Segmentation of image ensembles via latent atlases

Tammy Riklin-Raviv, Koen Van Leemput, Bjoern H. Menze, William M., III Wells, Polina Golland

Research output: Contribution to journalConference abstract in journalResearchpeer-review


Spatial priors, such as probabilistic atlases, play an important role in MRI segmentation. However, the availability of comprehensive, reliable and suitable manual segmentations for atlas construction is limited. We therefore propose a method for joint segmentation of corresponding regions of interest in a collection of aligned images that does not require labeled training data. Instead, a latent atlas, initialized by at most a single manual segmentation, is inferred from the evolving segmentations of the ensemble. The algorithm is based on probabilistic principles but is solved using partial differential equations (PDEs) and energy minimization criteria. We evaluate the method on two datasets, segmenting subcortical and cortical structures in a multi-subject study and extracting brain tumors in a single-subject multi-modal longitudinal experiment. We compare the segmentation results to manual segmentations, when those exist, and to the results of a state-of-the-art atlas-based segmentation method. The quality of the results supports the latent atlas as a promising alternative when existing atlases are not compatible with the images to be segmented.
Original languageEnglish
JournalMedical Image Analysis
Issue number5
Pages (from-to)654-665
Number of pages13
Publication statusPublished - 2010
Externally publishedYes
Event12th International Conference on Medical Image Computing and Computer-Assisted Intervention - Imperial College, London, United Kingdom
Duration: 20 Sep 200924 Sep 2009
Conference number: 12


Conference12th International Conference on Medical Image Computing and Computer-Assisted Intervention
LocationImperial College
CountryUnited Kingdom


  • Latent atlas
  • Segmentation
  • MRI
  • Level-sets

Fingerprint Dive into the research topics of 'Segmentation of image ensembles via latent atlases'. Together they form a unique fingerprint.

Cite this