Zu den Inhalten springen

Hauptnavigation:

Contact | Deutsch | Sitemap
| Home > Departments > Center for Experimental Medicine > Department of Medical Informatics > Research > representation and visualization of variability in a 3d anatomical atlas

Representation and Visualization of 3D Shape Variability of Organs in an Interactive Anatomical Atlas

Problem:


Computer-based 3D models of the human body have become an important tool for learning anatomy, treatment planning and clinical diagnostics. In contrast to the classical knowledge representation in books and printed atlases, computerized anatomic atlases allow interactive exploration of the human anatomy and make spatial relationships easier to comprehend.
However, in most cases such 3D atlases are derived from one single individual, and are therefore not representative of the human anatomy in general. Yet, anatomical structures exhibit major differences with respect to size and shape, which is partly due to natural variability, but also subject to factors like age, gender, diseases and habits.
Since the geometric variability across humans plays an important role in many medical applications, we intend to develop a framework of an anatomical atlas for representation and visualization of variability.

Method:


As basis for developing the variability atlas we use the VOXEL-MAN system. For modelling anatomic shapes and their variability we utilize “m-reps” (“medial representations”) which allow a compact representation of anatomical objects on the basis of their skeletons (fig. 1). A detailed description of the concept and application of m-reps can be found in the reference stated below. An m-rep model possesses a figural coordinate system, on the basis of which a geometric correspondence between shape variants can be established.

We developed a methodology for integrating shape models into an anatomical atlas. As an example we used a statistical model of the kidney that is based on 48 different variants. For linking the kidney models with the volume based model we have also built an m-rep model for the Visible Human kidney. It serves as a reference model and possesses the same skeletal structure as the rest of the kidneys, so that a geometric correspondence between the m-reps can be assured.



Fig. 1: An m-rep model of a kidney. Left: skeletal structure. Right: the corresponding boundary surface.

With the integration of the m-rep shape description into the VOXEL-MAN atlas it becomes possible to query and visualize different variations of an organ, in our example the kidney. In addition to the representation of individual shape variants, the average shape of a population can be displayed (see fig. 2).

Besides a surface based representation, a volume based representation of the kidney’s shape variants is also possible. It results from the deformation of the reference kidney of the volume based model according to the m-rep shape description. By this means a realistic visual representation of the shape variants and the visualization of the organ’s interior’s texture, made visible by cross sections, becomes possible. It is obvious that this is not reasonable for pathological variants that show substantial morphological differences to a healthy kidney.

Furthermore we are planning to examine and implement various techniques for visualizing the variability within a population and temporal changes of organ shapes.





Fig. 2: Visualization of shape variants of the kidney within the VOXEL-MAN atlas. The left kidneys (in the pictures on the right hand side) are surface representations on the basis of m-rep models.
Top: left kidney of the Visible Human. Bottom: average shape of a population of 48 kidneys.

Selected Publications:

  1. Silke Hacker, Heinz Handels: Representation and Visualization of Variability in a 3D anatomical Atlas using the Kidney as an Example. In: Cleary, K.R., Galloway, R.L. (eds.), Visualization, Image-Guided Procedures, and Display, SPIE Medical Imaging 2006, San Diego, SPIE Vol. 6141, 61410B-1-61410B-7, 2006

  2. Silke Hacker, Heinz  Handels, (2006): Repräsentation und Visualisierung von 3D-Formvarianten von Organen für die medizinische Ausbildung:
    In: Handels, H., Erhardt, J., Horsch, A., Meinzer, H.P., Tolxdorff, T. (Hrsg.), Bildverarbeitung für die Medizin 2006, Informatik aktuell, Springer Verlag, Berlin, 276-280.

  3. Silke Hacker, Ulf Tiede, Heinz Handels: Repräsentation von Variabilität in einem dreidimensionalen anatomischen Atlas am Beispiel der Niere.
    50. Jahrestagung der Deutschen Gesellschaft für Medizinische Informatik, Biometrie und Epidemiologie, Freiburg, Tagungsband, 16-18, 2005.

  4. Pizer S M, Fletcher P T, Joshi S C, Stough J, Thall A, Chen J Z, Fridman Y, Fritsch D S, Gash G, Glotzer J M, Jiroutek M R, Lu C, Muller K E, Tracton G, Yushkevich P A, Chaney E L. Deformable M-reps for 3D Medical Image Segmentation. IJVC, vol. 55, 2003.


Project Team:


Silke Hacker
Heinz Handels

Cooperation Partners:


Stephen M. Pizer,
Medical Image Display and Analysis Group (MIDAG), University of North Carolina in Chapel Hill, NC, USA

The kidney data was kindly provided by the radiooncology of the University of North Carolina in Chapel Hill, NC, USA.

Top of page    Print page

© Universitätsklinikum Hamburg-Eppendorf, Impressum
last update: Martin Dalladas, 09.06.2006

Version