Pre-operative Assistant Based on Data-Driven Approaches for Vascular Grafts Surgery

Short description of the business experiment

Aortic aneurysm surgeries require precise graft selection and sizing. In this context, where data-driven support tools for cardiovascular surgery remain limited, PANDORA aims to develop a Digital Twin integrating clinical data and numerical simulations. The University of Rennes will extract 3D anatomical models, INSA Lyon will test prosthetic materials, and RBF Morph will generate synthetic anatomies. LivGemini will create a surrogate model and integrate it into its software for preoperative planning. The project leverages High-Performance Computing (HPC) and AI-driven segmentation (U-Net) to enhance predictive modelling. Statistical shape modelling will be employed to generate synthetic datasets representing anatomical variability. The surrogate model will be derived through model order reduction and implemented as a Functional Mock-up Unit (FMU). By simulating patient-specific cases, PANDORA will enable accurate graft selection, reduce reintervention rates, and improve outcomes.

Organisations involved: 

End User: LivGemini srl 

ISV, HPC expert, Technology expert: RBF Morph srl

Aplication expert: University of Rennes and Institut National des Sciences Appliquées de Lyon