BMC Ophthalmology (Jan 2018)

Segmentation error in spectral domain optical coherence tomography measures of the retinal nerve fibre layer thickness in idiopathic intracranial hypertension

  • Anuriti Aojula,
  • Susan P Mollan,
  • John Horsburgh,
  • Andreas Yiangou,
  • Kiera A Markey,
  • James L Mitchell,
  • William J Scotton,
  • Pearse A Keane,
  • Alexandra J Sinclair

DOI
https://doi.org/10.1186/s12886-017-0652-7
Journal volume & issue
Vol. 17, no. 1
pp. 1 – 7

Abstract

Read online

Abstract Background Optical Coherence Tomography (OCT) imaging is being increasingly used in clinical practice for the monitoring of papilloedema. The aim is to characterise the extent and location of the Retinal Nerve Fibre Layer (RNFL) Thickness automated segmentation error (SegE) by manual refinement, in a cohort of Idiopathic Intracranial Hypertension (IIH) patients with papilloedema and compare this to controls. Methods Baseline Spectral Domain OCT (SDOCT) scans from patients with IIH, and controls with no retinal or optic nerve pathology, were examined. The internal limiting membrane and RNFL thickness of the most severely affected eye was examined for SegE and re-segmented. Using ImageJ, the total area of the RNFL thickness was calculated pre and post re-segmentation and the percentage change was determined. The distribution of RNFL thickness error was qualitatively assessed. Results Significantly greater SegE (p = 0.009) was present in RNFL thickness total area, assessed using ImageJ, in IIH patients (n = 46, 5% ± 0–58%) compared to controls (n = 14, 1% ± 0–6%). This was particularly evident in moderate to severe optic disc swelling (n = 23, 10% ± 0–58%, p < 0.001). RNFL thickness was unable to be quantified using SDOCT in patients with severe papilloedema. Conclusions SegE remain a concern for clinicians using SDOCT to monitor papilloedema in IIH, particularly in the assessment of eyes with moderate to severe oedema. Systematic assessment and manual refinement of SegE is therefore important to ensure the accuracy in longitudinal monitoring of patients.

Keywords