Texture analysis of biomedical images based on dual-tree complex wavelet transform

Abstract

In biomedical image processing, texture analysis is used in several applications from tracking structural changes of a tissue to segmentation of anatomical structures.The core of texture analysis is the extraction of texture features. Among several approaches that have been studied in the literature, a transform based method is chosen for texture feature extraction in this thesis. The transform that has been used is the dual-tree complex wavelet transform (DT-CWT) which is a recent wavelet method that has proved to give better results for texture analysis applications compared to previously suggested signal processing methods. The goal of this work is to introduce DT-CWT based texture analysis methods for biomedical images. These methods can be used both for 2-D and 3-D processing of biomedical images for several applications. This thesis explains three of them which are classification of tissues, segmentation of anatomical structures and estimation of volume texture orientations.