Segmentation of subretinal hyperreflective material and pigment epithelial detachment using kernel graph cut

Melinda Katona, Attila Kovács, Rózsa Dégi, L. Nyúl

Research output: Chapter in Book/Report/Conference proceedingChapter

Abstract

Optical Coherence Tomography (OCT) is one of the most advanced, non-invasive method of eye examination. Age-related macular degeneration (AMD) is one of the most frequent reasons of acquired blindness and it has two forms. Our aim is to develop automatic methods that can accurately identify and characterize biomarkers in SD-OCT images, related to wet AMD. Detection of biomarkers can be challenging because of their variable shape, size, location and reflectivity. In this paper, we present an automatic method to localize subretinal hyperreflective material (SHRM) and pigment epithelial detachment (PED) via kernel graph cut. The proposed method is evaluated using an annotated dataset by ophthalmologists. The Dice coefficient was 0.81 (±0.11) in the case of PED and 0.77 (±0.11) for SHRM. In many cases, the ophthalmologist cannot clearly determine the exact location and extent of the biomarkers, so our achieved results are promising.

Original languageEnglish
Title of host publicationAdvances in Intelligent Systems and Computing
PublisherSpringer Verlag
Pages98-105
Number of pages8
DOIs
Publication statusPublished - Jan 1 2020

Publication series

NameAdvances in Intelligent Systems and Computing
Volume977
ISSN (Print)2194-5357

    Fingerprint

Keywords

  • Age-related macular degeneration
  • Optical coherence tomography
  • Pigment epithelial detachment
  • Subretinal hyperreflective material

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Computer Science(all)

Cite this

Katona, M., Kovács, A., Dégi, R., & Nyúl, L. (2020). Segmentation of subretinal hyperreflective material and pigment epithelial detachment using kernel graph cut. In Advances in Intelligent Systems and Computing (pp. 98-105). (Advances in Intelligent Systems and Computing; Vol. 977). Springer Verlag. https://doi.org/10.1007/978-3-030-19738-4_11