April 2009
Volume 50, Issue 13
ARVO Annual Meeting Abstract  |   April 2009
Detection of Edges of Tear Break-Up Imaged by Different Methods
Author Affiliations & Notes
  • Z. Wu
    Optometry, Indiana University, Bloomington, Indiana
  • C. G. Begley
    Optometry, Indiana University, Bloomington, Indiana
  • N. Himebaugh
    Optometry, Indiana University, Bloomington, Indiana
  • Footnotes
    Commercial Relationships  Z. Wu, None; C.G. Begley, None; N. Himebaugh, None.
  • Footnotes
    Support  None.
Investigative Ophthalmology & Visual Science April 2009, Vol.50, 523. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Z. Wu, C. G. Begley, N. Himebaugh; Detection of Edges of Tear Break-Up Imaged by Different Methods. Invest. Ophthalmol. Vis. Sci. 2009;50(13):523.

      Download citation file:

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

  • Supplements

Purpose: : Tear instability is considered a core mechanism of dry eye. However, study of the phenomenon is limited by our ability to quantify it and investigate changes over time, especially using non-invasive methods for measuring tear instability. It was the purpose of this investigation to develop a method that quantifies tear break-up (TBU) with both fluorescein (FL) and the non-invasive method known as tear film retroillumination (RI).

Methods: : FL (2µl of 2%) was instilled into one randomly chosen eye of 10 non-contact lens wearers, who kept one eye open as long as possible while the fluorescent tear film was videotaped. The same procedure was used with 10 contact lens wearers, except no FL was instilled and the tear film was viewed by RI. Frames extracted from videotapes of the tear film viewed with FL and RI were extracted and a custom MATLAB program was used to detect the edges of tear break-up via pixel intensity changes over the image. The number of edges detected was quantified in each image and tracked over time. The image analysis results were compared to TBU graded (0-4 scale) by a masked observer.

Results: : Four subjects (2 FL and 2 RI) showed no TBU during the trials and the others showed varying levels, with a median grade of 3 at the end of the trial, and no significant difference between FL or RI groups (p=0.74, Mann-Whitney). The number of pixels detecting edges (EP) in each image ranged from 0-7652, with a average of 1370±1709. EP was highly correlated with the grade of TBU within each trial, with an average correlation of 0.66±0.36 among all trials and a range of 0 (subjects with no TBU) to 0.98 (subject with extensive TBU). EP was significantly higher with FL compared to RI images (p=0.028, t-test).

Conclusions: : The FL method for viewing tear instability is based on dye concentration differences in the tear film over the corneal surface, whereas RI is an edge detector for TBU, as are many other non-invasive methods for viewing the tear film. This novel method, designed for finding subtle edges in images effectively detects TBU in both FL and RI images and therefore can be used to quantify tear instability viewed by various methods.

Keywords: cornea: tears/tear film/dry eye • clinical research methodology • 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.