Herzog, Nitsa J (2023) Deep learning of brain asymmetry digital biomarkers to support early diagnosis of cognitive decline and dementia. PhD thesis, Birkbeck, University of London.
|
Text
Herzog N Final Thesis.pdf - Full Version Download (3MB) | Preview |
Abstract
Early identification of degenerative processes in the human brain is essential for proper care and treatment. This may involve different instrumental diagnostic methods, including the most popular computer tomography (CT), magnetic resonance imaging (MRI) and positron emission tomography (PET) scans. These technologies provide detailed information about the shape, size, and function of the human brain. Structural and functional cerebral changes can be detected by computational algorithms and used to diagnose dementia and its stages (amnestic early mild cognitive impairment - EMCI, Alzheimer’s Disease - AD). They can help monitor the progress of the disease. Transformation shifts in the degree of asymmetry between the left and right hemispheres illustrate the initialization or development of a pathological process in the brain. In this vein, this study proposes a new digital biomarker for the diagnosis of early dementia based on the detection of image asymmetries and crosssectional comparison of NC (normal cognitively), EMCI and AD subjects. Features of brain asymmetries extracted from MRI of the ADNI and OASIS databases are used to analyze structural brain changes and machine learning classification of the pathology. The experimental part of the study includes results of supervised machine learning algorithms and transfer learning architectures of convolutional neural networks for distinguishing between cognitively normal subjects and patients with early or progressive dementia. The proposed pipeline offers a low-cost imaging biomarker for the classification of dementia. It can be potentially helpful to other brain degenerative disorders accompanied by changes in brain asymmetries.
Metadata
| Item Type: | Thesis |
|---|---|
| Copyright Holders: | The copyright of this thesis rests with the author, who asserts his/her right to be known as such according to the Copyright Designs and Patents Act 1988. No dealing with the thesis contrary to the copyright or moral rights of the author is permitted. |
| Depositing User: | Acquisitions And Metadata |
| Date Deposited: | 09 Feb 2024 11:22 |
| Last Modified: | 09 Feb 2024 12:59 |
| URI: | https://eprints.bbk.ac.uk/id/eprint/52554 |
| DOI: | https://doi.org/10.18743/PUB.00052554 |
Statistics
Additional statistics are available via IRStats2.
Archive Staff Only (login required)
