Meshing Arterial Networks From Manually Extracted Centerlines

Méghane Decroocq (1, 2, 3, 4, 5) , Carole Frindel (1, 6) , Makoto Ohta (4, 3) , Guillaume Lavoué (2)

1 MYRIAD - Modeling & analysis for medical imaging and Diagnosis
2 Origami - Origami
3 ELyTMaX
4 IFS - Institute of Fluid Sciences [Sendai]
5 BME - Graduate School of Biomedical Engineering
6 CREATIS - Centre de Recherche en Acquisition et Traitement de l'Image pour la Santé

In large vascular networks with small vessels, the low resolution of medical images causes the segmentation algorithms to fail. Simplified models based on the vessel centerline are convenient to represent complex networks, but are associated with a loss of information. In this work, we propose a pipeline which overcomes the limitations of manually extracted centerlines (low sampling, noise) to reconstruct an anatomically accurate surface or volume mesh suitable for numerical simulation.

Domaines: Géométrie algorithmique [cs.CG]; Imagerie médicale; Modélisation et simulation

Méghane Decroocq, Carole Frindel, Makoto Ohta, Guillaume Lavoué. Meshing Arterial Networks From Manually Extracted Centerlines. International Conference on Flow Dynamics (ICFD 2020), Oct 2020, Sendai, Japan. ⟨hal-03772159⟩