Deformable M-Reps for 3D Medical Image Segmentation (2000)
| Citations: | 54 - 22 self |
BibTeX
@MISC{Pizer00deformablem-reps,
author = {Stephen M. Pizer and P. Thomas Fletcher and Yonatan Fridman and Daniel S. Fritsch and A. Graham Gash and John M. Glotzer and Sarang Joshi and Andrew Thall and Gregg Tracton and Paul Yushkevich and Edward L. Chaney},
title = {Deformable M-Reps for 3D Medical Image Segmentation},
year = {2000}
}
Years of Citing Articles
Bookmark
 |
 |
 |
 |
 |
 |
OpenURL
Abstract
M-reps (formerly called DSLs) are a multiscale medial means for modeling and rendering 3D solid geometry. They are particularly well suited to the modeling of anatomic objects, producing models that can be used to capture prior geometric information effectively in deformable models segmentation approaches. The representation is based on figural models, defined at coarse scale by a hierarchy of figures -- protrusions, indentations, neighboring figures, and included figures -- which represent solid regions and their boundaries simultaneously. The linked collection of figural components imply a fuzzy, i.e., probabilistically described boundary position with a width-proportional tolerance. At small scale these figural boundaries are made precise by displacing a dense sampling of the m-rep implied boundary. While these models can exist in 2D, we focus on models of 3D objects. A model for a single figure is made from a net, i.e., a mesh or chain, of medial atoms (hence the name m-reps), each...
