TY - GEN
T1 - Topology noise removal for curve and surface evolution
AU - Chen, Chao
AU - Freedman, Daniel
PY - 2011
Y1 - 2011
N2 - In cortex surface segmentation, the extracted surface is required to have a particular topology, namely, a two-sphere. We present a new method for removing topology noise of a curve or surface within the level set framework, and thus produce a cortical surface with correct topology. We define a new energy term which quantifies topology noise. We then show how to minimize this term by computing its functional derivative with respect to the level set function. This method differs from existing methods in that it is inherently continuous and not digital; and in the way that our energy directly relates to the topology of the underlying curve or surface, versus existing knot-based measures which are related in a more indirect fashion. The proposed flow is validated empirically.
AB - In cortex surface segmentation, the extracted surface is required to have a particular topology, namely, a two-sphere. We present a new method for removing topology noise of a curve or surface within the level set framework, and thus produce a cortical surface with correct topology. We define a new energy term which quantifies topology noise. We then show how to minimize this term by computing its functional derivative with respect to the level set function. This method differs from existing methods in that it is inherently continuous and not digital; and in the way that our energy directly relates to the topology of the underlying curve or surface, versus existing knot-based measures which are related in a more indirect fashion. The proposed flow is validated empirically.
UR - https://www.scopus.com/pages/publications/79951587637
U2 - 10.1007/978-3-642-18421-5_4
DO - 10.1007/978-3-642-18421-5_4
M3 - Conference contribution
AN - SCOPUS:79951587637
SN - 9783642184208
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 31
EP - 42
BT - Medical Computer Vision
T2 - Workshop on Medical Computer Vision, MCV 2010, Held in Conjunction with the 13th International Conference on Medical Image Computing and Computer - Assisted Intervention, MICCAI 2010
Y2 - 20 September 2010 through 20 September 2010
ER -