Uncertainty meets 3D-spatial feature in colonoscopic polyp-size determination

This paper presents a new 3D-spatial feature extraction from a 2D colonoscopic image for polyp-size estimation. The polyp-size estimation poses potential demands on colonoscopy, since an endoscopist's subjective estimation is apt to result in uncertain polyp-size determination. This uncertain d...

Full description

Saved in:

Bibliographic Details
Published in:	Computer methods in biomechanics and biomedical engineering. Vol. 10; no. 3; pp. 289 - 298
Main Authors:	Itoh, Hayato, Oda, Masahiro, Jiang, Kai, Mori, Yuichi, Misawa, Masashi, Kudo, Shin-Ei, Imai, Kenichiro, Ito, Sayo, Hotta, Kinichi, Mori, Kensaku
Format:	Journal Article
Language:	English
Published:	Taylor & Francis 04.05.2022
Subjects:	computer-aided diagnosis depth estimation polyp localisation Polyp-size estimation
ISSN:	2168-1163, 2168-1171
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Abstract	This paper presents a new 3D-spatial feature extraction from a 2D colonoscopic image for polyp-size estimation. The polyp-size estimation poses potential demands on colonoscopy, since an endoscopist's subjective estimation is apt to result in uncertain polyp-size determination. This uncertain determination is derived from the lack of 3D-spatial information. Even though a previous work clarified that depth estimation mitigates uncertainty in binary polyp-size classification, precise estimation of 3D structure from a colonoscopic image(s) remains an unsolved challenge in medical image analysis. This work proposes a 3D-spatial feature that expresses a polyp's precise 3D shape to mitigate uncertainty in polyp-size determination. First, we introduce an accurate depth estimation method to capture the 3D structure of a colon. Next, we integrate depth estimation and polyp localisation to extract a 3D polyp shape as a feature. Finally, we achieve polyp-size estimation by statistical learning of extracted features. The experimental results demonstrated the validity both of our depth estimation and 3D-spatial feature. Compared with deep RGB and RGB-D convolutional neural networks (CNNs), a shallow CNN with the proposed 3D-spatial feature achieved a more accurate polyp-size estimation with a mean absolute error of 1.36 mm, whereas the one of the deep CNN is 3.11 mm.
AbstractList	This paper presents a new 3D-spatial feature extraction from a 2D colonoscopic image for polyp-size estimation. The polyp-size estimation poses potential demands on colonoscopy, since an endoscopist's subjective estimation is apt to result in uncertain polyp-size determination. This uncertain determination is derived from the lack of 3D-spatial information. Even though a previous work clarified that depth estimation mitigates uncertainty in binary polyp-size classification, precise estimation of 3D structure from a colonoscopic image(s) remains an unsolved challenge in medical image analysis. This work proposes a 3D-spatial feature that expresses a polyp's precise 3D shape to mitigate uncertainty in polyp-size determination. First, we introduce an accurate depth estimation method to capture the 3D structure of a colon. Next, we integrate depth estimation and polyp localisation to extract a 3D polyp shape as a feature. Finally, we achieve polyp-size estimation by statistical learning of extracted features. The experimental results demonstrated the validity both of our depth estimation and 3D-spatial feature. Compared with deep RGB and RGB-D convolutional neural networks (CNNs), a shallow CNN with the proposed 3D-spatial feature achieved a more accurate polyp-size estimation with a mean absolute error of 1.36 mm, whereas the one of the deep CNN is 3.11 mm.
Author	Itoh, Hayato Hotta, Kinichi Ito, Sayo Oda, Masahiro Misawa, Masashi Imai, Kenichiro Jiang, Kai Mori, Yuichi Kudo, Shin-Ei Mori, Kensaku
Author_xml	– sequence: 1 givenname: Hayato surname: Itoh fullname: Itoh, Hayato email: hitoh@mori.m.is.nagoya-u.ac.jp organization: Graduate School of Informatics, Nagoya University – sequence: 2 givenname: Masahiro surname: Oda fullname: Oda, Masahiro organization: Nagoya University – sequence: 3 givenname: Kai surname: Jiang fullname: Jiang, Kai organization: Graduate School of Informatics, Nagoya University – sequence: 4 givenname: Yuichi surname: Mori fullname: Mori, Yuichi organization: Showa University Northern Yokohama Hospital – sequence: 5 givenname: Masashi surname: Misawa fullname: Misawa, Masashi organization: Showa University Northern Yokohama Hospital – sequence: 6 givenname: Shin-Ei surname: Kudo fullname: Kudo, Shin-Ei organization: Showa University Northern Yokohama Hospital – sequence: 7 givenname: Kenichiro surname: Imai fullname: Imai, Kenichiro organization: Shizuoka Cancer Center – sequence: 8 givenname: Sayo surname: Ito fullname: Ito, Sayo organization: Shizuoka Cancer Center – sequence: 9 givenname: Kinichi surname: Hotta fullname: Hotta, Kinichi organization: Shizuoka Cancer Center – sequence: 10 givenname: Kensaku surname: Mori fullname: Mori, Kensaku organization: National Institute of Informatics
BookMark	eNqFkM9KAzEQh4NUsNY-grAvsDWz2WS7eFHqX6h4seeQzU4gspssSUTWp7e11YMHncPMMPD9GL5TMnHeISHnQBdAl_SiALEEEGxR0AK2jZZlyY_IdHfPASqY_OyCnZB5jK90W0shmOBT8rRxGkNS1qUx6xFTzNhNHgeVrOoygyq9Bcysy7TvvPNR-8HqbPDdOOTRfmDWYsLQW7cFvDsjx0Z1EeeHOSObu9uX1UO-fr5_XF2vc80qkXKsy9qwpuFY8rZqFTIB0LR1Dao1VUOxwoKj0A03nFXQ6pIWyhRImwYYUzWbEb7P1cHHGNDIIdhehVEClTst8luL3GmRBy1b7vIXp236-jwFZbt_6as9bZ3xoVfvPnStTGrsfDBBOW2jZH9HfAKnNH6-
CitedBy_id	crossref_primary_10_1109_ACCESS_2025_3527720 crossref_primary_10_1055_a_2502_9733
Cites_doi	10.1053/j.gastro.2012.06.001 10.1007/s11548-019-01962-w 10.1007/s11548-021-02398-x 10.1016/j.dld.2010.12.015 10.1109/3DV.2016.32 10.1007/978-3-030-32254-0_64 10.1109/ICCV.2019.00393 10.1016/j.gie.2020.07.060 10.1007/978-3-030-00934-2_68 10.1109/CVPR.2018.00907 10.1049/htl.2019.0071 10.1016/j.gie.2015.06.058 10.1109/CVPR42600.2020.00475 10.1053/j.gastro.2006.03.012 10.1055/s-0042-112580 10.1109/ICCV.2015.304 10.1109/ICRA.2018.8460184 10.1109/CVPR.2017.195 10.1055/s-0042-108727 10.1109/CVPR.2017.700 10.1016/j.gie.2015.08.082 10.1109/CVPR.2017.238 10.1109/CVPR.2019.01252 10.1109/CVPR.2018.00216 10.1117/12.480417 10.1109/ICCV.2017.244 10.1055/s-0033-1344548 10.1016/j.media.2018.06.005
ContentType	Journal Article
Copyright	2021 Informa UK Limited, trading as Taylor & Francis Group 2021
Copyright_xml	– notice: 2021 Informa UK Limited, trading as Taylor & Francis Group 2021
DBID	AAYXX CITATION
DOI	10.1080/21681163.2021.2004445
DatabaseName	CrossRef
DatabaseTitle	CrossRef
DatabaseTitleList
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	2168-1171
EndPage	298
ExternalDocumentID	10_1080_21681163_2021_2004445 2004445
Genre	Research Article
GroupedDBID	0BK 30N 4.4 AAGDL AAJMT AALDU AAMIU AAPUL AAQRR ABLIJ ABPAQ ABXUL ABXYU ACGFS ADCVX ADGTB ADMLS AEISY AFRVT AGDLA AHDZW AIJEM AIYEW AKBVH AKOOK ALMA_UNASSIGNED_HOLDINGS ALQZU AQTUD ARCSS BLEHA CCCUG EBS EUPTU GTTXZ H13 HZ~ KYCEM LJTGL M4Z O9- RIG RNANH ROSJB RTWRZ SNACF SOJIQ TBQAZ TDBHL TEN TFL TFT TFW TTHFI TUROJ AAYXX CITATION
ID	FETCH-LOGICAL-c376t-e949f3bb5e45d7dae3611bd991adf7b0e7e25e6cb5f5371dc402af2e0bb133a93
IEDL.DBID	TFW
ISICitedReferencesCount	5
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000746807700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	2168-1163
IngestDate	Sat Nov 29 06:34:10 EST 2025 Tue Nov 18 21:03:47 EST 2025 Mon Oct 20 23:46:52 EDT 2025
IsPeerReviewed	true
IsScholarly	true
Issue	3
Language	English
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c376t-e949f3bb5e45d7dae3611bd991adf7b0e7e25e6cb5f5371dc402af2e0bb133a93
PageCount	10
ParticipantIDs	informaworld_taylorfrancis_310_1080_21681163_2021_2004445 crossref_citationtrail_10_1080_21681163_2021_2004445 crossref_primary_10_1080_21681163_2021_2004445
PublicationCentury	2000
PublicationDate	2022-05-04
PublicationDateYYYYMMDD	2022-05-04
PublicationDate_xml	– month: 05 year: 2022 text: 2022-05-04 day: 04
PublicationDecade	2020
PublicationTitle	Computer methods in biomechanics and biomedical engineering.
PublicationYear	2022
Publisher	Taylor & Francis
Publisher_xml	– name: Taylor & Francis
References	cit0011 cit0033 cit0012 cit0034 cit0031 cit0010 cit0032 cit0030 cit0019 cit0017 cit0018 Hartley R (cit0006) 2003 cit0015 cit0016 Goodfellow I (cit0005) 2014 cit0013 cit0014 cit0022 cit0001 cit0023 cit0020 cit0021 Szegedy C (cit0029) 2017 Saxena A (cit0028) 2006; 18 cit0008 cit0009 cit0007 cit0004 cit0026 cit0027 cit0002 cit0024 cit0003 cit0025
References_xml	– ident: cit0016 doi: 10.1053/j.gastro.2012.06.001 – ident: cit0026 doi: 10.1007/s11548-019-01962-w – volume-title: Multiple view geometry in computer vision year: 2003 ident: cit0006 – ident: cit0011 doi: 10.1007/s11548-021-02398-x – ident: cit0009 doi: 10.1016/j.dld.2010.12.015 – ident: cit0015 doi: 10.1109/3DV.2016.32 – ident: cit0018 doi: 10.1007/978-3-030-32254-0_64 – ident: cit0004 doi: 10.1109/ICCV.2019.00393 – ident: cit0027 – ident: cit0021 doi: 10.1016/j.gie.2020.07.060 – volume: 18 start-page: 1161 year: 2006 ident: cit0028 publication-title: Adv Neural Inf Process Syst – ident: cit0012 doi: 10.1007/978-3-030-00934-2_68 – ident: cit0034 doi: 10.1109/CVPR.2018.00907 – ident: cit0023 doi: 10.1049/htl.2019.0071 – ident: cit0001 doi: 10.1016/j.gie.2015.06.058 – ident: cit0020 doi: 10.1109/CVPR42600.2020.00475 – ident: cit0031 doi: 10.1053/j.gastro.2006.03.012 – ident: cit0008 doi: 10.1055/s-0042-112580 – start-page: 4278 volume-title: Proc. Thirty-First AAAI Conference on Artificial Intelligence year: 2017 ident: cit0029 – ident: cit0003 doi: 10.1109/ICCV.2015.304 – ident: cit0017 doi: 10.1109/ICRA.2018.8460184 – ident: cit0002 doi: 10.1109/CVPR.2017.195 – ident: cit0010 – start-page: 2672 volume-title: Advances in neural information processing systems year: 2014 ident: cit0005 – ident: cit0024 doi: 10.1055/s-0042-108727 – ident: cit0032 doi: 10.1109/CVPR.2017.700 – ident: cit0013 doi: 10.1016/j.gie.2015.08.082 – ident: cit0014 doi: 10.1109/CVPR.2017.238 – ident: cit0025 doi: 10.1109/CVPR.2019.01252 – ident: cit0030 doi: 10.1109/CVPR.2018.00216 – ident: cit0022 doi: 10.1117/12.480417 – ident: cit0033 doi: 10.1109/ICCV.2017.244 – ident: cit0007 doi: 10.1055/s-0033-1344548 – ident: cit0019 doi: 10.1016/j.media.2018.06.005
SSID	ssj0000866365
Score	2.22588
Snippet	This paper presents a new 3D-spatial feature extraction from a 2D colonoscopic image for polyp-size estimation. The polyp-size estimation poses potential...
SourceID	crossref informaworld
SourceType	Enrichment Source Index Database Publisher
StartPage	289
SubjectTerms	computer-aided diagnosis depth estimation polyp localisation Polyp-size estimation
Title	Uncertainty meets 3D-spatial feature in colonoscopic polyp-size determination
URI	https://www.tandfonline.com/doi/abs/10.1080/21681163.2021.2004445
Volume	10
WOSCitedRecordID	wos000746807700001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVAWR databaseName: Taylor and Francis Online Journals customDbUrl: eissn: 2168-1171 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0000866365 issn: 2168-1163 databaseCode: TFW dateStart: 20130301 isFulltext: true titleUrlDefault: https://www.tandfonline.com providerName: Taylor & Francis
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3JTsMwELVQxQEO7IiyKQeuhjpekhwRUHGAikMreou8jKVIJa2agFS-HjtL1R6AA3zAi5KZiWfGevMGoSthXBXCTYIFJRa7jM9wLE2MIysMj7TUTFWefooGg3g8Tl4aNmHR0Cp9D21roYjqrPY_t1RFy4i7CYmIiasjXHcX1i0eY37M3FX2PsaH_dflLYsr2AWt9kl6EPaodoznuwetJag1-dKVxNPf_YdX3kM7TdUZ3NZhso82ID9A2ytahIfoeeTcX9EDykXwBlAWAb3HhSdcO6SFSgA0yPLAy1znUz_NkulgNp0sZrjIPiEwLbHGu_oIjfoPw7tH3OxawNodMSWGhCWWKsWBcRMZCVQQooyrHqWxkepBBCEHoRW3nEbEaNd3ShtCTynX5cqEHqNOPs3hxE-BM-0yngAlLAu1kTQW1rgPlkQqJm0XsdbAqW6EyP0-jElKGr3S1lqpt1baWKuLrpewWa3E8RsgWfVeWlZXILbeV5LSH7Gnf8Ceoa3Qz0h4ViQ7R51y_g4XaFN_lFkxv6zi8wsJNeCw
linkProvider	Taylor & Francis
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3JTsMwELWgIAEHdkRZc-BqqOMlyREBVRFtT63oLfIqVSpp1QSk8vXYWar0ABzgA16UzDieRW_eAHDDlM1CqIogw8hAG_EJDLkKYWCYooHkkojc092g3w9Ho6g-C-Nola6GNoVQRH5Xu5_bNaMrStydj1iIbCJhyzu_qPEIoetgg9pY62h9g_brss9iU3aG842SDgUdrBrk-e5JKyFqRcC0Fnrae__x0vtgt0w8vfvipByANZ0cgp2aHOER6A3tCcgZAtnCe9M6Sz38CFPHubZIo3MNUG-ceE7pOpm6gZax9GbTyWIG0_Gn9lTFrXHePgbD9tPgoQPLdQtQ2lsmgzoikcFCUE2oChTXmCEklE0guTKBaOlA-1QzKaihOEBK2tKTG1-3hLCFLo_wCWgk00SfukFwIm3QY1owQ3ypOA6ZUfaDOeKCcNMEpLJwLEstcrcSYxKjUrK0slbsrBWX1mqC2yVsVohx_AaI6u6Ls7wLYoqVJTH-EXv2B-w12OoMet24-9x_OQfbvhuZcCRJcgEa2fxdX4JN-ZGN0_lVfli_AD6e5NE
linkToPdf	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwrV3JTsMwELWgIAQHdkRZc-BqwLHjJEdEiUCUqodW9BZ5lSKVNGoCUvl67CxVewAO8AEvSmYmnhnrzRsArqg0VYgnQ0gx0tBkfAIDJgPoayo9XzBBeOnprt_rBaNR2K_ZhHlNq7Q9tK6EIsqz2v7cmdQNI-7GRTRApo4w3Z1btXiEeKtgrRTHMiE9iF7n1yymYqe4XChpUdDCmjme7560lKGW9EsXMk-08w_vvAu267LTuaviZA-sqHQfbC2IER6Al6Hxf8kPKGbOm1JF7uAOzC3j2iC1KhVAnSR1rM51OrHjLIlwssl4lsE8-VSObJg11teHYBg9DO4fYb1sAQpzxhRQhSTUmHNPEU_6kilMEeLSlI9Map_fKl-5nqKCe9rDPpLCNJ5Mu-qWc9PmshAfgVY6SdWxHQMnwqQ8qjjVxBWS4YBqaT6YIcYJ021AGgPHolYitwsxxjGqBUsba8XWWnFtrTa4nsOySorjN0C46L24KO9AdLWwJMY_Yk_-gL0EG_1OFHefes-nYNO18xKWIUnOQKuYvqtzsC4-iiSfXpSh-gV1auN1
openUrl	ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Uncertainty+meets+3D-spatial+feature+in+colonoscopic+polyp-size+determination&rft.jtitle=Computer+methods+in+biomechanics+and+biomedical+engineering.&rft.au=Itoh%2C+Hayato&rft.au=Oda%2C+Masahiro&rft.au=Jiang%2C+Kai&rft.au=Mori%2C+Yuichi&rft.date=2022-05-04&rft.issn=2168-1163&rft.eissn=2168-1171&rft.volume=10&rft.issue=3&rft.spage=289&rft.epage=298&rft_id=info:doi/10.1080%2F21681163.2021.2004445&rft.externalDBID=n%2Fa&rft.externalDocID=10_1080_21681163_2021_2004445
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=2168-1163&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=2168-1163&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=2168-1163&client=summon