Optical flow based deformable volume registration using a novel second-order regularization prior

Sasa Grbic, Martin Urschler, Thomas Pock, Horst Bischof

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Nonlinear image registration is an initial step for a large number of medical image analysis applications. Optical flow based intensity registration is often used for dealing with intra-modality applications involving motion differences. In this work we present an energy functional which uses a novel, second-order regularization prior of the displacement field. Compared to other methods our scheme is robust to non-Gaussian noise and does not penalize locally affine deformation fields in homogeneous areas. We propose an efficient and stable numerical scheme to find the minimizer of the presented energy. We implemented our algorithm using modern consumer graphics processing units and thereby increased the execution performance dramatically. We further show experimental evaluations on clinical CT thorax data sets at different breathing states and on dynamic 4D CT cardiac data sets.
LanguageEnglish
Title of host publicationProc. SPIE 7623, Medical Imaging 2010: Image Processing
EditorsBenoit M. Dawant, David R. Haynor
PublisherSPIE
DOIs
StatusPublished - 2010

Fingerprint

Optical flows
Image registration
Image analysis
Graphics processing unit

Fields of Expertise

  • Information, Communication & Computing

Cite this

Grbic, S., Urschler, M., Pock, T., & Bischof, H. (2010). Optical flow based deformable volume registration using a novel second-order regularization prior. In B. M. Dawant, & D. R. Haynor (Eds.), Proc. SPIE 7623, Medical Imaging 2010: Image Processing SPIE. DOI: 10.1117/12.844549

Optical flow based deformable volume registration using a novel second-order regularization prior. / Grbic, Sasa; Urschler, Martin; Pock, Thomas; Bischof, Horst.

Proc. SPIE 7623, Medical Imaging 2010: Image Processing. ed. / Benoit M. Dawant; David R. Haynor. SPIE, 2010.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Grbic, S, Urschler, M, Pock, T & Bischof, H 2010, Optical flow based deformable volume registration using a novel second-order regularization prior. in BM Dawant & DR Haynor (eds), Proc. SPIE 7623, Medical Imaging 2010: Image Processing. SPIE. DOI: 10.1117/12.844549
Grbic S, Urschler M, Pock T, Bischof H. Optical flow based deformable volume registration using a novel second-order regularization prior. In Dawant BM, Haynor DR, editors, Proc. SPIE 7623, Medical Imaging 2010: Image Processing. SPIE. 2010. Available from, DOI: 10.1117/12.844549
Grbic, Sasa ; Urschler, Martin ; Pock, Thomas ; Bischof, Horst. / Optical flow based deformable volume registration using a novel second-order regularization prior. Proc. SPIE 7623, Medical Imaging 2010: Image Processing. editor / Benoit M. Dawant ; David R. Haynor. SPIE, 2010.
@inproceedings{6e7018d6edf44cd782f70fdc3f29a99d,
title = "Optical flow based deformable volume registration using a novel second-order regularization prior",
abstract = "Nonlinear image registration is an initial step for a large number of medical image analysis applications. Optical flow based intensity registration is often used for dealing with intra-modality applications involving motion differences. In this work we present an energy functional which uses a novel, second-order regularization prior of the displacement field. Compared to other methods our scheme is robust to non-Gaussian noise and does not penalize locally affine deformation fields in homogeneous areas. We propose an efficient and stable numerical scheme to find the minimizer of the presented energy. We implemented our algorithm using modern consumer graphics processing units and thereby increased the execution performance dramatically. We further show experimental evaluations on clinical CT thorax data sets at different breathing states and on dynamic 4D CT cardiac data sets.",
author = "Sasa Grbic and Martin Urschler and Thomas Pock and Horst Bischof",
year = "2010",
doi = "10.1117/12.844549",
language = "English",
editor = "Dawant, {Benoit M.} and Haynor, {David R.}",
booktitle = "Proc. SPIE 7623, Medical Imaging 2010: Image Processing",
publisher = "SPIE",
address = "United States",

}

TY - GEN

T1 - Optical flow based deformable volume registration using a novel second-order regularization prior

AU - Grbic,Sasa

AU - Urschler,Martin

AU - Pock,Thomas

AU - Bischof,Horst

PY - 2010

Y1 - 2010

N2 - Nonlinear image registration is an initial step for a large number of medical image analysis applications. Optical flow based intensity registration is often used for dealing with intra-modality applications involving motion differences. In this work we present an energy functional which uses a novel, second-order regularization prior of the displacement field. Compared to other methods our scheme is robust to non-Gaussian noise and does not penalize locally affine deformation fields in homogeneous areas. We propose an efficient and stable numerical scheme to find the minimizer of the presented energy. We implemented our algorithm using modern consumer graphics processing units and thereby increased the execution performance dramatically. We further show experimental evaluations on clinical CT thorax data sets at different breathing states and on dynamic 4D CT cardiac data sets.

AB - Nonlinear image registration is an initial step for a large number of medical image analysis applications. Optical flow based intensity registration is often used for dealing with intra-modality applications involving motion differences. In this work we present an energy functional which uses a novel, second-order regularization prior of the displacement field. Compared to other methods our scheme is robust to non-Gaussian noise and does not penalize locally affine deformation fields in homogeneous areas. We propose an efficient and stable numerical scheme to find the minimizer of the presented energy. We implemented our algorithm using modern consumer graphics processing units and thereby increased the execution performance dramatically. We further show experimental evaluations on clinical CT thorax data sets at different breathing states and on dynamic 4D CT cardiac data sets.

U2 - 10.1117/12.844549

DO - 10.1117/12.844549

M3 - Conference contribution

BT - Proc. SPIE 7623, Medical Imaging 2010: Image Processing

PB - SPIE

ER -