Meeting Banner
Abstract #1130

Highly Accelerated Cine-MRI in Mouse Hearts Using Compressed Sensing and Parallel Imaging at 9.4T

Tobias Wech1, Victoria Thornton2, Craig A. Lygate2, Stefan Neubauer2, Herbert Kstler1, Jrgen E. Schneider2

1Institute of Radiology, University of Wrzburg, Wrzburg, Germany; 2Cardiovascular Medicine, University of Oxford, Oxford, United Kingdom

Parallel Imaging (PI) and Compressed Sensing (CS) have been shown to individually provide a three-fold reduction in scan-time for cine MRI in mice whilst providing accurate measurement of the physiologically relevant parameters. We now sought to combine PI and CS to further accelerate cine-MRI in mouse hearts at 9.4T. Fully acquired cine-data sets were retrospectively undersampled (overall acceleration factors combined: 4 up to 16) and subjected to CS and PI reconstruction, followed by image analysis. It is shown that a 3x3-fold acceleration (RCS x RPI) still allows for accurate assessment of cardiac functional values.

Keywords

accelerate accelerated acceleration accuracy achieved acquisition adequate adult agreed aimed allowed animal applied assess axis blinded cardiac cardiovascular chamber channel channels chronic cine combination combine comparing compressed computer conducted consecutive consisting cycle determined detrimental diastole diastolic directly disease double elite equipped even exploiting fields fold foundation frame frames fully functional fundamentally funded gated generally generate gradient guide heart hearts highly horizontal hybrid impacting implement increasing indicates infarction institute intra kingdom lists long mass materials matrix medicine mice minute model models mouse myocardial network observer operator oxford parallel proceed progress proposed quality radiology receive recently reconstruction reconstructions reduce reduction reproducibility respectively retrospectively sampled scale scanner seemed sensing sets short significantly slice slices smaller space sparsity specifically speed steps subjected subsequently suggest system table temporal towards trend ultra unaccelerated utilized validate variability ventricular views volumes young