Bogdan Georgescu1,
Nicole Seiberlich2, Tommaso Mansi1, Xiaoguang Lu1,
Ali Kamen1, Vidya Nadig3, Dorin Comaniciu1,
Mark A. Griswold2
1Imaging
and Computer Vision, Siemens Corporation, Corporate Technology, Princeton,
NJ, United States; 2Biomedical Engineering, Case Western Reserve
University, Cleveland, OH, United States; 3MetroHealth Heart and
Vascular Center, Case Western Reserve University, Cleveland, OH, United
States
We propose a data-driven automated method to fit a parameterized 4D cardiac model of the left and right ventricles while simultaneously correcting for spatial slice mis-alignment. The method provides quantitative estimates of the ventricles morphology and dynamics and it can be applied to both standard breath-hold short axis CMR stacks as well as to a recently developed real-time free-breathing protocol using an undersampled radial trajectory and reconstructed using through-time radial GRAPPA. Initial results show successful slice re-alignment and quantitative measurements on both protocols are in good agreement with the ground-truth.