MIAL

Congratulations to Ahmed for passing his depth exam today

June 15, 2011

Details of his PhD Depth Examination:

TitleA SURVEY OF SEGMENTATION TECHNIQUES FOR DYNAMIC EMISSION TOMOGRAPHY IMAGES

Abstract
In this report, we will review a number of automatic and semi-automatic image segmentation methods proposed to identify functional regions of interest in dynamic emission tomography images (e.g. dynamic positron emission tomography and dynamic single photon emission computed tomography). We identified three main types of segmentation algorithms used: 1) matrix decomposition such as PCA, FA, ICA, and NMF, 2) clustering, and 3) segmentation with emphasis on spatial information such as Markov random field and levelsets. We will review the mathematical formulation for each type followed by the validation procedure used. We will show how prior knowledge had been encoded into these algorithms to overcome the low signal-to-noise ratio and partial volume effect often encountered in emission tomography images.

Ph.D. Depth Examining Committee
Dr. Torsten Moller, Sr. Supervisor
Dr. Ghassan Hamarneh, Sr. Supervisor
Dr. Vesna Sossi, Examiner
Dr. Petra Berenbrink, Chair


 


Site by Faculty of Applied Science web team, Simon Fraser University