June 2021
Volume 62, Issue 8
Open Access
ARVO Annual Meeting Abstract  |   June 2021
AI estimation of the non-pathological loss rate for the waist of the nerve fiber layer in the optic nerve head
Author Affiliations & Notes
  • Konstancija Kisonaite
    Neuroscience, Ophthalmology, Gullstrand lab, Uppsala Universitet, Uppsala, Uppsala, Sweden
  • Zhaohua Yu
    Neuroscience, Ophthalmology, Gullstrand lab, Uppsala Universitet, Uppsala, Uppsala, Sweden
  • Camilla Sandberg Melin
    Neuroscience, Ophthalmology, Gullstrand lab, Uppsala Universitet, Uppsala, Uppsala, Sweden
    Ophthalmology clinic, Gavle sjukhus, Gavle, Sweden
  • Gabriel Carrizo
    Biomedical engineering and health systems, Kungliga Tekniska Hogskolan, Stockholm, Sweden
  • Chunliang Wang
    Biomedical engineering and health systems, Kungliga Tekniska Hogskolan, Stockholm, Sweden
  • Per G Soderberg
    Neuroscience, Ophthalmology, Gullstrand lab, Uppsala Universitet, Uppsala, Uppsala, Sweden
  • Footnotes
    Commercial Relationships   Konstancija Kisonaite, None; Zhaohua Yu, None; Camilla Sandberg Melin, None; Gabriel Carrizo, None; Chunliang Wang, None; Per Soderberg, None
  • Footnotes
    Support  Föreningen Synskadades Vänner i Uppsala Län, Kronprinsessan Margaretas Arbetsnämnd för synskadade, Ögonfonden, The Uppsala University/Uppsala Läns Landsting’s ALF Research grants, Wallinders stiftelse, Erik Funks minnesfond, Vinova, AIDA
Investigative Ophthalmology & Visual Science June 2021, Vol.62, 999. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Konstancija Kisonaite, Zhaohua Yu, Camilla Sandberg Melin, Gabriel Carrizo, Chunliang Wang, Per G Soderberg; AI estimation of the non-pathological loss rate for the waist of the nerve fiber layer in the optic nerve head. Invest. Ophthalmol. Vis. Sci. 2021;62(8):999.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : There is agreement on pathological loss rate of nerve fiber layer (NFL) in the optic nerve head (ONH) in glaucomatous individuals, as well as non-pathological due to aging. Morphometric measures of the NFL allow for assessment of the loss rate. Previous studies of non-glaucomatous individuals have reported a normal loss rate ranging from -1 to -2 µm/year. So far, a fully automatic method has never been used in studies for morphometrical measurement of the NFL.
Our clinical experimental study was performed with the use of an in-house developed AI-algorithm for automatical annotation and measurement of the NFL waist along the ONH, to investigate whether there is an age-related NFL loss rate in non-glaucomatous subjects.

Methods : Altogether 16 non-glaucomatous individuals were enrolled and equally stratified according to sex and age in four age groups of [30 39], [40 49], [50 59], [60 69] years. Three left-eye volumes were obtained from each participant. In each volume 500 annotations of the central limit of pigmental epithelium and their shortest distances to the inner limit of the retina were detected by the AI algorithm. The average distance was calculated in each volume. Then, all three volumes were averaged for each subject.
ANOVA and simple linear regression analysis were performed to evaluate the obtained data.

Results : Automatic measurement of one volume took approximately 1 minute. The 95% CI for the mean in age groups of [30 39], [40 49], [50 59], [60 69] years were estimated as 376.0±33.5, 292.3±51.4, 364.6±61.6 and 344.3±49.9 µm.
Variation among volumes and subjects was 7.7 µm2 (d.f.= 32) and 2758 µm2 (d.f.= 12) respectively. In a linear regression graph a negative trend was found, equivalent to a NFL loss of -2.43 µm/year.

Conclusions : Our automatical algorithm successfully annotated and computed the NFL waist thickness and produced results in line with current research, although general conclusions about normal loss rate in the population are difficult to draw due to our small sample size.
A fully automated approach to track changes in NFL thickness over time for glaucoma follow-up is possible. Comparison of loss rate in the glaucomatous and non-glaucomatous population enables new glaucoma follow-up method to detect pathological changes earlier than current clinical perimetry.

This is a 2021 ARVO Annual Meeting abstract.

×
×

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.

×