Computational Texture Features and Mechanical Anisotropy Reflect Structure and Composition of Dystrophic Canine Skeletal Muscle

In Golden Retriever Muscular Dystrophy (GRMD) dogs, a relevant animal model of human Duchenne muscular dystrophy (DMD), skeletal muscles progressively inflame, necrose, and undergo fibrous and fatty deposition. These degenerative changes alter the composition and structure of muscle, which affects the degree of directional variation, or anisotropy, in tissue mechanical properties. Further, such degenerative changes alter the spatial distribution of mechanical property across muscles. In this work we investigate the relevance of mechanical anisotropy and texture, estimated using Viscoelastic Response (VisR) ultrasound, to tracking progressive dystrophic muscle degeneration in GRMD dogs, in vivo. In a year-long study of 20 dogs, degree of shear elastic anisotropy (DoA), was significantly higher in the GRMD semitendinosus (ST) muscle compared to control ST (GRMD: 1.23 +- 0.25, Control 1.13 +- 0.27, p<;0.05, Wilcoxon). Also, VisR-derived textural entropy and heterogeneity indices were significantly different at 6, 9, and 12 months of age in the rectus femoris (RF) muscle, and at 12 months of age in the vastus lateralis (VL) muscle, of GRMD versus control. These results suggest that VisR-derived anisotropy estimates and texture features may be clinically relevant biomarkers for longitudinally tracking dystrophic muscle degeneration, in vivo.