RESUMEN
PURPOSE: To introduce a novel method for skeletal muscle water T2 determination in fat-infiltrated tissues, using a tri-exponential fit of the global muscle signal decay. MATERIALS AND METHODS: In all, 48 patients with various neuromuscular diseases were retrospectively selected and their thigh muscles analyzed. Each patient was imaged using a multispin-echo (MSME) sequence with a 17-echo train. The transmit field (B1+) inhomogeneities were evaluated using the actual flip angle imaging method toward voxel sorting. Muscle water T2 was quantified using a tri-exponential signal decay model. The difference between water T2 of voxels within the same muscle but having different fat ratio was analyzed using nonparametric statistical tests. In addition, we evaluated the correlation between fat ratio and T2 values obtained using both a mono- and tri-exponential approach. RESULTS: The results showed that muscle water T2 values obtained using a tri-exponential approach combined with B1+ map-based voxel sorting were independent of the fat infiltration degree inside the muscle (R(2) < 0.03). This was not the case using the mono-exponential model, which measured different T2s between voxels of the same muscle but with various fat ratio (R(2) > 0.67; P < 10e(-4) ). CONCLUSION: The tri-exponential model is an accurate tool to monitor muscle tissue disease activity devoid of bias introduced by fat infiltration.