Abstract #0866
Synthetic CT generation from T2 weighted MRI using a hybrid regression and multi-atlas approach
S. Ghose 1 , D. Rivest Henault 1 , J. Mitra 1 , J. Sun 2 , P. Pichler 3 , P. Greer 3 , and J. Dowling 4
1
Australian e-Health Research Centre, CSIRO
Digital Productivity Flagship, Herston, QLD, Australia,
2
University
of Newcastle, NSW, Australia,
3
Department
of Radiation Oncology, University of Newcastle, NSW,
Australia,
4
Australian
e-Health Research Centre, CSIRO Digital Productivity
Flagship, QLD, Australia
In this work, we propose to use a hybrid atlas and
regression-based method to generate synthetic CT images
from MRI for MRI based radiation therapy planning for
prostate cancer treatment. Compared to a pure
regression-based method, an additional UTE sequence is
not required to segment the bone, and in comparison to a
pure atlas-based method, we use regression-based soft
tissue modeling, producing better soft tissue
estimation. The proposed hybrid approach achieves a 0.6%
dose difference compared to the use of original CT in
the dose plan.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
