Locally Adapted Spatio-temporal Deformation Model for Dense Motion Estimation in Periodic Cardiac Image Sequences

We recently introduced a continuous state space parametric model of spatio-temporal transformations and an algorithm, based on Kalman filtering, to represent motion in an image sequence describing a periodic phenomena. One advantage of this method is to simultaneously take into account all the sequence frames to robustly estimate the parameters of a unique spatial and periodic-temporal model. However, in 3D+time, a large number of parameters is required. In this paper, we propose a criterion based on motion energy to locally adapt the trajectory model and thus the temporal complexity of the model. The influence of the model order is illustrated on true 2D+time Magnetic Resonance Images (MRI) of the heart in order to motivate the proposed adaptative criteria. Quantitative results of the proposed adapted spatio-temporal motion model are given on synthetic 2D+time MRI sequences. Preliminary experiments show a significant impact notably regarding the parameter saving while preserving the accuracy of the motion estimates.

[1]  Taku Komura,et al.  Topology matching for fully automatic similarity estimation of 3D shapes , 2001, SIGGRAPH.

[2]  Fabrice Heitz,et al.  3-D deformable image registration: a topology preservation scheme based on hierarchical deformation models and interval analysis optimization , 2005, IEEE Transactions on Image Processing.

[3]  Lasse Riis Østergaard,et al.  Active Surface Approach for Extraction of the Human Cerebral Cortex from MRI , 2006, MICCAI.

[4]  Daniel Rueckert,et al.  Nonrigid registration using free-form deformations: application to breast MR images , 1999, IEEE Transactions on Medical Imaging.

[5]  S. Mallat.  A wavelet tour of signal processing , 1998 .

[6]  Amir A. Amini,et al.  Spatio-temporal tracking of myocardial deformations with a 4-D B-spline model from tagged MRI , 1999, IEEE Transactions on Medical Imaging.

[7]  Daniel Rueckert,et al.  Fast Spatio-temporal Free-Form Registration of Cardiac MR Image Sequences , 2004, FIMH.

[8]  Torsten Rohlfing,et al.  Volume-preserving nonrigid registration of MR breast images using free-form deformation with an incompressibility constraint , 2003, IEEE Transactions on Medical Imaging.

[9]  Michael Unser,et al.  Spatio-temporal nonrigid registration for ultrasound cardiac motion estimation , 2005, IEEE Transactions on Medical Imaging.

[10]  R. E. Kalman,et al.  A New Approach to Linear Filtering and Prediction Problems , 2002 .

[11]  Thomas W. Sederberg,et al.  Free-form deformation of solid geometric models , 1986, SIGGRAPH.

[12]  Patrick Clarysse,et al.  Two-dimensional spatial and temporal displacement and deformation field fitting from cardiac magnetic resonance tagging , 2000, Medical Image Anal..

[13]  Johan Montagnat,et al.  4D deformable models with temporal constraints: application to 4D cardiac image segmentation , 2005, Medical Image Anal..