Carotid artery bifurcation modelling from patient CT angiography and ultrasound technics

s Posters Cerebrovasc Dis 2013;35(suppl 2):1–77 47 P36 Carotid artery bifurcation modelling from patient CT angiography and ultrasound technics A. Santos, J. Tavares, L. Sousa, C. Castro, C. Antonio, R. Santos, P. Castro, E. Azevedo 1 Faculty of Medicine, University of Porto, Sao Joao Hospital Center; 2 Faculty of Engineering, University of Porto/ Porto, Portugal Background: Blood #ow simulation of the common carotid artery bifurcation is an undergoing research project with topography based on CT angiography and velocity information given by ultrasound techniques. Methods: *e segmentation algorithm is based on the echogenic characteristics of the le+ and right carotid arteries for the identi,cation of their boundaries, through all the slices acquired from a CT angiography. Initially, it requires the de,nition of two centroids, de,ning the region of the common carotid artery (CCA). *rough the application of morphological and segmentation operators, the structural information of the boundary walls of the CCA is saved and the algorithm proceeds with the calculation of two new centroids, based on the information resultant from the segmentation. *e calculated centroids will be used in the de,nition of a new region in the next slice, proceeding with the same calculations for the posterior segmentation. In the separation of the CCA into the internal and external arteries (ICA and ECA), the algorithm calculates the centroids for each feature, resulting in a total of four centroids that will be used in the posterior slice. *e calculations proceed till the last slice of the acquisition. Results: All the contours from each slice were saved and represented in a volume, obtaining a correct delineation of the bifurcation for both le+ and right carotid arteries in a total of 63 slices. Finally, the surface of the model of both le+ and right carotid arteries are built, by creating a Delaunay triangulation for all possible set of points considering from successive slices. A+er the meshing of the created volumes blood #ow simulation is performed using a developed ,nite element code. Conclusion: *e segmentation algorithm has proven to be highly e3cient allowing an accurate topography necessary for the correct interpretation of the simulated carotid artery bifurcation hemodynamics.