WO2008152555A2

WO2008152555A2 - Anatomy-driven image data segmentation

Info

Publication number: WO2008152555A2
Application number: PCT/IB2008/052249
Authority: WO
Inventors: Nicolas F. Villain; Roberto Ardon; Frans A. Gerritsen
Original assignee: Koninklijke Philips Electronics N.V.
Priority date: 2007-06-12
Filing date: 2008-06-09
Publication date: 2008-12-18
Also published as: WO2008152555A3

Abstract

The invention relates to a system (100) for delineating a first object in image data by using a model, the first object describing a first anatomical structure, the system comprising an adaptation unit (120) for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on a second object previously identified in the image data, the second object describing a second anatomical structure. Thus, the system (100) of the invention is arranged for adapting the model to the image data, based on the energy functional of the model, where energy terms describing forces acting on the model or on the model parts, model details, and adaptation algorithms are allowed to be dependent on the position of the second object. Hence, the system (100) is capable of using further information available in the image data, which information may be derived from the previously identified second object comprised in the image data.

Description

Anatomy-driven image data segmentation

FIELD OF THE INVENTION

The invention relates to the field of segmentation of image data and more specifically to model-based segmentation of image data.

BACKGROUND OF THE INVENTION

A system for segmenting image data is described in patent application WO 2004/075112 entitled "Image segmentation by assigning classes to adaptive mesh segmentation", hereinafter referred to as Ref. 1. This document describes a system for adapting a polygonal mesh to the image data, wherein each mesh primitive, e.g., a triangular face of a triangular mesh, is assigned to a class of a plurality of classes. The probability that a mesh primitive belongs to a class of mesh primitives is estimated using a feature vector derived from data sampled from the image data. An energy function corresponding to the class of the mesh primitive is used for adapting the mesh primitive to the image data.

SUMMARY OF THE INVENTION

It would be advantageous to have a system capable of using further information available in the image data. For example, useful information may be derived from a previously identified, e.g., delineated, object comprised in the image data and describing an anatomical structure. To better address this issue, in an aspect of the invention, a system for delineating a first object in image data by using a model is provided, the first object describing a first anatomical structure, the system comprising an adaptation unit for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on a second object previously identified in the image data, the second object describing a second anatomical structure.

Different parts of the image data are defined by their positions relative to the previously identified second object. These different parts of the image data may be described by different image data characteristics. Furthermore, properties of the first anatomical structure may be different in different parts of the first anatomical structure, described by a model adapted to said different parts of the image data. Thus, said different parts of the image data may require using different approaches to delineating respective parts of the first object, e.g., using different energy functional terms or force field components, employing different algorithms for minimizing the energy functional, and/or using different model properties, such as local resolution of a mesh comprised in the model, based on the part of the image data to which the model is adapted by the adaptation unit. Hence, the system is capable of using further information available in the image data, which information is derived from the position of the model being adapted to the image data relative to the position of the previously identified second object. In an embodiment of the system, the system further comprises an identification unit for identifying the second object in the image data. If the second object is comprised in the image data, identifying the second object may be carried out by the system of the invention. Identifying may comprise, for example, finding the geometrical center of the second object, finding a major axis of the second object, or delineating the second object. In an embodiment of the system, the adaptation unit is arranged for finding a minimum value of energy of the model and the aspect of the adaptation unit is based on one of the following: an energy functional and/or a force field of the model; and an algorithm for minimizing the energy functional and/or a force field of the model.

In a first part of the image data, a first type of image features may be used to attract the model. In a second part of the image data, using a second type of image features to attract the model may be more advantageous. Different feature types may require different feature detection algorithms and/or different energy functional terms. Further, in a third part of the image data, the surface of the first object may be known to be flat. In a fourth part of the image data, the surface of the first object may be known to be curved. Modeling flat surfaces with small curvature may require stronger regularization forces and/or fewer nodes of a mesh comprised in the model than modeling curved surfaces with large curvature. Furthermore, when the algorithm used for adapting the model to the image is an iterative algorithm, the energy functional of the model may be different in different iterations. For example, modeling flat surfaces with small curvature may require more steps with strong regularization forces and/or weak attraction forces than modeling curved surfaces with large curvature. In an embodiment of the system, the energy functional of the model comprises a regularization term and an image-feature term and at least one of these terms depends on the second object. The regularization term is responsible for preserving the shape of the model close to what is assumed to be a typical shape of the first anatomical structure while the image-feature terms are responsible for attracting the model parts to features in the image data. The system allows using different regularization terms at different parts of the model, based on the position of these parts relative to the previously identified second object. The system further allows using different image- feature terms for different features comprised in the image data, based on the position of these features relative to the previously identified second object.

In an embodiment of the system, the energy functional of the model comprises a second object interaction term describing interaction between the model and the second object. The second object interaction term allows the previously identified second object to directly influence the shape and/or position of the model. In an embodiment of the system, the first object describes a liver. The system is particularly useful for adapting the model to the first object describing the first anatomical structure, the first anatomical structure being the liver, because characteristics of the liver vary from part to part. For instance, the liver has a smooth convex part, with a substantially constant curvature, facing the diaphragm. To preserve the shape of this part, strong regularization interactions may be used in the model parts facing the diaphragm.

In an embodiment of the system, the second object describes a rib cage or a diaphragm. The rib cage and the diaphragm are easy to identify, e.g., using image segmentation, in the image data and may provide useful information for delineating the first object describing, e.g., the liver. In a further aspect of the invention, the system according to the invention is comprised in an image acquisition apparatus.

In a further aspect of the invention, the system according to the invention is comprised in a workstation.

In a further aspect of the invention, a method of delineating a first object in image data using a model is provided, the first object describing a first anatomical structure, the method comprising: an identification step for identifying a second object in the image data, the second object describing a second anatomical structure; and an adaptation step for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on the identified second object.

In a further aspect of the invention, a computer program product to be loaded by a computer arrangement is provided, the computer program product comprising instructions for delineating a first object in image data using a model, the first object describing a first anatomical structure, the computer arrangement comprising a processing unit and a memory, the computer program product, after being loaded, providing said processing unit with the capability to adapt the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on a second object previously identified in the image data, the second object describing a second anatomical structure.

It will be appreciated by those skilled in the art that two or more of the above- mentioned embodiments, implementations, and/or aspects of the invention may be combined in a useful way.

Modifications and variations of the image acquisition apparatus, of the workstation, of the method, and/or of the computer program product, which correspond to the described modifications and variations of the system, can be carried out by a skilled person on the basis of the present description. The skilled person will appreciate that the method may be applied to multidimensional image data acquired by various acquisition modalities such as, but not limited to, standard X-ray Imaging, Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Ultrasound (US), Positron Emission Tomography (PET), Single Photon Emission Computed Tomography (SPECT), and Nuclear Medicine (NM).

BRIEF DESCRIPTION OF THE DRAWINGS

These and other aspects of the invention will become apparent from and will be elucidated with respect to the implementations and embodiments described hereinafter and with reference to the accompanying drawings, wherein: Fig. 1 shows a block diagram of an exemplary embodiment of the system;

Fig. 2 shows views of surfaces of two objects delineated in CT image data;

Fig. 3 shows views of two surfaces delineating two vessel objects in CT angiography image data;

Fig. 4 shows a flowchart of an exemplary implementation of the method; Fig. 5 schematically shows an exemplary embodiment of the image acquisition apparatus; and

Fig. 6 schematically shows an exemplary embodiment of the workstation. The same reference numerals are used to denote similar parts throughout the Figures.

DETAILED DESCRIPTION OF EMBODIMENTS

Fig. 1 shows a block diagram of an exemplary embodiment of the system 100 for delineating a first object in image data using a model, the first object describing a first anatomical structure, the system comprising an adaptation unit 120 for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on a second object previously identified in the image data, the second object describing a second anatomical structure.

The exemplary embodiment of the system 100 further comprises the following optional units: an identification unit 110 for identifying the second object identified in the image data; a control unit 160 for controlling the workflow in the system 100; a user interface 165 for communicating with a user of the system 100; and - a memory unit 170 for storing data.

In an embodiment of the system 100, there are three input connectors 181, 182 and 183 for the incoming data. The first input connector 181 is arranged to receive data coming in from a data storage means such as, but not limited to, a hard disk, a magnetic tape, a flash memory, or an optical disk. The second input connector 182 is arranged to receive data coming in from a user input device such as, but not limited to, a mouse or a touch screen. The third input connector 183 is arranged to receive data coming in from a user input device such as a keyboard. The input connectors 181, 182 and 183 are connected to an input control unit 180.

In an embodiment of the system 100, there are two output connectors 191 and 192 for the outgoing data. The first output connector 191 is arranged to output the data to a data storage means such as a hard disk, a magnetic tape, a flash memory, or an optical disk. The second output connector 192 is arranged to output the data to a display device. The output connectors 191 and 192 receive the respective data via an output control unit 190. The skilled person will understand that there are many ways to connect input devices to the input connectors 181, 182 and 183 and the output devices to the output connectors 191 and 192 of the system 100. These ways comprise, but are not limited to, a wired and a wireless connection, a digital network such as, but not limited to, a Local Area Network (LAN) and a Wide Area Network (WAN), the Internet, a digital telephone network, and an analog telephone network.

In an embodiment of the system 100, the system 100 comprises a memory unit 170. The system 100 is arranged to receive input data from external devices via any of the input connectors 181, 182, and 183 and to store the received input data in the memory unit 170. Loading the input data into the memory unit 170 allows quick access to relevant data portions by the units of the system 100. The input data may comprise, for example, the image data. Optionally, the input data may comprise data describing the previously identified second object. The memory unit 170 may be implemented by devices such as, but not limited to, a Random Access Memory (RAM), a Read Only Memory (ROM), and/or a hard disk drive and a hard disk. The memory unit 170 may be further arranged to store the output data. The output data may comprise, for example, the delineated first object described by a polygonal mesh of the model adapted to the image data by the adaptation unit 120 of the system 100. The memory unit 170 is also arranged to receive data from and deliver data to the units of the system 100 comprising the identification unit 110, the adaptation unit 120, the control unit 160, and the user interface 165, via a memory bus 175. The memory unit 170 is further arranged to make the output data available to external devices via any of the output connectors 191 and 192. Storing the data from the units of the system 100 in the memory unit 170 may advantageously improve the performance of the units of the system 100 as well as the rate of transfer of the output data from the units of the system 100 to external devices. Alternatively, the system 100 may comprise no memory unit 170 and no memory bus 175. The input data used by the system 100 may be supplied by at least one external device, such as an external memory or a processor, connected to the units of the system 100. Similarly, the output data produced by the system 100 may be supplied to at least one external device, such as an external memory or a processor, connected to the units of the system 100. The units of the system 100 may be arranged to receive the data from each other via internal connections or via a data bus.

In an embodiment of the system 100, the system 100 comprises a control unit 160 for controlling the workflow in the system 100. The control unit may be arranged to receive control data from and provide control data to the units of the system 100. For example, after the second object is identified by the identification unit 110, the identification unit 110 may be arranged to pass a message "the second object is identified" to the control unit 160 and the control unit 160 may be arranged to provide a message "adapt the model to the image message" to the adaptation unit 120, requesting the adaptation unit 120 to adapt the model to the image data based on the identified second object. Alternatively, a control function may be implemented in another unit of the system 100.

In an embodiment of the system 100, the system 100 comprises a user interface 165 for communicating with the user of the system 100. The user interface 165 may be arranged to provide the user with means for initializing parameters of the model for adapting to the image data. Optionally, the user interface may receive a user input for selecting a mode of operation of the system 100, such as a mode for adapting the model to the image data, using mode-specific terms of the energy functional of the model. The skilled person will understand that more functions may be advantageously implemented in the user interface 165 of the system 100. One of the main tasks of image segmentation is the delineation of objects describing anatomical structures comprised in the image data. Model-based segmentation uses anatomical structure models, such as, but not limited to, contour models, surface models, and volumetric models, for delineating objects. A discrete representation of a contour, surface, or volumetric model may be a point distribution. During the adaptation of the model, model parameters, e.g., coordinates of points of the point distribution describing the contour or surface in the image data space, are adapted to the image data, thereby defining an object. The adaptation of the model is governed by an algorithm for finding a minimum of energy functional of the model. The energy functional is a function for computing the energy of the model for a given set of model parameters. The energy functional comprises image-feature terms typically describing attraction of model parts towards image features, e.g., edges, as well as regularity terms enforcing some desired geometric properties of the model, e.g., a constant curvature in an area of a model surface. The skilled person will understand that the energy functional may be also referred to as energy or as a cost function and that finding a minimum of the energy functional may be replaced by finding a maximum of another functional, or a root of an equation, for example. The skilled person will also understand that finding values of model parameters which minimize the energy may be alternatively carried out by finding values of model parameters which correspond to a substantially zero force field acting on the model, e.g., on the nodes of a mesh comprised in the model. Thus, the scope of the claims must not be construed limited by the use of the terms "energy functional", "minimum" and "minimization".

In an embodiment of the system 100, the system is used to delineate the first object describing the liver in CT image data. Because the shape of the liver is very complex and varies a lot from one patient to another, it is difficult to capture all liver shape characteristics in a deformable model of the liver. On the other hand, it is useful to observe some rules based on the human anatomy, which rules may help delineating the liver more robustly using the deformable model of the liver. For instance, the part of the liver that faces the diaphragm and the rib cage is typically convex, smooth, and has small constant curvature. To use this information, the identification unit 110 may be arranged to identify the rib cage and the diaphragm. Because the rib cage and diaphragm are relatively easy to identify in the image data, simple segmentation methods based on ray casting, for example, may be used. The position of the rib cage and the diaphragm as well as an estimate of the average Hounsfield unit (HU) value of the liver tissue is used to initialize the deformable model of the liver - a small sphere of a diameter of about 5 cm. The sphere surface may be described by a polygonal mesh, e.g., a simplex mesh.

In an embodiment of the system 100, the model comprises a simplex mesh for delineating the first object describing the liver in the CT image data. The simplex mesh may be described by coordinates of mesh vertices and a list of edges connecting pairs of vertices. The adaptation of the simplex mesh is based on deforming the simplex mesh in order to minimize the value of the energy functional of the model. The energy functional of the model is a function of the coordinates of the mesh vertices. A typical energy functional comprises regularization terms and image-feature terms. The regularization terms describe the potential energy of the mesh. The regularization terms typically depend on the positions of vertices of the simplex mesh, e.g., on the edge lengths and on the angles between pairs of edges sharing a vertex. A minimum of the potential energy is typically attained when the edge lengths and angles assume some typical values learned, for example, from a learning set of image data. The image-feature terms typically describe the energy of interaction of the mesh with image features, e.g., with edges present in the image. The model further comprises a method for detecting image features. The energy of interaction between a vertex and an image feature may depend on the strength of the image feature, e.g., on the value of the gradient of the image intensity at the location of the image feature such as an edge. Typically, this interaction is attraction of the vertex by the feature. The skilled person will know different regularization and image-feature terms. The article by H. Delingette entitled "General Object Reconstruction based on Simplex Meshes" in International Journal of Computer Vision, vol. 32, pages 11-142, 1999, hereinafter referred to as Ref. 2, describes the energy functional terms suitable for use in the case of simplex meshes.

The system 100 of the invention allows taking into account the location of vertices of the model simplex mesh relative to the rib cage and diaphragm. In an embodiment of the invention, three regions of the liver surface are identified: (1) the tips, (2) the convex region facing the diaphragm and rib cage, and (3) the rest of the liver surface. In region (3), default values of parameters of the model are used. In region (1) the mesh faces are smaller and therefore the mesh resolution is higher. Furthermore, the regularization forces between the mesh vertices are weaker than the default model regularization forces. This allows the mesh to be adapted to the image data and to delineate parts of the first object describing the sharp tips of the liver. In region (2) the regularization forces between the mesh vertices are stronger than the default model regularization forces. This makes the mesh more rigid and allows the mesh to preserve the surface characteristics learned from a learning set of image data. Thus, the mesh resolution and the forces between the mesh vertices are determined by the system 100, based on the position of the mesh vertices relative to the diaphragm and the rib cage. In order to minimize the energy functional, an iterative approach, described for example in Ref. 2, based on the Euler-Lagrange equation of the energy functional may be used. In order to obtain a smoother mesh surface in region (2), the energy functional of the model, and thus the Euler-Lagrange equation derived from said energy functional, is different in different iterations. An iteration based on the energy functional comprising both regularization and image-feature terms is followed by two iterations based on the energy functional comprising only regularization terms.

Fig. 2 shows views of surfaces of two objects delineated in CT image data. Both objects describe the liver and both objects are delineated by adapting a simplex mesh to the image data. The left object 201 is delineated using the default values of parameters of the model in region (2). The surface area 210 corresponding to region (2) comprises a cavity 211. This cavity corresponds to the location of a tumor. Because tumor tissue corresponds to different features in the image data, this tissue is not comprised in the delineated object. The right object 202 is delineated using the stronger regularization forces for adapting the model mesh to the surface area 210 corresponding to region (2). Hence, the surface area 210 in the right object 202 does not comprise a cavity because the non-cancerous features of the liver surrounding the cancerous features attract mesh vertices and these vertices pull the adjacent vertices due to strong regularization forces. Comparing the left object 201 and the right object 202 allows computing the volume of the liver consumed by cancer. The skilled person will understand that another useful model and the corresponding energy functional of the model may be utilized by the system 100. For example, the article by J. Weese et al. entitled "Shape constrained deformable models for 3D medical image segmentation" in Proc. IPMI, 380-387, Springer Verlag, 2001, pages 380-387, describes energy functional terms suitable for triangular meshes. Another class of models is based on level sets. Level set based segmentation methods are described by J. S. Suri et al. in the article entitled "Shape Recovery Algorithms Using Level Sets in 2-D/3-D Medical Imagery: A State-of-the-Art Review", IEEE Transactions On Information Technology In Biomedicine, Vol. 6, No. 1, March 2002, pages 8-28.

In an embodiment of the system 100, the system 100 is applied to delineate a blood vessel, the internal carotid, from the aortic arch up to the Circle-Of- Willis in CT angiography images. Delineating the vessel is difficult because the attenuation values along the vessel show huge variations ranging from 100 HU to 500 HU.

In an embodiment, the centreline of the vessel is used to describe anatomical information for delineating the vessel. The vessel is described by a tubular surface defined by the centreline and the vessel radius. This surface may be further described by a polygonal mesh or a level set. The surface is partitioned into three parts in order to take into account variations in the attenuations. To this end, the identification unit is arranged to identify the main bones comprised in the image data: the skull, the vertebrae, the clavicles, etc. The first part of the vessel model surface is the neck part, stretching from the shoulder to the lower part of the skull. The second part of the vessel model surface is the skull base part, stretching from the lower part of the skull base to the upper part of the clivus bone. The third part of the vessel model surface is the brain part above the upper part of the clivus. While computing the minimum of the energy functional, weak image-feature forces are used in the neck part and in the clivus part while strong image-feature forces are used in the skull base part. Fig. 3 shows views of two surfaces delineating two vessel objects in CT angiography image data. The first surface 301 correctly depicts the skull-base part 321 of the vessel while the neck part 311 and the brain part 331 of the vessel are incorrectly delineated. This is because low attenuations in the neck region and in the brain region result in faint features, which do not attract the respective model parts strong enough to overcome regularization interactions. The second surface 302 correctly depicts all three parts, the neck part 312, the skull-base part 322 and the brain part 332 of the vessel because of the increased attraction forces in the neck region and brain region of the image data.

It will be understood by those skilled in the art that other embodiments of the system 100 are also possible. It is possible, among other things, to redefine the units of the system and to redistribute their functions. For example, in an embodiment of the system 100, the functions of the adaptation unit 120 may be implemented in a number of units, e.g., an energy unit for computing the energy of the model, a model unit for determining local parameters of the model, based on the position of the model relative to the second object, and an optimization unit for computing the minimum of the energy functional. In a further embodiment of the system 100, there can be a plurality of adaptation units replacing the adaptation unit 120. Each adaptation unit of the plurality of adaptation units may be arranged to employ a different adaptation method. The method employed by the system 100 may be based on a user selection. The units of the system 100 may be implemented using a processor. Normally, their functions are performed under control of a software program product. During execution, the software program product is normally loaded into a memory, like a RAM, and executed from there. The program may be loaded from a background memory, like a ROM, hard disk, or magnetic and/or optical storage, or may be loaded via a network like the Internet. Optionally an application specific integrated circuit may provide the described functionality. Fig. 4 shows a flowchart of an exemplary implementation of the method 400 of delineating a first object in image data using a model, the first object describing a first anatomical structure. The method 400 begins with an identification step 410 for identifying a second object in the image data, the second object describing a second anatomical structure. After the identification step 410, the method 400 continues with an adaptation step 420 for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on the identified second object. After the adaptation step 420, the method 400 terminates.

Fig. 5 schematically shows an exemplary embodiment of the image acquisition apparatus 500 employing the system 100, said image acquisition apparatus 500 comprising an image acquisition unit 510 connected via an internal connection with the system 100, an input connector 501, and an output connector 502. This arrangement advantageously increases the capabilities of the image acquisition apparatus 500 by providing said image acquisition apparatus 500 with advantageous capabilities of the system 100. Examples of image acquisition apparatus comprise, but are not limited to, a CT system, an X-ray system, an MRI system, an US system, a PET system, a SPECT system, and a NM system.

Fig. 6 schematically shows an exemplary embodiment of the workstation 600. The workstation comprises a system bus 601. A processor 610, a memory 620, a disk input/output (I/O) adapter 630, and a user interface (UI) 640 are operatively connected to the system bus 601. A disk storage device 631 is operatively coupled to the disk I/O adapter 630. A keyboard 641, a mouse 642, and a display 643 are operatively coupled to the UI 640. The system 100 of the invention, implemented as a computer program, is stored in the disk storage device 631. The workstation 600 is arranged to load the program and input data into memory 620 and execute the program on the processor 610. The user can input information to the workstation 600, using the keyboard 641 and/or the mouse 642. The workstation is arranged to output information to the display device 643 and/or to the disk 631. The skilled person will understand that there are numerous other embodiments of the workstation 600 known in the art and that the present embodiment serves the purpose of illustrating the invention and must not be interpreted as limiting the invention to this particular embodiment. It should be noted that the above-mentioned embodiments illustrate rather than limit the invention and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim or in the description. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The invention can be implemented by means of hardware comprising several distinct elements and by means of a programmed computer. In the system claims enumerating several units, several of these units can be embodied by one and the same item of hardware or software. The usage of the words first, second and third, etc., does not indicate any ordering. These words are to be interpreted as names.

Claims

CLAIMS:

1. A system (100) for delineating a first object in image data by using a model, the first object describing a first anatomical structure, the system comprising an adaptation unit (120) for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on a second object previously identified in the image data, the second object describing a second anatomical structure.

2. A system (100) as claimed in claim 2, further comprising an identification unit (110) for identifying the second object in the image data.

3. A system (100) as claimed in claim 1, wherein the adaptation unit (120) is arranged for finding a minimum value of energy of the model and the aspect of the adaptation unit is based on one of the following: an energy functional and/or a force field of the model; and an algorithm for minimizing the energy functional and/or a force field of the model.

4. A system (100) as claimed in claim 1, wherein the energy functional of the model comprises a second object interaction term describing interaction between the model and the second object.

5. A system (100) as claimed in claim 1, wherein the first object describes a liver.

6. A system (100) as claimed in claim 5, wherein the second object describes a rib cage or a diaphragm.

7. An image acquisition apparatus (500) comprising a system (100) as claimed in claim 1.

A workstation (600) comprising a system (100) as claimed in claim 1.

9. A method (400) of delineating a first object in image data by using a model, the first object describing a first anatomical structure, the method comprising: - an identification step (410) for identifying a second object in the image data, the second object describing a second anatomical structure; and an adaptation step (420) for adapting the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on the identified second object.

10. A computer program product to be loaded by a computer arrangement, comprising instructions for delineating a first object in image data by using a model, the first object describing a first anatomical structure, the computer arrangement comprising a processing unit and a memory, the computer program product, after being loaded, providing said processing unit with the capability to adapt the model to the image data, thereby delineating the first object, wherein an aspect of the adaptation depends at least on a second object previously identified in the image data, the second object describing a second anatomical structure.