Title :
Left endocardium tracking via collaborative trackers and shape prior
Author :
Zhou, Yan ; Zhang, Shaoting ; Tsekos, Nikolaos ; Pavlidis, Ioannis ; Metaxas, Dimitris
Author_Institution :
Elekta Inc., Maryland Heights, MO, USA
Abstract :
Real-time MRI emerges as a feasible modality for guiding interventions and surgeries on the beating heart. One challenge is the path planning task for intra-cardiac procedures, which requires tracking the inner contour of the left endocardium on the beating heart. We propose a bottom-up approach to solve this tracking problem. On the bottom level, a coarse contour is generated based on a collaborative network of trackers at every time step. On the top level, PCA based shape prior is used to refine the coarse contours, and eventually obtain a smoothed contour of the left endocardium over time. The approach is validated on 14 MRI sequences of a beating heart, with comparisons to the state-of-the-art approaches. Experimental results show that our approach outperforms other approaches, providing a robust, accurate and fast solution to track the left endocardium of the beating heart.
Keywords :
biomedical MRI; cardiology; image sequences; medical image processing; principal component analysis; surgery; MRI sequences; PCA based shape prior; coarse contours; collaborative network trackers; collaborative trackers; heart beating; intracardiac procedures; left endocardium tracking; path planning task; real-time MRI; state-of-the-art approaches; surgeries; Collaboration; Heart; Image edge detection; Magnetic resonance imaging; Real time systems; Robustness; Shape; Collaborative tracking; shape prior;
Conference_Titel :
Biomedical Imaging (ISBI), 2012 9th IEEE International Symposium on
Conference_Location :
Barcelona
Print_ISBN :
978-1-4577-1857-1
DOI :
10.1109/ISBI.2012.6235665