Improved generation of probabilistic atlases for the expectation maximization classification

Jyrki Lötjönen, Robin Wolz, Juha Koikkalainen, Lennart Thurfjell, Roger Lundqvist, Gunhild Waldemar, Hilkka Soininen, Daniel Rueckert

Research output: Chapter in Book/Report/Conference proceedingConference article in proceedingsScientificpeer-review

7 Citations (Scopus)

Abstract

Probabilistic atlases present prior knowledge about the spatial distribution of various structures or tissues in a population, used commonly in segmentation. We propose three methods for generating probabilistic atlases: 1) the atlases are constructed in a template space using dense non-rigid transformations and transformed to the space of unseen data, 2) as the method 1 but atlas selection is performed in addition, and 3) the atlases are constructed directly in the space of the unseen data. The methods were evaluated in the segmentation of hippocampus from 340 ADNI cases. When comparing with manual segmentations, the Dice similarity indices were 0.84, 0.85 and 0.87 and the correlation coefficients for the volumes were 0.84, 0.92 and 0.96 for the three methods tested. Our results show clearly the importance of probabilistic atlases in segmentation.
Original languageEnglish
Title of host publication2011 IEEE International Symposium on Biomedical Imaging
Subtitle of host publicationFrom Nano to Macro
Place of PublicationPiscataway
PublisherIEEE Institute of Electrical and Electronic Engineers
Pages1839-1842
ISBN (Electronic)978-1-4244-4128-0
ISBN (Print)978-1-4244-4127-3
DOIs
Publication statusPublished - 2011
MoE publication typeA4 Article in a conference publication
Event8th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2011 - Chicago, IL, United States
Duration: 30 Mar 20112 Apr 2011

Conference

Conference8th IEEE International Symposium on Biomedical Imaging: From Nano to Macro, ISBI 2011
Abbreviated titleISBI 2011
CountryUnited States
CityChicago, IL
Period30/03/112/04/11

Keywords

  • segmentation
  • atlas
  • brain

Fingerprint Dive into the research topics of 'Improved generation of probabilistic atlases for the expectation maximization classification'. Together they form a unique fingerprint.

  • Cite this

    Lötjönen, J., Wolz, R., Koikkalainen, J., Thurfjell, L., Lundqvist, R., Waldemar, G., Soininen, H., & Rueckert, D. (2011). Improved generation of probabilistic atlases for the expectation maximization classification. In 2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro (pp. 1839-1842). IEEE Institute of Electrical and Electronic Engineers. https://doi.org/10.1109/ISBI.2011.5872765