Reproducibility of Retinal Vascular Phenotypes Obtained with Optical Coherence Tomography Angiography: Importance of Vessel Segmentation

Darwon Rashid*, Sophie Cai, Ylenia Giarratano, Calum Gray, Charlene Hamid, Dilraj S. Grewal, Tom MacGillivray, Sharon Fekrat, Cason B. Robbins, Srinath Soundararajan, Justin P. Ma, Miguel O. Bernabeu

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

Optical coherence tomography angiography (OCTA) is a non-invasive imaging method that can visualize the finest vascular networks in the human retina. OCTA image analysis has been successfully applied to the investigation of retinal vascular diseases of the eye and other systemic conditions that may manifest in the eye. To characterize and distinguish OCTA images from different pathologies, it is important to identify quantitative metrics and phenotypes that have high reproducibility and are not overly susceptible to the effects of imaging artifacts. This paper demonstrates the reproducibility of several recently demonstrated candidate OCTA quantitative metrics: mean curvature and tortuosity of the whole, foveal, superior, nasal, inferior, and temporal regions; foveal and parafoveal vessel skeleton density; and finally, foveal avascular zone area and acircularity index. This paper also highlights the importance of vessel segmentation choice on reproducibility using two different segmentation methods: optimally oriented flux and Frangi filter.

Original languageEnglish
Title of host publicationMedical Image Understanding and Analysis - 25th Annual Conference, MIUA 2021, Proceedings
EditorsBartłomiej W. Papież, Mohammad Yaqub, Jianbo Jiao, Ana I. Namburete, J. Alison Noble
PublisherSpringer
Pages238-249
Number of pages12
ISBN (Electronic)978-3-030-80432-9
ISBN (Print)9783030804312
DOIs
Publication statusPublished - 6 Jul 2021
Event25th Annual Conference on Medical Image Understanding and Analysis, MIUA 2021 - Virtual, Online
Duration: 12 Jul 202114 Jul 2021

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12722 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference25th Annual Conference on Medical Image Understanding and Analysis, MIUA 2021
CityVirtual, Online
Period12/07/2114/07/21

Keywords / Materials (for Non-textual outputs)

  • OCTA imaging
  • Reproducibility
  • Retinal vascular phenotype

Fingerprint

Dive into the research topics of 'Reproducibility of Retinal Vascular Phenotypes Obtained with Optical Coherence Tomography Angiography: Importance of Vessel Segmentation'. Together they form a unique fingerprint.

Cite this