AI and the Eye

  • Research type

    Research Study

  • Full title

    AI and the Eye – Integrating deep learning and in silico simulations to optimise diagnosis and treatment of wet macular degeneration

  • IRAS ID

    315821

  • Contact name

    Yalin Zheng

  • Contact email

    yzheng@liverpool.ac.uk

  • Sponsor organisation

    University of Liverpool

  • Duration of Study in the UK

    3 years, 11 months, 30 days

  • Research summary

    Wet age-related macular degeneration (wAMD), or neovascular AMD (nAMD), is a disease of increasing prevalence due to an aging population and is the leading cause of legal blindness worldwide. It is characterised by the growth of abnormal new blood vessels in the macula, which is the central part of the retina, and is responsible for sharp central vision. The mainstay of current treatment is repeated injections into the eye of drugs that inhibit the growth of blood vessels over a long period of time. However, the treatment's effectiveness which depends on a number of patient-related factors, including the sub-type of wAMD, varies greatly and represents a heavy burden for patients, the health sector and social care services. Sub-optimal response on standard treatment regimes is well-recognised.
    Optical Coherence Tomography Angiography (OCTA)is a new, exciting and non-invasive technology that captures 3-D images of the blood vessels in the macula. Interpreting OCTA scans requires special expertise and is time-consuming.
    In this study, we will collect OCTA scans of patients with wAMD attending St.Paul's Eye Department. Scans will be taken at given timepoints before and during treatment and for up to 2 years after the first injection. The study will not interfere with the normal care of the patient.
    We will develop deep learning algorithms to interpret OCTA scans and classify wAMD into its sub-types. Additionally, we will develop mathematical models of the retina in health and in wAMD with the prospect of simulating clinical trials on computers, sometimes known as in-silico trials. The OCTA data will be used to calibrate and validate those models. Once validated, we will simulate different treatment strategies based on the characteristics of individual eyes. Those simulations will shed light on new treatment strategy, potentially more personalised treatment regimes, aiming for better visual outcomes.

  • REC name

    London - London Bridge Research Ethics Committee

  • REC reference

    22/PR/1189

  • Date of REC Opinion

    27 Sep 2022

  • REC opinion

    Further Information Favourable Opinion