Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.537240
Title: Automated extraction of biomarkers for Alzheimer's disease from brain magnetic resonance images
Author: Wolz, Robin
Awarding Body: Imperial College London
Current Institution: Imperial College London
Date of Award: 2011
Availability of Full Text:
Access through EThOS:
Full text unavailable from EThOS. Please try the link below.
Access through Institution:
Abstract:
In this work, different techniques for the automated extraction of biomarkers for Alzheimer's disease (AD) from brain magnetic resonance imaging (MRI) are proposed. The described work forms part of PredictAD (www.predictad.eu), a joined European research project aiming at the identification of a unified biomarker for AD combining different clinical and imaging measurements. Two different approaches are followed in this thesis towards the extraction of MRI-based biomarkers: (I) the extraction of traditional morphological biomarkers based on neuronatomical structures and (II) the extraction of data-driven biomarkers applying machine-learning techniques. A novel method for a unified and automated estimation of structural volumes and volume changes is proposed. Furthermore, a new technique that allows the low-dimensional representation of a high-dimensional image population for data analysis and visualization is described. All presented methods are evaluated on images from the Alzheimer's Disease Neuroimaging Initiative (ADNI), providing a large and diverse clinical database. A rigorous evaluation of the power of all identified biomarkers to discriminate between clinical subject groups is presented. In addition, the agreement of automatically derived volumes with reference labels as well as the power of the proposed method to measure changes in a subject's atrophy rate are assessed. The proposed methods compare favorably to state-of-the art techniques in neuroimaging in terms of accuracy, robustness and run-time.
Supervisor: Rueckert, Daniel Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.537240  DOI: Not available
Share: