Analysing retinal fundus images with deep learning models

Simple item page
dc.contributor.advisorBah, Bubacarren_ZA
dc.contributor.advisorBrink, Willieen_ZA
dc.contributor.authorOfosu Mensah, Samuelen_ZA
dc.contributor.otherStellenbosch University. Faculty of Science. Dept. of Mathematical Sciences. Applied Mathematics Division.en_ZA
dc.date.accessioned2023-11-27T05:31:13Zen_ZA
dc.date.accessioned2024-01-08T11:11:35Zen_ZA
dc.date.available2023-11-27T05:31:13Zen_ZA
dc.date.available2024-01-08T11:11:35Zen_ZA
dc.date.issued2023-12en_ZA
dc.descriptionThesis (PhD)--Stellenbosch University, 2023.en_ZA
dc.description.abstractENGLISH ABSTRACT: Convolutional neural networks (CNNs) have successfully been used to classify diabetic retinopathy but they do not provide immediate explanations for their decisions. Explainability is relevant, especially for clinicians. To make results explainable, we use a post-attention technique called gradient-weighted class activation mapping (Grad- CAM) on the penultimate layer of deep learning models to produce localisation maps on retinal fundus images after using them to classify diabetic retinopathy. Moreover, the models were initialised using pre-trained weights obtained from training models on the ImageNet dataset. The results of this are fewer training epochs and improved performance. Next, we predict cardiovascular risk factors (CVFs) using retinal fundus images. In detail, we use a multi-task learning (MTL) model since there are several CVFs. The impact of using an MTL model is the advantage of simultaneously training for and predicting several CVFs rather than doing so individually. Also, we investigate the performance of the fundus cameras used to capture the retinal fundus images. We notice a superior performance of the desktop fundus cameras to the handheld fundus camera. Finally, we propose a hybrid model that fuses convolutions and Transformer encoders. This is done to harness the benefits of convolutions and Transformer encoders. We compare the performance of the proposed model with other attention-based models and observe on-par performance. en_ZA
dc.description.abstractAFRIKAANSE OPSOMMING: Geen opsomming beskikbaar.af_ZA
dc.description.versionDoctorateen_ZA
dc.format.extentxvii, 117 pages : illustrationsen_ZA
dc.identifier.urihttps://scholar.sun.ac.za/handle/10019.1/128788en_ZA
dc.language.isoen_ZAen_ZA
dc.language.isoen_ZAen_ZA
dc.publisherStellenbosch : Stellenbosch Universityen_ZA
dc.rights.holderStellenbosch Universityen_ZA
dc.subject.lcshDeep learning (Machine learning)en_ZA
dc.subject.lcshComputer visionen_ZA
dc.subject.lcshConvolutional neural networksen_ZA
dc.subject.lcshConvolutions (Mathematics)en_ZA
dc.subject.lcshNeural networks (Computer science)en_ZA
dc.subject.lcshDiabetic retinopathyen_ZA
dc.titleAnalysing retinal fundus images with deep learning modelsen_ZA
dc.typeThesis en_ZA
Files
Original bundle
Now showing 1 - 1 of 1
Thumbnail Image
Name:
ofosumensah_analysing_2023.pdf
Size:
47.65 MB
Format:
Adobe Portable Document Format
Description:
Download
Collections
Doctoral Degrees (Applied Mathematics)