Maxime Calka PhD on 06/08/2023

Phd defense of Maxime CALKA from TIMC BIOMECA team on 06/08/2023 :
" Finite element biomechanical modeling of the tongue: evaluation, speech production, and perspectives for computer-assisted tongue surgery "
-
Place: Amphithéâtre Boucherle, Facultés de Médecine & Pharmacie de l'Université Grenoble Alpes, site Santé, La Tronche
-
Broadcast: https://univ-grenoble-alpes-fr.zoom.us/j/94412939140?pwd=dnNKY0RIM3RKSTNyaGhHaHNaejIxZz09
Jury :
- Yohan PAYAN, Directeur de Recherche CNRS, TIMC, Supervisor
- Pascal PERRIER, Professeur des Universités, Université Grenoble Alpes, GIPSA-Lab, co-Supervisor
- Maud MARCHAL, Professeure des Universités, Université de Rennes, IRISA, reporter
- Aline BEL-BRUNON, Maitresse de conférences HDR, Université de Lyon, INSA Lyon, LaMCoS, reporter
- Florence BERTAILS-DESCOUBES, Directrice de Recherche, INRIA Grenoble, ELAN, examiner
- Emmanuel PROMAYON, Professeur des Universités, Université Grenoble Alpes, TIMC, examiner
- Christelle BOICHON-GRIVOT, Ingénieure de recherche (PhD), ANSYS France, examiner
- Muriel LALAIN, Chargée de recherche, CNRS, Laboratoire Parole et Langage, examiner
Keywords
Biomechanics, Finite Element Method, CAMI, Model Order Reduction, Tongue, Lingual surgery, Speech production
Abstract
The long-term goal of this thesis is to develop a tool to assist in the surgical planning of lingual surgery. This tool is based on a 3D biomechanical model of the tongue allowing to evaluate the functional consequences of a surgery on the lingual mobility. Our work attempts to lay the theoretical and methodological foundations for the creation of such a tool.
The first part of this thesis was devoted to the creation of a 3D biomechanical finite element (FE) model of the oral cavity. This model has a refined and converged mesh and a precise definition of the anatomical structures leading to more realistic and quality simulations.
The second part of the thesis focuses on obtaining a clinician-usable model of the language in interactive time using model reduction methods. The results of a model reduction method "DRB" based on machine learning applied to our 3D biomechanical model of the human tongue.
The last part consists of preliminary work concerning the creation of a prototype of pre-operative simulator.