May 2003
Volume 44, Issue 13
Free
ARVO Annual Meeting Abstract  |   May 2003
JPEG Compression of Stereoscopic Digital Images for the Tele-ophthalmology Diagnosis of Diabetic Retinopathy
Author Affiliations & Notes
  • M. Greve
    Ophthalmology, University of Alberta, Edmonton, AB, Canada
  • C.F. Baker
    Ophthalmology, University of Alberta, Edmonton, AB, Canada
  • C.J. Rudnisky
    Ophthalmology, University of Alberta, Edmonton, AB, Canada
  • M.T. Tennant
    Retina Service, Wills Eye Hospital, Philadelphia, PA, United States
  • B.J. Hinz
    Retina Service, Wills Eye Hospital, Philadelphia, PA, United States
  • A.R. DeLeon
    Mathematics, University of Calgary, Calgary, AB, Canada
  • Footnotes
    Commercial Relationships  M. Greve, None; C.F. Baker, None; C.J. Rudnisky, None; M.T.S. Tennant, None; B.J. Hinz, None; A.R. DeLeon, None.
Investigative Ophthalmology & Visual Science May 2003, Vol.44, 3989. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      M. Greve, C.F. Baker, C.J. Rudnisky, M.T. Tennant, B.J. Hinz, A.R. DeLeon; JPEG Compression of Stereoscopic Digital Images for the Tele-ophthalmology Diagnosis of Diabetic Retinopathy . Invest. Ophthalmol. Vis. Sci. 2003;44(13):3989.

      Download citation file:


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

      ×
  • Supplements
Abstract

Abstract: : Purpose: We investigated the application of Joint Photographic Experts Group (JPEG) compression to digital retinal images. The goal of this study was to identify the optimal level of JPEG compression that would minimize file size, while retaining diagnostic image information. Methods: 20 diabetic patients from a northern Alberta tele-ophthalmology clinic were enrolled. Following informed consent and mydriasis , seven 30° fields of each fundus were digitally photographed with a six megapixel digital camera and saved as uncompressed TIFF (Tag Image File Format) files. The files were then compressed approximately 50 and 100 times their original size using JPEG compression. The three file types were then randomly in a masked fashion. Each reviewer analyzed the digital photographs for image quality and specific diabetic retinal pathology in accordance with ETDRS (Early Treatment of Diabetic Retinopathy Study) standards. The level of diabetic retinopathy and recommended clinical follow-up was also recorded. Exact agreement and weighted kappa statistics were calculated for pathology identification and grading of diabetic retinopathy. Results: All patients were included in the final analysis. Correlation between both 50 and 100x compressed JPEG and TIFF images was good for all pathologies (kappa 0.4 > x > 0.75) and excellent (kappa > 0.75) for all except hard exudates and IRMA in the 50x JPEG images and cotton-wool spots and IRMA in the 100x JPEGs. With respect to the diagnosis of level of retinopathy, both levels of compression had excellent correlation with uncompressed TIFF images. Conclusions: JPEG compression of at least 100 times, and perhaps more, results in a significantly smaller file size and yet does not degrade the quality of a digital image with respect to the diagnosis of diabetic retinopathy.

Keywords: diabetic retinopathy • diabetes • 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.

×