På svenska

Presentation Information     2005-09-26 (15:15)   •  The seminar room at Vi2

Speaker Filip Malmberg
Type Master thesis presentation
Title 3D Live-wire: Semi-automatic segmentation of volume images in a haptic environment
Abstract Volume images have become increasingly common in biomedical research and clinical medicine. Most applications require the images to be segmented into objects of interest at some stage in the process. Segmentation is a difficult task, and sometimes it is not possible to achieve a good segmentation using automatic methods alone. In such cases it might be neccesary to utilize interactive, semi-automatic methods.

Efficient interaction with a 3D image is much harder to achieve than interaction with 2D images. We have used a system that simplifies 3D interaction in two ways: (1) by a stereoscopic display and (2) by adding haptic feedback through a sensing probe.

One interactive method that is popular for segmentation of 2D images is the Live-wire method. Various ways of extending this method to segment 3D images have been proposed in the literature. Most of these methods are based on using the 2D Live-wire method on a number of slices in the volume, and then reconstructing the entire object using this information. One problem with these methods is that they place restrictions on how the user may choose the slices in order for the reconstruction algorithm to work, forcing the user into a non-intuitive workflow. For example, the user might have to take into account the topology of the segmented object.

In this Master's thesis we describe a method where the user works directly with the 3D-volume instead of the 2D slices, and thus we are able to overcome some of the limitations of earlier approaches. We also discuss some problems with our method, and possible ways of solving them.