DOI

https://doi.org/10.25772/SJVR-9R50

Defense Date

2010

Document Type

Thesis

Degree Name

Master of Science

Department

Electrical Engineering

First Advisor

Yuichi Motai

Abstract

Computed tomographic (CT) colonography, or virtual colonoscopy, is a promising technique for screening colorectal cancers by use of CT scans of the colon. Current CT technology allows a single image set of the colon to be acquired in 10-20 seconds, which translates into an easier, more comfortable examination than is available with other screening tests. Currently, however, interpretation of an entire CT colonography examination is time-consuming, and the reader performance for polyp detection varies substantially. To overcome these difficulties while providing a high detection performance of polyps, researchers are developing computer-aided detection (CAD) schemes that automatically detect suspicious lesions in CT colonography images. The overall goal of this study is to achieve a high performance in the detection of polyps on CT colonographic images by effectively incorporating an appearance-based object recognition approaches into a model-based CAD scheme. Our studies are focused in developing a fast kernel feature analysis that can efficiently differentiate polyps from false positives and thus improve the detection performance of polyps. We have developed a novel method of selecting kernel functions that are appropriate for the given data set and then use their linear combination in the construction of Kernel Gram matrix which can then used for efficient reconstruction of feature space. The main contribution of this work lies in providing a Composite kernel Matrix that involves appearance-based approach to improve kernel feature analysis for the classification of texture-based features. We evaluated our proposed kernel feature analysis on texture-based features that were extracted from the polyp candidates generated by our shape-based CAD scheme.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

May 2010

Share

COinS