Artificial Intelligence for Ocular Imaging: PhD Studentship

UCL Department / Division:   UCL Department of Medical Physics and
                                                  Biomedical Engineering and
                                                  UCL Institute of Ophthalmology
Duration of Studentship:        Three years, starting Autumn 2024
PhD Title:                                 Artificial Intelligence for Ocular Imaging: An Eye on
                                                  Vision-robbing, Neurodegenerative, and Systemic
                                                  Diseases
Supervisor(s):                          Professor Marinko Sarunic and
                                                  Professor Pearse Keane

We are offering a full-time, three-year PhD studentship to investigate ocular diseases and develop artificial intelligence methods at the NIHR Biomedical Research Center at Moorfields Eye Hospital and UCL Institute of Ophthalmology. Our team is highly interdisciplinary: Professor Sarunic is cross-appointed between the Department of Medical Physics and Biomedical Engineering and the Institute of Ophthalmology, and Professor Keane is a clinician-scientist consultant ophthalmologist at Moorfields Eye Hospital and the Institute of Ophthalmology.  This research project will be in collaboration with the research and development team at Optos Inc., a world-leading provider of ocular imaging instrumentation and analysis tools, and the provider of the Optomap.

The eyes are an extension of the brain and a window to the whole body. Imaging the back of the eye, the light-sensitive tissue called the retina, is performed routinely at check-ups and vision tests. Recently, machine learning (ML) methods have been used to show that these images of the eye’s neuronal tissue and microvasculature can reveal information about a person’s overall health, including factors such as the risk of cardiovascular disease and ‘biological age’. The information-rich ocular images contain a wealth of information that motivates the development of new tools for the identification of novel biomarkers.

This project will initially focus on a vision-robbing disease called glaucoma, a neurodegenerative disease of the eye. Referred to as the ‘sneak thief of sight’, glaucoma progresses very slowly, and often goes unnoticed until significant irreversible vision loss has occurred. Our goal is to investigate ML methods to identify risk factors for glaucoma based on ocular images and to investigate the prediction of response to treatment. The data to be used for this project are Optomap images, which capture ultrawide field pictures of the retina. The student will develop ML approaches to investigate the classification of the Optomap images for patients with glaucoma and to identify the features that the networks use in making the prediction (‘explainable AI’). Our team has a wealth of data (thousands of labelled patient Optomap images) available to initiate the study.

Subsequent research may move beyond glaucoma to other ocular neurodegenerative diseases, including diabetic retinopathy (a neurovascular disease of the eye) and even diseases of the brain. This research also has significant potential to pave the way for future, larger-scale collaborative research projects in the area of novel instrumentation and software for ophthalmic vision health, and more generally neurovascular diseases with ocular biomarkers.

Person Specification

• High academic achievement in the area of specialisation 
• Demonstrable interest in working on translational medical artificial intelligence 
• Skill in programming and artificial intelligence
• Interest in developing skills in statistical methods and related software packages
• High proficiency in written and spoken English is required
• Very strong work ethic, with the ability to think creatively and work both individually and within a team
• Familiarity in imaging processing and image analysis is desired

Application Process

Informal enquiries should be made to Professor Marinko Sarunic (m.sarunic@ucl.ac.uk) or Professor Pearse Keane (p.keane@ucl.ac.uk). 

Find out more and apply