Accurate automatic segmentation of anatomical structures is usually considered a difficult problem to solve because of anatomical variability and varying imaging conditions. A prior description of the shape of the anatomical structure to be segmented can reduce the ambiguity associated with the segmentation task. However this prior information has to be prepared specifically for the structure of interest, usually supervised and under favorable imaging conditions. An alternative is to consider the shape of the object sequentially, along a particular dimension of the data. This is the approach taken here, ie on-line modeling of sequential shape information which is combined with sequential segmentation of the intensity distributions for the segmented structure and the surrounding region.
|Translated title of the contribution||Probabilistic Sequential Segmentation and Simultaneous On-Line Shape Learning of Multi-Dimensional Medical Imaging Data|
|Title of host publication||MICCAI Workshop on Probabilistic Models for Medical Image Analysis|
|Publication status||Published - 2009|
Bibliographical noteOther page information: -
Conference Proceedings/Title of Journal: MICCAI Workshop on Probabilistic Models for Medical Image Analysis
Other identifier: 2001032