Schwartz, Roy;
(2024)
Deep Learning and Genome-Wide Analysis of the UK Biobank
to Unravel Novel Pathways and Reticular Pseudodrusen’s
Role in Age-Related Macular Degeneration.
Doctoral thesis (Ph.D), UCL (University College London).
Preview |
Text
Roy Schwartz PhD thesis - final.pdf - Accepted Version Download (39MB) | Preview |
Abstract
Background: Age-related macular degeneration (AMD, the leading cause of vision loss in the Western world, is characterised by drusen, the prototypical lesion. In recent years, reticular pseudodrusen (RPD) have been identified as a distinct type of lesion that can accelerate the progression of AMD. Existing evidence suggests that RPD differ from drusen in several aspects. However, the genetic origin of RPD remains unclear. Previous genetic studies have yielded inconsistent results, primarily due to methodological limitations. We hypothesize that RPD have unique genetic associations that differ from those of drusen.// Objectives: To i) train a machine learning algorithm for detecting and quantifying RPD and drusen in optical coherence tomography (OCT) scans; ii)To deploy the algorithm on the UK Biobank (UKBB) to identify cohorts with pathology and controls for genetic analysis; iii) To validate the cohorts through human assessment; iv) to perform a genome-wide association study (GWAS) to identify genetic associations of RPD.// Methods: A deep learning (DL) framework consisting of two ungradability detection models, a drusen-RPD classification model, and a semantic segmentation model was developed. It was deployed on OCT images from the UKBB to identify eyes with RPD and/or drusen and controls. These cohorts were validated by five retina specialists. Quantification of RPD and drusen derived from the framework was used as variables in a GWAS.// Results: The models of the DL network performed strongly at their respective tasks, similar or close to human performance. A total of 1,037 controls, 361 participants with only drusen (pure drusen), 66 with pure RPD, and 323 mixed were included in the GWAS. The primary pure RPD GWAS yielded four genomewide significant loci, three of which are novel and the fourth differs from previous findings.// Conclusion: Confirming our initial hypothesis, RPD were found to have associations which differ from those of drusen-associated AMD.
Type: | Thesis (Doctoral) |
---|---|
Qualification: | Ph.D |
Title: | Deep Learning and Genome-Wide Analysis of the UK Biobank to Unravel Novel Pathways and Reticular Pseudodrusen’s Role in Age-Related Macular Degeneration |
Open access status: | An open access version is available from UCL Discovery |
Language: | English |
Additional information: | Copyright © The Author 2024. Original content in this thesis is licensed under the terms of the Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) Licence (https://creativecommons.org/licenses/by-nc/4.0/). Any third-party copyright material present remains the property of its respective owner(s) and is licensed under its existing terms. Access may initially be restricted at the author’s request. |
UCL classification: | UCL UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences UCL > Provost and Vice Provost Offices > School of Life and Medical Sciences > Faculty of Population Health Sciences > Institute of Health Informatics |
URI: | https://discovery.ucl.ac.uk/id/eprint/10199972 |
Archive Staff Only
![]() |
View Item |