Aaron Ward successfully defended his PhD Thesis
October 21, 2008
Congratulations to Aaron Ward who made a successful PhD defense today.
Thesis Title: Computational Intelligence Supporting Anatomical Shape Analysis and Computer-Aided Diagnosis
Medical imaging technologies allow the collection of remarkable, three-dimensional pictures of the inside of the body, and have led to novel, noninvasive means of disease diagnosis and treatment planning. However, the proliferation of medical imaging technology has resulted in the production of a huge amount of medical image data, increasing the demand on the radiology work force to a critical level. There is therefore an important need to provide a means of converting medical image data into useful, higher-level information that aids radiologists in performing diagnoses and developing treatment plans more efficiently and accurately. This dissertation focuses on the problem of processing medical image data in order to provide valuable, high-level information about the shapes of anatomical structures.
Such shape information is useful to medical researchers addressing questions about the relationship between anatomical shape and pathology, and is also useful to the development of computer-aided diagnosis systems based on shape information. This dissertation describes two studies relating shape to pathology of musculoskeletal structures in the human shoulder, and uses these studies to motivate research into further interesting questions in shape analysis. Techniques from computational intelligence, such as machine learning, graph matching, feature selection, manifold learning, optimization, and pattern recognition, are used in novel approaches to several components of an overall shape analysis pipeline designed to support medical research. We describe a machine learning-based approach to eliciting expert knowledge about feature saliency, for use in addressing the shape correspondence problem. We propose a novel approach to medial-based shape description that localizes shape deformations, and demonstrate a manifold learning-based approach to computing the basic building blocks of it and other medial-based shape descriptions. Finally, we propose a groupwise paradigm based on graph matching for the computation of a pruning order for the components of medial shape representations, in order to remove unwanted parts of the representation arising from noise. These contributions to the shape analysis pipeline enable more accurate and intuitively-understood shape analysis results, creating tools enabling medical researchers to gain further understanding into pathological processes in the human body.
Dr. Ghassan Hamarneh, Senior Supervisor
Dr. Stella Atkins, Supervisor
Dr. Richard Zhang, Supervisor
Dr. Mark Drew, Internal Examiner
Dr. Martin Styner, External Examiner
Dr. Greg Mori, Chair