Determining position of implant in hollow organ

Abstract

In order to determine the position of the implant (9) in the hollow organ, image data is acquired, which consists of one or more two-dimensional images representing the implant (9) in a first state in the hollow organ from a total of up to ten different viewing directions, respectively. A three-dimensional first reconstruction of the implant (9) in the first state is generated based on the image data. The image data comprises two to ten two-dimensional images having different viewing directions with respect to the implant (9). Furthermore, the image data comprises two or more two-dimensional first images (7, 7a, 7b, 7c) which represent the implant (9) in the first state in each case in a first viewing direction at different points in time. At least one averaged first image is generated based on the two or more first images (7, 7a, 7b, 7c), and the first reconstruction is generated based on the at least one averaged first image. The first reconstructed representation is displayed on a display device (6).

Description

Technical Field

[0001] This invention relates to a computer-implemented method for determining the position of an implant in a hollow organ, wherein position is understood as spatial location, orientation, orientation, and / or shape. The invention also relates to a data processing system for performing such a computer-implemented method, an imaging system having such a data processing system, and related computer program products. Background Technology

[0002] For many patients with cerebral aneurysms, the implantation of a stent for blood flow diversion is the preferred treatment. These stents are also known as flow diverters, flow diverting stents, or blood flow diversion implants. During endovascular surgery, this implant is inserted into the artery from which the aneurysm bulges to restore blood flow and reduce blood flow into the aneurysm itself, causing thrombosis within the aneurysm and eventually leading to its regression. The placement of the implant in the blood vessel is crucial to the success of the treatment. For example, angiography systems, especially X-ray angiography systems, can be used to determine the implant's location. For instance, biplane X-ray angiography systems can also be used to determine the implant's location, providing the surgeon with two views from different directions simultaneously. This is also known as biplane fluoroscopy.

[0003] Stents used for blood flow diversion come in various sizes, especially different lengths and diameters. Selecting the appropriate implant and its placement can be challenging and requires extensive experience. Typically, implant selection and proper placement planning in the artery are based on pre-interventional 3D angiographic images, CTA images, or cone-beam computed tomography (CTA) images.

[0004] It is generally desirable to determine the location of an implant in a blood vessel using an angiography system, as implant insertion is often assisted by angiography, allowing for immediate location determination using the same system without changing equipment. Furthermore, location determination based on cone-beam CTA images is generally undesirable due to its relatively time-consuming nature; instead, it is preferred to determine location based on a limited number of images generated by the angiography system from a limited number of views.

[0005] When using single-plane angiography systems to determine location, it is difficult for therapists to assess the three-dimensional orientation and / or shape of the implant. In biplane fluoroscopy, although two real-time image streams are available, the therapist's experience in estimating the implant's three-dimensional orientation and / or shape from two two-dimensional images remains a limiting factor. For flow diversion stents, another complicating factor is that these stents are typically initially placed in a catheter, which is then gradually removed and the stent is released, thus dynamically deploying until it reaches its final orientation and shape.

[0006] Similar issues arise with other endovascular implants, such as so-called aneurysm coils, intraaneural flow diverters, or other neural stents used to treat narrowed intracranial vessels. Furthermore, similar situations may occur with applications outside the brain, such as renal artery stents and bile duct stents. Finally, similar issues may arise when implanting implants in other hollow organs. This approach is not necessarily limited to endovascular implantation. Implantation in other hollow organs may present similar challenges.

[0007] Using a CT system to perform a complete three-dimensional determination of the implant location can reliably assess the implant's posture and / or shape, but this is usually not considered due to the high cost and associated X-ray dose.

[0008] In the article "In Vivo Validation of CAAS QCA-3D Coronary Reconstruction Using Fusion of Angiography and Intravascular Ultrasound (ANGUS)", Catheterization and Cardiovascular Interventions 73:620–626 (2009), JCH Schuurbiers et al. described how to generate a three-dimensional approximate reconstruction of blood vessels based on only two two-dimensional images with different views. Summary of the Invention

[0009] The technical problem to be solved by the present invention is to provide visual support for the therapist in determining the position of an implant in a hollow organ, which provides higher reliability and does not require full three-dimensional imaging.

[0010] Location should be understood as the spatial position, orientation, orientation, and / or shape of the implant within the hollow organ. Therefore, determining location requires good visibility and identifiability of the implant. Visibility and identifiability of the implant require sufficiently high image quality in its representation. Therefore, another technical problem this invention aims to solve is to improve the quality of the representation of the implant within the hollow organ without using full three-dimensional imaging.

[0011] This technical problem is solved by the technical solution of the independent claim. Advantageous extended designs and preferred embodiments are the technical solutions of the dependent claims.

[0012] This invention is based on the idea of ​​generating a three-dimensional reconstruction of an implant from a limited number of two-dimensional images, up to ten different views, and displaying a representation of the reconstruction on a display device. The number of views is two to ten, and / or the reconstruction is also generated based on a geometric and / or mechanical model of the implant.

[0013] According to one aspect of the invention, a computer-implemented method is provided for determining the position of an implant in a hollow organ. Here, image data is acquired, consisting of one or more two-dimensional images, wherein these images represent the implant in a first state within the hollow organ from a total of up to ten different views. A first three-dimensional reconstruction of the implant in the first state is generated based on the image data. The image data includes two to ten two-dimensional images having different views of the implant, and / or the first reconstruction is generated based on the image data and based on a preset geometric and / or mechanical model of the implant. The representation of the first reconstruction is displayed on a display device.

[0014] According to the present invention, the image data includes two or more two-dimensional first images, which respectively represent the implant in a first state at different time points according to a first viewing direction. An averaged first image is generated based on the two or more first images. A first reconstruction is generated based on the averaged first image.

[0015] Each of the first images corresponds to the same viewpoint, i.e., the first viewpoint. This does not preclude the possibility that the image data may contain additional images according to other viewpoints.

[0016] By averaging two or more first images, noise in the images can be effectively suppressed while preserving, enhancing, or more clearly representing the contours of the implant depicted in the two or more first images. In other words, the signal-to-noise ratio is improved. Therefore, to generate a first reconstruction, the shape and / or orientation of the implant in the corresponding averaged images can be determined or extracted more accurately and reliably, resulting in a more accurate and reliable first reconstruction.

[0017] Unless otherwise stated, all steps of the computer-implemented method can be performed by a data processing system comprising at least one data processing device. In particular, the at least one data processing device is configured or adapted to perform the steps of the computer-implemented method. For this purpose, the at least one data processing device may, for example, store a computer program containing instructions that, when executed by the at least one data processing device, cause the at least one data processing device to perform the computer-implemented method. The computer-implemented method can also be implemented wholly or partially in hardware. The terms "data processing system" and "at least one data processing device" are used interchangeably herein and hereinafter. This also applies to their derived terms.

[0018] If the at least one data processing device comprises two or more data processing devices, then a particular step performed by the at least one data processing device can also be understood as different steps being performed by different data processing devices or different parts of a step. It is particularly unnecessary that each data processing device performs these steps. In other words, the execution of these steps can be distributed among two or more data processing devices.

[0019] By including the corresponding steps for generating image data, corresponding implementations of non-purely computer-implemented methods for determining the location of an implant in a hollow organ are derived from each implementation of the computer-implemented method.

[0020] The position of the implant in the hollow organ is determined by generating a first reconstruction once, or as necessary, multiple times, or continuously, and displaying the representation of the corresponding first reconstruction on a display device. Thus, the position of the implant in the hollow organ can be repeatedly determined for different consecutive time points.

[0021] Implants can change their position, orientation, and / or orientation in three-dimensional space and / or their geometry at different points in time. Especially for blood flow diversion implants used to guide blood flow in hollow organs away from aneurysms or other implants inserted into cerebral blood vessels, the implant can initially be introduced into the hollow organ in a compressed state, for example, surrounded by a vascular catheter, and gradually guided to a predetermined position and / or orientation on the one hand, and unfolded on the other hand, especially by retraction catheters, so that the implant can take on its final shape.

[0022] Therefore, the state of the implant can be understood, for example, as its current posture and / or geometry in three-dimensional space. Thus, the two-dimensional images of the image data, representing the implant in the first state, can be understood as having the same or substantially the same posture and / or shape in the two-dimensional images of the image data. Since the deployment of the implant within the hollow organ typically occurs on a timescale significantly longer than the acquisition time of two-dimensional images, such as X-ray images, it is possible to reasonably approximate the implant as being in a stationary or quasi-stationary state in these images of the image data when multiple images are acquired within a short timeframe compared to the implant deployment time.

[0023] The image data represents the implant from a maximum of ten different views; in other words, it represents the implant from a number of views ranging from one to ten. Here, the image data can consist of one to ten two-dimensional images, each corresponding to a different view. In the extreme case of a single view, the image data may also consist of exactly one two-dimensional image with that view. However, it is also possible that, for a number of views between one and ten, the image data contains two or more two-dimensional images for each view.

[0024] In particular, the first three-dimensional reconstruction can be generated based on the image data without using additional image data representing the implant at different subsequent time points. However, this does not preclude the use, in particular, additional use of, image data from before the intervention to generate the first reconstruction. The difference between the computer-implemented method according to the invention and full three-dimensional monitoring, for example, via a CT system or CBCT system, is that the number of different views is significantly reduced compared to the techniques mentioned later.

[0025] Therefore, the first reconstruction can be understood as an approximate three-dimensional reconstruction of the implant. For precise or explicit reconstruction, such as according to the principles of computed tomography, depiction from a large number of views is required. However, this application can utilize the fact that implants cannot present arbitrary shapes in three-dimensional space. The final or approximate final shape of the implant, as well as its initial or approximate initial shape, are actually known. Interpolation can be performed accordingly for the state between the initial and final shapes, or these states may also be known. This approximate three-dimensional reconstruction is also called the symbolic reconstruction of the implant.

[0026] Therefore, unlike traditional CT or CBCT methods, the solution space is significantly restricted when generating the first reconstruction. Specifically, the solution space can be restricted using the known shape of the implant or certain aspects of that known shape, such as symmetry, length, diameter, etc. Thus, even using a small number of views, and in extreme cases, only a single view, can generate a good approximation of the implant's three-dimensional shape in the form of the first 3D reconstruction. However, if only a single view is available, a geometric and / or mechanical model of the implant is required, specifying geometric characteristics such as the implant's shape, symmetry, diameter, or length.

[0027] The geometric and / or mechanical model may also include the symmetry of the implant, particularly rotational symmetry. A geometric and / or mechanical model of the implant is not mandatory if two or more views are available, but in some implementations it can improve the accuracy of the first reconstruction.

[0028] In one variation of the computer-implemented method according to the invention, the image data comprises two to ten two-dimensional images having different views of the implant, and a first reconstruction is generated based on the image data and independent of the geometric and / or mechanical model of the implant. In another variation, the image data comprises two to ten two-dimensional images having different views of the implant, and a first reconstruction is generated based on the image data and a preset geometric and / or mechanical model of the implant. According to yet another variation, all images of the image data correspond to the same view, and the first reconstruction is generated based on the image data and the preset geometric and / or mechanical model.

[0029] If the image data contains two to ten two-dimensional images with different viewpoints, it can be understood that all images in these two to ten two-dimensional images have different viewpoints. However, the image data may also contain more than one two-dimensional image for each viewpoint.

[0030] The first reconstruction can be generated using methods such as those described in the aforementioned article by Schourbieers et al. However, it should be noted that in that article, the corresponding reconstruction is generated for the blood vessel itself, not for the implant or similar object within the blood vessel. The inventors recognized that, due to the defined shape of the implant, this algorithm for symbol reconstruction of blood vessels can also be used for symbol reconstruction of implants.

[0031] Depending on the design of the method, the two-dimensional image can be an image from different sources. It is typically an image generated from a medical imaging modality. For example, it can be an X-ray image, especially a two-dimensional X-ray projection image. Similarly, two-dimensional images from magnetic resonance imaging or other imaging modalities can also be used.

[0032] The computer-implemented method according to the invention provides a representation of a first three-dimensional reconstruction to the physician, enabling the physician to assess a first state of the implant and, consequently, to dynamically assess another state of the implant by repeating the method steps accordingly. This provides, in particular, a direct three-dimensional representation of the implant without requiring the physician to make a more or less reliable interpretation of the image data from two or more simultaneously displayed two-dimensional images to infer the current shape and / or orientation of the implant. Because the invention uses only a very limited number of views and correspondingly a very small number of images for the operation, the radiation load on the patient is significantly reduced, in the case of X-ray imaging, compared to the full three-dimensional monitoring that could theoretically be considered using CT or CBCT. Furthermore, the computational load for generating the first reconstruction is also kept within limits, as only a small number of views need to be considered instead of hundreds or even thousands of views to generate the first reconstruction.

[0033] According to at least one embodiment, the hollow organ is a vascular structure, i.e., in particular one or more interconnected blood vessels, and the implant is a vascular implant.

[0034] In particular, the hollow organ can be a vascular structure in the brain, i.e., a cerebral vascular structure, especially a cerebral artery, and / or the implant can be a blood flow diversion implant used to guide blood flow in the hollow organ through an aneurysm, especially passing alongside the aneurysm sac. Alternatively, the implant can be a so-called aneurysm coil, an intra-aneurysm flow diverter, or other neural stent. It can also be a stent used in the bile duct, a renal artery stent, etc.

[0035] According to at least one embodiment, the implant is a blood flow diversion implant for guiding blood flow in a hollow organ away from an aneurysm in the hollow organ, i.e., in particular a blood flow guiding device, and the hollow organ is a blood vessel in the brain, in particular.

[0036] In such embodiments, the solution according to the invention is particularly advantageous for various reasons. Unlike in cardiovascular applications, such as those using stents or other vascular implants, implants used as flow diverters in brain applications are much more sophisticated in structure. This is partly because the radial force exerted on peripheral blood vessels by a cardiac stent to maintain blood flow through the vessel is many times greater than the radial force exerted outward on cerebral blood vessels by a flow diverter, especially since the flow diverter does not involve opening stenosis or the like but rather preventing blood flow from the aorta into an aneurysm. Furthermore, the radial reaction force exerted by cerebral arteries is generally many times lower than the reaction force exerted by blood vessels in the heart or coronary arteries, etc. Therefore, flow diverters can be designed to be much more sophisticated, but this makes it significantly more difficult for therapists to identify these flow diverters on a basic two-dimensional image in fluoroscopy, such as X-ray fluoroscopy. Therefore, it is particularly advantageous to provide three-dimensional reconstruction to determine location.

[0037] According to at least one embodiment, the image data consists of one or more two-dimensional images, each representing the implant in a first state within the hollow organ and corresponding to up to five different views, particularly two or three different views.

[0038] For each of the maximum five different views, i.e., one to five views, the image data contains at least one two-dimensional image. However, depending on the implementation, the image data may also contain two or more corresponding images for one or more or all of the existing views. It is not mandatory that the number of images in the image data for the different views be the same.

[0039] In such an implementation, the first reconstruction can already reproduce the shape and / or orientation of the implant with very good approximation, wherein the computational cost remains very low even when considering a maximum of five, preferably two or three, different views. Furthermore, in the case of X-ray-based imaging used to generate two-dimensional images, the lower radiation dose implied by the small number of views is particularly advantageous.

[0040] According to at least one embodiment, the geometric and / or mechanical model of the implant includes the length and / or diameter of the implant.

[0041] The length and / or diameter of the implant can be, in particular, the final length or final diameter after the implant is fully deployed. Alternatively, the geometric and / or mechanical model of the implant can include different lengths and / or diameters for different states of the implant, especially different deployed states, and, if necessary, diameters existing at different longitudinal positions along the longitudinal direction of the implant. This approach further reduces the solution space considered for generating a 3D reconstruction based on image data, thereby further reducing the computational cost for generating the first reconstruction.

[0042] Alternatively or additionally, the geometric and / or mechanical model may also include one or more radial forces for the final state, initial state, or other state between the final and initial states of the implant. Radial forces may, for example, correspond to the maximum outward radial force that the implant can exert on surrounding tissues, particularly hollow organs.

[0043] Such an implementation can further reduce the solution space used to generate the first reconstruction and has the advantages described above.

[0044] According to at least one implementation, the first reconstruction is based on image data and pre-intervention representation of hollow organs, particularly three-dimensional pre-intervention representation generation.

[0045] Pre-interventional representations of hollow organs can be generated, for example, based on pre-interventional image data, such as CT or CBCT image data, acquired before surgery, i.e., before the implant is inserted into the hollow organ. The pre-interventional representation of the hollow organ specifically provides the topological and geometric characteristics of the hollow organ. Once the implant is positioned as intended within the hollow organ, the solution space used to generate the first reconstruction can be further constrained by considering the pre-interventional representation, and the aforementioned advantages are maintained.

[0046] To more precisely or specifically constrain the solution space, the pre-intervention representation of the hollow organ can also be generated from the initial or undeformed representation of the hollow organ, taking into account the deformation of the hollow organ caused by the introduction of the implant or, if necessary, the introduction of the corresponding tool for the introduction of the implant.

[0047] To average two or more first images, it can be specified that the two or more first images are registered with each other, that is, the two or more first images are first registered or aligned with each other, and then averaged based on the aligned or registered first images. This has the advantage, for example, that it can compensate for minor movements of the implant between different time points when the two or more first images were acquired, thereby enabling a more accurate determination or extraction of the implant's contour. This is particularly advantageous when implant movement is anticipated in a specific application.

[0048] In other applications, especially for brain applications, the first image can be averaged without correspondingly aligning or registering the first images to each other, without obtaining suboptimal results due to motion artifacts, because in such applications, motion caused by cardiac motion or implant motion caused by respiratory or peristaltic motion of the patient being examined can be almost ignored.

[0049] In a preferred embodiment, the two or more two-dimensional first images consist of three to five first images.

[0050] In this implementation, the computational cost for determination is very low, and the noise suppression effect of averaging is already significant.

[0051] According to at least one embodiment, the image data includes two or more two-dimensional second images, which represent the implant in a first state at different time points according to a second view different from the first view. An averaged second image is generated based on the two or more second images. A first reconstruction is generated based on the averaged second image, particularly based on the averaged first image and the averaged second image.

[0052] Similar implementations are also applicable to all other views among the maximum ten views.

[0053] In other words, for each of the maximum ten view directions, a corresponding averaged image can be generated based on two or more images corresponding to the view direction at different time points of the image data, and the first reconstruction is generated based on these averaged images.

[0054] According to at least one embodiment, in each of the two or more first images, a marker position is determined for a marker indicating the location of the implant. Here, the marker may be placed on the implant. However, the marker may also be placed on an auxiliary tool for inserting the implant into a hollow organ. The two or more first images are aligned relative to each other, i.e., registered, according to the marker position. The averaged first image is generated based on the aligned two or more first images.

[0055] In particular, in order to generate an averaged first image, the two or more aligned first images are averaged.

[0056] The location of the marker can be given in the image space of the two or more first images. The auxiliary tool may be, for example, a catheter or guidewire for introducing the implant into a blood vessel.

[0057] The markers can be used to identify possible movements of implants or assistive devices between individual acquisitions and to compensate for these movements by aligning the first images with each other. In particular, the first images can be aligned relative to each other such that the marker positions overlap in the aligned images.

[0058] In a further embodiment, in each of the two or more first images, corresponding marker positions are determined for two or more markers of the implant or assistive device. The two or more first images are aligned relative to each other based on these marker positions. An averaged first image is generated based on the aligned two or more first images.

[0059] In this implementation, not only can the translation of the marker position of a single marker in different first images be compensated, but also the rotation of the implant or assistive device across two or more first images can be compensated. This enables a more accurate representation of the implant in the averaged first images.

[0060] The corresponding implementation is also feasible with respect to the second image corresponding to the second view, or it is also applicable to all averaging processes of corresponding images for different view directions.

[0061] According to at least one embodiment, the two or more first images correspond to an image sequence, particularly to consecutive images within that image sequence. The image sequence has an image acquisition rate in the range of three to fifteen images per second.

[0062] Given that the deployment of the implant occurs on a significantly longer timescale, it can be assured with high reliability that all first images—which represent the implant in the first state at different points in time with a first viewpoint—actually represent the same first state of the implant, or the deviation is negligible.

[0063] Similar implementations are also applicable to the two or more second images and / or other images that are correspondingly averaged from different views.

[0064] According to at least one embodiment, the contour of the implant is determined by segmenting an averaged first image, and a first reconstruction is generated based on the contour of the implant.

[0065] This method enables the generation of a first-dimensional 3D reconstruction with exceptionally high reliability.

[0066] Similar implementations are also applicable to the two or more second images and / or other images that are correspondingly averaged from different views.

[0067] According to at least one embodiment, additional image data is acquired, which consists of one or more two-dimensional additional images representing the implant in a second state within the hollow organ from up to ten different views. A three-dimensional second reconstruction of the implant in the second state is generated based on the additional image data. The additional image data includes two to ten two-dimensional additional images having different views of the implant, and / or the second reconstruction is generated based on the additional image data and the geometric and / or mechanical model of the implant. The representation of the second reconstruction is displayed on a display device.

[0068] Here, in particular, a second state exists after the implant is in the first state. Therefore, the obtained second reconstructed representation can be understood, for example, as updating the displayed representation from the first reconstruction to the second reconstruction. In particular, corresponding image data for other states of the implant can also be obtained similarly, and corresponding three-dimensional reconstructions can be generated and displayed as described.

[0069] Therefore, it is possible to display a near real-time representation of the implant's reconstruction. This allows medical personnel to more quickly assess the implant's position within the hollow organ.

[0070] According to at least one embodiment, the representation of the first reconstruction includes the superposition of the first reconstruction and the reference representation of the hollow organ, and / or the representation of the second reconstruction includes the superposition of the second reconstruction and the reference representation of the hollow organ.

[0071] The reference representation of the hollow organ can be based on a preoperative three-dimensional dataset, such as a CT dataset or a CBCT dataset. The reference representation can also be modified where necessary, for example, to account for deformations caused by the introduction of implants or assistive devices into the hollow organ.

[0072] The representation through the first or second reconstruction also includes a reference representation, allowing therapists to better understand the three-dimensional pose and shape of the implant without having to rely on additional displays of the reference representation.

[0073] According to another aspect of the present invention, a data processing system is provided, which is configured to perform a computer-implemented method according to the present invention.

[0074] In this disclosure, the terms "data processing system" and "at least one data processing device" are used interchangeably. A data processing device can be particularly understood as a data processing device containing processing circuitry. Therefore, a data processing device can, in particular, process data to perform computational operations. The computational operations may also include, where necessary, operations for performing index accesses to data structures, such as look-up tables (LUTs), and data processing procedures implemented in hardware.

[0075] Data processing devices may, in particular, include one or more computers, one or more microcontrollers, and / or one or more integrated circuits, such as one or more application-specific integrated circuits (ASICs), one or more field-programmable gate arrays (FPGAs), and / or one or more system-on-a-chip (SoCs). Data processing devices may also include one or more processors, such as one or more microprocessors, one or more central processing units (CPUs), one or more graphics processing units (GPUs), and / or one or more signal processors, particularly one or more digital signal processors (DSPs). Data processing devices may also include a physical or virtual collection of computers or other such units.

[0076] In different embodiments, the data processing device includes one or more hardware and / or software interfaces and / or one or more storage units.

[0077] The storage cell can be designed as volatile data memory, such as dynamic random access memory (DRAM) or static random access memory (SRAM), or as non-volatile data memory, such as read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), flash memory or flash EEPROM, ferroelectric random access memory (FRAM), magnetoresistive random access memory (MRAM), or phase-change random access memory (PCRAM).

[0078] According to another aspect of the invention, an imaging system is provided, which includes a data processing system and an imaging mode according to the invention. The imaging mode is configured to generate image data.

[0079] For example, the imaging modality can be an X-ray imaging modality, such as an X-ray angiography system or a C-arm X-ray system.

[0080] According to another aspect of the invention, a computer program is provided having instructions that, when executed by a data processing system, cause the data processing system to perform a computer-implemented method according to the invention.

[0081] The instructions may exist, for example, as program code. The program code may be provided, for example, in the form of binary code or assembler and / or source code of a programming language, such as C, and / or program scripts, such as Python.

[0082] According to another aspect of the invention, there is a further computer program having additional instructions that, when executed by an imaging system according to the invention, particularly by a data processing system of the imaging system, cause the data processing system to perform a computer-implemented method according to the invention and to generate image data from the imaging modality.

[0083] The additional instructions may exist, for example, as program code. The program code may be provided, for example, in the form of binary code or assembler and / or source code of a programming language, such as C, and / or program scripts, such as Python.

[0084] According to another aspect of the invention, a computer-readable storage medium, particularly a physical and / or non-volatile computer-readable storage medium, is provided that stores a computer program according to the invention and / or another computer program according to the invention.

[0085] The computer program, the additional computer program, and the computer-readable storage medium are respectively computer program products having the instructions and / or the additional instructions.

[0086] Other features and combinations of features of the invention are derived from the accompanying drawings and description, as well as the claims. In particular, other embodiments of the invention do not necessarily include all features of any one of the claims. Other embodiments of the invention may have features or combinations of features not mentioned in the claims. Attached Figure Description

[0087] The present invention will now be explained in more detail with the aid of specific embodiments and accompanying diagrams. In the drawings, identical or functionally identical elements may be provided with the same reference numerals. Identical or functionally identical elements may not necessarily be described repeatedly in different figures if necessary.

[0088] In the attached diagram:

[0089] Figure 1 A schematic diagram illustrating an exemplary embodiment of the imaging system according to the present invention is shown;

[0090] Figure 2 A schematic flowchart illustrating an exemplary embodiment of a computer-implemented method for determining the position of an implant in a hollow organ according to the present invention;

[0091] Figure 3 A schematic diagram showing a two-dimensional image illustrates the implant and the assistive tools used to insert the implant into a hollow organ;

[0092] Figure 4 A schematic flowchart illustrating another exemplary embodiment of a computer-implemented method for determining the position of an implant in a hollow organ according to the present invention;

[0093] Figure 5 A schematic flowchart illustrating another exemplary embodiment of a computer-implemented method according to the present invention for determining the position of an implant in a hollow organ; and

[0094] Figure 6 The diagram shows two-dimensional images illustrating the implant and the assistive tools used to insert the implant into the hollow organ. Detailed Implementation

[0095] exist Figure 1 The diagram illustrates an exemplary embodiment of the imaging system 1 according to the present invention. The imaging system 1 can be used to determine the location of an implant 9, such as a blood flow guiding device, in a hollow organ, such as a cerebral artery of a patient 5.

[0096] The imaging system 1 has imaging modes, such as X-ray imaging modes 3 and 4. The imaging mode settings are used to generate image data 7, 7a, 7b, 7c, 8, 8a, 8b, 8c, which consist of one or more two-dimensional images 7, 7a, 7b, 7c, 8, 8a, 8b, 8c, which respectively represent the implant 9 in a first state within the hollow organ and correspond to up to ten different views.

[0097] Imaging system 1 has a data processing system 2 equipped with a display device 6. The data processing system 2 is configured to execute a computer-implemented method for determining the position of an implant 9, such as a blood flow guiding device, in a hollow organ, such as a cerebral artery, particularly based on image data 7, 7a, 7b, 7c, 8, 8a, 8b, 8c.

[0098] If the imaging mode is designed as X-ray imaging mode 3 or 4, the imaging mode includes a source unit 3 with an X-ray source, a detector unit 4 with an X-ray detector, and a control system configured to control the X-ray source and the X-ray detector to generate two-dimensional images 7, 7a, 7b, 7c, 8, 8a, 8b, and 8c.

[0099] Figure 2 A schematic flowchart illustrating an exemplary embodiment of a computer-implemented method according to the present invention for determining the position of an implant 9 in a hollow organ.

[0100] In step 100, image data 7 is acquired, particularly from an imaging modality. Based on image data 7, in step 200, a three-dimensional first reconstruction of the implant 9 in a first state is generated. Image data 7 comprises two to ten two-dimensional images 7 having different views of the implant 9, and / or the first reconstruction is generated based on image data 7 and a preset geometric and / or mechanical model of the implant 9. In step 300, a representation of the first reconstruction is displayed on a display device 6.

[0101] Steps 100 to 300 can be repeated multiple times, wherein in each repetition, different states of the implant can exist or be represented by image data 7. This enables dynamic real-time representation of the implant 9.

[0102] Figure 3 A schematic two-dimensional image 7 is shown, illustrating an implant 9, such as a stent or blood flow guide, and an auxiliary tool 10, such as a guidewire, for introducing the implant 9 into a hollow organ.

[0103] Figure 4 Showing based on Figure 2 A schematic flowchart illustrating an exemplary embodiment of a computer-implemented method according to the present invention.

[0104] exist Figure 4 In this implementation, image data 7 and 8 include a first two-dimensional image 7 and a second two-dimensional image 8, the first two-dimensional image showing the implant along a first viewing direction and the second two-dimensional image showing the implant along a second viewing direction. The imaging modality can therefore be designed, for example, as a dual-plane X-ray imaging modality 3 and 4. By considering two different viewing directions, a first reconstruction can be generated with high accuracy.

[0105] Figure 5 Showing based on Figure 4 A schematic flowchart illustrating an exemplary embodiment of a computer-implemented method according to the present invention.

[0106] exist Figure 4In the implementation of the method, the image data 7a, 7b, 7c, 8a, 8b, 8c include multiple, such as three two-dimensional first images 7a, 7b, 7c and multiple, such as three two-dimensional second images 8a, 8b, 8c. The multiple first images show the implant 9 along a first viewing direction at different time points, and the multiple second images show the implant along a second viewing direction at different time points.

[0107] In step 200, an averaged first image is generated based on the first images 7a, 7b, and 7c, and an averaged second image is generated based on the second images 8a, 8b, and 8c. The first reconstruction is generated based on the averaged first image and the averaged second image.

[0108] Optionally, one or more markings 11a, 11b, such as two markings 11a, 11b, may be provided on the implant 9 or the assistive tool 10. These markings are designed to be opaque to X-rays, for example. This is in Figure 6 The diagram illustrates this schematically. For example, the mark position of each mark 11a, 11b is determined in each of the first images 7a, 7b, 7c. The first images 7a, 7b, 7c are aligned relative to each other according to the mark positions, specifically such that the corresponding marks in the different aligned first images 7a, 7b, 7c coincide. An averaged first image is generated based on the aligned first images 7a, 7b, 7c, and an averaged second image is generated based on the aligned second images 8a, 8b, 8c.

[0109] In the present invention, especially regarding Figure 4 In the various embodiments described, visual enhancement or highlighting of the implant 9 and noise suppression are achieved through averaging. For this purpose, an improved, particularly sharpened version of the implant 9 can be calculated periodically by averaging based on a sliding time window that covers a series of consecutively acquired two-dimensional images 7, 7a, 7b, 7c, 8, 8a, 8b, 8c. In some embodiments, images 7, 7a, 7b, 7c, 8, 8a, 8b, 8c can be first registered according to markers 11a, 11b, and then averaged according to the current position of the sliding time window.

[0110] Since motion is usually not a major concern in brain imaging, averaged images can be generated even without registration if necessary. This also means that in many cases, averaging a small number of consecutive images 7, 7a, 7b, 7c, 8, 8a, 8b, 8c within the current position of the sliding time window is sufficient.

[0111] In some implementations, a symbolic reconstruction method is used to generate a first reconstruction of implant 9. In this case, a very limited number of views are sufficient. Symbolic reconstruction is particularly suitable for objects with sufficiently simple geometry, and therefore also suitable for implant 9, especially blood flow guiding devices.

[0112] In some implementations, markers 11a and 11b are tracked for two different views. This can be achieved, for example, by an algorithm based on a trained machine learning model, such as a trained neural network. Two averaged images are generated for the two views. The implant is segmented in each averaged image, for example using known analysis or data control methods. Symbolic reconstruction, particularly real-time reconstruction, is then performed on the current state of the implant 9, and an updated 3D representation of the reconstruction results is displayed.

[0113] In some implementations, instead of simply visualizing the implant 9 based on its three-dimensional reconstruction, this reconstruction can be integrated with and displayed in real time with the pre-interventional planning to better assess whether the expected deployment result is sufficiently close to the planned outcome. If the current state deviates significantly from the pre-interventional planning result, the implant 9 can be withdrawn and insertion restarted, for example. Alternatively, a different implant 9 can be selected for the insertion. This may occur, for example, when the pre-interventional planning result is based on inaccurate vascular geometry data or when the estimation of vascular geometry deformation caused by the implant 9 is inaccurate.

[0114] In some implementations, particularly in cases of geometric ambiguity, pre-interventional 3D angiographic images can be used to algorithmically constrain the solution space during symbol reconstruction. These pre-interventional 3D angiographic images can be updated using appropriate deformation correction techniques, such as those known in the field as endovascular aneurysm repair guidance (EVAR).

[0115] In some implementations, a geometric and / or mechanical model of the implant 9 may be used to constrain the solution space in the symbolic reconstruction step.

[0116] Other advantages of this invention include improved accuracy of implant placement 9 and a potential reduction in the number of implants 9 required to achieve the desired outcome, thereby reducing equipment requirements. Both improve treatment outcomes and reduce surgical costs, particularly material costs.

[0117] In the above description, regardless of the grammatical gender of a particular term, people of male or female identity should be included.

Claims

1. A computer-implemented method for determining the location of an implant (9) in a hollow organ, wherein: - Acquire image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c), which consists of one or more two-dimensional images (7, 7a, 7b, 7c, 8, 8a, 8b, 8c), which represent the implant (9) in a first state within the hollow organ from a total of up to ten different views. - Based on image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c), generate a three-dimensional first reconstruction of the implant (9) in the first state; - Among them, the image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c) contains two to ten two-dimensional images (7, 7a, 7b, 7c, 8, 8a, 8b, 8c) with different views of the implant (9). - Among them, the image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c) includes two or more two-dimensional first images (7, 7a, 7b, 7c), which represent the implant (9) in a first state at different time points according to a first viewing direction; - Generate at least one averaged first image based on the two or more first images (7, 7a, 7b, 7c); - Generate the first reconstruction based on the at least one averaged first image; and - Display the representation of the first reconstruction on the display device (6).

2. The computer-implemented method according to claim 1, wherein, The first reconstruction is generated based on image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c) and a pre-defined geometric and / or mechanical model of the implant (9).

3. The computer-implemented method according to any one of the preceding claims, wherein, Hollow organs are vascular structures, and implants (9) are vascular implants.

4. The computer-implemented method according to claim 3, wherein, The implant (9) is a blood flow diversion implant used to guide blood flow in a hollow organ to pass next to an aneurysm in the hollow organ.

5. The computer-implemented method according to any one of the preceding claims, wherein, The image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c) consists of one or more two-dimensional images (7, 7a, 7b, 7c, 8, 8a, 8b, 8c), which represent the implant (9) in a first state within the hollow organ and correspond to a maximum of five different views.

6. The computer-implemented method according to any one of the preceding claims, wherein, The geometric and / or mechanical model of the implant (9) includes the length and / or diameter of the implant (9).

7. The computer-implemented method according to any one of the preceding claims, wherein, The first reconstruction is generated based on image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c) and the pre-intervention representation of the hollow organ.

8. The computer-implemented method according to any one of the preceding claims, wherein - In each of the two or more first images (7, 7a, 7b, 7c), a marker position is determined for the marker (11a, 11b) showing the position of the implant (9); - Align two or more first images (7, 7a, 7b, 7c) relative to each other according to the marked positions; and - The averaged first image is generated based on two or more aligned first images (7, 7a, 7b, 7c).

9. The computer-implemented method according to any one of the preceding claims, wherein, The two or more first images (7, 7a, 7b, 7c) correspond to images in an image sequence with an image acquisition rate ranging from three images per second to 15 images per second.

10. The computer-implemented method according to any one of the preceding claims, wherein, The contour of the implant (9) is determined by segmenting the averaged first image, and a first reconstruction is generated based on the contour of the implant (9).

11. The computer-implemented method according to any one of the preceding claims, wherein - Acquire additional image data consisting of one or more two-dimensional additional images representing the implant in a second state within the hollow organ from up to ten different views (9). - Generate a three-dimensional second reconstruction of the implant (9) in the second state based on the additional image data; - The additional image data includes two to ten additional two-dimensional images with different views of the implant (9), and / or a second reconstruction is generated based on the additional image data and the geometric and / or mechanical model of the implant (9); and - Display the second reconstruction representation on the display device (6).

12. The computer-implemented method according to any one of the preceding claims, wherein, The representation of the first reconstruction includes the superposition of the first reconstruction and the reference representation of the hollow organ.

13. A data processing system (2, 6) configured to perform a computer-implemented method according to any one of the preceding claims.

14. An imaging system (1) having a data processing system (2, 6) according to claim 13 and an imaging mode (3, 4) configured for generating image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c).

15. A computer program product having: - Instructions, which, when executed by the data processing system (2, 6), cause the data processing system (2, 6) to perform the computer-implemented method according to any one of the preceding method claims; or - Additional instructions, which, when executed by the imaging system (1) according to claim 14, cause the data processing system (2, 6) to perform the computer-implemented method according to any one of the preceding method claims and cause the imaging mode (3, 4) to generate image data (7, 7a, 7b, 7c, 8, 8a, 8b, 8c).

Images