Ultrahigh‐Field Multimodal MRI Assessment of Muscle Damage

Abstract

Magnetic resonance imaging (MRI) is a powerful tool to assess skeletal muscle damage both in injured athletes and patients with neuromuscular diseases. Alterations in muscle have been assessed with T2-weighted, 1 diffusion tensor, and sodium MRI. However, these sequences require a high signal-to-noise ratio. Considering these limitations, one could expect that measurements performed at ultrahigh-field strength (eg, 7T) would provide a more accurate assessment of skeletal muscle diffusivity properties and sodium concentration so that relevant information regarding changes in biochemical and structural tissues properties related to muscle damage may be obtained. A 33-year-old man (188 cm, 82 kg) volunteered to participate in the study after providing informed consent. Calf muscles were injured during a high-intensity electrostimulation exercise performed using a commercial device (Compex Performance, Djo global, France). Muscle MRI was performed using a whole-body scanner (Magnetom 7T, Siemens Healthcare, Erlangen, Germany). Five days after the damaging exercise, a multimodal imaging session was performed. Highresolution T1-weighted images (a gradient-recalled echo [GRE] with parameters previously detailed and a turbo spin-echo, repetition time [TR] = 3500 msec, echo time [TE] = 34 msec, in-plane field of view [FoV] = 180 × 180 mm2, matrix = 640 × 640, 15 slices, ST = 6 mm, gap between slice = 6 mm, bandwidth = 230 Hz/px, TA = 240 sec) were acquired and the corresponding images were used as anatomical reference images. A multiecho GRE sequence (TR = 43 msec, TEs = 3.57/8.51/13.60/18.69/23.78/28.87/33.96/39.05 msec, flip angle = 10 , in-plane FoV = 180 × 180 mm2, matrix = 448 × 448, 60 slices, ST = 3 mm, bandwidth = 370 Hz/px, TA = 456 sec) was used to produce T2 maps, while T2 maps were generated from T2-weighted images acquired with a segmented (15 segments) echo planar imaging sequence with TEs = 8/18/28/38/48/58 msec. Other acquisition parameters were as follows: FoV = 180 × 180 mm2; matrix = 192 × 192; TR = 4800 msec; 20 slices; ST = 6 mm; gap between slices = 6 mm, short-tau inversion-recovery for fat saturation; TA for each TE = 132 sec. Axial multidirectional diffusionweighted images with fat saturation were acquired to characterize diffusion tensor imaging derived parameters as previously described. A 3D radial sequence was used to obtain sodium images using the following parameters: TR = 110 msec, TE = 0.2 msec, FoV = 180 × 180 × 180 mm, voxel resolution = 2 × 2 × 2 mm, 10 radial spokes, bandwidth = 276 Hz/px, TA = 1102 sec. High-resolution T1-weighted images were used as reference images (GRE, TR = 13 msec, TE = 5 msec, in-plane FoV = 180 × 180 mm2, matrix = 512 × 512, 88 slices, ST = 2 mm, bandwidth = 260 Hz/px, TA = 587 sec). Sodium images were acquired in the presence of six calibration phantoms with different sodium concentrations (10, 30, and 50 mM) and placed within the FoV. Images were then corrected from B1 inhomogeneities using precalibrated B1 sensitivity maps from a phantom acquisition measured using the phase sensitive method. Additional investigations were performed at 1.5T (Avanto, Siemens Healthcare) in order to quantify muscle T2 values averaged on all slices 3 (D3), 5 (D5), 7 (D7), and 14 (D14) days after the damaging exercise, as previously described. Control muscle T2 values were determined 6 months (M6) after the damaging exercise. As illustrated in Fig. 1a, 1.5T measurements showed an increased T2 value in gastrocnemius medialis reaching the highest intensity 5 days after the damaging exercise and being resolved after a 6-month period. A slight T2 increase was observed in the soleus at D3 but no other changes could be seen in antagonist muscles. Ultrahigh-field MRI measurements confirmed the T2 increase (Fig. 1b) and extended our knowledge, showing concomitant local changes in T2 and sodium concentration (in comparison to unaffected muscles, eg, lateral compartment of the lower leg) in addition to transverse diffusivity (λ3) and fractional anisotropy (FA) as compared to published control values. In addition, diffusion parameters were increased in other areas, ie, FA and sodium concentration in the gastrocnemius lateralis, eigenvalues (λ1, λ2, λ3), mean diffusivity (MD), T2, and sodium concentration in the anterior compartment of the lower leg muscles (Fig. 1c). The 7T MRI measurements showed additional changes in the same muscle regarding diffusion properties and sodium concentration together with changes in agonist and antagonist muscles that were not visible from the 1.5T measurements. Studies have previously reported T2 changes resulting from damaging exercise and assigned them to inflammatory/edematous processes. Changes in DTI metrics have been reported in the early phase of muscle damage, illustrating subtle alterations of muscle diffusion properties. Sodium MRI can provide further information regarding biochemical changes in the damaged muscle, as recently shown in Achilles tendon and knee cartilage. From 1.5T MRI measurements, a large T2 increase was quantified in the gastrocnemius medialis muscle as a result of the damaging muscle stimulation. Interestingly, measurements performed at higher field (7T) confirmed the T2 changes but also illustrated changes in agonist and antagonist muscles that were not detected at 1.5T. Larger diffusion parameters values were measured in the anterior compartment of the lower leg muscles, thereby suggesting subtle changes in the diffusion properties. These changes could be