June 2013
Volume 54, Issue 15
Free
ARVO Annual Meeting Abstract  |   June 2013
Inter- and Intra-Observer Agreement and Repeatability of Imaging the Meibomian Glands with the OCULUS Keratograph 4 and Keratograph 5M
Author Affiliations & Notes
  • William Ngo
    Centre for Contact Lens Research, School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada
  • Sruthi Srinivasan
    Centre for Contact Lens Research, School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada
  • Marc-Matthias Schulze
    Centre for Contact Lens Research, School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada
  • Lyndon Jones
    Centre for Contact Lens Research, School of Optometry and Vision Science, University of Waterloo, Waterloo, ON, Canada
  • Footnotes
    Commercial Relationships William Ngo, None; Sruthi Srinivasan, None; Marc-Matthias Schulze, Abbott Medical Optics (R); Lyndon Jones, Alcon (F), Alcon (R), Allergan (F), Abbott Medical Optics (R), Bausch & Lomb (R), Ciba Vision (F), Ciba Vision (R), CooperVision (F), Johnson & Johnson (F), Johnson & Johnson (R)
  • Footnotes
    Support None
Investigative Ophthalmology & Visual Science June 2013, Vol.54, 3569. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      William Ngo, Sruthi Srinivasan, Marc-Matthias Schulze, Lyndon Jones; Inter- and Intra-Observer Agreement and Repeatability of Imaging the Meibomian Glands with the OCULUS Keratograph 4 and Keratograph 5M. Invest. Ophthalmol. Vis. Sci. 2013;54(15):3569.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose: To compare the inter- and intra-observer agreement and repeatability of using a 4-point scale to grade meibography images taken from the OCULUS Keratograph 4 (K4) and OCULUS Keratograph 5M (K5M) instruments.

Methods: The superior and inferior eyelids from a single eye of 40 participants were imaged 3 times on both the K4 and K5M. The images were organized into 3 sets each containing 160 images (2 devices x 40 images each for inferior superior eyelid). Arita et al’s 4-point grading scale (0=no dropout, 1=less than 1/3 total area dropout, 2=1/3 to 2/3 total area dropout, 3=>2/3 total area dropout) was used to grade meibomian gland dropout in the images. The first set of images served to train two observers, O1 and O2; the second and third sets were presented to both observers (24 hours apart) on a 50” high definition monitor in a darkened room. All images in each set were presented in a random order.

Results: For both devices combined, the mean difference in ratings between O1 and O2 (mean±sd) was 0.08±0.55 on day 1 (D1), and 0.13±0.50 on day 2 (D2); the concordance correlation coefficient (CCC) between O1 and O2 was CCC=0.79 on D1 and CCC=0.81 on D2. When looking at both devices individually, the inter-observer mean differences in ratings when using the K4 was -0.01±0.58 on D1 and 0.08±0.52 on D2; while the mean differences when using the K5M was -0.16±0.51 on D1, and 0.18±0.47 on D2. Inter-observer CCCs when using the K5M (D1=0.79, D2=0.81) were slightly higher than with the K4 (D1=0.78, D2=0.80). When comparing ratings between sessions for each observer, the intra-observer mean difference in ratings for O1 was -0.08±0.49 and -0.01±0.58 for the K5M and K4, respectively; mean difference for O2 was -0.09±0.51 and -0.10±0.68 for the K5M and K4, respectively. The intra-observer CCC for the K5M (O1=0.82, O2=0.80) was higher than for K4 (O1=0.76, O2=0.68).

Conclusions: The repeatability of the K5M was slightly better than the K4 when using a 4-point grading scale. In all cases, both observers graded within -1 to +1 grades of each other and against themselves.

Keywords: 550 imaging/image analysis: clinical • 526 eyelid • 486 cornea: tears/tear film/dry eye  
×
×

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.

×