Institute Homepage DE Sign In

Back

Empirical Inference Conference Paper 2009

Incorporating Prior Knowledge on Class Probabilities into Local Similarity Measures for Intermodality Image Registration

Thumb ticker sm thumb ilja bezrukov
Empirical Inference
Ilja Bezrukov
Thumb ticker sm l1170153
Empirical Inference
Bernhard Schölkopf
  • Director
no image
Empirical Inference
Matthias Hofmann

We present a methodology for incorporating prior knowledge on class probabilities into the registration process. By using knowledge from the imaging modality, pre-segmentations, and/or probabilistic atlases, we construct vectors of class probabilities for each image voxel. By defining new image similarity measures for distribution-valued images, we show how the class probability images can be nonrigidly registered in a variational framework. An experiment on nonrigid registration of MR and CT full-body scans illustrates that the proposed technique outperforms standard mutual information (MI) and normalized mutual information (NMI) based registration techniques when measured in terms of target registration error (TRE) of manually labeled fiducials.

Author(s): Hofmann, M. and Schölkopf, B. and Bezrukov, I. and Cahill, ND.
Book Title: Proceedings of the MICCAI 2009 Workshop on Probabilistic Models for Medical Image Analysis
Pages: 220-231
Year: 2009
Month: September
Day: 0
Editors: W Wells and S Joshi and K Pohl
Bibtex Type: Conference Paper (inproceedings)
Event Name: PMMIA 2009
Event Place: London, UK
Electronic Archiving: grant_archive
Language: en
Organization: Max-Planck-Gesellschaft
School: Biologische Kybernetik
Links:

BibTex 

@inproceedings{6040,
  title = {Incorporating Prior Knowledge on Class Probabilities into Local Similarity Measures for Intermodality Image Registration},
  booktitle = {Proceedings of the MICCAI 2009 Workshop on Probabilistic Models for Medical Image Analysis  },
  abstract = {We present a methodology for incorporating prior knowledge
  on class probabilities into the registration process. By using knowledge
  from the imaging modality, pre-segmentations, and/or probabilistic atlases,
  we construct vectors of class probabilities for each image voxel. By
  defining new image similarity measures for distribution-valued images,
  we show how the class probability images can be nonrigidly registered in
  a variational framework. An experiment on nonrigid registration of MR
  and CT full-body scans illustrates that the proposed technique outperforms
  standard mutual information (MI) and normalized mutual information
  (NMI) based registration techniques when measured in terms of
  target registration error (TRE) of manually labeled fiducials.},
  pages = {220-231},
  editors = {W Wells and S Joshi and K Pohl},
  organization = {Max-Planck-Gesellschaft},
  school = {Biologische Kybernetik},
  month = sep,
  year = {2009},
  slug = {6040},
  author = {Hofmann, M. and Sch{\"o}lkopf, B. and Bezrukov, I. and Cahill, ND.},
  month_numeric = {9}
}