Optimization of a mutual shape based on the Fréchet-Nikodym metric for 3D shapes fusion
Stephanie Jehan-Besson  1@  , Patrick Clarysse  2  , Regis Clouard  3  , Fréderique Frouin  4  
1 : Centre de Recherche en Acquisition et Traitement de lÍmage pour la Santé
Université Jean Monnet [Saint-Etienne], Hospices Civils de Lyon, Institut National des Sciences Appliquées de Lyon, Université Claude Bernard Lyon 1, Centre National de la Recherche Scientifique : UMR5220, Institut National de la Santé et de la Recherche Médicale : U1206
2 : Centre de Recherche en Acquisition et Traitement de lÍmage pour la Santé
Université Jean Monnet [Saint-Etienne], Hospices Civils de Lyon, Institut National des Sciences Appliquées de Lyon, Université de Lyon, Institut National des Sciences Appliquées, Université Claude Bernard Lyon 1, Université de Lyon : UMR5220, Centre National de la Recherche Scientifique : U1206, Institut National de la Santé et de la Recherche Médicale
7 avenue Jean Capelle, Bat Blaise Pascal, 69621 Villeurbanne Cedex -  France
3 : Normandie Univ, UNICAEN, ENSICAEN, CNRS, GREYC
Normandie Univ, UNICAEN, ENSICAEN CNRS
4 : Laboratoire dÍmagerie Translationnelle en Oncologie
Institut Curie [Paris], Institut National de la Santé et de la Recherche Médicale : U1288

In the field of delineation of 2D or 3D regions of interest (ROI) in medical imaging, and especially due to the development of multimodal and multiparametric image acquisition devices, the combination of segmentations of anatomical structures from different sources is interesting. It is also essential to accurately assess the variability between delineation experts or algorithms with different parameters. In this work, we propose to estimate a mutual shape defined as the optimum of a statistical criterion based on information theory. The mutual shape is computed using shape optimization tools through the computation of shape gradients. Compared to our previous work, we propose to interpret the mutual shape as a sum of distances of the Fréchet family. Moreover, we extend our framework to 3D shape fusion and provide a synthetic example to demonstrate the difference between mutual shape, average shape and shape union.


Personnes connectées : 11 Vie privée
Chargement...