An important part of object oriented image processing procedure is image segmentation which, is a method of separating an image into regions of interest. Our contributions are as follows: (i) We propose a novel method to apply the random walker method to segment non-scalar diffusion tensor magnetic resonance imaging (DT-MRI) data. We also extend the implementation by including a non-parametric probability density model to enable the segmentation of disconnected objects. (ii) We apply the random walker method to both second and fourth order DT-MR data and demonstrate the advantages of performing segmentations on higher order data. (iii) We use a DTI segmented atlas to investigate tissue discrimination in the brain, which serves to evaluate diffusion anisotropy measures. (iv) Finally, we propose a novel method for the segmentation of the breast from mammograms. The method automatically identifies intensity values that are used to define a probability distribution used in the segmentation.
Copyright is held by the author.
The author granted permission for the file to be printed, but not for the text to be copied and pasted.
Supervisor or Senior Supervisor
Thesis advisor: Atkins, Stella
Member of collection