PhD defence by Torkan Gholamalizadeh
![Portrait of Torkan Gholamalizadeh Portrait of Torkan Gholamalizadeh](/forsidebegivenheder/begivenheder-2022/phd-defence-by-torkan-gholamalizadeh/torkanedit.png)
The defence will be held at 3Shape and online:
https://ucph-ku.zoom.us/j/65962214494
3Shape requires that you register as a visitor to enter. Please inform Peter Søndergaard (Peter.Soendergaard@3shape.com) of your participation by email before 23 November if you wish to avoid potential queues.
Title
Computational modeling and analysis of tooth movement
Abstract
Computational models obtained from medical data, such as finite-element (FE) methods, are beneficial for predicting values when measuring them directly in vivo is invasive, costly, or difficult. However, developing an accurate patient-specific FE model of a human jaw is non-trivial and involves time-consuming and labor-intensive segmentation and re-meshing processes. Therefore, almost all studies in the field are limited to single-model analysis, which raises the question of the generalizability of their results to a larger population with significant geometrical variations. In this thesis, we investigate the influence of geometrical variations on teeth movements by conducting intra- and inter-patient analyses to show that a combination of two clinical biomarkers can affect tooth displacements. Moreover, we develop a pipeline that nearly automates the model generation process, making it attainable to develop the largest dataset of clinically-validated FE models of human jaws. For the reproducibility of our work, we share our research materials, including the segmented geometries, quality conformal volumetric meshes, and the pipeline for generating them, as an open-access GitHub repository named as Open-Full-Jaw (https://github.com/diku-dk/Open-Full-Jaw). We believe this unique repository, including a high-geometrical variation, will pave the way for future population studies by providing reproducible FE models and allowing researchers to easily produce simulation-ready models for their data without spending extensive time on conventional approaches.
Supervisor
Principal Supervisor Kenny Erleben
Assessment Committee
Associate Professor Jens Petersen, Department of Computer Science
Professor John Rasmussen, Aalborg University, Dept. Materials and Production
Associate Professor Line Katrine Harder Clemmensen, DTU Compute
Moderator of defence: Associate Professor Melanie Ganz-Benjaminsen, Department of Computer Science
For a digital copy of the thesis, please visit https://di.ku.dk/english/research/phd/