Wavelet analysis for multiresolution tissue characterization in intracoronary optical images

Abstract

The rupture of atherosclerotic plaques in coronaries is responsible for most acute coronary events. A new imaging modality called intravascular optical coherence tomography (IOCT) enables a detailed high resolution visualization of micro-structural changes of the in vivo arterial wall and provides an accurate representation of the thickness of a fibrous cap and a useful tool to assess the vulnerability in lipid-rich plaques. This article describes a fully automated method for tissue characterization in IOCT human coronary images. In this study the continous wavelet transform is used to analyze plaque tissue in IOCT frames in polar coordinates, enabling a quantification of fibrous cap thickness. In-vivo human coronary images were used in experimental tests for detecting fibrous cap automatically, where the results were compared with manual segmentations from the same expert in two different instants separated by 2 months.