%0 Thesis
%T Detection and Alignment of Vascular Structures in Intravascular Ultrasound using Pattern Recognition Techniques
%A Marina Alberti
%E Simone Balocco
%E Petia Radeva
%D 2013
%I Ediciones Graficas Rey
%F Marina Alberti2013
%O MILAB
%O exported from refbase (http://refbase.cvc.uab.es/show.php?record=2215), last updated on Fri, 17 Dec 2021 13:48:05 +0100
%X In this thesis, several methods for the automatic analysis of Intravascular Ultrasound(IVUS) sequences are presented, aimed at assisting physicians in the diagnosis, the assessment of the intervention and the monitoring of the patients with coronary disease.The basis for the developed frameworks are machine learning, pattern recognition andimage processing techniques.First, a novel approach for the automatic detection of vascular bifurcations inIVUS is presented. The task is addressed as a binary classication problem (identifying bifurcation and non-bifurcation angular sectors in the sequence images). Themultiscale stacked sequential learning algorithm is applied, to take into account thespatial and temporal context in IVUS sequences, and the results are rened usinga-priori information about branching dimensions and geometry. The achieved performance is comparable to intra- and inter-observer variability.Then, we propose a novel method for the automatic non-rigid alignment of IVUSsequences of the same patient, acquired at dierent moments (before and after percutaneous coronary intervention, or at baseline and follow-up examinations). Themethod is based on the description of the morphological content of the vessel, obtained by extracting temporal morphological proles from the IVUS acquisitions, bymeans of methods for segmentation, characterization and detection in IVUS. A technique for non-rigid sequence alignment – the Dynamic Time Warping algorithm -is applied to the proles and adapted to the specic clinical problem. Two dierent robust strategies are proposed to address the partial overlapping between framesof corresponding sequences, and a regularization term is introduced to compensatefor possible errors in the prole extraction. The benets of the proposed strategyare demonstrated by extensive validation on synthetic and in-vivo data. The resultsshow the interest of the proposed non-linear alignment and the clinical value of themethod.Finally, a novel automatic approach for the extraction of the luminal border inIVUS images is presented. The method applies the multiscale stacked sequentiallearning algorithm and extends it to 2-D+T, in a rst classication phase (the identi-cation of lumen and non-lumen regions of the images), while an active contour modelis used in a second phase, to identify the lumen contour. The method is extendedto the longitudinal dimension of the sequences and it is validated on a challengingdata-set.
%9 theses
%9 Ph.D. thesis