Meeting Banner
Abstract #3835

Iterative K-T PCA with Motion Corrected Training Regularization for 3D Myocardial Perfusion Imaging

Johannes F.M. Schmidt1, Lukas Wissmann1, Robert Manka1, 2, Sebastian Kozerke1, 3

1Institute for Biomedical Engineering, University and ETH Zurich, Zurich, Switzerland; 2University Hospital Zurich, Zurich, Switzerland; 3Imaging Sciences and Biomedical Engineering, King's College London, London, United Kingdom

In this work, an iterative k-t PCA algorithm is proposed where an additional spatial transformation is used to further sparsify the data. Training data based regularization is performed in a motion corrected x-pc domain where each time frame is warped to a reference respiratory position. Spatial transformations are derived from frame-by-frame composite images using atlas-based image registration. Using 3D perfusion data acquired in vivo it is demonstrated that this approach successfully corrects for incomplete unfolding due to respiratory bulk motion.

Keywords

accelerated acceleration acquisition additional address affine allowed applicable artifacts asked atlas avoid back best binary biomedical blood bottom breathing bulk cardiac central clearly college comparable compartment compartments compliant component composite constrain constrained contrast converted corrected correction corrects corrupted curves dashed defined delay derived diagnostic diagonal dimensional domain dotted dynamic dynamics engineering enhanced eventual example exploit fields fold frame fully general generation geometry gradient graph half healthy heart hence in vivo incomplete indicated individual intensity interrupted issue iterations iterative king larger masks math minimization model motion myocardial myocardium newton nominal occurs orientation particular pass peak perfusion pool position principal program proposed quality reconstructed reconstruction recovery redundancy registration regularization regularizes removal replicas required resolution respiratory rigid sampled saturation sciences segmentation segmented sense several shallow sheared since slice slices sparse spatial speed spline step studies subject subjects subsequent survey sustain system training transform transformation transformations unable undergo unfold unfolding upslope various ventricular versa warped warping window