Congrats to Hengameh for her successful depth exam
April 16, 2012
Congratulations to Hengameh for her successful depth exam today! Further details below.
Title: FUNDAMENTALS OF GRAPH BASED APPROACHES FOR SKIN IMAGE ANALYSIS
Abstract: Malignant melanoma (MM) is one of the fastest growing cancers among the world’s white population. The mechanism of melanoma development is not fully understood. However, early diagnosis and complete removal of MM are critical for the containment of the malignancy. For early detection of MM, as noninvasive methods, dermatologists advocate: i) dermoscopic images allowing a better visualization of skin structures, and ii) total body photography for high-risk patients to identify newly-appearing, disappearing, and changing pigmented skin lesions (PSLs) by tracking them in the images collected during regular examinations. Recently, a considerable amount of research has been done toward preparing a computer aided diagnosis system for early detection of MM. Designing such a system would contain various parts including: skin segmentation, PSL detection/segmentation, PSL registration, and feature extraction and classification. In this report, we review previous works on these steps, with a specific focus on the PSL registration and graph based point matching algorithms.
Ph.D. Depth Examining Committee
Dr. Ghassan Hamarneh, Sr. Supervisor
Dr. Tim K. Lee, Supervisor
Dr. Stella Atkins, Examiner