Current HDR Students

Mahdi Heravian Shandiz | Jason Kugelman | Yoel Garcia Marin | Ignacio Escalona Viedma | Zachery Quince

Mahdi Heravian Shandiz

Thesis Title: Towards a more efficient platform for analysing retinal images via machine learning

Mahdi commenced his PhD candidature as a part-time external student in May 2021. Mahdi’s Principal Supervisor is Dr David Alonso-Caneiro, with Associate Supervisors, Professor Michael Collins and Associate Professor Scott Read. His research project will investigate the use of machine learning methods for the analysis and quantification of medical images of the human eye, with a particular focus on the posterior segment of the eye.

Back to top

Jason Kugelman

Thesis Title: Application of Generative Deep Learning Models in Ophthalmic Imaging

Jason begins his PhD candidature at QUT in January 2021. His research project will investigate generative deep learning methods such as generative adversarial networks (GANs), and their application to ophthalmic images as well as the potential extensibility of new and modified techniques beyond this domain. Applications include data augmentation, super resolution, de-noising + de-artifacting, domain translation, image segmentation and anomaly detection, among others. The aim is to use these methods to improve image analysis in ophthalmology, supporting research, clinical practice and diagnostics, while the potential of applying new methods beyond this domain is likely to have general positive benefits in the area of deep learning-based image analysis.

Back to top

Yoel Garcia Marin

Thesis Title: Deep learning methods applied to anterior eye optical coherence tomography images

Yoel’s research project is focused on the development of effective and automatic deep learning algorithms to be applied in anterior segment optical coherence tomography (AS-OCT) images. This covers a range of tasks including segmentation, classification and image quality assessment/improvement.

Since currently there are only a limited number of studies that have addressed the use of deep learning methods in anterior segment images, the procedures that will be developed represent a novel and innovative solution to current problems found in AS-OCT image analysis. Given the potential that deep learning methods have demonstrated in a number of imaging modalities, including the in posterior segment OCT images, this thesis aims to create effective and powerful methods that can be applied to both clinical and research applications.

Publications (arising from thesis):

Garcia Marin, YoelSkrok, MartaSiedlecki, DamianVincent, Stephen J.Collins, Michael J., & Alonso-Caneiro, David (2021) Segmentation of anterior segment boundaries in swept source OCT images. Biocybernetics and Biomedical Engineering41(3), pp. 903-915.

Back to top

Ignacio Escalona Viedma

Thesis Title: Automatic layer segmentation in retinal OCT imaging using deep learning methods

Ignacio’s research program involves the development of image analysis methods for Optical Coherence Tomography (OCT) images of the posterior segment of the eye (retina). He aims to develop accurate methods of segmentation using supervised and unsupervised Deep Learning approaches as well as pre-processing task that can support the segmentation task. The developed tools will provide automatic segmentation methods that can be used in a range of clinical and research studies. He is interested in exploring novel methods that have not been used for such an application, as well as the development of techniques that can be used in a range of different instruments (instrument-independent image analysis methods).

Back to top

Zachery Quince

Thesis Title: Optical Coherence Elastography for the Measurement of Corneal and Scleral Biomechanical Properties

Zachery’s doctoral research is focused on a developing a clinical method to quantify tissue biomechanics for the anterior segment of the eye, specifically the Young’s Modulus of the cornea and sclera. This method combines static compression of the tissue under examination with Spectral Domain optical coherence tomography imaging, in a technique called optical coherence elastography. This project requires the development of both hardware and software techniques necessary to perform the experiment and to analyse the images in order to extract biomechanical properties. As there is currently limited number of methods that can assess the eye mechanical properties, the proposed method provides a novel approach of quantifying this vital parameter of the eye. Using the proposed method, may allow for early detection, and monitoring of ocular diseases including keratoconus and myopia.

Theory re. novel approach of quantifying vital parameter of the eye
Results (from Experiment 1)









Publications (arising from thesis):

Quince, ZacheryAlonso-Caneiro, DavidRead, Scott A., & Collins, Michael J. (2021) Static compression optical coherence elastography to measure the mechanical properties of soft contact lenses. Biomedical Optics Express12(4), pp. 1821-1833.

Back to top