Toward Data-Driven Handheld Fundus Videos Analysis

The School of EECS is hosting the following HDR Progress Review 3 Seminar:

Toward Data-Driven Handheld Fundus Videos Analysis

Speaker: Hongwei Sheng

Abstract:

Fundus videos enable direct observation of retinal vascular dynamics, particularly venous pulsation around the optic disc, which is clinically relevant to raised intracranial pressure states such as idiopathic intracranial hypertension and papilledema, and has also been associated with glaucoma and its severity. Handheld fundus video places this examination in routine clinical pathways, including timely review, referral support, screening, telemedicine, and longitudinal follow-up, so its value lies not only in what can be observed, but also in how readily that observation can be brought into care. However, the clinical promise of handheld fundus video is still constrained by the lack of an objective and scalable way to interpret it. At present, clinically relevant retinal dynamics are still assessed mainly through specialist review, whereas their use in screening, referral, telemedicine, and follow-up requires analysis that is more systematic, reproducible, and deployable. Our work is motivated by this gap, and aims to address it through data-driven analysis of handheld fundus videos.

Bio:

Hongwei Sheng is a PhD candidate at the School of Electrical Engineering and Computer Science, The University of Queensland. His research focuses on computer vision and deep learning for retinal medical imaging, with a particular interest in video ophthalmology and handheld retinal videos.