The distributed lambda (λ) model (DLM): a 3-D, finite-element muscle model based on Feldman's λ model; assessment of orofacial gestures

Machine-Learning based model order reduction of a biomechanical model of the human tongue

BACKGROUND AND OBJECTIVES

This paper presents the results of a Machine-Learning based Model Order Reduction (MOR) method applied to a complex 3D Finite Element (FE) biomechanical model of the human tongue, in order to create a Digital Twin Model (DTM) that enables real-time simulations. The DTM is designed for future inclusion in a computer assisted protocol for tongue surgery planning.

METHODS

The proposed method uses an "a posteriori" MOR that allows, from a limited number of simulations with the FE model, to predict in real time mechanical responses of the human tongue to muscle activations.

RESULTS

The MOR method is evaluated for simulations associated with separate single tongue muscle activations. It is shown to be able to account with a sub-millimetric spatial accuracy for the non-linear dynamical behavior of the tongue model observed in these simulations.

CONCLUSION

Further evaluations of the MOR method will include tongue movements induced by multiple muscle activations. At this stage our MOR method offers promising perspectives for the use of the tongue model in a clinical context to predict the impact of tongue surgery on tongue mobility. As a long term application, this DTM of the tongue could be used to predict the functional consequences of the surgery in terms of speech production and swallowing.

Quick compensatory mechanisms for tongue posture stabilization during speech production

The human tongue is atypical as a motor system since its movement is determined by deforming its soft tissues via muscles that are in large part embedded in it (muscular hydrostats). However, the neurophysiological mechanisms enabling fine tongue motor control are not well understood. We investigated sensorimotor control mechanisms of the tongue through a perturbation experiment. A mechanical perturbation was applied to the tongue during the articulation of three vowels (/i/, /e/, /ε/) under conditions of voicing, whispering, and posturing. Tongue movements were measured at three surface locations in the sagittal plane using electromagnetic articulography. We found that the displacement induced by the external force was quickly compensated for. Individual sensors did not return to their original positions but went toward a position on the original tongue contour for that vowel. The amplitude of compensatory response at each tongue site varied systematically according to the articulatory condition. A mathematical simulation that included reflex mechanisms suggested that the observed compensatory response can be attributed to a reflex mechanism, rather than passive tissue properties. The results provide evidence for the existence of quick compensatory mechanisms in the tongue that may be dependent on tunable reflexes. The tongue posture for vowels could be regulated in relation to the shape of the tongue contour, rather than to specific positions for individual tissue points.NEW & NOTEWORTHY This study presents evidence of quick compensatory mechanisms in tongue motor control for speech production. The tongue posture is controlled not in relation to a specific tongue position, but to the shape of the tongue contour to achieve specific speech sounds. Modulation of compensatory responses due to task demands and mathematical simulations support the idea that the quick compensatory response is driven by a reflex mechanism.

IMAGE-BASED SKELETAL MUSCLE COORDINATION: CASE STUDY ON A SUBJECT SPECIFIC FACIAL MIMIC SIMULATION

Facial muscle coordination is a fundamental mechanism for facial mimics and expressions. The understanding of this complex mechanism leads to better diagnosis and treatment of facial disorders like facial palsy or disfigurement. The objective of this work was to use magnetic resonance imaging (MRI) technique to characterize the activation behavior of facial muscles and then simulate their coordination mechanism using a subject specific finite element model. MRI data of lower head of a healthy subject were acquired in neutral and in the pronunciation of the sound [o] positions. Then, a finite element model was derived directly from acquired MRI images in neutral position. Transversely-isotropic, hyperelastic, quasi-incompressible behavior law was implemented for modeling facial muscles. The simulation to produce the pronunciation of the sound [o] was performed by the cumulative coordination between three pairs of facial mimic muscles (Zygomaticus Major (ZM), Levator Labii Superioris (LLS), Levator Anguli Oris (LAO)). Mean displacement amplitude showed a good agreement with a relative deviation of 15% between numerical outcome and MRI-based measurement when all three muscles are involved. This study elucidates, for the first time, the facial muscle coordination using in vivo data leading to improve the model understanding and simulation outcomes.

Atlas-Based Automatic Generation of Subject-Specific Finite Element Tongue Meshes

Generation of subject-specific 3D finite element (FE) models requires the processing of numerous medical images in order to precisely extract geometrical information about subject-specific anatomy. This processing remains extremely challenging. To overcome this difficulty, we present an automatic atlas-based method that generates subject-specific FE meshes via a 3D registration guided by Magnetic Resonance images. The method extracts a 3D transformation by registering the atlas' volume image to the subject's one, and establishes a one-to-one correspondence between the two volumes. The 3D transformation field deforms the atlas' mesh to generate the subject-specific FE mesh. To preserve the quality of the subject-specific mesh, a diffeomorphic non-rigid registration based on B-spline free-form deformations is used, which guarantees a non-folding and one-to-one transformation. Two evaluations of the method are provided. First, a publicly available CT-database is used to assess the capability to accurately capture the complexity of each subject-specific Lung's geometry. Second, FE tongue meshes are generated for two healthy volunteers and two patients suffering from tongue cancer using MR images. It is shown that the method generates an appropriate representation of the subject-specific geometry while preserving the quality of the FE meshes for subsequent FE analysis. To demonstrate the importance of our method in a clinical context, a subject-specific mesh is used to simulate tongue's biomechanical response to the activation of an important tongue muscle, before and after cancer surgery.