May 2004
Volume 45, Issue 13
Free
ARVO Annual Meeting Abstract  |   May 2004
Identification and follow–up of diabetic retinopathy in rural health in Australia: an automated screening model
Author Affiliations & Notes
  • A.P. Luckie
    Albury Eye Clinic, Albury, Australia
  • H. Jelinek
    Charles Sturt University, School of Community Health, Albury, Australia
  • M. Cree
    Department of Physics and Engineering, University of Waikato, Hamilton, New Zealand
  • R. Cesar
    Albury Eye Clinic, Albury, Australia
  • Jr
    Department of Physics, Vision Research Group, University Sao Paulo, Sao Paulo, Brazil
  • J. Leandro
    Department of Physics, Vision Research Group, University Sao Paulo, Sao Paulo, Brazil
  • C. McQuellin
    Albury Eye Clinic, Albury, Australia
  • P. Mitchell
    Department of Clinical Ophthalmology, University of Sydney, Sydney, Australia
  • Footnotes
    Commercial Relationships  A.P. Luckie, None; H. Jelinek, None; M. Cree, None; R. Cesar, Jr, None; J. Leandro, None; C. McQuellin, None; P. Mitchell, None.
  • Footnotes
    Support  none
Investigative Ophthalmology & Visual Science May 2004, Vol.45, 5245. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      A.P. Luckie, H. Jelinek, M. Cree, R. Cesar, Jr, J. Leandro, C. McQuellin, P. Mitchell; Identification and follow–up of diabetic retinopathy in rural health in Australia: an automated screening model . Invest. Ophthalmol. Vis. Sci. 2004;45(13):5245.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Abstract: : Purpose: To provide an automated diabetic retinopathy screening programme for use by primary health care providers in rural and remote areas. Methods: Using Morlet wave transformation, mathematical morphology operations, and fractal analysis developed initially with fluorescein angiography, an automated assessment of non–mydriatic images of the posterior pole of the retina was developed, to provide on–the–spot feedback of diabetic retinopathy from early signs of microaneurysms to extensive vascular proliferation. All images were pre–processed to select possible microaneurysms and provide the vessel network for analysis. Results: Non–proliferative retinopathy is characterised by the presence of microaneurysms and absence of new vessel growth. Our mathematical procedures identified 88% of microaneurysms present in the retinal fundus with 5 false positives per image on average. This compares favourably with reported results by ophthalmologists. The graph below shows the sensitivity of our methods in detecting proliferative retinopathy, characterised by new vessel growth. Of importance is the section between a fractal dimension of 1.14 and 1.82, which indicates the mean increase of blood vessels in proliferative retinopathy. Conclusion: A practical and innovative approach to contribute to the national health priority area of diabetes is to provide computer tools in conjunction with non–mydriatic photography that can identify non–proliferative and proliferative retinopathy and provide the allied health care worker with an effective assessment tool for use in rural and remote Australian communities.  

Keywords: diabetic retinopathy • clinical (human) or epidemiologic studies: health care delivery/economics/manpower • imaging/image analysis: clinical 
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×