Imperial College London

ProfessorDanielRueckert

Faculty of EngineeringDepartment of Computing

Head of Department of Computing
 
 
 
//

Contact

 

+44 (0)20 7594 8333d.rueckert Website

 
 
//

Location

 

568Huxley BuildingSouth Kensington Campus

//

Summary

 

Publications

Citation

BibTex format

@inproceedings{Pszczolkowski:2012:10.1007/978-3-642-35428-1_28,
author = {Pszczolkowski, S and Pizarro, L and O'Regan, DP and Rueckert, D},
doi = {10.1007/978-3-642-35428-1_28},
pages = {226--233},
title = {Gradient projection learning for parametric nonrigid registration},
url = {http://dx.doi.org/10.1007/978-3-642-35428-1_28},
year = {2012}
}

RIS format (EndNote, RefMan)

TY  - CPAPER
AB - A potentially large anatomical variability among subjects in a population makes nonrigid image registration techniques prone to inaccuracies and to high computational costs in their optimisation. In this paper, we propose a new learning-based approach to accelerate the convergence rate of any chosen parametric energy-based image registration method. From a set of training images and their corresponding deformations, our method learns offline a projection from the gradient space of the energy functional to the parameter space of the chosen registration method using partial least squares. Combined with a regularisation term, the learnt projection is subsequently used online to approximate the optimisation of the energy functional for unseen images. We employ the B-spline approach as underlying registration method, but other parametric methods can be used as well. We perform experiments on synthetic image data and MR cardiac sequences to show that our approach significantly accelerates the convergence -in number of iterations and total computational cost- of the chosen registration method, while achieving similar results in terms of accuracy. © 2012 Springer-Verlag.
AU - Pszczolkowski,S
AU - Pizarro,L
AU - O'Regan,DP
AU - Rueckert,D
DO - 10.1007/978-3-642-35428-1_28
EP - 233
PY - 2012///
SN - 0302-9743
SP - 226
TI - Gradient projection learning for parametric nonrigid registration
UR - http://dx.doi.org/10.1007/978-3-642-35428-1_28
ER -