July 2019
Volume 60, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2019
Morphometric detection of angular loss of nerve fiber bundles in glaucoma patients
Author Affiliations & Notes
  • Per G Soderberg
    Neuroscience, Uppsala university , Uppsala, Sweden
  • Zhaohua Yu
    Neuroscience, Uppsala university , Uppsala, Sweden
  • Camilla Sandberg Melin
    Neuroscience, Uppsala university , Uppsala, Sweden
    Ophthalmology, Gävle regional hospital, Gävle, Sweden
  • Footnotes
    Commercial Relationships   Per Soderberg, None; Zhaohua Yu, None; Camilla Sandberg Melin, None
  • Footnotes
    Support  Föreningen Synskadades Vänner i Uppsala Län, Gun och Bertil Stohnes Stiftelse, Kronprincessan Margaretas Arbetsnämnd för synskadade, Ögonfonden, The Uppsala university/Uppsala Läns Landsting’s ALF Research grants, Wallinders gåva, Erik Funks Minnesfond, Karin Sandqvists Stiftelse
Investigative Ophthalmology & Visual Science July 2019, Vol.60, 5575. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Per G Soderberg, Zhaohua Yu, Camilla Sandberg Melin; Morphometric detection of angular loss of nerve fiber bundles in glaucoma patients. Invest. Ophthalmol. Vis. Sci. 2019;60(9):5575.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : To develop a strategy for morphometric detection of angular loss of nerve fiber bundles in the optic nerve head (ONH) in glaucoma patients based on OCT detection of angular loss of Pigment Epithelium Inner limit of the retina - Minimal Distance (PIMD).

Methods : 3D representations of the ONH were captured with Topcon OCT-2000 from one eye of each of totally 39 subjects with early stage primary open angle glaucoma. Altogether, 3 volumes were captured at two occasions within 3 months and PIMD was semi-automatically segmented in each volume in 500 equally spaced angular segments distributed over 2p radians in the frontal plane.
Within each patient, all angular recordings of PIMD as a function of angle in the frontal plane were rotation adjusted to the average rotation using cross correlation. Then, the circumference of the ONH was split into 250 pairs of adjacent PIMDS and the average variance for occasions, volumes, segmentations and adjacent PIMDS were estimated with a mixed model analysis of variance.

Results : The relative magnitudes of the variances for occasions, volumes, segmentations and adjacent PIMDS were 5, 24, 35 and 1 respectively, indicating that averaging over PIMDS does not reduce the variability at each angular segment significantly but as many volumes and segmentations as possible should be averaged. A subject specific critical limit for excessive loss of PIMD at a defined angle between two occasions can be statistically defined as a tolerance limit on the condition that the variance for PIMD at each angle is known with a sufficient precision. The variance for PIMD at each angle was estimated for each subject (d.f. = 19 500) assuming that it is clinically feasible to capture 3 volumes at each occasion and segment each volume once. The 90 % tolerance limit below the reference level of PIMD at a defined angle varied among subjects in the interval [7;26] mm or [0.03;0.12] of the average PIMD. It is suggested that the first recording(s) of angular PIMD is used as a reference for comparison with subsequent occasions until PIMD becomes lower than the critical limit at any angle. Then, the contrast between sector(s) of suspect PIMD reduction compared to the average of sectors with no apparent reduction is followed over time to detect loss of bundles of ganglion cell axons.

Conclusions : A strategy for early morphometric detection of loss of bundles of nerve fibers in glaucoma patients has been developed.

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

×
×

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.

×