fusion space to form a closed-loop navigation mechanism. This solves the problems of inaccurate target area positioning and interference of cardiac motion on the puncture path caused by single ultrasound guidance, and reduces operational risks and the probability of repeated punctures.

[0112] The targeted puncture navigation method of this application will be described in detail below with reference to application examples of the scheme in the embodiments of this application.

[0113] Taking hypertrophic obstructive cardiomyopathy (HOCM) as an example, the targeted puncture navigation method of this application is applied to the percutaneous apical approach pulsed electric field ablation system for HOCM patients. The entire system consists of a preoperative high-resolution image acquisition and intelligent target area annotation module, a non-rigid elastic registration module for intraoperative ultrasound and preoperative images, a real-time six-degree-of-freedom pose tracking module for the puncture needle, a needle tip dynamic visualization and motion compensation module in fusion space, and a target confirmation and ablation trigger closed-loop control module. These modules work collaboratively to achieve precise, safe, and efficient targeted ablation. The specific implementation process of the technical solution is detailed below, in conjunction with each module: 1. Preoperative high-resolution image acquisition and intelligent target area annotation module.

[0114] Patients can undergo delayed-enhanced MRI (LGE-MRI) of the heart one week prior to surgery, using a 3T MRI scanner with a slice thickness of 1.5 mm, covering the entire left ventricle. The acquired 3D image data is input into an automatic segmentation network based on the nnU-Net architecture. This network has been trained on a multicenter dataset containing 200 patients with HOCM and can simultaneously output voxel-level semantic labels for the left ventricular myocardium, right ventricular cavity, interventricular septum, aortic valve annulus, and the adjacent region of the His bundle. Physicians can interactively correct the automatic segmentation results in a 3D visualization interface. For example, when nnU-Net misclassifies some fibrotic tissue as normal myocardium, physicians can manually adjust the boundaries using a drawing tool.

[0115] Based on this, the system calculates the wall thickness at each point of the interventricular septum: for each voxel Its wall thickness Defined as the shortest Euclidean distance from that point to the surface of the right ventricular cavity. If If a target is identified, it is marked as a candidate target. Simultaneously, the system invokes a hemodynamic simulation module based on the Lattice Boltzmann Method (LBM) to estimate the left ventricular outflow tract pressure gradient (LVOT gradient) and overlays high-gradient regions (>30 mmHg) onto the candidate target map. Finally, cardiac electrophysiology experts manually delineate the final ablation target area on the 3D model and set safety boundaries.

[0116] In some embodiments, the security boundary must satisfy the following constraints: Distance from the aortic valve annulus ≥ 2 mm; Distance from the His bundle path ≥ 3 mm (path reconstructed by diffusion tensor imaging DTI fiber tracing, or average anatomical coordinates from cited literature); Avoid the left bundle branch trunk area (based on the fusion of coronary CTA and prior conduction system maps for localization).

[0117] It should be noted that all the above annotation information is embedded into the original MRI coordinate system in the form of a voxel mask to generate a structured preoperative model. This is for use in subsequent registration.

[0118] 2. Non-rigid elastic registration module for intraoperative ultrasound and preoperative images.

[0119] On the day of surgery, the patient was placed in the left lateral decubitus position, and the ultrasound probe was placed in the apical region to acquire apical four-chamber ultrasound video (30 fps). The system selected one keyframe every 5 frames. The procedure was performed. Simultaneously, the preoperative model... The data is input into the differentiable ultrasound simulator DeepUS, which generates synthetic ultrasound images from the corresponding viewpoint based on a ray projection and scattering model. .

[0120] Subsequently, a dual-branch feature extraction network was constructed using an improved VoxelMorph architecture (adapted to ultrasonic modalities): the two branches process the features separately. and Output multi-scale feature maps The registration loss function is defined as:

[0121] in, The first term is the non-rigid transformation field parameterized by Thin Plate Spline (TPS), the second term is the normalized cross-correlation used to measure image similarity, the third term is the deformation smoothing regularization term, and the fourth term is the temporal continuity constraint (Frobenius norm of deformation difference between adjacent frames).

[0122] In a single registration process, the system calculates the deformation field of the current frame. This is then applied to the preoperative model to generate a registered dynamic model. Verification showed that, at the end of diastole, the registration error (the average distance from the target surface point to the corresponding ultrasound structure) was 0.87 mm, meeting the sub-millimeter accuracy requirement.

[0123] 3. Real-time tracking module for six degrees of freedom pose of the puncture needle.

[0124] A miniature 6-DOF electromagnetic sensor (NDI Aurora Planar Probe, 0.8 mm in diameter) can be embedded 2 cm from the rear end of the handle of the pulsed electric field ablation needle used (e.g., 15 cm in length and 1.2 mm in diameter). The electromagnetic field generator is fixed to the right side of the operating table, approximately 30 cm from the heart. To suppress metallic interference, the system employs an adaptive magnetic field distortion compensation algorithm: this algorithm dynamically corrects the electromagnetic sensor readings based on a pre-calibrated metallic interference field template library and combined with real-time field strength residuals.

[0125] In some embodiments, the system can acquire needle pose data at a frequency of 100 Hz. ,in, Let be the position of the needle tip in the electromagnetic coordinate system. The orientation is represented by a quaternion.

[0126] In this embodiment, the hardware synchronization signal (TTL pulse) can be strictly aligned with the ultrasound frame to ensure... and The timestamps are consistent. Subsequently, a rigid transformation matrix from the preoperatively calibrated electromagnetic coordinate system to the ultrasound coordinate system is used. ,Will The signal is converted to ultrasound space and then output through the inverse transformation of the registration module. Mapping to the preoperative model space, the needle tip is obtained. precise location .

[0127] 4. Integration of needle tip dynamic visualization and motion compensation module in space.

[0128] The system displays three views simultaneously on the operation interface: (1) original ultrasound image superimposed with a virtual model of the needle body; (2) any section of the preoperative MRI three-dimensional model, showing the target area and contraindicated structures; (3) target area thermogram, with color depth reflecting wall thickness and hemodynamic risk.

[0129] The key to this step lies in the dynamic compensation mechanism of cardiac motion. Specifically, the system continuously monitors the registration deformation field for five consecutive frames. The input is fed into a lightweight ConvLSTM network to predict the next time step. Local displacement field When the needle tip is detected to be less than 5 mm from the target area boundary and at a velocity of less than 2 mm / s, the system determines that it has entered the "approaching the target area" state, immediately freezes the current cardiac phase (using R-wave synchronized ECG signal for assistance), and initiates motion compensation to update the needle tip display position to:

[0130] This counteracts the target displacement caused by myocardial contraction.

[0131] For example, in a mid-systolic puncture, the actual target point shifted upward by 2.3 mm due to myocardial shortening, while the system predicted a displacement of 2.1 mm. After compensation, the error was only 0.2 mm.

[0132] In some embodiments, the system may also integrate respiratory gating logic: estimating the overall cardiac displacement by the change in the position of the diaphragm in the ultrasound image. If the displacement is >2 mm, a "Please hold your breath" prompt will appear on the interface, and navigation guidance will be paused until the displacement falls back within the threshold.

[0133] 5. Target confirmation and ablation trigger closed-loop control module.

[0134] When the needle tip is in the compensated position The target area is maintained within the preset target area for 1.2 seconds (with a pose standard deviation of <0.5 mm within the sliding window), and the Euclidean distance from the nearest contraindicated structure (such as the aortic valve annulus) is [missing information]. The system illuminates the green "Target Ready" indicator light and highlights the target area with a green semi-transparent circle on the ultrasound image.

[0135] In some embodiments, when the operator presses a foot switch, the system sends an authorization signal to the pulsed electric field generator. The energy control system of the pulsed electric field generator only responds upon receiving and verifying this authorization signal in real time. High-voltage pulses are released only while the target area is still within the target area (typical parameters: 1800 V, 8 pulses, pulse width 100 μs).

[0136] In some embodiments, if a pose drift >1.5 mm is detected during the ablation process (lasting approximately 5 seconds) (e.g., due to a patient coughing or a rapid heartbeat), the system immediately cuts off the energy output and records the event log.

[0137] In the embodiments of this application, an intelligent navigation system is constructed that integrates intelligent annotation of high-resolution preoperative images, non-rigid elastic registration of intraoperative ultrasound and preoperative models, real-time six-degree-of-freedom tracking of the puncture needle, dynamic compensation of cardiac motion, and closed-loop control for target confirmation, so as to realize the closed-loop management of the entire process of precise target positioning and safe ablation in pulsed electric field ablation.

[0138] Based on the same principles as the methods shown in the embodiments of this application, the embodiments of this application also provide a signal processing system, which includes the targeted puncture navigation system provided in the above embodiments.

[0139] In an alternative embodiment, a signal processing system or electronic device, such as... Figure 3 As shown, Figure 3The signal processing system 20 shown includes a processor 21 and a memory 23. The processor 21 is communicatively connected to the memory 23, for example, via a bus 22. Optionally, the signal processing system 20 may also include a transceiver 24, which can be used for data interaction between the signal processing system and other electronic devices, such as sending and / or receiving data. It should be noted that in practical applications, the transceiver 24 is not limited to one type, and the structure of this signal processing system 20 does not constitute a limitation on the embodiments of this application.

[0140] Processor 21 may be a CPU (Central Processing Unit), general-purpose processor, DSP (Digital Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA (Field-Programmable Gate Array), or other programmable logic device, transistor logic device, hardware component, or any combination thereof, including the chip or targeted puncture navigation system described in any of the above embodiments. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. Processor 21 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of DSP and microprocessor, etc.

[0141] Bus 22 may include a pathway for transmitting information between the aforementioned components. Bus 22 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 22 may be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0142] The memory 23 may be a ROM (Read-Only Memory) or other type of static storage device capable of storing static information and instructions, RAM (Random Access Memory) or other type of dynamic storage device capable of storing information and instructions, or it may be an EEPROM (Electrically Erasable Programmable Read Only Memory), a CD-ROM (Compact Disc Read-Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0143] The memory 23 is used to store computer programs that execute the embodiments of this application, and the execution is controlled by the processor 21. The processor 21 is used to execute the computer programs stored in the memory 23 to implement the steps shown in the foregoing method embodiments.

[0144] In one alternative embodiment, such as Figure 3 The signal processing system shown can also be described as an electronic device. This electronic device may include a memory, a processor, and a computer program stored in the memory, the processor executing the computer program to implement the steps and corresponding content of the foregoing method embodiments.

[0145] This application provides a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it can implement the steps and corresponding content of the aforementioned method embodiments.

[0146] This application also provides a computer program product, including a computer program that, when executed by a processor, can implement the steps and corresponding content of the aforementioned method embodiments.

[0147] In the description of this application, the terms "center," "upper," "lower," "front," "rear," "left," "right," "vertical," "horizontal," "top," "bottom," "inner," and "outer," etc., indicate directions or positional relationships based on the exemplary directions or positional relationships shown in the accompanying drawings. They are used to facilitate the description or simplification of the embodiments of this application and are not intended to indicate or imply that the device or component referred to must have a specific orientation or be constructed and operated in a specific orientation. Therefore, they should not be construed as limitations on this application.

[0148] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0149] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "installation," "connection," and "joining" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal communication between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0150] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0151] The above description is only a partial implementation of this application. It should be noted that for those skilled in the art, other similar implementation methods based on the technical concept of this application, without departing from the technical concept of this application, also fall within the protection scope of the embodiments of this application.

Claims

1. A targeted puncture navigation method, characterized in that, include: Based on the image data of the target biological organ obtained before surgery, a three-dimensional model of the target biological organ is constructed, and the target area and safety boundary are marked on the three-dimensional model; The first image acquired during the operation is spatially registered with the three-dimensional model to obtain a fused coordinate system. The first image is an ultrasound image of the first region of the target biological organ. The pose data of the puncture needle in the surgical field space is mapped to the fused coordinate system, and the first position of the tip of the puncture needle in the three-dimensional model is superimposed on the first image. Based on the displacement estimated from multiple consecutive frames of the first image, the first position of the needle tip is dynamically compensated to obtain the second position of the needle tip. When the second position of the needle tip remains within the target area for a first duration and the constraint conditions of the safety boundary are met, the control pulse electric field generator triggers a pulse electric field for ablation.

2. The targeted puncture navigation method according to claim 1, characterized in that, The step of spatially registering the first image acquired during the operation with the three-dimensional model to obtain a fused coordinate system includes: The first image is obtained by selecting keyframe images from the ultrasound images of the first region acquired in real time during the operation; Based on the three-dimensional model, a simulated ultrasound image corresponding to the ultrasound image is generated; A feature extraction network is used to extract features from the first image and the simulated ultrasound image respectively to obtain multi-scale feature maps; Based on the multi-scale feature map, the deformation field for registering the first image to the coordinate system of the three-dimensional model is calculated, and the deformation field is used to characterize the non-rigid displacement of the first image relative to the three-dimensional model. Based on the deformation field, the first image and the three-dimensional model are elastically registered to obtain the fused coordinate system.

3. The targeted puncture navigation method according to claim 2, characterized in that, The step of calculating the deformation field for registering the first image to the coordinate system of the 3D model based on the multi-scale feature map includes: The multi-scale feature map is input into the registration network, and the initial deformation field is obtained through feature fusion and regression. The initial deformation field and the registration network are iteratively optimized based on the registration loss function to obtain the deformation field. The registration loss function includes: Data fidelity items used to measure the similarity between the first image and the simulated ultrasound image; A regularization term used to constrain the deformation smoothness of the deformation field; and, Temporal continuity constraints are used to limit the difference between the deformation field of the current frame and the deformation field of the previous frame.

4. The targeted puncture navigation method according to claim 2, characterized in that, The step of elastically registering the first image and the 3D model based on the deformation field to obtain the fused coordinate system includes: Based on the deformation field, a non-rigid spatial transformation is performed on the three-dimensional model to generate a dynamic three-dimensional model corresponding to the current first image. A unified coordinate system is established based on the three-dimensional space where the dynamic three-dimensional model is located, and the first image is mapped to the unified coordinate system to achieve flexible registration between the first image and the dynamic three-dimensional model; The unified coordinate system is defined as the fused coordinate system.

5. The targeted puncture navigation method according to claim 2, characterized in that, The process of mapping the pose data of the puncture needle in the surgical field space to the fused coordinate system includes: Acquire the pose data of the puncture needle in the surgical field space, and the pose data has a synchronous temporal relationship with the first image; Based on the preset spatial transformation relationship between the external tracking device coordinate system and the ultrasound image coordinate system, the pose data is converted to the ultrasound image space; Based on the inverse transformation of the deformation field, the transformed pose data is mapped onto the fused coordinate system to determine the real-time position of the puncture needle in the fused coordinate system.

6. The targeted puncture navigation method according to any one of claims 1-5, characterized in that, The process of dynamically compensating the first position of the needle tip based on the displacement estimated from multiple consecutive frames of the first image to obtain the second position of the needle tip includes: Based on the registration deformation field between multiple consecutive frames of the first image, the motion phase of the target biological organ at the current moment is determined. Using a pre-trained neural network model, the motion trend vector of the target biological organ at the next moment is predicted based on the motion phase and historical displacement information. When it is determined that the needle tip has entered the state of approaching the target area, the first position of the needle tip is compensated for by reverse motion based on the motion trend vector to obtain the second position of the needle tip.

7. The targeted puncture navigation method according to claim 6, characterized in that, Also includes: If the distance between the needle tip and the boundary of the target area is less than a first threshold and the moving speed of the needle tip is less than a second threshold, it is determined that the needle tip has entered the state of approaching the target area.

8. The targeted puncture navigation method according to claim 6, characterized in that, Also includes: Based on the positional changes of the diaphragm in multiple consecutive frames of the first image, the overall displacement of the target biological organ is estimated. When the overall displacement exceeds a preset safety threshold, a breathing control prompt message is output, and the current navigation guidance is paused; When the overall displacement of the target biological organ is detected to return to within the safety threshold, the navigation guidance is restored.

9. The targeted puncture navigation method according to any one of claims 1-5, characterized in that, Before controlling the pulse electric field generator to trigger pulses for electric field ablation, the following steps are also included: Light up the status indicator light to a preset color to indicate that the target is ready; On the ultrasound images displayed on the user interface, areas that meet the safety distance constraints are marked by highlighting and overlaying graphics.

10. The targeted puncture navigation method according to any one of claims 1-5, characterized in that, The constraints of the security boundary include at least one of the following: The distance between the second position of the needle tip and the aortic valve annulus is greater than or equal to the third threshold; The distance between the second position of the needle tip and the His bundle travel path is greater than or equal to the fourth threshold. The second position of the needle tip avoids the left bundle branch trunk region.

11. The targeted puncture navigation method according to any one of claims 1-5, characterized in that, Controlling the pulsed electric field generator to trigger a pulsed electric field for ablation includes: Send a control signal to the pulse electric field generator; When the pulse electric field generator receives the control signal and verifies in real time that the second position of the needle tip is within the target area, it outputs a pulse.

12. The targeted puncture navigation method according to any one of claims 1-5, characterized in that, Also includes: During the ablation process, if the positional drift of the puncture needle is continuously detected to be greater than the fifth threshold within the second time period, the pulse electric field generator is controlled to stop pulse output.

13. A targeted puncture navigation system, characterized in that, include: The processor, controller, pulsed electric field generator, and image acquisition unit, among which, The image acquisition device is used to acquire image data of the target biological organ; The processor is used to perform the following operations: Based on the image data of the target biological organ acquired by the image acquisition device before surgery, a three-dimensional model of the target biological organ is constructed, and the target area and safety boundary are marked on the three-dimensional model; The first image acquired during the operation by the image acquisition device is spatially registered with the three-dimensional model to obtain a fused coordinate system. The first image is an ultrasound image of the first region of the target biological organ. The pose data of the puncture needle in the surgical field space is mapped to the fused coordinate system, and the first position of the tip of the puncture needle in the three-dimensional model is superimposed on the first image. Based on the displacement estimated from multiple consecutive frames of the first image, the first position of the needle tip is dynamically compensated to obtain the second position of the needle tip. The controller is configured to control the pulse electric field generator to trigger a pulse electric field for ablation when the second position of the needle tip remains within the target area for a first duration and the constraint conditions of the safety boundary are met.

14. An electronic device comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the method according to any one of claims 1 to 12.

15. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it implements the method according to any one of claims 1 to 12.

16. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by a processor, it implements the method according to any one of claims 1 to 12.

Images