Mass preserving image registration: results of Evaluation of Methods for Pulmonary Image Registration 2010 Challenge

Vladlena Gorbunova, Jon Sporring, Pechin Chien Pau Lo, Asger Dirksen, Marleen de Bruijne

Abstract

The paper presents results the mass preserving image registration method in the Evaluation of Methods for Pulmonary Image Registration 2010 (EMPIRE10) Challenge. The mass preserving image registration algorithm was applied to the 20 image pairs. Registration was evaluated using four different scores: lung boundary alignment, major fissure lignment, landmark alignment and transform singularity scores. The registration algorithm achieved an average landmark alignment score of 2.20 } 2.05 mm and the median of 1.29 mm. In 19 out of 20 image pairs, the method produced invertible deformations. Overall, the masspreserving image registration method was ranked 20th out of 34 participants
OriginalsprogEngelsk
TitelMedical image analysis for the clinic: a grand challenge : workshop proceedings from the 13th International Conference on Medical Image Computing And Computer Assisted Intervention
RedaktørerBram van Ginneken, Keelin Murphy, Tobias Heimann, Vladimir Pekar, Xiang Deng
Antal sider10
ForlagCreateSpace Independent Publishing Platform
Publikationsdato2010
Sider155-164
ISBN (Trykt)978-1453759394
StatusUdgivet - 2010
BegivenhedMICCAI 2010 Workshop on Medical Image Analysis for the Clinic: a grand challenge - Beijing, Kina
Varighed: 24 sep. 201024 sep. 2010

Konference

KonferenceMICCAI 2010 Workshop on Medical Image Analysis for the Clinic
Land/OmrådeKina
ByBeijing
Periode24/09/201024/09/2010

Fingeraftryk

Dyk ned i forskningsemnerne om 'Mass preserving image registration: results of Evaluation of Methods for Pulmonary Image Registration 2010 Challenge'. Sammen danner de et unikt fingeraftryk.

Citationsformater