Combined magnetic resonance and diffusion tensor imaging analyses provide a powerful tool for in vivo assessment of deformation along human muscle fibers

摘要

Muscle fiber direction strain provides invaluable information for characterizing muscle function. However, methods to study this for human muscles in vivo are lacking. Using magnetic resonance (MR) imaging based deformation analyses and diffusion tensor (DT) imaging based tractography combined, we aimed to assess muscle fiber direction local tissue deformations within the human medial gastrocnemius (GM) muscle. Healthy female subjects (n=5, age = 27 +/- 1 years) were positioned prone within the MR scanner in a relaxed state with the ankle angle fixed at 90 degrees. The knee was brought to flexion (140.8 +/- 3.0 degrees) (undeformed state). Sets of 3D high resolution MR, and DT images were acquired. This protocol was repeated at extended knee joint position (177.0 +/- 1.0 degrees) (deformed state). Tractography and Demons nonrigid registration algorithm was utilized to calculate local deformations along muscle fascicles. Undeformed state images were also transformed by a synthetic rigid body motion to calculate strain errors. Mean strain errors were significantly smaller then mean fiber direction strains (lengthening: 0.2 +/- 0.1% vs. 8.7 +/- 8.5%; shortening: 3.3 +/- 0.9% vs. 7.5 +/- 4.6%). Shortening and lengthening (up to 23.3% and 116.7%, respectively) occurs simultaneously along individual fascicles despite imposed GM lengthening. Along fiber shear strains confirm the presence of much shearing between fascicles. Mean fiber direction strains of different tracts also show non-uniform distribution. Inhomogeneity of fiber strain indicates epimuscular myofascial force transmission. We conclude that MR and DT imaging analyses combined provide a powerful tool for quantifying deformation along human muscle fibers in vivo. This can help substantially achieving a better understanding of normal and pathological muscle function and mechanisms of treatment techniques. (C) 2016 Elsevier Ltd. All rights reserved.