HK1150894B - Automatic geometrical and mechanical analyzing method and system for tubular structures - Google Patents

Description

Automated geometric and mechanical analysis method and system for tubular structures

Technical Field

The present invention relates to the field of diagnostic systems, and more particularly to computer-based diagnostic systems for hollow structures, such as elongated hollow structures, such as tubular structures, for example vascular structures such as including vascular tissue. The diagnostic system provides, for example, analysis and information data regarding the geometry and mechanical properties of the elongated hollow structure.

Background

Many procedures, such as interventions and diagnostics on vascular tissue, have to be performed at internal anatomical sites. The physician's information about these medical procedures is enriched by image data acquired with an imaging device (modality), such as a Computer Tomography (CT) or Magnetic Resonance (MR) based scanning apparatus. Typically, this provides a plurality of two-dimensional (2D) images, also referred to as slices, of the patient's anatomy. Some scanning apparatuses include computer hardware and software for constructing a three-dimensional (3D) dataset of multiple 2D images.

Furthermore, computer system implemented models for visualizing specific isolated regions of a patient's body (such as a patient's organ) and for virtually measuring the dimensions of the patient's anatomy are known. These known models are mainly based on threshold methods and therefore require a moderately high image quality in order to function properly. For example, high image resolution and high contrast between objects to be separated are required. For CT scanning systems, this is in many cases achieved only by using contrast agents (contrast media) and/or high X-ray radiation doses, both of which are burdens on the patient that are desired to be reduced or avoided.

A particular model is described in Olabariaga SD, Rouet J-M, Fradkin M, Breeuwer M and NiessenlWJ, (2005) Segmentation of Thrombus in Abdominal atomic atmospheric analytical from CTA With Nonparametric Statistical information Grey Level application modeling. IEEE Trans Med 24.477-485. The model disclosed in this paper is based on the segmentation of an endoluminal thrombus (ILT) in an Abdominal Aortic Aneurysm (AAA), which uses the principles of deformable models. However, the enablement of this model requires the presence of an ILT and also a high threshold between the lumen and the ILT, and does not take into account the vessel branching, which is often required for analysis. Therefore, this model has very limited applicability in practice.

For some medical procedures, such as when assessing the risk of rupture of an AAA, or when identifying the vulnerability of an arterial stenosis, the mechanical loading of vascular tissue is useful to enrich the physician's information for planning the treatment. This information about the type of mechanical loading situation cannot be provided by the imaging system, but rather a structural analysis based on a post-processing of the data they provide can facilitate this. However, no automated and integrated system is currently available which integrates all structurally relevant anatomical objects and provides information about the mechanical load situation of e.g. vascular tissue. For example, in the case of AAA, the arterial wall and ILT present in almost all clinically relevant AAAs are relevant structural components.

An anatomical visualization and measurement system and method is disclosed in U.S. patent application US2006/0100502-A1 to Chen DT et al, which is hereby incorporated by reference in its entirety. According to this disclosure, a suitable set of 2D slice images obtained by scanning a blood vessel is used to determine the risk of blood vessel rupture. The method comprises the following steps: generating a mesh model of the vessel using the set of 2D slice images; performing finite element stress analysis on the mesh model to calculate stress levels at different locations on the mesh model; and determining a risk of vessel rupture based on the calculated stress levels at different locations on the mesh model. However, the method and system disclosed in US2006/0100502-A1 is limited to providing a single surface mesh of the vessel lesion, and thus only the shell-like structural effect of the vessel wall can be considered, e.g. using shell finite elements. Details regarding this Finite Element (FE) formulation are disclosed, for example, in Zienkiewicz OC and Taylor RL, (2005) The FiniteElement Method for Solid and Structural Mechanics, Butterworth Heinemann, 6th edition, which is incorporated herein in its entirety. Thus, in the case of AAA, ignoring the structural impact of the ILT results in impractical and unreliable mechanical predictions for AAA. In documents such as the Effect of organic sulfur on wallstreess in patent-specific models of exogenous origin in Wang et al (2002), J Vasc Surg.36, p.598-604), several studies have highlighted this Effect. Therefore, unreliable diagnoses and predictions of the risk of rupture may be made, which is unsatisfactory at least from the point of view of patient safety. Furthermore, the method and system disclosed in US2006/0100502-A1 require manual intervention, for example to remove unwanted segmentation elements. In addition, the method and system disclosed in US2006/0100502-A1 uses a number of different software products (one for separating 2D images, one for meshing the surface (representing the vessel wall), and one for performing Finite Element (FE) analysis) that are included to predict the mechanical load state. This is inconvenient for the user because of technical difficulties, since it is necessary to ensure the interface between the different software products in a reliable and secure manner, which is difficult to guarantee in practice.

Another method is described in automatic method for determination of stress distribution in human adoptive analysis by Raghaven et al (2005, J biomechEng.127, p.868-71). In this disclosure, the geometry data from the 3D visualization system is post-processed. Again, this method is limited to modeling the outer surface of the AAA with shell-like structure effects and ignoring thick-walled structures (or volume effects) of the vessel wall.

To avoid the "mesh-based" stress artifact of FE analysis, a suitably high quality computational grid (mesh) is required, as well as applying different types of mesh smoothing strategies. However, the methods and systems disclosed in US2006/0100502-A1 or Raghaven et al (2005) apply a mesh smoothing strategy that changes the geometry of the object, so that the (external) geometry of the vessel body cannot be accurately captured, i.e. there is a mismatch between the model geometry and the underlying image data. Accordingly, there is a need for at least one improved method and/or system that facilitates accurate geometric and mechanical modeling of hollow structures, such as elongated hollow structures, e.g. tubular structures, such as vascular bodies, in order to provide reliable data regarding their geometric properties and their mechanical loading conditions.

For example, Dimitrios E.Kiousis, T.Christian Gasser and Gerharda.Holzapfel, A Numerical Model to Study the Interaction of VascalStents with Human Atherotic Length, Ann Biomed Eng.2007; 35(11): 1857-69 represent prior art finite element modeling of vascular bodies, which, while taking into account the volumetric effect of the structure by discretizing the tissue into hexahedral cells, is a semi-automated solution. The principle proposed therein requires expert knowledge of the structural modeling and also involves several steps, which are inconvenient due to the difficulty of applying the technology.

The semi-automated reconstruction scheme disclosed in an Biomed end of kiusis et al (2007) basically applies three steps to generate a computational grid for structural analysis:

1) the in-plane segmentation is performed using, for example, a NURBS representation of the edges, wherein a deformable model is used on a single image slice in order to ignore out-of-plane information of the image dataset. This principle can only be applied to a sub-class of geometries and excludes, for example, saccular aneurysms, see fig. 0, which have significant clinical relevance. Moreover, this principle cannot be applied to vessel branching.

2) A solid model is generated based on edge information defined by the segmentation. Smoothing of the segmentation curve is always required here, in particular to avoid scattering in the out-of-plane direction. This naturally changes the geometry and thus the geometry of the vessel volume as defined by the image dataset cannot be maintained.

3) The solid model is formed into a mesh, which for the true (clinically relevant) geometry of the vessel body needs to be subdivided into smaller volumes, which are simple enough to allow automatic generation of the mesh. This is often a time consuming task requiring engineering expert knowledge regarding grid generation and structural analysis. Most importantly, if the geometry is too complex, modification of the solid model may even be required to facilitate generation of the mesh for the structure. Thus, the geometry of the vessel volume as defined by the image dataset cannot be maintained.

In summary, the currently known methods are characterized by severe manual intervention and require engineering expert knowledge of the user, which makes them unusable for clinical applications. Likewise, it is emphasized that the image data used in the Ann Biomed end of kiusis et al (2007) cited above is based on in vitro nuclear magnetic resonance, where naturally better image quality can be obtained than from a standard clinical imaging dataset.

Therefore, a fully automated solution is needed so that detailed structural analysis of tubular bodies, such as vascular bodies, becomes clinically applicable. There is a need for a clinically applicable system for clinical users without the background of engineering experts to perform this protocol. Furthermore, this approach should be applicable or useful for in vivo 3D image data of clinically available patients with lower resolution than ex vivo 3D image data.

Disclosure of Invention

Accordingly, embodiments of the present invention preferably aim to mitigate, alleviate or eliminate one or more deficiencies, disadvantages or issues in the art, such as the above-identified, singly or in any combination by providing a system, a method, a computer program, a medical workstation and a graphical user interface according to the appended patent claims.

The present invention uses a combination of 3D image reconstruction and hexahedral mesh generation. This enables a fast and robust generation of a finite element mesh for structural analysis of the tubular body. This principle is significantly different from other methods such as those mentioned and cited by Ann Biomed Eng et al (2007) cited above.

The difference between the present invention and the existing methods makes a fully automated approach feasible so that detailed structural analysis of the vascular body becomes clinically applicable. These differences enable the development of systems that can be operated even by users without engineering background.

The fully automatic 3D method presented by the present invention does not distinguish out-of-plane directions so that subsequent smoothing is not required and the precise 3D geometry of the vascular body is maintained.

Finally, the invention defines a digitally robust and efficient method which is applicable to clinical image dataset recordings.

Embodiments of the present invention include methods and systems for analyzing vascular bodies for their geometric characteristics and mechanical loading. To this end, the method or system generates a geometric and structural model of the vessel body from a standard image dataset. The method or system works automatically and the blood vessel volume is analyzed by clinical personnel (i.e., users without engineering expertise) within clinically relevant time. There is typically no engineering background for clinical personnel who manipulate such systems. Most critical in this sense is the integration of new volumetric mesh generation and 3D segmentation techniques. The resulting geometric and structural models distinguish structurally related types of tissue, for example for abdominal aortic aneurysms, vessel walls and endoluminal thrombi are considered separately. The structural study of the vascular body is based on detailed nonlinear finite element analysis. Here, the derived geometric model of the vascular tissue, the in vivo boundary/load condition, and the finite deformation constitutive description present (render) structural biomechanical problems. Different visualization principles are provided which allow efficient and detailed study of the derived geometric and mechanical data. In addition, this information is pooled, and the statistical properties derived therefrom, can be used to analyze the vascular body of interest.

The relevant clinical time, where results can be provided by the method or system, is in the range of a few minutes with current computing power typically clinically available.

According to a first aspect of the invention, a method is disclosed for automatically analyzing the geometrical properties and mechanical load conditions of a tubular body, such as a vascular body.

The method is for analyzing a substantially tubular body having a wall with a wall thickness. The method comprises the following steps: 3D reconstructing at least one component of at least a part of the tubular body and/or at least one element associated therewith from the image dataset; generating quadrilateral and/or hexahedral Finite Element (FE) meshes for the members and/or units; performing structural nonlinear finite element analysis on at least one member and/or element of the tubular body; and thereby providing at least information data about the geometrical properties and the internal mechanical loads of at least a sub-portion of said portion of the tubular body for analysis of the tubular body.

This method can be applied to the entire portion of the tubular body including the branches and side branches.

Alternatively, the geometric characteristics and internal load data may be provided separately for further processing. The geometric properties, i.e. the data representing the geometric structure, are associated to its local mechanical properties. Both the geometric and mechanical properties are provided as a 3D data set for further processing.

According to a second aspect of the invention, a system for automated analysis of geometrical and mechanical load conditions of a tubular structure, such as a vascular body, is disclosed.

The system is used for analyzing a substantially tubular body having a wall with a wall thickness. The system comprises: means for 3D reconstruction of at least one member of at least a portion of the tubular body and/or at least one unit related thereto from the image dataset; means for generating a quadrilateral 2D and/or hexahedral 3D Finite Element (FE) mesh for the members and/or cells; means for performing a structural nonlinear finite element analysis on the at least one component and/or element; and means for thereby providing at least information data about the geometrical properties and the internal mechanical loads of at least a sub-portion of said portion of the tubular body for analysis of the tubular body.

According to a third aspect of the present invention there is provided a computer program for processing by a computer. The computer program comprises a code segment for a medical workstation providing an automatic analysis of geometrical properties and mechanical load conditions of a tubular structure, such as a vessel body.

The computer program is for processing, with a computing device, to analyze a substantially tubular body having a wall with a wall thickness. The computer program includes: a first code segment for 3D reconstruction of at least one member of at least a part of the tubular body and/or at least one unit related thereto from an image data set; a second code segment for generating quadrilateral and/or hexahedral Finite Element (FE) meshes for said members and/or elements; a third code segment for performing structural nonlinear finite element analysis on the at least one component and/or element; and a fourth code segment for thereby providing at least information data about the geometrical properties and the internal mechanical loads of said at least a portion of the tubular body for analysis of the tubular body.

A member in this context is a structural member of an anatomical structure.

According to yet another aspect of the present invention, a graphical user interface is provided for visualizing the geometric properties and internal mechanical loads of vascular bodies using charts, 2D and 3D contour plots (contour plots), and 3D color coded geometric objects.

In an embodiment, the graphical user interface may allow for interpretation of the geometric and mechanical information of the vessel body with respect to information from the pooled data.

In another aspect of the invention, a method is provided for analyzing vascular bodies for their geometric characteristics and mechanical loading. The method comprises the following steps: generating at least one geometric and structural model of the at least one vessel body from the at least one image patient data set; differentiating between structurally related types of tissue in the geometric and structural models, such as between vessel walls and endoluminal thrombi in the case of abdominal aortic aneurysms; carrying out structural research on the blood vessel body based on nonlinear finite element analysis; structural biomechanical problems are presented based on a constitutive description of finite deformation of the vessel wall, in vivo boundary/loading conditions, structural models to provide geometric and mechanical data thereof.

Further embodiments of the invention are defined in the dependent claims, wherein features of the second and subsequent aspects of the invention are suitably modified with respect to the case of the first aspect.

Embodiments of the present invention differ significantly in several respects from the prior art, for example as mentioned in the "background" section. Most importantly, some embodiments of the present invention provide for integrating all post-step patient scans into a single (stand-alone) system and thereby provide information about the patient's specific vascular lesions, i.e. their geometric characteristics and their mechanical load situation, within a clinically acceptable time. The gist of an embodiment of the present invention may work fully automatically, which makes its clinical application and/or clinical acceptance feasible, and its application without expert knowledge, e.g. in engineering.

Furthermore, the present invention in some embodiments uses the principles of deformable models to reconstruct the geometry of the vessel body, and therefore lower image quality can also be processed and improved results compared to reconstructions based on thresholding methods. Deformable models have several advantages over threshold-based methods, especially when applied to medical images, see for example Suri et al a reviewon 30MR spatial image processing: skeletton versus nonsskeletoproppoches: part II. (2002, IEEE Trans Inf Technol biomed.6338-50). The method applied by some embodiments of the invention directly affects the safety of the patient, for example for image data from a CT scan the amount of contrast agent and/or x-ray radiation burden may be reduced. Notably, while the method described in Olabarriaga et al (2005) uses the principles of deformable models, it requires a high threshold for its initialization.

Some embodiments of the invention provide 3D accurate image segmentation based on deformable models. The applied principle presents a robust method and the reconstructed and discretized (mesh generated) object can be directly used as geometrical input for finite element analysis.

Some embodiments of the invention also provide automated quadrilateral mesh generation of the surface of the body of the blood vessel.

Some embodiments of the invention also provide automatic hexahedral dominated meshing of the volume of the body of the blood vessel and thus allow the application of efficient mixed finite elements, for example the so-called Q1P0 formulation, see Simo and Taylor, 1991, Quasi-compressive finite elastic mechanics in principal protocols, content basis and numerical algorithms, comprehensive application Mech engrg.85.273-310. This is essential to express the incompressible nature of the vascular tissue in a digitally efficient and adequate way.

Some embodiments of the invention also provide automatic 2D and 3D mesh smoothing and optimization to improve the quality of the finite element mesh and thus the prediction results.

Some embodiments of the present invention also provide for full 3D structural analysis of tubular bodies, such as vascular bodies, in which different types of tissue are treated separately.

It should be emphasized that the term "comprises/comprising" when used in this specification is taken to specify the presence of stated features, integers, steps or components but does not preclude the presence or addition of one or more other features, integers, steps, components or groups thereof. As used herein, the term "and/or" includes any and all combinations of one or more of the associated listed items.

Drawings

These and other aspects, features and advantages of the present invention will become apparent from and elucidated with reference to the following description of embodiments of the invention, taken in conjunction with the accompanying drawings, in which

FIG. 0 is a schematic view of a saccular aneurysm;

fig. 1 is a flow diagram illustrating automated geometric and mechanical analysis of a vessel body, in accordance with an embodiment, wherein a general system for performing a method of implementation includes a medical workstation, and wherein some embodiments in the form of a computer program for implementing the method are stored on a computer readable medium for execution by the medical workstation;

FIG. 2 is a pictorial view of an image viewer of a graphical user interface of the system that enables a user to explore a loaded image patient data set and define regions of interest, for example, by mouse interaction;

fig. 3 is an illustration of initialization of reconstruction of a 2D image of patient data, wherein a user may, for example, place a circular dot within an arterial lumen for initialization;

FIG. 4 is a flow chart showing the algorithmic formulation of a snake model that utilizes a Finite Element (FE) problem and an iterative strategy to solve the resulting nonlinear numerical problem;

FIG. 5 is an illustration of a lumen in a 2D image of patient data for which a segmentation has been performed using the snake model of FIG. 4, in which the branched (in this non-limiting example renal) arteries are truncated, so as to reduce the geometric complexity of the problem, so that a finite element analysis of the entire vessel volume becomes feasible;

fig. 6(a) and (b) are schematic diagrams of a subdivision strategy by introducing node lines between adjacent snake nodes, where (a) shows refinement without subdivision (tessellation) and (b) shows refinement with subdivision;

FIGS. 7(a), (b), and (c) are schematic diagrams of applying a strategy to locally refine a mesh, where (a) shows removing a quadrilateral at the boundary of a surface, (b) shows locking the collapse of the quadrilateral, and (c) shows improving the sick cell;

FIG. 8 is a diagram of a 3D reconstructed luminal surface of an AAA object, wherein a mesh is generated for the surface of the AAA object with optimized quadrilateral cells;

fig. 9 is a schematic diagram of the definition of a (hexahedral) volume segmentation, which is used as a basis for generating a mesh for a complex-shaped vessel volume;

fig. 10(a) is a diagram of a strategy for generating a mesh for an artery wall, based on the definition of volume segmentation;

FIG. 10(b) is a graph showing the functional relationship between the thickness of the ILT and the arterial wall thickness;

FIG. 11 is a schematic diagram of a strategy for generating meshes for ILTs based on the definition of volume segmentation and its main generation of hexahedral cells;

FIG. 12 is a graphical representation of the definition of the principal material axes of the stress field pre-calculated from the structure of the vascular body; and

fig. 13 is a graphical representation of a 3D visualization of vmeses stress (left) and risk of fracture (right) for a particular AAA wall, where this information is color coded.

Detailed Description

Specific embodiments of the present invention will now be described with reference to the accompanying drawings. This invention may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the invention to those skilled in the art. The terminology used in the detailed description of the embodiments illustrated in the accompanying drawings is not intended to be limiting of the invention. In the drawings, like reference numerals designate like elements.

The following description focuses on embodiments applicable to analyzing vascular lesions and in particular to analyzing AAA or carotid stenosis. However, it will be appreciated that the invention is not limited to this particular application, but may be applied in some embodiments to: many other tubular internal organs including, for example, other blood vessels, the trachea, the urethra, the esophagus, the intestine, the fallopian tubes, the brain, the atrial appendage including the Left Atrial Appendage (LAA), the coronary vessels, etc.; or an external part of the body, such as a limb, including a leg, arm, finger, etc. Furthermore, some embodiments of the invention may also be applied to tubular portions of organs (e.g., heart, bone, etc.). Finally, it is also noted that some embodiments of the present invention may also be applied to generally tubular structures, such as pipelines and the like.

While stress analysis of engineered structures is well established, and several analytical tools are commercially available, the present invention enables such analysis to be applied in new ways for medical applications. Biological organs may have complex geometries compared to engineered (artificial) structures, and their 3D reconstruction is another challenging aspect of the present invention.

In fig. 0, a geometry that cannot be reconstructed with the planar method is shown. Horizontal line 200 represents a scan slice. Fig. 0 illustrates the limitations of 2D segmentation compared to the full 3D approach. In detail, schematic geometries (e.g. representing saccular hemangiomas) are outlined which are hardly segmented with 2D methods. Here, the parallel horizontal lines 200 represent image slices, and the methods described and referenced in Ann Biomed end of kiusis et al (2007) (incorporated herein) are unable to segment such relevant clinical geometries. Further, it is noted that one problem with planar reconstruction is that because image information is not taken into account in the reconstruction, significant dispersion along the out-of-plane direction results. Thus, a large amount of smoothing is required and the exact reconstruction geometry cannot be maintained. Embodiments of the present invention overcome these and other disadvantages and the like. This is particularly advantageous for analysis of tubular structures such as branched structures (e.g. vessel branches, irregular structures such as saccular aneurysms, brain or bowel bends, etc.).

In some embodiments, the present invention is a method for analyzing a substantially tubular body having a wall with a wall thickness, wherein the method comprises: 3D reconstructing a component of at least a portion of the tubular body and/or a unit associated therewith from an image dataset; generating quadrilateral and/or hexahedral Finite Element (FE) meshes for the members and/or units; performing structural nonlinear finite element analysis on the component and/or unit; and thereby providing at least information data about the geometrical properties and the internal mechanical loads of at least a portion of the tubular body for the analysis of the tubular body.

In fig. 1, these different (algorithmic) steps are described in more detail using an embodiment of the present invention, are shown, and their features are described below.

1) Start of

This step allows the user to start the analysis system. Alternatively, this step may also be entered automatically or upon request from other routines of the medical system, imaging device, or medical workstation associated therewith.

2) Loading image data

During this step, the user loads patient-specific data, for example in the form of a standardized DICOM image patient 3D dataset comprising a plurality of 2D slice image patient data, into an analysis system, for example a medical workstation. For this purpose, a graphical user interface may be used, as well as storing image data and additionally or alternatively other patient-specific information data in a specific file of the analysis system.

3) Defining a region of interest (ROI)

In an embodiment, the medical workstation may have an image data viewer to analyze the loaded image data set and to define a region of interest (ROI) using a human interface device, e.g. using mouse actions, see fig. 2. To this end, the ROI is framed by defining the minimum (201) and maximum (202) axis coordinates of the image dataset, i.e. the axial limits of the reconstruction process. Here, the GLUT and openGL may be used to handle user interaction. Alternatively or additionally, the ROI may be detected automatically or semi-automatically (for confirmation or adjustment by a user of the medical workstation) using a suitable image recognition method, for example based on a suitable object segmentation or recognition method.

4) Initial reconstruction

For example, a specific 2D image slice of the 3D image patient data set at the bottom of the ROI is used to define the initialization of the automatic reconstruction, i.e. where the reconstruction algorithm starts in space. To this end, a substantially circular point, denoted 301 in fig. 3, may be drawn by the user on the image slice in order to identify a tubular structure, such as a vessel wall in the 2D image slice. For fast tracking of lumen boundaries, the point should be as large as possible, but completely within the lumen of a particular artery. Alternatively, or in addition, this boundary delineation of the tubular structure may be performed automatically or semi-automatically, using suitable boundary detection algorithms known in the art. For geometrically complex shaped lesions, such as pseudoaneurysms, it may be more convenient to position the initialization region of the reconstruction to be performed on a slice within the ROI rather than on its boundary. Some embodiments of the invention provide for initialization on any 2D image slice within the ROI, and may use GLUT and openGL to handle user interaction.

5) 3D reconstruction of lumen of tubular structure

A series of method steps is applied to arrive at an accurate 3D reconstruction of the surface of the tubular structure (i.e. the tubular structure is a cavity in this embodiment). The reconstructed surface defines a cavity boundary. The lumen boundaries can be used sequentially, for example in a finite element model, and therefore it is more critical to exclude elements that disturb subsequent steps, such as small branch vessels and image artifacts.

5.1) Snake model

Initialization, as input in step 4), is used to define an initial configuration of the snake model, which itself can be used to segment the lumen from the remaining anatomical information on the current image slice. Here, one or more snake models are used on a particular 2D image slice, depending on the number of cavities to be segmented. The underlying snake model is driven with internal forces due to snake flexion, shear and extension, and with internal forces due to image second order gradients and intensity dependent pressure-like loading. To this end, the image intensity in the vicinity of the pixel of interest is approximated analytically using a quadratic surface. The image intensity is defined using least squares fitting, and the second order gradient at the pixel of interest is calculated using second order differentiation with respect to spatial coordinates.

To calculate the internal and external forces acting on the snake model, the snake is discretized and represented with beam elements (beam elements) in the context of a finite element method, see, for example, Zienkiewicz and Taylor, 2005 cited above. Thus, the snake is approximated with a plurality of nodes connected by the beam unit. Finally, this presents a nonlinear mechanical problem, formulated in a typical finite element scheme, see fig. 4. Fig. 4 is a flow chart showing the algorithmic formulation of a snake model that utilizes a Finite Element (FE) problem and an iterative strategy to solve the resulting nonlinear numerical problem.

In more detail, in some embodiments, a breadth optimization of a global stiffness matrix (global stiffness matrix)400 is performed to render an efficient and stable digital system, as well as to increment the load onto the snake model according to a finite number of time steps 401. At each time step a newton iteration scheme 402 is applied, i.e. in a loop of newton steps, the equation 407 of the linearized system is solved until the equilibrium of snake is determined for the current time step. Finally, a (linearized) global system of algebraic equations is assembled 408 within the loop 403 over all snake units, where first and second order image gradients 404, external nodal forces and stiffness 405, and internal nodal forces and stiffness 406 are computed.

In general, the nonlinear snake problem is solved iteratively until the snake successfully segments the lumen from the remaining anatomical information on the image slice.

To deal with the (mechanical) problems generated, a viscosity can be added which substantially stabilizes the snake movement. To achieve faster convergence, the amount of viscosity is correlated to the norm of the image gradient.

The snake model is applied iteratively until all lumen boundaries in the ROI on all slices are segmented. During this iterative process, the geometric information is stored in the RAM of a computer system (e.g., a medical workstation), and the snake model is initialized with the lumen boundaries on the previous (already segmented) image slices. Again, one or more snake models are used on a particular image slice. If two snake models overlap, as is the case with branches, they are combined into a single model. In order to obtain an equilibrium distribution of nodes representing the segmented lumen, the number of snake nodes is taken from image slice to image slice according to a predetermined distance.

For illustrative purposes, the cavity on slice No.534 of the loaded CT dataset has been segmented using a snake model and is shown in fig. 5. Note that with the snake model represented by 501, the aortic lumen is segmented, and the branch (renal) lumen is truncated, in order to reduce anatomical complexity and thus enable finite element analysis of the entire vessel body. The basic principles presented in this section have similarities to the principles of other image-based reconstruction methods, for example as outlined and cited in the above cited document Kiousis et al (2007), but with significant differences, for example as with the formulation of the finite element problem of the present invention, with the advantage of significantly improved efficiency. It is emphasized that the remaining steps of image segmentation set forth the basic novel principles of the present invention. More importantly, the method of the present invention enables fast hexahedral dominated mesh generation for complex and even multi-branched tubular objects, which will be discussed later. This has not heretofore been effectively achieved.

5.2) refinement

Geometrical information in terms of logically arranging snake nodes (point clouds), as provided by the segmentation in step 5.1, is used to refine the lumen surface (tessellate). Here, a grid of cavity boundaries is generated, wherein a geometrical object is represented using quadrilaterals. Alternatively, triangles may be used and the information may be derived for use by other computer programs, for example, in STL format. Here, it is to be noted that in order to make the refinement with quadrilaterals feasible, the applied classification principle is essential. Prior to the present invention, refinement using quadrilaterals has traditionally been more computationally demanding and resulted in clinically unacceptable computational time.

The underlying algorithm considers a point-by-point description of two continuous lumen boundaries, as segmented with a snake model at step 5.1, and the principles of dynamic programming are used to compute the optimization refinement. In particular, a cost function, such as refining the area of the surface, is minimized.

At the branch, the algorithm refines (combines) two lumen surfaces on one image slice with a single lumen surface on an adjacent slice. To this end, a single cavity surface is divided and both portions are uniquely associated with two cavity surfaces on adjacent slices. Thus, the same algorithm as defined above can be applied to refine the main part of the vessel branch, and the remaining (open) part of the surface can be refined in a single second step. This principle presents an efficient approach, which is linear with respect to the multiple (snake) points used to describe the cavity surface.

If the distance between the associated points becomes too large, additional nodes are introduced in order to avoid cells with poor aspect ratios. To this end, an additional node line is introduced, see fig. 6, in such a way that the above refinement algorithm can be applied again. Finally, it is emphasized that integrated surface mesh generation as proposed by embodiments of the present invention is again fundamentally different from the method proposed in Kiousis et al (2007) cited above, and crucial for an automatic reconstruction scheme.

5.3) smoothing the mesh

The refinement described in 5.2 keeps the nodes of the cavity fixed and, naturally, the generated mesh includes poor-condition surface elements, so that direct application in the finite element method can lead to large local errors. Thus, the surface mesh needs to be improved, and here, mesh smoothing and local cell improvement are applied iteratively until the mesh is optimized. Here, in some embodiments, Laplacian (Laplacian) smoothing may be used, for example, and the strategy for the considered local cell improvement is shown in fig. 7.

In particular, in fig. 7(a) it is shown how the difference quadrangles at the boundaries of the surface may be removed, in fig. 7(b) it is shown how the quadrangles locked to each other during the smoothing algorithm collapse, and in fig. 7(c) it is shown how the sick cells may be improved. Finally, it is noted that surface mesh smoothing can be applied to both types of meshes, namely quadrilateral and triangular meshes, while local cell refinement is only performed on quadrilateral meshes.

5.4) air bag model (Balloon model)

Note that the smoothing of the surface mesh discussed in step 5.3 changes its topology and therefore does not describe the lumen as precisely as given by the image dataset. (this is also a disadvantage of the currently available reconstruction schemes discussed in the "background" section). To account for this, the optimized surface mesh from step 5.3 may be used to initialize the balloon model. The balloon model accurately segments the lumen boundary by taking into account the complete 3D information of the 3D image data set.

For this purpose, the structural effect of the balloon is modeled using shell finite elements, wherein any suitable type of formulation may be applied, for example discrete Kirchhoff (Kirchhoff) may be applied, see for example Zienkiewicz and Taylor, 2005 cited above. Such as internal forces due to membrane deformation and sheet bending and shearing and external forces due to pressure-like loading, e.g. due to second order gradients and intensity of the image, drive the balloon model, which is again formulated as a finite element problem, i.e. similar to that outlined above for the snake model with reference to fig. 4. The image intensities in the vicinity of the voxel (voxel) of interest are analytically approximated using a quadratic hypersurface. It is defined using least squares fitting and the second order gradient at the voxel of interest is calculated using the second order differential in relation to the spatial coordinates. The resulting nonlinear finite element problem is solved iteratively until the lumen is segmented and the geometric information is saved in the RAM of a computer system, such as a medical workstation. Again, viscosity can be added to the digital system to stabilize it, where the amount of viscosity is correlated to the norm of the image gradient.

The method is fully implemented in 3D and the typical outcome of the segmentation algorithm is shown with the quadrilateral surface mesh of the lumen boundary of the particular AAA in fig. 8. In fig. 8, the luminal surface of a 3D reconstruction of an AAA object is shown, wherein the surface of the AAA object is gridded with optimized quadrilateral cells 800 and includes aortic branches 801.

Finally, it is emphasized that this approach presents a completely 3D solution that does not distinguish in an out-of-plane direction as is common to other approaches such as proposed in the Kiousis et al document (2007) cited above. Most importantly, this method does not require smoothing to avoid spreading of the reconstruction in the out-of-plane direction and therefore enables more accurate results to be obtained.

6) 3D reconstruction of the outside of a tubular structure

In this step, the luminal surface, as segmented in step 5 above, may be replicated and thus used as an initialization for another balloon model used for segmenting the outside of the object (i.e. the tubular structure), such as a vessel wall. Thus, the lumen (or the inside surface of the tubular structure) as well as the outside of the segmented tubular structure, such as the vessel body, is represented by the associated mesh, i.e. the pair of lumen and outside points, which can be uniquely defined. This is an essential feature of the applied principle and enables direct forward mesh generation as discussed in subsequent steps 7) and 8) to be further described below for the entire volume.

6.1) air bag model

In order to apply the balloon model to segment the outer side, some modifications to the balloon model discussed in step 5.4) are required. The most critical is to nullify the high image gradients at the lumen boundary (deactivation). To this end, in some embodiments the lumen (as represented within the image dataset) is replaced with the mean intensity (grey value) of the volumes of the "outer neighbors" of the segmented lumen. Finally, the constraint that the distance between the relevant lumen and the lateral point is greater than a predetermined minimum thickness of the blood vessel may be met in some embodiments by a post-geometry correction step. Again, the balloon model is formulated as a nonlinear finite element problem and solved iteratively until the outside of the object is segmented and the geometric information data is available, e.g. saved in the RAM of a computer system such as a medical workstation.

7) Generating meshes for walls of tubular structures such as arterial walls

Embodiments of the present invention use the relevant luminal and lateral meshes, i.e. each node on the luminal border has a copy at the lateral border, which results in a direct forward volumetric mesh generation of the arterial wall. For this purpose, a (hexahedral) volume segment can be defined, as shown in fig. 9. This subdivision of the vessel volume serves as the basis for the mesh generation algorithm for the arterial wall shown in figure 10(a), where for simplicity a single cell across the wall thickness is used.

Again, a full 3D solution is presented which does not require subsequent smoothing compared to e.g. the method proposed in the Kiousis et al document (2007) cited above, and is thus able to reconstruct the exact outer geometry of the tubular body.

An endoluminal thrombus (ILT) is considered to be present in some embodiments if the distance between the relevant lumen and the lateral point is greater than a predetermined maximum thickness. In this case, the data from the segmentations (i.e., steps 5) and 6)) is enriched with predetermined information about the wall thickness. From the reported data in the literature, see, for example, Kazi et al, of the influx of the intracellular polymeric and cellular composition of the extracellular atomic emission wall (2003, J Vasc Surg 38, p.1283-1292), it can be assumed that the wall thickness is substantially dependent on the thickness of the underlying ILT. This is evaluated in some embodiments using the distance between the relevant lumen and the lateral point. The functional relationship between vessel wall thickness and ILT thickness, such as shown in fig. 10(b), may be used in some embodiments to define a mesh of arterial walls. Here, h0 and h1 represent the thickness of the arterial wall without the ILT and covering the ILT, respectively.

A better wall mesh may be obtained by introducing another balloon model defining the interface between the ILT and the wall. Most important in this respect is that the balloon model must be a copy of the outer balloon model in order to be able to define pair-wise nodes again. Penalty forces may be applied at the nodes of the balloon model until a predetermined wall thickness is reached, for example as defined by the above-cited document Kazi et al (2003). This in turn presents a structural problem that is addressed using the finite element technique discussed above.

Finally, it is emphasized that the structural mesh (to some extent) enables anisotropic mesh subdivision, i.e. is different in thickness and circumferential/axial direction. This is a desirable advantageous property because the desired stress gradient in the thickness direction may be significantly different from the desired stress gradient in the circumferential/axial direction.

8) Generating meshes for endoluminal thrombosis (ILT)

Again, volumetric mesh generation of the ILT is easy to implement because some embodiments of the invention use related luminal and lateral meshes. A step-by-step volume mesh generation algorithm that mainly generates hexahedral volume units (breakelements) is applied to generate the mesh for the ILT. The algorithm starts on the outside of the ILT (which is the inside of the artery wall) and gradually generates a mesh towards the lumen boundary of the subject. As long as the (hexahedral) volume part is not completely gridded, see fig. 9, it remains activated (active) and the gridding is done (stepwise) for all activated volume parts from the outside to the luminal side. The volume segments are connected to each other with their radial edges and thus the connectivity of the mesh is enhanced via these edges.

A mesh generation scheme is shown in fig. 11, where only two cells in the thickness direction are considered for simplicity. Note that the algorithm mainly generates hexahedral cells (only exactly the cavity cells may be degenerated hexahedral cells, where two or four nodes are collapsed), and the radial dimensions of the volume mesh can be controlled independently in order to generate a suitable (anisotropic) mesh. Alternatively, the generated mesh may be divided into tetrahedral meshes, as it may be available, for example, for importing meshes into other programs for some reason.

Again, by moving the balloon model stepwise from the ILT wall interface towards the cavity, a better mesh can be obtained; similar to the discussion made in step 7 above. The statements about mesh subdivision are valid similarly as discussed in section 7.

8.1) smooth volume mesh

The (mainly) hexahedral lattice of blocks of the vessel body as generated at steps 7) and 8) needs to be smoothed (e.g. using the Laplacian with constraints method) to be used as geometrical input for the finite element analysis. Here, the surfaces representing the body of the blood vessel or parts thereof, such as the luminal surface, the outer surface and the interface between different types of tissue, are constrained and, therefore, their precise geometry is maintained. Furthermore, the most distorted cell can be improved by moving the connected nodes according to the cell type and optimizing the quality criteria.

9) Output geometry

During the previous step, the geometry of the tubular body (here the vascular body) has been completely defined (from the point of view of finite element dispersion), and this step is used to output the critical geometry. To this end, scalars such as ILT volume, outer diameter of the infrarenal aorta, maximum outer diameter, maximum local ILT thickness, maximum local ILT area, luminal and abluminal minimum and maximum radii of curvature (min. Additionally or alternatively, geometric quantities may be plotted on top of the geometric objects themselves, such as ILT thickness, lumen major radius (principal radii of the lumen), outer major radius of curvature, etc., on the luminal or outer surface of the vessel body. For this purpose, the developed characteristics are color-coded or alternatively contour maps are used. GLUT and openGL may be used here, and the user can explore the data through mouse interaction. For example, standard mouse actions may be used to rotate and zoom in on the model, and select the amount or area to be developed, e.g., from a drop down menu.

10) Defining finite element problem

The volumetric mesh generated in steps 7) and 8) is used as a computational finite element grid for structural analysis. To present the complete finite element problem, the geometric information (finite element mesh) is enriched by the boundary/loading conditions and the compositional properties of the involved vessel tissue.

10.1) Q1P0 Unit

Considering the incompressible nature of e.g. vascular tissue, hybrid finite element methods are followed, and volume viewing (hooking) phenomena of finite element models are avoided. In particular, in some embodiments, see Simo and Taylor, 1991, cited above, a hybrid finite element Q1P0 may be used, which the inventors have found in practical implementations to be a very efficient finite element formulation under the current circumstances.

10.2) composition model

A constitutive description of the type of tissue involved is a crucial part of a reliable prediction of the internal mechanical load (stress field) of a blood vessel. Tissue-driven formulation can be applied to the arterial wall, which enables isotropic or anisotropic non-linear description of the wall, such as in the Review by Gasser et al: described in the Hyperelastic modeling of individual layers with distributed collagen fiber orientations (2006, J R Soc Interface, 3, p.15-35), which is hereby incorporated in its entirety. For example, to model AAA, The set of material parameters involved in The compositional formulation may be defined using least squares fitting of experimental data, such as that given in The effects of anerysm on The biological mechanical behavior of human abdominal aborta (2006a, J biomech.39, p.1324-1334) by Vande Geest et al.

The application of anisotropic compositional models requires the definition of principal material axes throughout the arterial wall (where anisotropy can be correlated locally). This directional information can be generated using structural pre-calculations, where pressure can be applied on the inside of the artery wall, and a single isotropic compositional model, such as neoHookean, can be used. The calculated stress field, which may quantitatively have nothing in common with the true stress state, is used to define the material principal axis. In detail, the principal stress direction is assumed to coincide with the principal material axis. This always gives a true prediction of the main material axis as long as the arterial wall is thin (e.g. compared to the diameter of the vessel body), as shown in fig. 12. Here, the line unit 1201 is used to visualize a main axis and observe the blood vessel branch 1200(one way of the blood flow direction into a vascular branch 1200) along the direction of blood flow, where the mark 1202 indicates the iliac artery.

For ILT organization, a parametric model of the Oggen (Ogden) type is used in the examplesWherein ψ and λ_iAnd i is 1, 2 and 3 respectively represent a free energy function and main stretching. The material parameter c involved can be defined by least squares fitting of available experimental data in the literature (e.g. available experimental data found in A planar biological diagnostic correlation of the valuable test JP et al for the synthetic layer of intra-nuclear in anaerobic atomic sources systems (2006b, J biomech. 39.2347-2354)), which is incorporated herein in its entirety.

10.3) boundary/load conditions

Two different boundary/loading conditions may be applied, namely (i) the displacement at the nodes of the computational grid fixed at the top and bottom boundaries of the ROI, or (ii) the node fixed at one boundary of the ROI, and the axial loading applied at the node of the other boundary of the ROI depending on the in vivo (blood) pressure and the area of the cavity there. In vivo (blood) pressure loads, in terms of deformation-related driven loads, may be applied to the luminal surface of the vascular object. The pressure considered may be predetermined or possibly modified by the user of the system.

11) Solving finite element problems

Step 10) presents the 3D structural finite element problem of the vascular body to be studied in its entirety. In standard finite deformation finite element calculations, a reference configuration is given, and the deformed configuration (in terms of applied external load) is unknown, i.e. it needs to be calculated. However, in the present case, the reconstructed geometry already accounts (state) for the deformed configuration caused by the in vivo loading situation, and its reference configuration is unknown and needs to be calculated. For this, an iterative solution similar to the nonlinear standard finite element method is applied, in which the external load is increased stepwise until the required load level is reached. However, instead of the current configuration, the reference configuration may be updated iteratively during the loading step. Once the mechanical problem is solved, the internal mechanical loads in terms of the six components of the stress tensor are stored, for example, in a system-specific file format. The most time consuming step to solve the numerical problem is to solve the resulting linearized system of equations and, therefore, requires a profile optimization scheme and/or a sparse storage scheme for the direct solver and appropriate pre-processing for the iterative solver. Furthermore, parallel solving strategies for both types of solvers can be applied to shorten the computation time.

12) Output mechanical characteristics

From the calculated mechanical stress tensor, for example, the mechanical quantities to be visualized or used for further processing such as automatic diagnosis (e.g. vMises stress, maximum principal stress, maximum shear stress, etc.) are derived. The mechanical quantity may be visualized (e.g. color coded) as a contour on top of the visualization of the rendered geometric 3D object itself. Here, GLUT and openGL may be used, and the user may conveniently explore the data by means of mouse interaction, as discussed above in step 9). In fig. 13, an example of such a visualization is shown with a color-coded image representing vmeses stress (left) and fracture risk index (right) of the AAA wall. Here, the red areas represent high mechanical stress 1301 or high risk of fracture 1302, wherein their amount is given by a specific color code, namely 1303 for stress and 1304 for risk of fracture.

Finally, the mechanical stress may be correlated to the local strength of the object, for example to the strength of the walls of the AAA and the ILT, and developed to assess the likelihood of its failure (cracking). For this purpose, the local strength of the walls of e.g. the AAA and the ILT can be calculated from e.g. the following documents: 2006c, Towards a non innovative method for determination of patent-specific wall string distribution in adominal atomic emission analysis, Ann.biomed.Eng., 34, of Vande Geest et al: 1098-. The color-coded development of the risk of cracking is shown in fig. 13 (right panel).

13) Exchanging information using a database

The user is able to upload and download a computational model of the vessel body, i.e. its discretized 3D geometry as generated at step 7), and the mechanical data as generated at step 11). Thus, the geometric and mechanical data of the vascular body is aggregated and stored in a database, and the user can access this information using a file transfer protocol. Furthermore, a statistical distribution of key quantities, such as ILT volume, maximum wall stress, maximum ILT stress, maximum diameter, maximum ILT thickness, etc., is derived and stored from the aggregated model. Users can download this statistical information to analyze the computational model of their vessel volume.

14) End up

This step enables the user to terminate the analysis system. Alternatively, other steps may follow, such as branching to other image analysis and treatment software, analysis of new structures or new patients, and so forth. Geometric and mechanical data of the body of the blood vessel can be provided for further processing, such as virtual planning of the surgical procedure. The surgical procedure may include virtual planning for locating a suitable medical implant. A medical graft may be patient configured based on this virtual plan. The virtual plan may then provide data for manufacturing a real medical graft. Method for manufacturing a medical implant, such as a graft vessel, comprising the above method of providing geometric and mechanical data of a tubular body, the above method for virtually planning a surgical procedure, and generating a real medical implant based on the data provided by the latter method.

In some embodiments, the method comprises: loading and preprocessing patient image data; viewing the image dataset; defining a region of interest (ROI); starting a reconstruction process; segmenting (separating) the lumen of the geometric object from the remaining anatomical information of the image dataset; performing 2D and 3D deformable models (e.g., snake and balloon models) to segment the image dataset; surface refinement of the logically arranged point cloud; smoothing, defining, optimizing and solving finite element problems of the 2D and 3D meshes; segmenting (separating) the outer side of the geometric object from the remaining anatomical information of the image dataset; generating volumetric meshes of different vascular tissues for finite element analysis; analyzing the geometric characteristics and internal mechanical load of the vascular body; a prompt message; changing software related characteristics and saving data to a computer readable medium; and uploading and downloading information to and from the database.

In this way, regions with specific mechanical properties can be easily identified. For example, the risk of rupture of the AAA may be determined or diagnosed. This diagnosis can be done manually by an experienced medical staff by developing the analysis; or may be made semi-automatically, e.g. by the system based on the determined mechanical properties, by an indicator (indicator) giving the risk of rupture of a certain area; or automatically by a suitable algorithm that determines the risk of rupture and/or the estimated time to rupture. Statistical distribution of critical quantities can facilitate diagnostics, such as primary or secondary aspects of diagnostics. Thus, by means of embodiments of the present invention, an efficient and reliable diagnosis of a tubular structure and its mechanical load may be provided in a convenient manner. Based on such a diagnosis, appropriate treatment may be initiated to prevent rupture of the AAA, for example, during a medical procedure that augments the area of the AAA with an appropriate medical graft. The surgical procedure may be virtually planned based on such diagnosis, as explained above in section 14). The medical workstation described above includes typical computer components such as a Central Processing Unit (CPU), memory, interfaces, and the like. Furthermore, it is provided with suitable software for processing data received from a data input source, such as data obtained from an imaging device or data from a suitable data carrier, for example in DICOM format. The software may be stored, for example, on a computer readable medium accessible by a medical workstation. The computer-readable medium may comprise software in the form of a computer program comprising suitable code segments for performing the method according to the above-described embodiments. The medical workstation further comprises e.g. a monitor for displaying the visualization of the presentation and suitable human interface means, such as a keyboard, a mouse, etc., e.g. for manual fine-tuning of the automated diagnosis additionally provided by the software.

As used herein, the singular forms "a", "an" and "the" are intended to include the plural forms as well, unless expressly stated otherwise. It will be further understood that the terms "comprises" and "comprising," when used in this specification, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. It will be understood that when an element is referred to as being "connected" or "coupled" to another element, it can be directly connected or coupled to the other element or intervening elements may be present.

Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and will not be interpreted in an idealized or overly formal sense unless expressly so defined herein.

As will be appreciated by one skilled in the art, the present invention may be embodied as a system, method or computer program product. Accordingly, the present invention may take the form of an entirely hardware embodiment, a software embodiment or an embodiment combining software and hardware aspects all generally referred to herein as a "code segment" or a "component (unit)". Furthermore, the present invention may take the form of a computer program product on a computer-usable storage medium having computer-usable program code embodied in the medium. Any suitable computer readable medium may be utilized including hard disks, CD-ROMs, optical storage devices, a transmission media such as those supporting the internet or an intranet, or magnetic storage devices.

The invention has been described above with reference to specific embodiments. However, other embodiments than the above described are equally possible within the scope of the invention. Different method steps than those described above, performing the method by hardware or software, may be provided within the scope of the invention. The different features and steps of the invention may be combined in other combinations than those described. The scope of the invention is only limited by the appended claims.

Claims (20)

1. A method for analyzing a substantially tubular blood vessel body having a wall with a wall thickness, the method comprising:

performing a 3D reconstruction of at least one structural member of at least a part of the vessel body from the image dataset,

generating a quadrilateral and/or hexahedral finite element mesh for the at least one structural member,

performing a structural nonlinear finite element analysis on the at least one structural member, an

Thereby providing information data on geometrical properties and internal mechanical loads of at least a sub-part of said portion of said vessel body for said analysis of said vessel body, wherein said generating said quadrilateral and/or hexahedral finite element mesh comprises using a relevant luminal and an abluminal mesh of said wall of said vessel body, wherein each node on the luminal inside boundary of said wall has a copy at its abluminal outside boundary for volumetric mesh generation of said wall, and

determining a distance between each of the replica nodes as the thickness of the wall at the replica node,

and wherein the mesh is used as a geometric input for the finite element mesh generation.

2. The method of claim 1, wherein said providing information data about the geometric characteristics and internal mechanical loading of at least a sub-portion of said vascular body comprises: automatically analyzing the information data regarding the geometric properties and internal mechanical loads of at least a portion of the vascular body.

3. The method of claim 1, further comprising:

the patient image data is loaded and pre-processed,

the image data set is viewed in the form of a digital image,

a region of interest is defined which is,

the re-establishment procedure is initiated and,

the information of the image data set is artificially enriched,

the lumen of the geometric object is segmented from the remaining anatomical information of the image dataset,

the 2D and 3D deformable models are performed to segment the image data set,

the logical arrangement point cloud is subjected to surface refinement,

the 2D and 3D mesh is smoothed and optimized,

defining, optimizing and solving a finite element problem,

the outer side of the geometric object is segmented from the remaining anatomical information of the image dataset,

quadrilateral and hexahedral meshes are generated for different vascular tissues for finite element analysis,

the geometric characteristics and the internal mechanical load of the blood vessel body are analyzed,

prompting messages, changing software-related characteristics, and saving data to a computer-readable medium,

uploading and downloading information to and from the database.

4. A method according to claim 3, wherein the 2D and 3D deformable models are snake models or balloon models.

5. The method according to any preceding claim, comprising integrating all post-step patient scans into a single system and providing information about a patient's specific vascular lesion, i.e. its geometrical properties and its mechanical load status.

6. The method of claim 5, comprising using a stand-alone system as the system.

7. The method of claim 1, wherein the method comprises using a deformable model for reconstructing the geometry of the vessel body.

8. The method according to claim 1, wherein the 3D reconstruction of at least one structural member comprises 3D accurate image segmentation based on deformable models that exhibit robust methods, and wherein the reconstructed and discretized member object can be directly used as geometrical input for the finite element analysis.

9. The method of claim 1, wherein the method comprises providing quadrilateral mesh generation of at least one surface of the vessel body.

10. The method of claim 1, wherein the method comprises providing hexahedral dominated mesh generation of the volume of the vessel body, applying mixed finite elements for the finite element analysis.

11. The method of claim 1, comprising a full 3D structural analysis of the vascular body, wherein different types of materials are processed separately.

12. The method of claim 11, wherein different types of vascular tissue are treated separately.

13. The method of claim 1, wherein the method comprises providing access to pooled data of vascular bodies.

14. A computer-implemented method of processing, by a computing device, a substantially tubular blood vessel body for analyzing a wall having a wall thickness, the computer-implemented method comprising:

performing a 3D reconstruction of at least one structural member of at least a part of the vessel body from the image dataset,

generating a quadrilateral and/or hexahedral finite element mesh for the structural member,

performing a structural nonlinear finite element analysis on the at least one structural member, an

Thereby providing at least information data on geometrical properties and internal mechanical loads of at least a part of the vessel body for the analysis of the vessel body, wherein the generating the quadrilateral and/or hexahedral finite element mesh comprises using related luminal and abluminal meshes of the wall of the vessel body, wherein each node on the luminal inside boundary of the wall has a replica at its abluminal side boundary for volumetric mesh generation of the wall, and determining the distance between each of the replica nodes as the thickness of the wall at the replica node,

and wherein the mesh is used as a geometric input for the finite element mesh generation.

15. The computer-implemented method of claim 14, further comprising:

the image data is loaded and pre-processed,

the image data set is viewed in the form of a digital image,

the information of the image data set is artificially enriched,

defining a region of interest, an

A reconstruction process is initiated.

16. The computer-implemented method of claim 15, further comprising:

the lumen of the geometric object is segmented from the remaining anatomical information of the image dataset,

the 2D and 3D deformable models are performed to segment the image data set,

triangular and/or quadrilateral surface refinement is performed on the logically arranged point cloud,

the 2D and 3D mesh is smoothed and optimized,

defining, optimizing and solving a finite element problem,

the outer side of the geometric object is segmented from the remaining anatomical information of the image dataset,

surface meshes were generated for different vascular tissues for finite element analysis,

volumetric meshes are generated for different vascular tissues for finite element analysis,

the geometric characteristics and the internal mechanical load of the blood vessel body are analyzed,

prompting messages, changing software-related characteristics, and saving data to a computer-readable medium, an

Uploading and downloading information to and from the database.

17. The computer-implemented method of claim 16, wherein the 2D and 3D deformable models are snake models or balloon models.

18. A system for analyzing a substantially tubular blood vessel body having a wall with a wall thickness, the system comprising:

means for 3D reconstruction of at least one structural member of at least a part of the vessel body from an image dataset,

means for generating a quadrilateral and/or hexahedral finite element mesh for said structural member,

means for performing a structural nonlinear finite element analysis on said at least one structural member, and

means for thereby providing at least information data on geometrical properties and internal mechanical loads of at least a sub-portion of said vessel body for said analysis of said vessel body, wherein said generating said quadrilateral and/or hexahedral finite element mesh comprises using related luminal and abluminal meshes of said wall of said vessel body, wherein each node on the luminal inside boundary of said wall has a replica at its abluminal outside boundary for volumetric mesh generation of said wall, and determining the distance between each said replica node as said thickness of said wall at said replica node,

and wherein the mesh is used as a geometric input for the finite element mesh generation.

19. The system of claim 18, wherein the system is configured to analyze at least one of the vascular body for geometric characteristics and mechanical loading conditions of the vascular body.

20. Use of the method according to claim 1, wherein the method is performed on manipulation of the system according to claim 18 by clinical personnel without engineering expert knowledge.