@inproceedings{d59e9229c25944e080c0f58c0946d839,
title = "Multi-object spring level sets (MUSCLE)",
abstract = "A new data structure is presented for geometrically modeling multi-objects. The model can exhibit elastic and fluid-like behavior to enable interpretability between tasks that require both deformable registration and active contour segmentation. The data structure consists of a label mask, distance field, and springls (a constellation of disconnected triangles). The representation has sub-voxel precision, is parametric, re-meshes, tracks point correspondences, and guarantees no self-intersections, air-gaps, or overlaps between adjacent structures. In this work, we show how to apply existing registration algorithms and active contour segmentation to the data structure; and as a demonstration, the data structure is used to segment cortical and subcortical structures (74 total) in the human brain.",
keywords = "Active contour, Level set, Mesh, Registration, Segmentation, Tracking",
author = "Lucas, {Blake C.} and Michael Kazhdan and Taylor, {Russell H.}",
note = "Publisher Copyright: {\textcopyright} Springer-Verlag Berlin Heidelberg 2012.; 15th International Conference on Medical Image Computing and Computer Assisted Intervention, MICCAI 2012 ; Conference date: 01-10-2012 Through 05-10-2012",
year = "2012",
doi = "10.1007/978-3-642-33415-3_61",
language = "English (US)",
isbn = "9783642334146",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "495--503",
editor = "Nicholas Ayache and Herve Delingette and Polina Golland and Kensaku Mori",
booktitle = "Medical Image Computing and Computer-Assisted Intervention, MICCAI2012 - 15th International Conference, Proceedings",
}