Skip to Main content Skip to Navigation

Utilisation d'atlas anatomiques et de modèles déformables pour la segmentation de la prostate en IRM

Sébastien Martin 1
1 TIMC-GMCAO - Gestes Medico-chirurgicaux Assistés par Ordinateur
TIMC - Techniques de l'Ingénierie Médicale et de la Complexité - Informatique, Mathématiques et Applications, Grenoble - UMR 5525
Abstract : In this thesis we focus on the development of tools for the automatic segmentation of the prostate gland in Magnetic Resonance Images. In a near future, these tools will be used to help clinicians for prostate brachytherapy. Our method is based on the use of both an anatomical atlas and deformable models to achieve an automatic segmentation. We use a hierachical approach composed of three steps. In the first step, a pre-processing of input images is performed to correct the inhomogeneity of images acquired with an endorectal MRI coil. In the second step, the target patient is registered onto an anatomical atlas ; this step allows for a first segmentation of the prostate and of the rectum. In the last step, two deformable models evolve to refine segmentations obtained in the previous step. The segmentation of the rectum is executed first. In a second time, a deformable model evolves to segment the prostate. The segmentation of the rectum is used to constrain the evolution of the prostate deformable model. The anatomical atlas is built using a hybrid registration method. The later both exploits geometric and intensity-based information for mapping a population onto a standard patient. We have shown that this strategy improves the accuracy of obtained automatic segmentations.
Document type :
Complete list of metadatas

Cited literature [106 references]  Display  Hide  Download
Contributor : Sébastien Martin <>
Submitted on : Saturday, February 13, 2010 - 3:35:52 PM
Last modification on : Tuesday, November 24, 2020 - 4:18:04 PM
Long-term archiving on: : Wednesday, November 30, 2016 - 1:49:03 PM


  • HAL Id : tel-00374342, version 2



Sébastien Martin. Utilisation d'atlas anatomiques et de modèles déformables pour la segmentation de la prostate en IRM. Informatique [cs]. Université Joseph-Fourier - Grenoble I, 2008. Français. ⟨tel-00374342v2⟩



Record views


Files downloads