The object of this thesis is to describe tissue classification software that was developed specifically for the identification of cerebral white matter lesions found in patients with multiple sclerosis (MS). Multiple sclerosis is a debilitating disease of the central nervous system (CNS) which results in a pathology observable macroscopically by magnetic resonance imaging (MRI). Lesion volumes are used as a surrogate of disease progression in clinical trials for MS treatment. The availability of accurate, reliable and reproducible measurements is invaluable in the advancement of patient care and research into this disease. This thesis documents a comprehensive approach to achieve such results which to date has proven to be challenging. A novel fully automated Bayesian classifier was developed which utilizes a variety of techniques to more accurately model brain tissue, and performs lesion identification at a level comparable to human experts. A new validation model that affords a better estimation of ground truth is introduced, providing a better means to measure classification accuracy. It is hoped that this document will be practical in nature, so that people who continue this work will have a step upon which to start.
Identifer | oai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:QMM.82231 |
Date | January 2004 |
Creators | Francis, Simon J. |
Publisher | McGill University |
Source Sets | Library and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada |
Language | English |
Detected Language | English |
Type | Electronic Thesis or Dissertation |
Format | application/pdf |
Coverage | Master of Science (Division of Neuroscience.) |
Rights | All items in eScholarship@McGill are protected by copyright with all rights reserved unless otherwise indicated. |
Relation | alephsysno: 002199584, proquestno: AAIMR12442, Theses scanned by UMI/ProQuest. |
Page generated in 0.002 seconds