A 2D Fourier Approach to Deformable Model Segmentation of 3D Medical Images

Anatomical shapes present a unique problem in terms of accurate representation and medical image segmentation. Three-dimensional (3D) statistical shape models have been extensively researched as a means of autonomously segmenting and representing models. We present a method based on a principal component analysis of a stack of 2D contours represented as Fourier descriptors (FDs). A training set for the shape model is generated directly from the FDs of the perimeters of the segmented regions on each image after a transformation into a canonical coordinate frame. We apply our shape model to the segmentation of CT and MRI images of the distal femur via an iterative method based on active contours. Results of the application of our method demonstrate its ability to accurately capture shape variations and guide segmentation.

• Type: Conference
• Year: 2004
• Where: ECCV
• Authors: Eric Berg, Mohamed Mahfouz, Christian Debrunner, William Hoff
• Keywords: Computer Vision, ECCV 2004, Medical Image Segmentation, Shape Model, Statistical Shape Models