Title :
Interactive polygons in region-based deformable contours for medical images
Author :
Zhu, Yaoyao ; Shen, Tian ; Lopresti, Daniel ; Huang, Xiaolei
Author_Institution :
Dept. of Comput. Sci. & Eng., Lehigh Univ., Bethlehem, PA, USA
fDate :
June 28 2009-July 1 2009
Abstract :
A new user interaction method called interactive polygons is presented in this paper. These interaction polygons are designed for use with the Active Volume Model segmentation method, which deforms with constraints from both Region Of Interest (ROI) and image gradient information. The two kinds of interaction polygons we apply are "merge polygons" and "split polygons" which identify the foreground and background, respectively. Users are allowed to draw these polygons to correct the original AVM segmentation results. These interactive polygons are used to update the region statistics in the original model and help the model deform to the desirable boundaries.
Keywords :
image segmentation; medical image processing; statistical analysis; active volume model segmentation method; image gradient information; interactive polygons; medical images; merge polygons; region statistics; region-based deformable contours; split polygons; Active contours; Biomedical engineering; Biomedical imaging; Computer science; Data mining; Deformable models; Image analysis; Image segmentation; Solid modeling; Statistics; Segmentation; active volume models; interactive segmentation; snakes; user interaction;
Conference_Titel :
Biomedical Imaging: From Nano to Macro, 2009. ISBI '09. IEEE International Symposium on
Conference_Location :
Boston, MA
Print_ISBN :
978-1-4244-3931-7
Electronic_ISBN :
1945-7928
DOI :
10.1109/ISBI.2009.5192977
