uu.seUppsala University Publications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Multi-atlas based segmentation using probabilistic label fusion with adaptive weighting of image similarity measures
Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Radiology, Oncology and Radiation Science, Section of Medical Physics.
Uppsala University, Disciplinary Domain of Medicine and Pharmacy, Faculty of Medicine, Department of Radiology, Oncology and Radiation Science, Section of Medical Physics.
2013 (English)In: Computer Methods and Programs in Biomedicine, ISSN 0169-2607, E-ISSN 1872-7565, Vol. 110, no 3, 308-319 p.Article in journal (Refereed) Published
Abstract [en]

Label fusion multi-atlas approaches for image segmentation can give better segmentation results than single atlas methods. We present a multi-atlas label fusion strategy based on probabilistic weighting of distance maps. Relationships between image similarities and segmentation similarities are estimated in a learning phase and used to derive fusion weights that are proportional to the probability for each atlas to improve the segmentation result. The method was tested using a leave-one-out strategy on a database of 21 pre-segmented prostate patients for different image registrations combined with different image similarity scorings. The probabilistic weighting yields results that are equal or better compared to both fusion with equal weights and results using the STAPLE algorithm. Results from the experiments demonstrate that label fusion by weighted distance maps is feasible, and that probabilistic weighted fusion improves segmentation quality more the stronger the individual atlas segmentation quality depends on the corresponding registered image similarity. The regions used for evaluation of the image similarity measures were found to be more important than the choice of similarity measure.

Place, publisher, year, edition, pages
2013. Vol. 110, no 3, 308-319 p.
Keyword [en]
Segmentation, Atlas based segmentation, Deformable registration, Multi-atlas segmentation, Radiotherapy prostate, Label fusion
National Category
Medical and Health Sciences
Identifiers
URN: urn:nbn:se:uu:diva-202898DOI: 10.1016/j.cmpb.2012.12.006ISI: 000319178500009OAI: oai:DiVA.org:uu-202898DiVA: diva2:634639
Available from: 2013-07-01 Created: 2013-07-01 Last updated: 2017-12-06Bibliographically approved

Open Access in DiVA

No full text

Other links

Publisher's full text

Authority records BETA

Sjöberg, CarlAhnesjö, Anders

Search in DiVA

By author/editor
Sjöberg, CarlAhnesjö, Anders
By organisation
Section of Medical Physics
In the same journal
Computer Methods and Programs in Biomedicine
Medical and Health Sciences

Search outside of DiVA

GoogleGoogle Scholar

doi
urn-nbn

Altmetric score

doi
urn-nbn
Total: 616 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • ieee
  • modern-language-association
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf