Automatic prediction of tongue muscle activations using a finite element model Ian Stavness a,c,n , John E. Lloyd b , Sidney Fels b a Department of Bioengineering, Clark Center, Room S221, Stanford University, Mail Code 5448, 318 Campus Drive, Stanford, CA 94305, USA b Department of Electrical and Computer Engineering, University of British Columbia, Canada c Department of Computer Science, University of Saskatchewan, 110 Science Place, Saskatoon, SK S7N 5C9 Canada, Canada article info Article history: Accepted 22 August 2012 Keywords: Tongue Muscle function Finite-element methods Forward-dynamics tracking simulation Muscular-hydrostat modeling abstract Computational modeling has improved our understanding of how muscle forces are coordinated to generate movement in musculoskeletal systems. Muscular-hydrostat systems, such as the human tongue, involve very different biomechanics than musculoskeletal systems, and modeling efforts to date have been limited by the high computational complexity of representing continuum-mechanics. In this study, we developed a computationally efficient tracking-based algorithm for prediction of muscle activations during dynamic 3D finite element simulations. The formulation uses a local quadratic-programming problem at each simulation time-step to find a set of muscle activations that generated target deformations and movements in finite element muscular-hydrostat models. We applied the technique to a 3D finite element tongue model for protrusive and bending movements. Predicted muscle activations were consistent with experimental recordings of tongue strain and electromyography. Upward tongue bending was achieved by recruitment of the superior longitudinal sheath muscle, which is consistent with muscular-hydrostat theory. Lateral tongue bending, however, required recruitment of contralateral transverse and vertical muscles in addition to the ipsilateral margins of the superior longitudinal muscle, which is a new proposition for tongue muscle coordina- tion. Our simulation framework provides a new computational tool for systematic analysis of muscle forces in continuum-mechanics models that is complementary to experimental data and shows promise for eliciting a deeper understanding of human tongue function. & 2012 Elsevier Ltd. All rights reserved. 1. Introduction Muscular-hydrostats are complex biomechanical systems found in nature as tongues, tentacles, and trunks. These solely muscular organs are biomechanically distinct because they generate deformation, articulation, and movement without the mechanical support of a rigid skeletal structure (Kier and Smith, 1985). Instead, mechanical support is thought to be achieved by the incompressible nature of muscle tissue, as well as synergistic activation of orthogonally oriented muscle fibers (Gilbert et al., 2007). Muscular-hydrostats share common architectural features for which a biomechanical explanation has been posited: longitudinal muscle fibers are arranged at the peripheral margins of the organ to create bending movements through unilateral recruitment (see Kier and Smith, 1985, Fig. 7), and muscle fibers are arranged perpendicular to the longitudinal direction (either transverse, vertical, radial or circular) to create longitudinal elongation by reducing the transverse cross-sectional area of the organ (see Kier and Smith, 1985, Fig. 5). While muscle architecture appears linked to biomechanics in muscular-hydrostat systems, the way in which muscle forces are coordinated to control movement and shape deformation remains unclear. Computational modeling has been widely used to analyze musculo-skeletal biomechanics; however, modeling muscular- hydrostat systems entails a number of specific challenges as compared to musculoskeletal systems. Musculoskeletal models typically represent muscles as massless springs and neglect the effects of soft-tissues (Pai, 2010), whereas muscular-hydrostat models require a continuum mechanics approach to represent soft-tissue deformations. Finite-element (FE) methods are com- monly used to represent the large number of degrees-of-freedom associated muscular-hydrostat movements, including bending, twisting, grooving, elongating, and shortening. A number of macro-scale FE models have been developed to simulate 3D deformations of tentacles (Yekutieli et al., 2005; Liang et al., 2006) and tongues (Wilhelms-Tricarico, 1995; Buchaillard Contents lists available at SciVerse ScienceDirect journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com Journal of Biomechanics 0021-9290/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jbiomech.2012.08.031 n Corresponding author. Tel.: þ1 306 966 7995; fax: þ1 306 966 4884. E-mail addresses: stavness@gmail.com (I. Stavness), lloyd@cs.ubc.ca (J.E. Lloyd), ssfels@ece.ubc.ca (S. Fels). URL: http://www.cs.usask.ca/faculty/stavness (I. Stavness). Journal of Biomechanics 45 (2012) 2841–2848