Background
In-vivo diffusion tensor CMR (DT-CMR) is an emerging technique for microstructural tissue
characterisation in the myocardium. Most studies are performed at 3T, where higher signal
to noise ratio (SNR) should benefit this signal starved method. However, a few studies have
suggested that DT-CMR is possible at 1.5T, where EPI artefacts may be less severe and 1.5T
hardware is more widely available.
Methods
We recruited 20 healthy volunteers and performed mid-ventricular short axis DT-CMR at
1.5T and 3T. Acquisitions were performed at peak systole and end-diastole using both
stimulated echo acquisition mode (STEAM) and motion compensated spin-echo (MCSE)
sequences at matched spatial resolutions. DT-CMR parameters were averaged over the LV
and compared between 1.5T and 3T sequences using both datasets with and without the
blow reference data included.
Results
Eleven (1.5T) and 12 (3T) diastolic MCSE acquisitions were rejected as the helix angle (HA)
demonstrated <50% normal appearance circumferentially or the acquisition was abandoned
due to poor image quality; a maximum of one acquisition was rejected for other datasets.
Subjective HA map quality was significantly better at 3T than 1.5T for STEAM (p<0.05), but
not for MCSE and other DT-CMR quality measures were consistent with improvements in
STEAM at 3T over 1.5T. When blow data was excluded, no significant differences in mean
diffusivity were observed between field strengths, but fractional anisotropy was significantly
higher at 1.5T than 3T for STEAM systole (p<0.05). Absolute second eigenvector orientation
(E2A, sheetlet angle) was significantly higher at 1.5T than 3T for MCSE systole and STEAM
diastole, but significantly lower for STEAM systole (all p<0.05). Transmural HA distribution
was less steep at 1.5T than 3T for STEAM diastole data (p<0.05). SNR in the blow images was
higher at 3T than 1.5T for all acquisitions (p<0.05).
Conclusion
While 3T provides benefits in terms of SNR, both STEAM and MCSE can be performed at
1.5T. However, MCSE is unreliable in diastole at both field strengths and STEAM benefits
from the improved SNR at 3T over 1.5T. Future clinical research studies may be able to
leverage the wider availability of 1.5T CMR hardware where MCSE acquisitions are
desirable.