Summary: A method for the three-dimensional statistical shape analysis of
the bony orbit
Hans Lamecker, Lukas Kamer, Antonia Wittmers, Stefan Zachow, Thomas
Kaup, Alexander Schramm, Hansrudi Noser, Beat Hammer
Introduction. 2D CT or CBCT (Cone Beam CT) reconstructions are routinely
used for the assessment and treatment of orbital fractures  and deformities.
Routine 3D rendering of such image data is unreliable due to the conjunction
of the so called partial volume effects during imaging and the very thin
nature of the orbital walls.
However, 3D CT data have the potential to improve scientific shape analysis
and radiological diagnostics as well as surgical planning and treatment. To be
able to generate a precise 3D model of the orbit, manual or semi-automatic
segmentation tools have to be employed. Automatic 3D reconstruction of the
orbit would make 3D models more applicable for tasks such as generating
virtual templates or for optimizing the shape of orbital implants.
We adapted a method for the 3D statistical assessment of normal orbital
shape and its variability to accomplish automatic 3D reconstruction.
Methods. The main challenge for performing statistical analysis of three-
dimensional shapes lies in the identification of corresponding points on the
surfaces of different individual shapes. We adopt the method of consistent