Quantitative angiography: densitometry guided edge detection using mean field annealing

Abstract

Coronary angiograms are degraded by blur and noise from the imaging system. Previously the authors proposed a method of deblurring and edge-preserving smoothing using Mean Field Annealing. Here, they show the results of the application of DMFA technique on clinical cineangiography. Also, they propose a new edge detection method which uses densitometric information as well as gradient information in the estimation of the edges. These techniques are applied to 13 synthetic vessel images and several clinical angiograms and the results are compared to those of the conventional gradient-based techniques. These methods may improve QCA measures of vessel diameters.<>