Diffusion MRI of the human heart provides unique insight into myocardial microstructure but has been hampered by cardiac and respiratory motion, short T2 of the heart muscle, and limited gradient strength. Recent advances in ultra-strong gradient technology not only help to overcome these technical challenges but also allow higher diffusion weighting (i.e., b-values) with clinically compatible echo times.
The authors demonstrate how this enabled in vivo diffusion kurtosis imaging (DKI) and q-space trajectory imaging (QTI) in the beating human heart, therefore moving beyond the Gaussian assumptions of diffusion tensor imaging (DTI).
These advances may pave the way for more sensitive biomarkers of pathological changes of the myocardium and bring microstructural imaging closer to clinical application.
Key points
• Ultra-strong gradients (300 mT/m) make cardiac diffusion MRI feasible at higher b-values.
• In vivo cardiac diffusion kurtosis imaging and q-space trajectory imaging (QTI) were demonstrated with clinically compatible echo times.
• Kurtosis and QTI metrics reveal non-Gaussian diffusion, offering access to new imaging biomarkers of myocardial microstructure.
• Translation to clinical systems is within reach with new 200 mT/m gradient scanners.
Shoutout and thank you to the co-authors:
Lars Mueller, Ph.D.; Jürgen E Schneider, Ph.D. (Biomedical Imaging Science Department, Leeds Institute of Cardiovascular and Metabolic Medicine, University of Leeds, Leeds, UK)
Derek K Jones, Ph.D. (Cardiff University Brain Research Imaging Centre (CUBRIC), School of Psychology, Cardiff University, Cardiff, UK)
Filip Szczepankiewicz, Ph.D. (Department of Medical Radiation Physics, Lund University, Lund, Sweden)
Fabrizio Fasano, Ph.D. (Siemens Healthineers)
Unlocking the Heart’s #Microstructure: Cardiac #Diffusion #MRI with Ultra-Strong Gradients by Maryam Afzali, PhD; et al. (@universityofleeds.bsky.social).
marketing.webassets.siemens-healthineers.com/0267aa50bc95...
@deekayjay.bsky.social
#dMRI #RadSky #CardioSky #MagnetomWorld