Robust 3D Segmentation of Anatomical Structures with Level Sets

This paper is focused on the use of the level set formalism to segment anatomical structures in 3D images (ultrasound ou magnetic resonance images). A closed 3D surface propagates from an initial position towards the desired boundaries through the iterative evolution of a 4D implicit function. The major contribution of this work is the design of a robust evolution model which does not require a fine tuning of a set of parameters. It involves adaptive parameters depending on the data and on the current state of the process. The iteration step and the external propagation force, both usually constant, are automatically computed at each iteration. In addition, region-based information rather than gradient is used, via an estimation of intensity probability density functions over the image. As a result, the method can be applied to very different kinds of data. Results on brain MR images and 3D echographies of carotid arteries are presented and discussed. key words: 3D segmentation, defor...

• Type: Conference
• Year: 2000
• Where: MICCAI
• Authors: C. Baillard, Christian Barillot
• Keywords: Closed 3D Surface, Level Sets, Medical Imaging, MICCAI 2000, Probability Density Functions