March 2012
Volume 53, Issue 14
Free
ARVO Annual Meeting Abstract  |   March 2012
Combining Visual Field and Frequency Domain OCT Data Improves Detection of Glaucoma
Author Affiliations & Notes
  • Ali S. Raza
    Psychology,
    Neurobiology and Behavior,
    Columbia University, New York, New York
  • Xian Zhang
    Psychology,
    Columbia University, New York, New York
  • Carlos G. De Moraes
    Ophthalmology, NYU School of Medicine, New York, New York
  • Charles A. Reisman
    Topcon Adv Biomed Imaging Lab, Topcon Medical Systems, Oakland, New Jersey
  • Jeffrey M. Liebmann
    Ophthalmology, NYU School of Medicine, New York, New York
    Ophthalmology, New York Eye & Ear Infirmary, New York, New York
  • Robert Ritch
    Ophthalmology, New York Eye & Ear Infirmary, New York, New York
    Ophthalmology, New York Medical College, Valhalla, New York
  • Donald C. Hood
    Psychology,
    Ophthalmology,
    Columbia University, New York, New York
  • Footnotes
    Commercial Relationships  Ali S. Raza, None; Xian Zhang, Topcon (C); Carlos G. De Moraes, None; Charles A. Reisman, Topcon, Inc. (E); Jeffrey M. Liebmann, Topcon (F), Topcon, Zeiss (C); Robert Ritch, Topcon (F); Donald C. Hood, Topcon (F, C)
  • Footnotes
    Support  NIH Grant R01-EY002115 and Topcon, Inc
Investigative Ophthalmology & Visual Science March 2012, Vol.53, 702. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Ali S. Raza, Xian Zhang, Carlos G. De Moraes, Charles A. Reisman, Jeffrey M. Liebmann, Robert Ritch, Donald C. Hood; Combining Visual Field and Frequency Domain OCT Data Improves Detection of Glaucoma. Invest. Ophthalmol. Vis. Sci. 2012;53(14):702.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose: : A cluster criterion is often used to assess whether a visual field (VF) is abnormal, while frequency domain OCT (fdOCT) data are often analyzed based on averages over relatively large regions. Here, we compare cluster analyses for fdOCT and VF data, and also assess the benefit of combining these data to improve detection of glaucoma.

Methods: : One eye from each of 54 controls (53.2 yrs) and 156 eyes from 113 glaucoma patients and suspects (55.7 yrs, MD -3.0 ± 3.6 dB) were tested with VFs (24-2, HFA, Zeiss, Inc) and fdOCT (3DOCT-2000, Topcon, Inc) cube scans (128 Bscans) of the macula and optic disc. The thicknesses of the retinal nerve fiber layer at the optic disc (dRNFL) and macula (mRNFL) as well as the retinal ganglion cell plus inner plexiform layer at the macula (mRGCPL) were calculated for each fdOCT scan based on hand-correction of a previously validated automated algorithm.[1] Probability plots were then calculated for the fdOCT thicknesses (as in [2]), which were then divided into a grid, and clusters for both the fdOCT and VF were determined based on a cluster criteria of 3 contiguous, significant points at 5%-5%-1%, respecting the midline. Combinatorial probabilities of fdOCT and VF abnormalities in correspondent locations were also calculated.

Results: : By itself, the VF (cluster analysis) provided sensitivity (SN) of 62% and specificity (SP) of 93%. Optimal measures derived independently from the fdOCT resulted in similar values (e.g., mRGCPL cluster, SN=60%, SP=91%). By combining the VF and fdOCT measures, using a simple "or" criteria, SN increased to 78% with a slight decrease in SP to 87%. Using a more nuanced combination of VF and fdOCT criteria, making use of correspondent regions of damage (e.g. inferotemporal optic disc RNFL and superior VF both have significant clusters), yielded a SN of 74% while maintaining a SP of 94%.

Conclusions: : Individually, using cluster criteria for either the VF or the fdOCT data resulted in high specificity and relatively high sensitivity, especially given that the glaucoma group included suspects and had a low average VF MD (-3 dB). Better sensitivity/specificity scores might be obtained by combining the information in the VF and fdOCT in an optimal (e.g. spatially correspondent) manner.[1] Yang Q et al. Opt Exp. 2010 [2] Hood DC & Raza AS. Biomed Opt Exp. 2011

Keywords: imaging methods (CT, FA, ICG, MRI, OCT, RTA, SLO, ultrasound) • nerve fiber layer • visual fields 
×
×

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.

×