Publication Details
Ronak J. Dholakia, Fotios Drakopoulos, Chander Sadasivan, Xiangmin Jiao,, David J. Fiorella, Henry H. Woo, Baruch B. Lieber and Nikos Chrisochoides.
Published in IEEE International Symposium on Biomedical Imaging: From Nano to Macro, April, 2015
Abstract
A new class of stents called flow diverters are currently being implanted to treat large and complex-shaped brain aneurysms. These devices engender flowstasis within the aneurysm sac leading to ‘curative’ resorption of most aneurysms over a period of about one year. Due to this long ‘cure’ time and as some aneurysms do not get resolved, computational fluid dynamics (CFD) analysis can substantially contribute to treatment planning.Flow diverters generally comprise 48 to 96 wires of about 30 micron diameter helically inter-wound to form tubular meshes that are implanted across aneurysms as large as 25 mm in diameter in tortuous vascular fluid domains of around 100 mm (for computational purposes) in length. Mesh generation for such multi-scale domains can thus be extremely complex. Fig. 1B-1D and Fig. 2 show the current state-of-the-art procedure for vascular segmentation, stentgeometry generation and boolean combination to produce an appropriate mesh for computational analysis. We propose an automatic and high fidelity image-to-mesh (I2M) conversion scheme that replaces several of these steps (and the existing software pipeline) with resultant improvements in fidelity, accuracy and easy-of-use i.e., efficiency and labor.
