Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm

Abstract In this study we developed a deep learning (DL) algorithm that detects errors in retinal never fibre layer (RNFL) segmentation on spectral-domain optical coherence tomography (SDOCT) B-scans using human grades as the reference standard. A dataset of 25,250 SDOCT B-scans reviewed for segment...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Alessandro A. Jammal, Atalie C. Thompson, Nara G. Ogata, Eduardo B. Mariottoni, Carla N. Urata, Vital P. Costa, Felipe A. Medeiros
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2019
Materias:	Medicine R Science Q
Acceso en línea:	https://doaj.org/article/ff59cbf4e58c4006b56258a938b3adfb
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:ff59cbf4e58c4006b56258a938b3adfb
record_format	dspace
spelling	oai:doaj.org-article:ff59cbf4e58c4006b56258a938b3adfb2021-12-02T15:09:00ZDetecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm10.1038/s41598-019-46294-62045-2322https://doaj.org/article/ff59cbf4e58c4006b56258a938b3adfb2019-07-01T00:00:00Zhttps://doi.org/10.1038/s41598-019-46294-6https://doaj.org/toc/2045-2322Abstract In this study we developed a deep learning (DL) algorithm that detects errors in retinal never fibre layer (RNFL) segmentation on spectral-domain optical coherence tomography (SDOCT) B-scans using human grades as the reference standard. A dataset of 25,250 SDOCT B-scans reviewed for segmentation errors by human graders was randomly divided into validation plus training (50%) and test (50%) sets. The performance of the DL algorithm was evaluated in the test sample by outputting a probability of having a segmentation error for each B-scan. The ability of the algorithm to detect segmentation errors was evaluated with the area under the receiver operating characteristic (ROC) curve. Mean DL probabilities of segmentation error in the test sample were 0.90 ± 0.17 vs. 0.12 ± 0.22 (P < 0.001) for scans with and without segmentation errors, respectively. The DL algorithm had an area under the ROC curve of 0.979 (95% CI: 0.974 to 0.984) and an overall accuracy of 92.4%. For the B-scans with severe segmentation errors in the test sample, the DL algorithm was 98.9% sensitive. This algorithm can help clinicians and researchers review images for artifacts in SDOCT tests in a timely manner and avoid inaccurate diagnostic interpretations.Alessandro A. JammalAtalie C. ThompsonNara G. OgataEduardo B. MariottoniCarla N. UrataVital P. CostaFelipe A. MedeirosNature PortfolioarticleMedicineRScienceQENScientific Reports, Vol 9, Iss 1, Pp 1-9 (2019)
institution	DOAJ
collection	DOAJ
language	EN
topic	Medicine R Science Q
spellingShingle	Medicine R Science Q Alessandro A. Jammal Atalie C. Thompson Nara G. Ogata Eduardo B. Mariottoni Carla N. Urata Vital P. Costa Felipe A. Medeiros Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm
description	Abstract In this study we developed a deep learning (DL) algorithm that detects errors in retinal never fibre layer (RNFL) segmentation on spectral-domain optical coherence tomography (SDOCT) B-scans using human grades as the reference standard. A dataset of 25,250 SDOCT B-scans reviewed for segmentation errors by human graders was randomly divided into validation plus training (50%) and test (50%) sets. The performance of the DL algorithm was evaluated in the test sample by outputting a probability of having a segmentation error for each B-scan. The ability of the algorithm to detect segmentation errors was evaluated with the area under the receiver operating characteristic (ROC) curve. Mean DL probabilities of segmentation error in the test sample were 0.90 ± 0.17 vs. 0.12 ± 0.22 (P < 0.001) for scans with and without segmentation errors, respectively. The DL algorithm had an area under the ROC curve of 0.979 (95% CI: 0.974 to 0.984) and an overall accuracy of 92.4%. For the B-scans with severe segmentation errors in the test sample, the DL algorithm was 98.9% sensitive. This algorithm can help clinicians and researchers review images for artifacts in SDOCT tests in a timely manner and avoid inaccurate diagnostic interpretations.
format	article
author	Alessandro A. Jammal Atalie C. Thompson Nara G. Ogata Eduardo B. Mariottoni Carla N. Urata Vital P. Costa Felipe A. Medeiros
author_facet	Alessandro A. Jammal Atalie C. Thompson Nara G. Ogata Eduardo B. Mariottoni Carla N. Urata Vital P. Costa Felipe A. Medeiros
author_sort	Alessandro A. Jammal
title	Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm
title_short	Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm
title_full	Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm
title_fullStr	Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm
title_full_unstemmed	Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm
title_sort	detecting retinal nerve fibre layer segmentation errors on spectral domain-optical coherence tomography with a deep learning algorithm
publisher	Nature Portfolio
publishDate	2019
url	https://doaj.org/article/ff59cbf4e58c4006b56258a938b3adfb
work_keys_str_mv	AT alessandroajammal detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm AT ataliecthompson detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm AT naragogata detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm AT eduardobmariottoni detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm AT carlanurata detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm AT vitalpcosta detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm AT felipeamedeiros detectingretinalnervefibrelayersegmentationerrorsonspectraldomainopticalcoherencetomographywithadeeplearningalgorithm
_version_	1718387977644670976

Detecting Retinal Nerve Fibre Layer Segmentation Errors on Spectral Domain-Optical Coherence Tomography with a Deep Learning Algorithm

Ejemplares similares