Analysis of Doppler blood ﬂow velocity in carotid arteries for the detection of atherosclerotic plaques
Tuesday 4th October 2011
In this thesis, a technique for nearly automated detection of Doppler ﬂow velocity proﬁle is developed. From this, it is possible to estimate the period of the velocity waveform throughout an innovative iterative method, based on Fourier smoothing and non linear least squares. The estimate of the period enables the reconstruction of a function representing the velocity of the red blood cells, along the common carotid artery and at different distances from the bifurcation. Considering these functional data, it is possible to reduce the dimension of the problem performing a functional principal component analysis. The aim is that of exploring the blood ﬂow along the common carotid artery, before the plaque is reached, and searching for features of the curve that indicate the presence of the plaque downstream through a linear discriminant analysis. This permits to compare the blood velocity ﬁeld in healthy people with the blood ﬂow of stenotic patients, who will undergo TEA surgery.