Automated analysis of ophthalmic surgical videos

  • Research type

    Research Study

  • Full title

    Automated segmentation and analysis of cataract and vitreoretinal surgical videos

  • IRAS ID

    280040

  • Contact name

    Riaz Asaria

  • Contact email

    r.asaria@ucl.ac.uk

  • Sponsor organisation

    University College London (Joint Research Office)

  • Duration of Study in the UK

    1 years, 11 months, 31 days

  • Research summary

    This research involves the automated analysis of ophthalmic surgical procedures.

    Cataract surgery is the most common operation in the United Kingdom. Vitreoretinal surgery, defined as any surgery involving the back of the eye, is also commonly performed, and surgeons who perform this require sub-specialty training. Both operations are relatively low risk, but complications can occur and, rarely, result in loss of vision.

    In this project, we will be collecting videos of cataract surgery and vitreoretinal surgery. Any patients over the age of 16 who are able to given consent are eligible. We will also collect an additional type of intra-operative imaging, optical coherence tomography (OCT), for vitreoretinal procedures. All surgical video data will be anonymised before any analysis. Surgical data will be collected from participating hospitals within Royal Free London NHS Foundation Trust. We aim to collect surgical videos over a 2 year period, aiming for 1500 cataract surgeries and 500 vitreoretinal procedures. We will also be collecting basic demographic data and surgical outcome data from participants electronic medical records.

    The analysis and algorithm develop will broadly involve segmentation of surgical phases, detection of surgical instruments and detection of surgical actions. Within this work, we also plan to develop a system that can identify causative factors relating to intra-operative complications. If successful, these algorithms will enable the automatic analysis and interpretation of surgical videos, thereby providing quantitative data for research into intra-operative risk prediction and support for surgical decision making. Overall, this information may be used to improve how trainee surgeons learn to operate and make surgery safer for patients.

  • REC name

    East Midlands - Nottingham 1 Research Ethics Committee

  • REC reference

    23/EM/0039

  • Date of REC Opinion

    16 May 2023

  • REC opinion

    Further Information Favourable Opinion