Meeting Banner
Abstract #4502

Helix Angle (HA) Healthy Statistical Average Technique for HA Quantification in vivo Cardiac Diffusion Tensor Imaging

Pedro Ferreira1, Sonia Nielles-Vallespin2, Peter David Gatehouse1, Ranil de Silva1, Jennifer Keegan1, Peter Speier3, Thorsten Feiweier4, Timothy G. Reese5, Tevfik Ismail1, Andrew Scott1, Choukri Mekkaoui5, David E. Sosnovik5, David N. Firmin1

1BRU, Royal Brompton Hospital, London, United Kingdom; 2National Heart Lung and Blood Institute (NHLBI), National Institutes of Health (NIH), Bethesda, MD, United States; 3MR Application & Workflow Development, Siemens, Erlangen, Germany; 4MR Application & Workflow Development, Siemens AG, Erlangen, Germany; 5Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Charlestown, MA, United States

A healthy average HA map was synthesised from a dataset of 10 healthy volunteers, matching the myocardial shape of the DTI data being analysed. This map allows for a direct pixel-by-pixel comparison of the cardiac fibres orientation in vivo. This technique was subsequently applied to evaluate the improved accuracy of in vivo cardiac DTI data sets based on numbers of averages, together with mean diffusivity and fractional anisotropy quantitation.

Abstract PDF Presentation Video

Keywords

accurate allows although apical appear application applied assessed assuming asymptotic atlas audience automatic averaging basal become biomedical blood breath cardiac care circumferential clinicians correlation cross curve datasets decreases decreasing degrees depict derived developed development deviation diagnosis difficult difficulty diffusion diffusivity direct disease divided eigenvalues example examples expected field finding fractional function gatehouse general generate generated great health healthy heart hearts helix holds hospital human illustrated illustrates in vivo increasing insertion institutes inter invest likely linear lung magnitude maps match matching measured muscle myocardial myocardium national negative normality numerical orientation overcome papillary particularly pattern peter pixel plotted poor post previous previously print processed protocol quality quantification quantify quantitative registration relatively remove reproducibility resemble resolution respective respectively royal scanned scanning segment segments sensitivity shape shot simulations slice slices slope spatial stages statistical studies subject substantially superior systole taken target tensor tensors timothy together tolerates trans trend underestimation useful varies vary varying ventricular volunteer volunteers weaker