Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation

Deep learning techniques for 3D brain vessel image segmentation have not been as successful as in the segmentation of other organs and tissues. This can be explained by two factors. First, deep learning techniques tend to show poor performances at the segmentation of relatively small objects compare...

ver descrição completa

Detalhes bibliográficos
Autores: Ngoc Dang, Vien, Galati, Francesco, Cortese, Rosa, Di Giacomo, Giuseppe, Marconetto, Viola, Mathur, Prateek, Lekadir, Karim, 1977-, Lorenzi, Marco, Prados, Ferran, Zuluaga, Maria A.
Formato: artículo
Estado:Versión publicada
Fecha de publicación:2022
País:España
Recursos:Universidad de Barcelona
Repositorio:Dipòsit Digital de la UB
OAI Identifier:oai:diposit.ub.edu:2445/190550
Acesso em linha:https://hdl.handle.net/2445/190550
Access Level:acceso abierto
Palavra-chave:Aprenentatge automàtic
Processament digital d'imatges
Diagnòstic per la imatge
Machine learning
Digital image processing
Diagnostic imaging
id ES_bba2ae69fb71e93f1b5912df5131b6fb
oai_identifier_str oai:diposit.ub.edu:2445/190550
network_acronym_str ES
network_name_str España
repository_id_str
spelling Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentationNgoc Dang, VienGalati, FrancescoCortese, RosaDi Giacomo, GiuseppeMarconetto, ViolaMathur, PrateekLekadir, Karim, 1977-Lorenzi, MarcoPrados, FerranZuluaga, Maria A.Aprenentatge automàticProcessament digital d'imatgesDiagnòstic per la imatgeMachine learningDigital image processingDiagnostic imagingDeep learning techniques for 3D brain vessel image segmentation have not been as successful as in the segmentation of other organs and tissues. This can be explained by two factors. First, deep learning techniques tend to show poor performances at the segmentation of relatively small objects compared to the size of the full image. Second, due to the complexity of vascular trees and the small size of vessels, it is challenging to obtain the amount of annotated training data typically needed by deep learning methods. To address these problems, we propose a novel annotation-efficient deep learning vessel segmentation framework. The framework avoids pixel-wise annotations, only requiring weak patch-level labels to discriminate between vessel and non-vessel 2D patches in the training set, in a setup similar to the CAPTCHAs used to differentiate humans from bots in web applications. The user-provided weak annotations are used for two tasks: (1) to synthesize pixel-wise pseudo-labels for vessels and background in each patch, which are used to train a segmentation network, and (2) to train a classifier network. The classifier network allows to generate additional weak patch labels, further reducing the annotation burden, and it acts as a second opinion for poor quality images. We use this framework for the segmentation of the cerebrovascular tree in Time-of-Flight angiography (TOF) and Susceptibility-Weighted Images (SWI). The results show that the framework achieves state-of-the-art accuracy, while reducing the annotation time by ∼77% w.r.t. learning-based segmentation methods using pixel-wise labels for training.Elsevier2022info:eu-repo/semantics/articleinfo:eu-repo/semantics/publishedVersionapplication/pdfhttps://hdl.handle.net/2445/190550Articles publicats en revistes (Matemàtiques i Informàtica)reponame:Dipòsit Digital de la UBinstname:Universidad de BarcelonaInglésReproducció del document publicat a: https://doi.org/10.1016/j.media.2021.102263Medical Image Analysis, 2022, vol. 75, num. 102263https://doi.org/10.1016/j.media.2021.102263cc-by (c) Ngoc Dang, Vien et al., 2022http://creativecommons.org/licenses/by/3.0/es/info:eu-repo/semantics/openAccessoai:diposit.ub.edu:2445/1905502026-05-27T06:46:51Z
dc.title.none.fl_str_mv Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
title Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
spellingShingle Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
Ngoc Dang, Vien
Aprenentatge automàtic
Processament digital d'imatges
Diagnòstic per la imatge
Machine learning
Digital image processing
Diagnostic imaging
title_short Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
title_full Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
title_fullStr Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
title_full_unstemmed Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
title_sort Vessel-CAPTCHA: An efficient learning framework for vessel annotation and segmentation
dc.creator.none.fl_str_mv Ngoc Dang, Vien
Galati, Francesco
Cortese, Rosa
Di Giacomo, Giuseppe
Marconetto, Viola
Mathur, Prateek
Lekadir, Karim, 1977-
Lorenzi, Marco
Prados, Ferran
Zuluaga, Maria A.
author Ngoc Dang, Vien
author_facet Ngoc Dang, Vien
Galati, Francesco
Cortese, Rosa
Di Giacomo, Giuseppe
Marconetto, Viola
Mathur, Prateek
Lekadir, Karim, 1977-
Lorenzi, Marco
Prados, Ferran
Zuluaga, Maria A.
author_role author
author2 Galati, Francesco
Cortese, Rosa
Di Giacomo, Giuseppe
Marconetto, Viola
Mathur, Prateek
Lekadir, Karim, 1977-
Lorenzi, Marco
Prados, Ferran
Zuluaga, Maria A.
author2_role author
author
author
author
author
author
author
author
author
dc.subject.none.fl_str_mv Aprenentatge automàtic
Processament digital d'imatges
Diagnòstic per la imatge
Machine learning
Digital image processing
Diagnostic imaging
topic Aprenentatge automàtic
Processament digital d'imatges
Diagnòstic per la imatge
Machine learning
Digital image processing
Diagnostic imaging
description Deep learning techniques for 3D brain vessel image segmentation have not been as successful as in the segmentation of other organs and tissues. This can be explained by two factors. First, deep learning techniques tend to show poor performances at the segmentation of relatively small objects compared to the size of the full image. Second, due to the complexity of vascular trees and the small size of vessels, it is challenging to obtain the amount of annotated training data typically needed by deep learning methods. To address these problems, we propose a novel annotation-efficient deep learning vessel segmentation framework. The framework avoids pixel-wise annotations, only requiring weak patch-level labels to discriminate between vessel and non-vessel 2D patches in the training set, in a setup similar to the CAPTCHAs used to differentiate humans from bots in web applications. The user-provided weak annotations are used for two tasks: (1) to synthesize pixel-wise pseudo-labels for vessels and background in each patch, which are used to train a segmentation network, and (2) to train a classifier network. The classifier network allows to generate additional weak patch labels, further reducing the annotation burden, and it acts as a second opinion for poor quality images. We use this framework for the segmentation of the cerebrovascular tree in Time-of-Flight angiography (TOF) and Susceptibility-Weighted Images (SWI). The results show that the framework achieves state-of-the-art accuracy, while reducing the annotation time by ∼77% w.r.t. learning-based segmentation methods using pixel-wise labels for training.
publishDate 2022
dc.date.none.fl_str_mv 2022
dc.type.none.fl_str_mv info:eu-repo/semantics/article
info:eu-repo/semantics/publishedVersion
format article
status_str publishedVersion
dc.identifier.none.fl_str_mv https://hdl.handle.net/2445/190550
url https://hdl.handle.net/2445/190550
dc.language.none.fl_str_mv Inglés
language_invalid_str_mv Inglés
dc.relation.none.fl_str_mv Reproducció del document publicat a: https://doi.org/10.1016/j.media.2021.102263
Medical Image Analysis, 2022, vol. 75, num. 102263
https://doi.org/10.1016/j.media.2021.102263
dc.rights.none.fl_str_mv cc-by (c) Ngoc Dang, Vien et al., 2022
http://creativecommons.org/licenses/by/3.0/es/
info:eu-repo/semantics/openAccess
rights_invalid_str_mv cc-by (c) Ngoc Dang, Vien et al., 2022
http://creativecommons.org/licenses/by/3.0/es/
eu_rights_str_mv openAccess
dc.format.none.fl_str_mv application/pdf
dc.publisher.none.fl_str_mv Elsevier
publisher.none.fl_str_mv Elsevier
dc.source.none.fl_str_mv Articles publicats en revistes (Matemàtiques i Informàtica)
reponame:Dipòsit Digital de la UB
instname:Universidad de Barcelona
instname_str Universidad de Barcelona
reponame_str Dipòsit Digital de la UB
collection Dipòsit Digital de la UB
repository.name.fl_str_mv
repository.mail.fl_str_mv
_version_ 1869418043578253312
score 15,300719

Registos relacionados