Rekonstrukce 3D modelu aorty z CT snímků

Abstract

This work deals with the reconstruction of a 3D model of the aorta from noisy CT images. The walls of the aorta are reconstructed based on Bayesian estimation of the parameters of the NURBS representations of the boundary curves of the aortic regions in each of the CT images. Through the prior distribution of the parameters of these curves the natural smoothness of the boundary curves and their similarity in neighboring layers is reflected in the resulting model. Finding the final estimate involves determining the argument of the maxima of the posterior density function, which combines information from the observed data with information propagated by the prior distribution. For this purpose, a sampling algorithm based on Markov Chain Monte Carlo methods is used.