Muscle–fat MRI: 1.5 tesla and 3.0 tesla versus histology


  • Andrew C. Smith is supported by the NIH-funded Training Program in the Neurobiology of Movement and Rehabilitation Sciences at Northwestern University, supported by the Eunice Kennedy Shriver National Institute Of Child Health & Human Development Grant T32 HD057845. Funding for the study was provided by internal sources and James Elliott is supported by the National Center for Research Resources (NCRR) and the Northwestern University Clinical and Translational Sciences Institute (NUCATS), NIH KL2 RR025740


Introduction: We evaluated muscle/fat fraction (MFF) accuracy and reliability measured with an MR imaging technique at 1.5 Tesla (T) and 3.0T scanner strengths, using biopsy as reference. Methods: MRI was performed on muscle samples from pig and rabbit species (n = 8) at 1.5T and 3.0T. A chemical shift based 2-point Dixon method was used, collecting in-phase and out-of-phase data for fat/water of muscle samples. Values were compared with MFFs calculated from histology. Results: No significant difference was found between 1.5T and 3.0T (P values = 0.41–0.96), or between histology and imaging (P = 0.83) for any muscle tested. Conclusions: Results suggest that a 2-point Dixon fat/water separation MRI technique may provide reliable quantification of MFFs at varying field strengths across different animal species, and consistency was established with biopsy. The results set a foundation for larger scale investigation of quantifying muscle fat in neuromuscular disorders. Muscle Nerve 50:170–176, 2014