Retinal Vasculitis I

• Research type

  Research Study

• Full title

  Retinal Vasculitis I: Validating Clinician Grading and Developing Automated Grading

• IRAS ID

  270123

• Contact name

  Nicholas Beare

• Contact email

  nbeare@liverpool.ac.uk

• Sponsor organisation

  University of Liverpool

• Duration of Study in the UK

  2 years, 0 months, 1 days

• Research summary

  Summary of Research
  Retinal vasculitis (RV) is inflammation of the small blood vessels supplying the tissue that detects light and images in the eye (the retina). It is an important component of inflammatory eye disease (uveitis). It leads to visual loss by causing swelling of the retina (macular oedema), loss of blood supply to the retina (retinal ischaemia) and haziness of the clear gel filling the eye (vitreous haze). It can be difficult to detect by looking into the eye, but is clearly detected by retinal photographs using an intravenous dye (fluorescein angiogram). Wide-field fluorescein angiograms (WFFA) to look at all the retinal blood vessels are commonly performed in clinical practice to assess retinal vasculitis.

  However there is no method of grading retinal vasculitis for severity, unlike other manifestations of uveitis. A validated grading scheme is necessary for clinical evaluation of retinal vasculitis, and for it to be properly used in clinical studies of new treatments for uveitis. Ideally this should be an automated grading system done by computer and potentially incorporated into the image capturing devices.

  This study aims firstly to test the reliability of a novel grading scheme for retinal vasculitis between different graders with different levels of experience. We will do this by finding existing WFFAs on patients with retinal vasculitis and getting four people to grade them: 2 consultant specialists, an ophthalmic image grader and ophthalmic specialist trainee. Appropriate statistics (Kappa test) will compare their results. Secondly our team includes experts in computer analysis of ophthalmic images and they will work on developing a computerised analysis of WFFA images. This will be able to include more detailed and comprehensive information on the WFFA than is possible using time-pressed human graders. To do this they need a similar set of images and a definitive human expert grade (ground truth).

  Summary of Results
  We set out to develop and validate a new grading scheme for retinal vasculitis which is currently lacking for assessing patients with inflammatory eye disease. We recruited 50 patients with inflammation of the blood vessels at the back of their eyes (retinal vasculitis) who had previously had images taken of the back of their eyes. We developed a new grading scheme to assess the severity of the retinal vasculitis based on scientific literature and expert opinion. Four people graded the images using this grading scheme: two expert eye doctors, one eye doctor in training and a specialist eye image grader. We found that the different graders all gave similar results using the grading scheme, and that the eye image grader gave similar results when they graded the same images one month apart. These results demonstrate that the grading scheme is reliable and consistent. We also found that the results of the grading scheme were correlated to the vision of the patients both at the time they had the imaging and a year later. This suggests that the new grading scheme for retinal vasculitis is a useful measure of its severity

• REC name

  North of Scotland Research Ethics Committee 2

• REC reference

  19/NS/0186

• Date of REC Opinion

  2 Dec 2019

• REC opinion

  Favourable Opinion