A new framework for analysis of three-dimensional shape and architecture of human skeletal muscles from in vivo imaging data.

A new framework is presented for comprehensive analysis of the three-dimensional shape and architecture of human skeletal muscles from magnetic resonance and diffusion tensor imaging data. The framework comprises three key features: ) identification of points on the surface of and inside a muscle that have a correspondence to points on and inside another muscle, ) reconstruction of average muscle shape and average muscle fiber orientations, and ) utilization of data on between-muscle variation to visualize and make statistical inferences about changes or differences in muscle shape and architecture. The general use of the framework is demonstrated by its application to three case studies. Analysis of data obtained before and after 8 wk of strength training revealed there was little regional variation in hypertrophy of the vastus medialis and vastus lateralis and no systematic change in pennation angle. Analysis of passive muscle lengthening revealed heterogeneous changes in shape of the medial gastrocnemius and confirmed the ability of the methods to detect subtle changes in muscle fiber orientation. Analysis of the medial gastrocnemius of children with unilateral cerebral palsy showed that muscles in the more-affected limb were shorter, thinner, and less wide than muscles in the less-affected limb and had slightly more pennate muscle fibers in the central and proximal part of the muscle. Among other applications, the framework can be used to explore the mechanics of muscle contraction, investigate adaptations of muscle architecture, build anatomically realistic computational models of skeletal muscles, and compare muscle shape and architecture between species. Muscle architecture is conventionally measured using simple scalar metrics such as muscle volume and average fascicle lengths. Here, a new framework is proposed for analysis of complex changes in three-dimensional architecture of whole human muscles from magnetic resonance and diffusion tensor imaging data. The general use of the framework is demonstrated through visualization, quantification, and statistical analysis of the effect of strength training, passive lengthening and cerebral palsy on three-dimensional muscle shape and architecture.