Detecting Dementia in the Retina; a Big Data Machine Learning Approach

  • Research type

    Research Study

  • Full title

    Detection Dementia from Retinal Morphology: a Big Data Machine Learning based Retrospective Case-Control Study

  • IRAS ID

    233974

  • Contact name

    Sabeena Johal

  • Contact email

    khpctofacilitator@kcl.ac.uk

  • Sponsor organisation

    Moorfields Eye Hospital NHS Foundation Trust

  • Duration of Study in the UK

    1 years, 0 months, 1 days

  • Research summary

    The proposed research project will apply a series of computational analyses to a set of over 1 million eye scans, which have been linked to the corresponding Hospital Episode Statistics from the National Health Service (NHS). Some evidence from small studies exists to suggest that the retina, the light detecting structure at the back of the eye, changes in appearance as a patient develops Alzheimer’s disease. At Moorfields Eye Hospital, a large database of ‘optical coherence tomography’ (OCT) images of patients’ retinas has been collected. By linking these images to patients diagnostic data we will compare the eyes of those who developed dementia to the eyes of patients who do not suffer from dementia. The primary research question, will be to identify on these scans of the patient’s retinas, the morphological features associated with a diagnosis of dementia. The secondary research question will focus on how retinal morphology evolves with time.

    This is a retrospective cohort study and requires no active engagement by participants. Any patient over the age of 40 who has had an eye scan at Moorfields will be eligible; currently this comprises 257,000 patients. These patients’ NHS numbers will be sent to NHS Digital in order to extract the corresponding diagnostic data. These records along with the eye images will be pseudonymised in collaboration with NHS digital, and then linked by the data processors at the UCL institute of Ophthalmology.

  • REC name

    London - Central Research Ethics Committee

  • REC reference

    18/LO/1163

  • Date of REC Opinion

    1 Aug 2018

  • REC opinion

    Favourable Opinion