A Component-Based Architechture for Medical Image Processing

Abstract

<p>This thesis discusses the development of a component-based architecture for the processing of medical imaging data. Such an architecture enables a wider range of analysis techniques as applied to Magnetic Resonance Imaging (MRI) data than is currently the norm. Differences in image segmentation results on images obtained using 1.5T vs. 3T MRI machines were investigated using the architecture and tools presented in this thesis. This thesis presents an analysis of the design, implementation, and sample results of such an architecture processing actual MRI data. Studies on images acquired from a test object (phantom) and the brain of a human subject reveal that a component-based architecture can assist in developing and testing novel algorithms in the computational laboratory, and serve as a platform for examining the effects of various image processing procedures in a radiology suite.</p>