Quick Fire Session
SCMR 22nd Annual Scientific Sessions
Kevin Moulin, PhD
Post-doctoral Fellow
Stanford University
Ilya Verzhbinsky, BSc
Research Associate
Stanford University
Nyasha Maforo
Graduate Student Researcher
University of California, Los Angeles
Luigi Perotti, PhD
Associate Project Scientist
University of California, Los Angeles
Pierre Croisille, MD, PhD
Professor
Univ Lyon, UJM-Saint-Etienne, INSA, CNRS UMR 5520, INSERM U1206, CREATIS
Magalie Viallon, PhD
MR Physicist
CREATIS UMR CNRS 5220 - INSERM U1206/ Hopital Universitaire de Saint-Etienne
Daniel Ennis, PhD
Associate Professor
Stanford University
Background: Myocardial microstructure critically underlies regional cardiac function. This has motivated the development of cardiac diffusion tensor imaging (cDTI) to analyze myocardial microstructure at different phase of contraction. Single-shot spin-echo planar imaging (SS-EPI) is one cDTI that can be faster with higher SNR than its stimulated-echo counterpart (1). However the acquisition of SS-EPI is still challenging in vivo because it involves using bulk motion compensation techniques (2) and can only be acquired at specifics trigger delay (TD) in the cardiac cycle (3). In this work, we have designed a fast TD scout sequence that permits quickly estimating a patient specific TD for SS-EPI cDTI, which was used to evaluate the cardiac cycle dependence of mean diffusivity (MD) and fraction of anisotropy (FA) in healthy volunteers.
Methods:
Seven healthy volunteers (N=7) were imaged on at 3T (Prisma, Siemens). Both TD scout and cDTI were acquired using the same image parameters in a single mid-ventricular short-axis slice using a second order motion compensated diffusion encoding waveform (3) (TE=61ms, 1.6x1.6x8mm interpolated to 0.8x0.8x8mm). Thirty cardiac phases, three directions at b-value 350s/mm² were first obtained using the TD scout sequence during free-breathing (TR=~2000ms, 90 img/~3 min total).
For the cDTI acquisition, a total of three cardiac phases, early systole (Syst1), late systole (Syst2) and diastole (Diast) were defined using the TDs observed during the TD scout (Figure 1). Two b-values (0-350s/mm²), twelve diffusion encoding directions, and five averages were obtained using an end-respiratory navigator trigger (TR=~4000ms, 250 img/~5 min of free-breathing per phase, 15 min total).
Results:
Example TD scout acquisitions are shown in Figure 1. The cDTI was successfully acquired during Syst1 and Syst2 in all seven volunteer (Success=100%). The diastolic acquisitions were only possible in four volunteers (Success=57%) as shown in the TD scout example given Figure 1-C. Figure 2 shows an example of DWI, MD, and FA maps obtained in one volunteer at the three different phases. Figure 3 represents the total distribution of MD and FA across volunteers. A lower MD and FA were found Syst2 compared to Syst1 and Dias. Using ANOVA, statistical differences were found between cardiac phases for the distribution of MD (p<0.001) or FA (p<0.001).
Conclusion:
The dynamic range of applicability of SS-EPI DWI approach was assessed using a TD scout sequence, which facilitates high quality cDTI acquisitions. The TD scout was then used to define the TDs for three cDTI multi-phase acquisitions. Herein systolic images were successfully obtained in both systolic phases, but a strong patient dependence remains in diastole (4). Significant differences were found for both MD and FA which thus depend on the cardiac cycle timing.
Funding – NIH HL131975 and HL131823.