June 2017
Volume 58, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2017
Individualized normal limits of retinal nerve fiber layer measurement improve diagnostic performance of optical coherence tomography
Author Affiliations & Notes
  • Clemens Vass
    Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria
  • Ivania Pereira
    Center for Medical Statistics, Informatics and Intelligent Systems, Vienna, Austria
    Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria
  • Hemma Resch
    Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria
  • Lorenz Wassermann
    Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria
  • Stephan Holzer
    Department of Ophthalmology and Optometry, Medical University of Vienna, Vienna, Austria
  • Georg Fischer
    Center for Medical Statistics, Informatics and Intelligent Systems, Vienna, Austria
  • Footnotes
    Commercial Relationships   Clemens Vass, Carl Zeiss Meditec (F), Carl Zeiss Meditec (R); Ivania Pereira, Carl Zeiss Meditec (F); Hemma Resch, None; Lorenz Wassermann, None; Stephan Holzer, None; Georg Fischer, None
  • Footnotes
    Support  WWTF- LS11-046
Investigative Ophthalmology & Visual Science June 2017, Vol.58, 3988. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Clemens Vass, Ivania Pereira, Hemma Resch, Lorenz Wassermann, Stephan Holzer, Georg Fischer; Individualized normal limits of retinal nerve fiber layer measurement improve diagnostic performance of optical coherence tomography. Invest. Ophthalmol. Vis. Sci. 2017;58(8):3988.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Recently we developed a compensation model based on multiple linear regression analysis that reduces the interindividual variation of retinal nerve fiber layer (RNFL) measurement on healthy subjects. This model can be applied to generate patient specific individualized normal limits (INL). The purpose of this study is to analyze whether the sensitivity and/or the specificity are improved when applying INL as opposed to standard age-corrected normal limits (ANL).

Methods : This prospective study compares 58 patients with early to moderate glaucomatous damage (mean MD ±standard deviation = -3.80 ±2.83) with previously collected data of 202 healthy subjects. All subjects underwent optical coherence tomography (OCT) (Cirrus®, Carl Zeiss Meditec) of the macula and the circumpapillary area and the retinal vessels were automatically segmented in the 2D projection images. Vessel thickness values were measured at a 3.4 mm diameter circle and used to calculate the retinal vessel density function (RVD), as previously published. The model was recalculated using 281 healthy subjects of the Cirrus normative database (CND) and considering 8 independent factors: RVD, distance between fovea and optic disk (OD) center, fovea angle, OD size, ratio and angle and refractive error. The normal limits were calculated twice: 1) the ANL based on CND using intercept, slope of age and variance; 2) the INL using intercept, slopes and variance of the multivariate model. We evaluated the number of subjects with at least one clock hour sector below the 5% normal limit (NL5), or the 1% normal limit (NL1). Sensitivities and specificities were compared between the INL and the ANL using McNemar tests.

Results : Specificity was significantly improved for the INL1 compared to the ANL1 (p=0.019). Sensitivity showed a non-significant trend towards improvement. There was no improvement for the INL5 over the ANL5.

Conclusions : Application of patient specific INL, based on anatomical parameters, resulted in an improved specificity compared to standard ANL while sensitivity was preserved, thus improving the diagnostic value of the OCT.

This is an abstract that was submitted for the 2017 ARVO Annual Meeting, held in Baltimore, MD, May 7-11, 2017.

 

Tab.1: Sensitivities and specificities comparing INL and ANL

Tab.1: Sensitivities and specificities comparing INL and ANL

×
×

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.

×