2298

Diffusion tensor and Dixon imaging for guiding proton magnetic resonance spectroscopic evaluation of intra and extra myocellular lipids in skeletal muscle
Jiming Zhang1, Claudio Arena1, Afis Ajala2, Luning Wang3, Rajagopal Viswanatah Sekhar4, and Raja Muthupillai1

1Diagnostic and Interventional Radiology, CHI St Luke's Health, Houston, TX, United States, 2Physics, Univeristy of Houston, Houston, TX, United States, 3Philips Healthcare, Gainesville, FL, United States, 4Diabetes, Endocrinology and Metabolism, Baylor College of Medicine, Houston, TX, United States

Synopsis

A different level of overlap between intramyocellular and extramyocellular lipids signal (EMCL and IMCL) depends on muscle fiber orientation and interstitial/fascial muscle fat tissue. We proposed diffusion tensor imaging and Dixon reconstructed fiber orientation map and maximum intensity projection fat maps to guide MR spectroscopy acquisition by carefully positioning the voxel so as to minimize these two effects. This method can be used to improve the consistency and accuracy of using MRS to quantify the intra skeletal muscle lipids and minimize variability in longitudinal studies such as evaluating disease progression or monitoring treatment progress

Introduction

Quantification of skeletal muscle lipids - intramyocellular (IMCL) and extramyocellular (EMCL)) – via magnetic resonance spectroscopy (MRS) is often used as markers for disease progression in metabolic syndrome and muscle dystrophy(1). Spectroscopic evaluation of EMCL fat deposits is complicated by the additional susceptibility induced chemical shift (δ) that depends on the muscle fiber orientation with the main magnetic field(2), and can have significant impact on the accuracy of skeletal muscle lipid content. Some groups have proposed the use of diffusion tensor imaging (DTI) to calculate the pennation angle of the muscle fiber to determine δ and use this as prior information during MRS signal analysis(3). A systematic approach to minimize error associated with skeletal muscle lipid evaluation would be desirable.

Purpose: In this work, we propose and test the feasibility of using DTI and Dixon derived maximum intensity projection (MIP) fat images to guide the prescription of MRS voxel, and evaluate the impact of fiber orientation within same muscle group (Soleus) for quantifying relative concentrations of IMCL and EMCL.


Materials and Methods

Patient population: MRI and MRS images of the calf muscle in 10 subjects (7 male, 72.1 ± 3.8 years) with biopsy confirmed non-alcoholic fatty liver disease (NAFLD), were imaged with informed consent at 3.0T on a commercial MR imager (Ingenia, Philips Healthcare) using a 16 channel Transmit/Receive coil.

MR acquisition: Muscle lipid content of three voxels (~ 1 cm3), one positioned in tibialis anterior (TA) and two on the soleus (Sol) muscle groups was evaluated using Point RESolved Spectroscopy (PRESS) with the following parameters: TR/TE = 3s/30ms; 2048 samples; bandwidth of 2kHz; number of signals averaged = 32-48; spectra acquired at two echo times, 30 and 60 ms, were used to calculate the T2 relaxation time. Unsuppressed water spectra was used as internal reference to quantify relative lipids concentration. Further, Dixon images, and DTI (16 direction with b = 0 and 500, TR/TE = 89ms/49ms, FOV( AP/FH/RL=160/120/160mm, acquired voxel size = 2.3x2.3x2.3mm3) covering the calf muscle were acquired to guide spectroscopic voxel positioning.

Voxel Positioning: Orthogonal thin slice maximum intensity projections (MIPs) reconstructed from fat images acquired using Dixon were used to guide the placement of MRS voxels to avoid contamination of interstitial or fascial fat. The DTI color coded maps depicting fiber orientation were used to position spectroscopic voxels in the tibial TA region (blue, indicating principal diffusivity in the FH direction), and two voxels positioned in the Sol region with principal diffusivities in the RL (red) and AP (green) directions.

Data Analysis: The primary eigenvalue (λ1) and eigenvector (ε1) of muscle fibers within each MRS voxel, were used to calculate the pennation angle (Φ = acos(ε1 · [0,0,1]), and the induced chemical shift δ. The MRS data were exported into JMRUI(v5) to quantify the chemical shift (δ) of EMCL with respect to IMCL in TA and Sol. The amplitude of water estimated from unsuppressed water spectrum, IMCL, and EMCL were corrected by their respective T2 relaxation times (4). Bland-Altman (BA) analysis was performed to test the agreement of relative concentration of IMCL and EMCL estimated within same muscle group but with different degree of fiber orientation.

Results and Discussion

Positioning of MRS voxel using information from DIXON fat derived MIPs, color-coded principal diffusion direction maps, and fiber tracks is shown in Fig.1. The widest separation between IMCL and EMCL was seen in TA with fiber more parallel to B0, but there was substantial overlap between IMCL and EMCL peaks in the soleus (Sol) muscle group, if the muscle fiber orientation was in the AP direction compared to RL direction. There is substantial difference in IMCL content between the TA and Sol muscle groups (Table 1). BA analysis shows that unlike IMCL, EMCL of soleus muscle can have substantial bias in measurements obtained from voxels positioned at locations with different fiber orientations (Figure 2). The δ of EMCL in Sol muscle fibers oriented in the RL direction estimated in this study is different from published data and requires further investigation (5).

Conclusions

(i) Muscle fiber orientation can significantly affect the quantification of EMCL and knowledge about the muscle fiber orientation could help to minimize variability in longitudinal studies. (ii) Dixon and DTI acquisitions can provide fiber orientation information for positioning MRS voxels for accurate estimation of IMCL and EMCL lipids in skeletal muscle, and can aid in consistent positioning of spectroscopic voxels within similar fiber orientation group.

Acknowledgements

No acknowledgement found.

References

1. Virkamaki A, Korsheninnikova E, Seppälä-Lindroos A, et al. Intramyocellular lipid is associated with resistance to in vivo insulin actions on glucose uptake, antilipolysis, and early insulin signaling pathways in human skeletal muscle. Diabetes 2001;50(10):2337-2343.

2. Chu SCK, Xu Y, Balschi JA, Springer CS. Bulk Magnetic-Susceptibility Shifts in Nmr-Studies of Compartmentalized Samples - Use of Paramagnetic Reagents. Magnet Reson Med 1990;13(2):239-262. 3. Valaparla SK, Gao F, Daniele G, Abdul-Ghani M, Clarke GD. Fiber orientation measurements by diffusion tensor imaging improve hydrogen-1 magnetic resonance spectroscopy of intramyocellular lipids in human leg muscles. J Med Imaging (Bellingham) 2015;2(2):026002.

4. Wang L, Salibi N, Wu Y, Schweitzer ME, Regatte RR. Relaxation times of skeletal muscle metabolites at 7T. J Magn Reson Imaging 2009;29(6):1457-1464.

5. Boesch C, Slotboom J, Hoppeler H, Kreis R. In vivo determination of intra-myocellular lipids in human muscle by means of localized 1H-MR-spectroscopy. Magn Reson Med 1997;37(4):484-493.

Figures

Figure 1.Three MRS voxels from three fiber orientation (TA: blue, Sol: red and Sol: green) were carefully positioned in the area with least contamination from interstitial muscle fat under the guidance of fat MIPs (top left) and principle diffusion direction image (top right). The corresponding fiber track of these three voxels shown in bottom left. These three spectrum demonstrate the lipids peaks (bottom right). The black dashed line shows the IMCL peak at 1.3ppm and cyan dashed line shows 0.2ppm from IMCL peak. Separation of IMCL from EMCL is largest in TA group while most overlap in Sol with fiber orient more in AP direction.

Table1. The mean and standard deviation of relative concentration ratio of IMCL and EMCL normalized to Cr and water, chemical shift of EMCL with respect to IMCL, and pennation angle were calculated from 10 subjects: mean ± SD

Figure 2. Bland-Altman (top) and bar plot (bottom) analysis of the relative concentration of IMCL(left) and EMCL (right) with internally referenced to water in same soleus muscle group are shown. A negligible of bias between red and green fiber orientation. Substantial bias and larger variation was shown two fiber orientation (red and green) in same type of muscle group (soleus).

Proc. Intl. Soc. Mag. Reson. Med. 27 (2019)
2298