Summary: Automatic Segmentation of the Liver for
Preoperative Planning of Resections
Hans Lamecker, Thomas Lange, Martin Seebaß,
Sebastian Eulenstein, Malte Westerhoff, HansChristian Hege
Takustr. 7, 14195 Berlin, Germany
Abstract. This work presents first quantitative results of a method for automatic liver
segmentation from CT data. It is based on a 3D deformable model approach using
a-priori statistical information about the shape of the liver gained from a training set.
The model is adapted to the data in an iterative process by analysis of the grey value
profiles along its surface normals after nonlinear diffusion filtering. Leaveoneout
experiments over 26 CT data sets reveal an accuracy of 2.4 mm with respect to the
Individual preoperative surgical planning for resections of tumors in the liver requires seg-
mentation of the liver tissue . Reliable image segmentation is essential for the correct
prediction of the blood circulation regions.
Semi-automatic methods may reduce the user interaction time for segmentation. However in
the clinical routine automatic methods are desirable. 3D statistical shape models are promis-
ing for robust and automatic segmentation of medical images.