Sarah Keller1, Jerry Zhiyue Wang2, Amir Golsari3, Adam Gerhard1, Hendrick Kooijman4, Mathias Gelderblom3, and Jin Yamamura1
1Diagnostic and Interventional Radiology, University Medical Center Hamburg Eppendorf (UKE), Hamburg, Germany, 2Radiology, University of Texas Soutwestern Medical Center, Dallas, TX, United States, 3Neurology, University Medical Center Hamburg Eppendorf (UKE), Hamburg, Germany, 4MRI, Philips GmbH, Hamburg, Germany
Synopsis
MRI-Diffusion tensor imaging (DTI) based fiber tracking is an emerging
tool for the evaluation of alterations in the skeletal muscle architecture
caused by trauma and various inflammatory or hereditary diseases. It remains
still a matter of debate, whether dystrophic conditions of the skeletal muscle,
which are frequently associated with fatty infiltration, can be reliable
assessed by DTI, as previous studies showed the potential biasing effect of the
fat fraction (%FF) and the concomitant decrease of the signal to noise ratio
(SNR) [1, 2]. The
goal of this study was to analyze the DTI based apparent diffusion coefficient
(ADC), fractional anisotropy (FA) and tractography data in various conditions
of muscular disease, either with or without increase of the FF in comparison to
healthy controls on a 3T system.
Purpose
To analyze the DTI based apparent diffusion coefficient (ADC), fractional anisotropy (FA) and tractography data in various conditions of muscular disease, either with or without increase of the MFF in comparison to controls on a 3T system.
Methods
NINE PATIENTS (age 43±19.9y, m:f 5:4) suffering of hereditary muscular dystrophies (Duchenne, Becker and Limb Girdle
Muscular Dystrophy Type IIa (DMD, BMD, LGMD), Central Core Disease (CCD) and
Spinal Muscular Atrophy Type III (SMA)) and chronic inflammatory myopathy
(IMNM), with involvement of the limb girdle muscles and
dystrophic changes upon histology and 10 controls were included (Table 1). DTI-based diffusion and
tractography data were assessed in combination with the 2p-mDixon muscle fat fraction (MFF%) in
the rectus femoris (RF), semitendinosus (ST), biceps femoris (BF) and gracilis
(G) muscle in patients and controls.
MRI-scans were performed on a 3.0-Tesla (T) MRI
(Ingenia, Software Release 5.1.7, Philips, Best, Netherlands). Subjects were examined
in supine position, feet first using a 28-channel sensitivity encoding torso
array coil. After a gradient echo localizer
sequence (TR/TE 8.0/2.3ms; FOV: RL 450mm, AP 115mm, FH 450mm; slice thickness
15mm), an axial T2w turbo spin echo sequence with Dixon fat-water separation
(T2w mDixon TSE), (TR 8111ms; TE 80 ms; FoV 270 x 221 x 300mm with a voxel of
0.65x0.65x3 mm3, TSE factor 13) with post processed images (in-phase-
and water images) was performed. For the DTI sequence an axial fat-suppressed
multislice spin-echo single-shot echo planar imaging sequence (EPI) (TR/TE
2479/43ms, NSA 8, gradient directions 15,
b-values 0, 250 and 500s/mm2, slice thickness 6mm, 1.5mm
in-plane resolution) was used. This was followed by an axial water-fat 3D FFE T1-weighted Dixon Sequence (T1w
mDixon) with following parameters: TR/TE/ΔTE
4.4 ms/1.18ms/2.6ms, flip angle 3°, FoV 250x250x250mm with voxel 2.0x1.0x0.5 mm3.
Post processing produced in- and opposed- phase water and fat images.
IMAGE ANALYSIS The ROI localization was chosen for the maximal transverse crosssectional area of
each muscle and ROIs were drawn circumferentially around the muscle using a
freehand technique in the RF, ST, BF and G muscle. DTI datasets were analysed using the
manufacturers software (FiberTrak;
Philips Healthcare). For the fiber tracking analysis, a single ROI line
propagation technique with the following parameters was used: FA threshold
0.15, direction threshold 6.75°. Quantitative ADC and FA-values were generated.
Muscle fat fraction% (FF) was obtained using two-point Dixon-based MRI
(2pt-MRIDIXON) with chemical shift selective reconstruction of fat- and
watersignal [3,
4]. Three ROIs,
covering the whole muscle were drawn in the (SIFAT) and water-only
(SIWATER) image of the proximal, middle and distal part of the RF,
ST, BF and G muscle using Osirix (version 6.5). The Muscle Fat Fraction (MFF) was
calculated using the algorithm:
MFF% = meanSIFAT/ (meanSIFAT+ meanSIWATER) x 100
STATISTICAL ANALYSIS
was performed using GraphPad Prism 6.0f (GraphPad Software Inc., La Jolla,
California, USA). Correlation and significance of DTI values from controls and
patients was tested by the Student’s T-test and parametric Pearson correlation
analyses. The dif-ference was considered statistically significant if the
significance level α = 0.05 was reached.
Results
Fiber tractography of affected muscles groups resulted in a significant shortened fiber track lengths, which was predominantly evident in the RF and ST muscle (RF: 28.5±3.3 mm vs. 76.5±7.1 mm; p=0.002; ST: 34.4±5.9 mm vs. 81.8±10.2 mm; p=0.004). The mean ADC was
significantly reduced in patient muscle and accompanied by a concomitant
increase of the FA (ST: ADC 1.055±0.479 mm2/s*10-3, FA
0.498±0.145; G: ADC 1.056±0.357 mm2/s*10-3, FA 0.532±0.126) (Figure 2). MFF resulted to be elevated in
patient muscles (median 11.6%; range 2.9-91.8%) versus control (median 4.8%;
range 1.5-9.6%). Using Pearson analysis there was an inverse
correlation of the RF, ST, BF and G muscle ADC and FA (rs= -0.91; p<0.0001),
and a good correlation of the ADC and fiber length (rs= 0.4551; p=0.005). FA
and ADC data of controls and patients correlated significantly to the MFF (FA
r= 0.5335; 95%CI 0.248-0.7335; p=0.0008; ADC r=-0.589; 95%CI -0.749-0.2822; p=0.0004) (Figure 3).
Automatic calculation of the fiber count revealed a higher variability in fiber
count and a tendency of fiber reduction in affected muscles, despite not
significant.
Discussion/Conclusion
Our data
confirm previous studies, presuming a biasing effect of the MFF% and low SNR on
DTI parameters, reflected by an overestimation of the FA and an underestimation
of the ADC in highly dystrophic muscles. These confounding factors are not of
major concern in muscle disease without concomitant increase of the fatty
infiltration and could be circumvented by a ROI selective analysis of muscle
tissue without overt fatty infiltration and a high SNR cut-off value.
Acknowledgements
Many thanks to the Department of Neurology for the clinical support and patients contact.
There is no actual or potential conflict of interest in relation to this abstract.
References
1. Froeling, M., et al., DTI of human skeletal muscle: the effects of diffusion encoding
parameters, signal-to-noise ratio and T2 on tensor indices and fiber tracts.
NMR Biomed, 2013. 26(11): p.
1339-52.
2. Hooijmans, M.T.,
et al., Evaluation of skeletal muscle DTI
in patients with duchenne muscular dystrophy. NMR Biomed, 2015.
3. Dixon, W.T., Simple proton spectroscopic imaging.
Radiology, 1984. 153(1): p. 189-94.
4. Ma, J., Dixon techniques for water and fat imaging.
J Magn Reson Imaging, 2008. 28(3):
p. 543-58.