ESR 15: Bayesian Geometrical Uncertainty Quantification for Soft Tissue Biomechanics Simulations
Early Stage Researcher Vasileios Krokos
Host institution: Synopsys – Simpleware Product Group (UK)
Supervisor: Senior Lecturer Pierre Kerfriden, Cardiff University(UK)
Co-supervisor: Viet Bui Xuan, R&D Manager of the Simpleware Product Group at Synopsys
Clinical expertise: Prof. Frank Hertel, Centre Hospitalier de Luxembourg (UL, LU)
Further institution involved: Mines ParisTech (FR)
Objectives
Develop the first systematic approaches to quantifying uncertainty due to the geometry of the organ in biomechanics simulations. Discretise the body using NURBS and use the control points and their weights as geometrical descriptors of the shape of the organ. This will allow us to quantify the uncertainties due to the lack of knowledge of the geometry of the organ. Collaboration with ESR08 (membrane cutting) and ESR12 (bulk cutting simulations based on NURBS approximations). The work will address the automated creation of simulation models for biomechanics, at the geometrical level, by empowering the definition of patient-specific geometrical descriptions of organs for biomedical simulations.).
Expected Results
A toolbox to quantify uncertainties in biomechanics simulations.
Planned research stays
- Synopsys – Simpleware Product Group (14 months): Training in using segmentation toolbox and handle patient specific geometric data.
- Mines ParisTech (9 months): Develop the theoretical aspects of the machine learning approach
- Synopsys – Simpleware Product Group (13 months): Finalize deliverables, thesis writing.
