Title :
Image registration using elastic contours and internal landmarks
Author :
Mitsa, Theophano ; Qian, Jiang
Author_Institution :
Dept. of Electr. Eng., Iowa Univ., Iowa City, IA, USA
Abstract :
In this paper, we present a regional image registration method. The basic idea is to use the contours of the regions and internal anatomical landmarks to assist the registration. Our method registers images by warping one image into the other. The warping transforms are derived from point based interpolation that passes through the contour and anatomical landmarks. The registration performance depends on the interpolation method used for warping. Two interpolation methods are investigated in the paper: the inverse distance weighted approach and the radial basis function approach. We also studied the effect of different combinations of internal and external points used for registration. CT images and mesh images are presented to illustrate the choice of interpolation methods and landmarks on the results of the image registration
Keywords :
feedforward neural nets; image matching; image registration; interpolation; CT images; anatomical landmarks; elastic contours; internal landmarks; inverse distance weighted approach; mesh images; point based interpolation; radial basis function approach; regional image registration; warping; Acceleration; Anatomical structure; Biomedical imaging; Cities and towns; Computed tomography; Image analysis; Image registration; Interpolation; Shape;
Conference_Titel :
Instrumentation and Measurement Technology Conference, 1998. IMTC/98. Conference Proceedings. IEEE
Conference_Location :
St. Paul, MN
Print_ISBN :
0-7803-4797-8
DOI :
10.1109/IMTC.1998.679827