rss
Br J Ophthalmol 2007;91:1512-1517 doi:10.1136/bjo.2007.119453
  • Extended report
    • Clinical science - Extended reports

The efficacy of automated “disease/no disease” grading for diabetic retinopathy in a systematic screening programme

  1. S Philip1,
  2. A D Fleming2,
  3. K A Goatman2,
  4. S Fonseca3,
  5. P Mcnamee4,
  6. G S Scotland4,
  7. G J Prescott3,
  8. P F Sharp2,
  9. J A Olson5
  1. 1
    Biomedical Physics and Grampian Retinal Screening Programme, University of Aberdeen, Foresterhill, Aberdeen
  2. 2
    Biomedical Physics, University of Aberdeen, Foresterhill, Aberdeen
  3. 3
    Department of Public Health, University of Aberdeen, Foresterhill, Aberdeen
  4. 4
    Health Economics Research Unit, University of Aberdeen, Foresterhill, Aberdeen
  5. 5
    Retinal Screening, David Anderson Building, Foresterhill Road, Aberdeen
  1. Dr John A Olson, Clinical Director, Diabetes Retinal Screening Service, David Anderson Building, Foresterhill Road, Aberdeen AB25 2ZP; John.olson{at}nhs.net
  • Accepted 4 May 2007
  • Published Online First 15 May 2007

Abstract

Aim: To assess the efficacy of automated “disease/no disease” grading for diabetic retinopathy within a systematic screening programme.

Methods: Anonymised images were obtained from consecutive patients attending a regional primary care based diabetic retinopathy screening programme. A training set of 1067 images was used to develop automated grading algorithms. The final software was tested using a separate set of 14 406 images from 6722 patients. The sensitivity and specificity of manual and automated systems operating as “disease/no disease” graders (detecting poor quality images and any diabetic retinopathy) were determined relative to a clinical reference standard.

Results: The reference standard classified 8.2% of the patients as having ungradeable images (technical failures) and 62.5% as having no retinopathy. Detection of technical failures or any retinopathy was achieved by manual grading with 86.5% sensitivity (95% confidence interval 85.1 to 87.8) and 95.3% specificity (94.6 to 95.9) and by automated grading with 90.5% sensitivity (89.3 to 91.6) and 67.4% specificity (66.0 to 68.8). Manual and automated grading detected 99.1% and 97.9%, respectively, of patients with referable or observable retinopathy/maculopathy. Manual and automated grading detected 95.7% and 99.8%, respectively, of technical failures.

Conclusion: Automated “disease/no disease” grading of diabetic retinopathy could safely reduce the burden of grading in diabetic retinopathy screening programmes.

Footnotes

  • Competing interests: Implementation in Scotland is being considered. If this occurs it is likely that there will be some remuneration for the University of Aberdeen, NHS Grampian and the Scottish Executive.

  • Funding: This project was funded by the Chief Scientist Office, Scottish Executive Health Department (grant number CZH/4/76).

  • Ethics approval: Ethics approval was obtained from the Grampian Medical Research Ethics Committee for the use of the anonymised images and grading data.

  • Abbreviation:
    DH/MA

    dot haemorrhage/microaneurysm

This Article

  1. Abstract
  2. Full text
  3. PDF
  4. All Versions of this Article:
    1. bjo.2007.119453v1
    2. 91/11/1512 most recent

Responses

  1. Submit a response
  2. No responses published

Social bookmarking

Register for free content


Free sample
 This recent issue is free to all users to allow everyone the opportunity to see the full scope and typical content of BJO.
View free sample issue >>

Free archive
The full back archive is now available for BJO. Institutional subscribers may access the entire archive as part of their subscription. Personal subscribers will also have access to all content when logged in. Non-subscribers who register have free access to all articles published before 2006, back to volume 1 issue 1.
Register to access the free archive >>

Don't forget to sign up for content alerts so you keep up to date with all the articles as they are published.