19 articles – 48 references  [version française]
HAL: hal-00726197, version 1

Detailed view  Export this paper
IEEE Transactions on Medical Imaging 31, 8 (2012) 1651-1660
Nonsupervised Ranking of Different Segmentation Approaches: Application to the Estimation of the Left Ventricular Ejection Fraction From Cardiac Cine MRI Sequences
Jessica Lebenberg1, 2, I. Buvat3, Alain Lalande4, Patrick Clarysse5, Christopher Casta5, Alexandre Cochet4, Constantin Constantinidés1, 2, Jean Cousty6, Alain De Cesare1, Stéphanie Jehan-Besson7, Muriel Lefort1, Laurent Najman6, Elodie Roullot2, Laurent Sarry8, Christophe Tilmant9, Mireille Garreau10, Frédérique Frouin1

A statistical methodology is proposed to rank several estimation methods of a relevant clinical parameter when no gold standard is available. Based on a regression without truth method, the proposed approach was applied to rank eightmethods without using any a priori information regarding the reliability of each method and its degree of automation. It was only based on a prior concerning the statistical distribution of the parameter of interest in the database. The ranking of the methods relies on figures of merit derived from the regression and computed using a bootstrap process. The methodology was applied to the estimation of the left ventricular ejection fraction derived from cardiac magnetic resonance images segmented using eight approaches with different degrees of automation: three segmentations were entirely manually performed and the others were variously automated. The ranking of methods was consistent with the expected performance of the estimation methods: the most accurate estimates of the ejection fraction were obtained using manual segmentations. The robustness of the ranking was demonstrated when at least three methods were compared. These results suggest that the proposed statistical approach might be helpful to assess the performance of estimation methods on clinical data for which no gold standard is available.
1:  LIF - Laboratoire d'Imagerie Fonctionnelle
3:  IMNC - Imagerie et Modélisation en Neurobiologie et Cancérologie
4:  Le2i - Laboratoire Electronique, Informatique et Image
5:  CREATIS - Centre de recherche en applications et traitement de l'image pour la santé
6:  LIGM - Laboratoire d'Informatique Gaspard-Monge
7:  GREYC - Groupe de Recherche en Informatique, Image, Automatique et Instrumentation de Caen
8:  ISIT - Image Science for Interventional Techniques
9:  LASMEA - Laboratoire des sciences et matériaux pour l'électronique et d'automatique
10:  LTSI - Laboratoire Traitement du Signal et de l'Image
Computer Science/Medical Imaging
Bootstrap process – cardiac image analysis – left ventricular ejection fraction – nonsupervised segmentation methods ranking – regression without truth
Attached file list to this document: 
Lebenberg_TMI_2012.pdf(1.1 MB)