Click here to


Are you sure ?

Yes, do it No, cancel

Improved Repeatabilityof Intra and Extra Myocellular Lipids Quantification Using Diffusion Tensor Imaging (DTI) and Dixon Imaging Guided Proton MR Spectroscopy in Skeletal Muscle

J Zhang1,2, C Arena2, A Ajala3, E Buko3, M Dotson2, R Muthupillai2,3*, (1) Department of Medical Physics and Radiation Oncology,University of Vermont Medical Center, Burlington, VT, (2)Department of Physics and Interventional Radiology,CHI St Luke's Health Baylor St Luke's Medical Center, Houston,TX(3) Department of Physics, University of Houston, Houston, TX,


(Monday, 7/13/2020) 4:30 PM - 5:30 PM [Eastern Time (GMT-4)]

Room: Track 1

Proton MR spectroscopic (MRS) measurement of intra and extra myocellular lipids (IMCL and EMCL) can serve as biomarkers of skeletal muscle metabolism. However, reproducibility of IMCL/EMCL concentration in vivo is hampered by: (1) muscle fiber orientation dependent frequency shift of IMCL/EMCL lipid peaks, and (2) inadvertent inclusion of fascial fat streaks within MRS voxel. We hypothesize that information about muscle fiber orientation from diffusion tensor (DTI), and about fascial fat streaks using Dixon imaging at the time of prescribing MRS voxel, can improve the reproducibility of IMCL/EMCL measurements.

MRI: IMCL/EMCL concentrations in a single (~1 cm3) voxel in soleus muscle was measured in six volunteers (36.7±7.8yrs) using MRS in two imaging sessions (two weeks apart) at 3.0T using transmit/receive knee coil. In each imaging session, MRS voxel was prescribed using: (ScoutMethod) Scout images only, (MIPmethod) Maximum intensity projections (MIP) of Dixon derived fat maps (to avoid extra-cellular fascial fat streaks), and (MIP+DTImethod) DTI (16 directions, b=0,500 s/mm2) and Dixon fat maps to position the MRS voxel in an area devoid of fascial fat where the fibers were predominantly in the R-L direction.
Data Analysis: The coefficient of variation (CoV), and bias/confidence interval (Bland-Altman) between corresponding pairs of repeated measurements was calculated.

1. The CoV of normalized IMCL|Cr and EMCL|Cr from voxel positioned using MIP+DTImethod was lower (0.08 and 0.24, respectively) than that from ScoutMethod (0.30 and 0.44 respectively) or from MIPmethod (0.57 and 0.94 respectively).
2. IMCL|Cr and EMCL|Cr had lower bias and narrower confidence interval (Bland-Altman analysis) for MIP+DTImethod (-0.42 ±1.37 and 0.51±3.68 respectively) when compared to ScoutMethod or MIPmethod

Muscle fiber orientation derived from DTI and Dixon-derived MIP maps can be used to improve the repeatability of MRS quantification of lipids content in longitudinal studies of muscle metabolism.


MRS, Quantitative Imaging, Quality Assurance


IM- MRI : Spectroscopy

Contact Email