The proposed project aims to develop machine learning based toolsets and techniques to evaluate and rehabilitate visual field loss through applied game conditions. The project will utilise and be supported with various facilities within GCU, including the Artificial Intelligence, Games eMotion UX laboratories and Vision Science facilities. The project builds upon current Visual Science research of measuring patient visual field loss and will contribute to developing a new method of measurement, evaluation and improved visual rehabilitation.
Machine learning is revolutionising technological applications with improved accuracy, reduced computational requirements and enabling technological solutions that were previously unobtainable across various sectors, including health and vision.
Stroke is the main cause of acquired adult disability in high-income countries. Around 50% of those who survive a stroke suffer from visual field loss, most commonly homonymous visual field defects (HVFD). HVFD leads to considerable difficulties, particularly with visual exploration and page navigation, which results in persistent and severe reading difficulties and impaired mobility. This significantly affects social functioning, mental health, and creates vision-specific dependency and a substantial reduction in health-related quality of life.
Current treatment methods for HVFD include visual field restitution and therapy to retrain eye-movement control. While rigorously controlled trials that clearly establish the efficacy of any method are lacking, there is increasing evidence in favour of treatment protocols that involve the systematic and repetitive practice of specific eye movements. Currently available methods for eye-movement training include: (1) individualised therapy developed empirically by clinicians, (2) simple paper based scanning exercises and (3) apps or web based training tools. Critical disadvantages, which are inherent to all current methods of HVFD rehabilitation are objective quantification of any treatment effect is difficult or impossible to obtain and feedback is provided based on a clinicians subjective observations. In comparison, apps and web based tools are relatively new approaches, however lack the technological models to effectively monitor or provide worthwhile rehabilitation.
The aim of this proposal is to develop novel methods in machine learning and signal/image processing which can be utilised to assess and rehabilitate HVFD of individuals undertaking an applied serious game, which utilises the above GCU perceptual learning paradigm. There will be three main components of the project.
1. The first is to investigate signal processing based measures of eye tracking and head movement with the serious games based stimuli from optical imaging and infrared eye tracking systems.
2. The second will be to develop machine learning based tools and models to investigate correlations between user interaction and HVFD from serious game interactions.
3. The final stage is to provide a machine learning based feedback loop, which aims to maximise the users interaction with the machine learning based applied serious game to try and maximise the HVFD rehabilitation improvement with objective measures previously unobtainable.
The successful candidate is expected to have a solid mathematical background, strong programming skills (in C++/Python/Matlab), and keen interest in undertaking high-impact research work. Due to the nature of the project the candidate should also possess the ability to work cooperatively in a multi-disciplinary setting. Candidates from Artificial Intelligence, Computing, Big Data, Games, Vision Science and Electronic Engineering based backgrounds will be considered.
Research Strategy and Research Profile
Glasgow Caledonian University’s research is framed around the United Nations Sustainable Development Goals, We address the Goals via three societal challenge areas of Inclusive Societies, Healthy Lives and Sustainable Environments. For more. This project is part of the research activity of the Research Group – Artificial Intelligence Research Lab
How to Apply
Applicants will normally hold a UK honours degree 2:1 (or equivalent); or a Masters degree in a subject relevant to the research project. Equivalent professional qualifications and any appropriate research experience may be considered. A minimum English language level of IELTS score of 6.5 (or equivalent) with no element below 6.0 is required. Some research disciplines may require higher levels.
Candidates are encouraged to contact the research supervisors for the project before applying. Applicants should complete the online GCU Research Application Form, stating the Project Title and Reference Number (listed above).
Please also attach to the online application, copies of academic qualifications (including IELTS if required), 2 references and any other relevant documentation.
Please send any enquiries regarding your application to: [email protected]
Applicants shortlisted for the PhD project will be contacted for an interview.
For more information on How to apply please go to https://www.gcu.ac.uk/research/postgraduateresearchstudy/applicationprocess/
Dr. Ryan Gibson, [email protected]