July 2018
Volume 59, Issue 9
Open Access
ARVO Annual Meeting Abstract  |   July 2018
Semi-Automated Segmentation of Hyperreflective Foci on Optical Coherence Tomography
Author Affiliations & Notes
  • Philip DeSouza
    Ophthalmology, Wake Forest School of Medicine, Winston-Salem, North Carolina, United States
  • Scott Walter
    Ophthalmology, Duke University Medical Center, Durham, North Carolina, United States
  • David Cunefare
    Biomedical Engineering, Duke University, Durham, North Carolina, United States
  • Rajiv Shah
    Ophthalmology, Wake Forest School of Medicine, Winston-Salem, North Carolina, United States
  • Sina Farsiu
    Biomedical Engineering, Duke University, Durham, North Carolina, United States
    Ophthalmology, Duke University Medical Center, Durham, North Carolina, United States
  • Glenn J Jaffe
    Ophthalmology, Duke University Medical Center, Durham, North Carolina, United States
  • Footnotes
    Commercial Relationships   Philip DeSouza, None; Scott Walter, None; David Cunefare, None; Rajiv Shah, None; Sina Farsiu, None; Glenn Jaffe, None
  • Footnotes
    Support  Research to Prevent Blindness
Investigative Ophthalmology & Visual Science July 2018, Vol.59, 4855. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Philip DeSouza, Scott Walter, David Cunefare, Rajiv Shah, Sina Farsiu, Glenn J Jaffe; Semi-Automated Segmentation of Hyperreflective Foci on Optical Coherence Tomography. Invest. Ophthalmol. Vis. Sci. 2018;59(9):4855.

      Download citation file:


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

      ×
  • Supplements
Abstract

Purpose : Hyperreflective foci (HF) are frequently observed in optical coherence tomography (OCT) images of patients with diabetic macular edema (DME) and may represent aggregated microglial cells or subvisible hard exudates. Reduction of HF has been observed as early as one month following intravitreal anti-vascular endothelial growth factor (anti-VEGF) injection. Previous investigators relied upon subjective manual grading of HF on a single foveal OCT line scan. We tested the validity of a semi-automated segmentation method for quantitation of HF on OCT raster volume scans.

Methods : Macular OCT images were obtained on 20 patients with DME at baseline and following 52 weeks of protocol-guided anti-VEGF treatment. All patients were concurrently imaged with both 7 and 49-line raster scanning protocols. Image grading was performed retrospectively by two independent, masked analysts using a custom-designed graphical user interface (GUI). Each grader marked the foveal center and adjusted the threshold intensity for each individual line scan to optimally detect HF within a 1 mm diameter region of interest (ROI) centered on the fovea. We calculated the HF area within the ROI on all scans. Intra- and inter-grader analysis included Pearson correlations and paired t-tests.

Results : 548 individual B-scans within the ROI from 80 OCT raster image stacks were analyzed by each grader. HF were detected on 81% of individual B-scans. Cumulative HF area was greater on 49 versus 7-line raster images (0.018 vs. 0.012 μm2, p=0.00009). Nevertheless, there was high intra-grader correlation between 49 and 7-line scans for HF areas (r2 = 0.70). Additionally, there was high inter-grader agreement for both HF area and change in HF area (Table 1). The average change in HF area was -4200 μm2 which represented a 24% reduction from baseline.

Conclusions : Our GUI allows for semi-automated, quantitative, volumetric assessment of HF in DME and was validated on two different OCT raster scanning protocols with high intra- and inter-grader reproducibility. While 49-line scans were more sensitive to detect HF, 7-line scans showed better inter-grader reproducibility. A 24% reduction in HF area was observed over 1 year of anti-VEGF treatment. These results will serve as the basis for a more extensive study examining the relationship between HF, anti-VEGF treatment, and visual function in DME.

This is an abstract that was submitted for the 2018 ARVO Annual Meeting, held in Honolulu, Hawaii, April 29 - May 3, 2018.

 

×
×

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.

×