The main objective of this thesis is to develop a reconstruction system: on the one hand, to fully control the process using graphical modeling methods; on the other hand, to supplement and process a much larger dataset that can guide the reconstruction process based on the medical data collected. To this end, a sophisticated reconstruction method will be developed, utilizing topological data structures enriched with medical data and thus better suited to a detailed mathematical analysis of the reconstructed models.

More specifically, the 3D modeling work planned for this project falls under several themes, involving the development of a process dedicated to 3D model reconstruction, including:

• A method for detailed reconstruction based on data provided by the medical community;
• A modeling and visualization system with multiple levels of detail;
• A calculation of multimodal parameters that provides structured data as input for both the visualization model and the mathematical evaluation model.

Data with varying resolutions and slice thicknesses pose a significant challenge for the reconstruction process, particularly when creating a “dense” model (super-resolution, etc.) and for achieving greater precision in data extraction, which is not possible with current “sparse” systems.
Such reconstruction requires preliminary studies and geometric calculations in order to reconstruct a 3D shape in a reasonable manner. Shape properties such as boundaries, smooth regions, and local connectivity can be inferred with high confidence from the segmentation stage of 2D MRI sequences to enrich an incomplete 3D model.