Label Cleaning Multiple Instance Learning: Refining Coarse Annotations on Single Whole-Slide Images

Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing o...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	IEEE transactions on medical imaging Ročník 41; číslo 12; s. 3952 - 3968
Hlavní autoři:	Wang, Zhenzhen, Saoud, Carla, Wangsiricharoen, Sintawat, James, Aaron W., Popel, Aleksander S., Sulam, Jeremias
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	United States IEEE 01.12.2022 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Témata:	Algorithms Annotations Breast cancer Breast Neoplasms - pathology Cancer Cleaning coarse annotations Colorectal carcinoma Female Humans Image segmentation Impact prediction label cleaning Learning algorithms Liver cancer Lymph nodes Lymphatic Metastasis Machine Learning Medical imaging Medical research Metastases multiple instance learning Pathology Prediction models Predictive models Refining Training Tumors Whole-slide image segmentation
ISSN:	0278-0062, 1558-254X, 1558-254X
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Abstract	Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing only coarse and approximate annotations is a much easier task, less costly, and it alleviates pathologists' workload. In this paper, we study the problem of refining these approximate annotations in digital pathology to obtain more accurate ones. Some previous works have explored obtaining machine learning models from these inaccurate annotations, but few of them tackle the refinement problem where the mislabeled regions should be explicitly identified and corrected, and all of them require a - often very large - number of training samples. We present a method, named Label Cleaning Multiple Instance Learning (LC-MIL), to refine coarse annotations on a single WSI without the need for external training data. Patches cropped from a WSI with inaccurate labels are processed jointly within a multiple instance learning framework, mitigating their impact on the predictive model and refining the segmentation. Our experiments on a heterogeneous WSI set with breast cancer lymph node metastasis, liver cancer, and colorectal cancer samples show that LC-MIL significantly refines the coarse annotations, outperforming state-of-the-art alternatives, even while learning from a single slide. Moreover, we demonstrate how real annotations drawn by pathologists can be efficiently refined and improved by the proposed approach. All these results demonstrate that LC-MIL is a promising, lightweight tool to provide fine-grained annotations from coarsely annotated pathology sets.
AbstractList	Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing only coarse and approximate annotations is a much easier task, less costly, and it alleviates pathologists’ workload. In this paper, we study the problem of refining these approximate annotations in digital pathology to obtain more accurate ones. Some previous works have explored obtaining machine learning models from these inaccurate annotations, but few of them tackle the refinement problem where the mislabeled regions should be explicitly identified and corrected, and all of them require a – often very large – number of training samples. We present a method, named Label Cleaning Multiple Instance Learning (LC-MIL), to refine coarse annotations on a single WSI without the need for external training data. Patches cropped from a WSI with inaccurate labels are processed jointly within a multiple instance learning framework, mitigating their impact on the predictive model and refining the segmentation. Our experiments on a heterogeneous WSI set with breast cancer lymph node metastasis, liver cancer, and colorectal cancer samples show that LC-MIL significantly refines the coarse annotations, outperforming state-of-the-art alternatives, even while learning from a single slide. Moreover, we demonstrate how real annotations drawn by pathologists can be efficiently refined and improved by the proposed approach. All these results demonstrate that LC-MIL is a promising, lightweight tool to provide fine-grained annotations from coarsely annotated pathology sets. Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing only coarse and approximate annotations is a much easier task, less costly, and it alleviates pathologists' workload. In this paper, we study the problem of refining these approximate annotations in digital pathology to obtain more accurate ones. Some previous works have explored obtaining machine learning models from these inaccurate annotations, but few of them tackle the refinement problem where the mislabeled regions should be explicitly identified and corrected, and all of them require a - often very large - number of training samples. We present a method, named Label Cleaning Multiple Instance Learning (LC-MIL), to refine coarse annotations on a single WSI without the need for external training data. Patches cropped from a WSI with inaccurate labels are processed jointly within a multiple instance learning framework, mitigating their impact on the predictive model and refining the segmentation. Our experiments on a heterogeneous WSI set with breast cancer lymph node metastasis, liver cancer, and colorectal cancer samples show that LC-MIL significantly refines the coarse annotations, outperforming state-of-the-art alternatives, even while learning from a single slide. Moreover, we demonstrate how real annotations drawn by pathologists can be efficiently refined and improved by the proposed approach. All these results demonstrate that LC-MIL is a promising, lightweight tool to provide fine-grained annotations from coarsely annotated pathology sets.Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing only coarse and approximate annotations is a much easier task, less costly, and it alleviates pathologists' workload. In this paper, we study the problem of refining these approximate annotations in digital pathology to obtain more accurate ones. Some previous works have explored obtaining machine learning models from these inaccurate annotations, but few of them tackle the refinement problem where the mislabeled regions should be explicitly identified and corrected, and all of them require a - often very large - number of training samples. We present a method, named Label Cleaning Multiple Instance Learning (LC-MIL), to refine coarse annotations on a single WSI without the need for external training data. Patches cropped from a WSI with inaccurate labels are processed jointly within a multiple instance learning framework, mitigating their impact on the predictive model and refining the segmentation. Our experiments on a heterogeneous WSI set with breast cancer lymph node metastasis, liver cancer, and colorectal cancer samples show that LC-MIL significantly refines the coarse annotations, outperforming state-of-the-art alternatives, even while learning from a single slide. Moreover, we demonstrate how real annotations drawn by pathologists can be efficiently refined and improved by the proposed approach. All these results demonstrate that LC-MIL is a promising, lightweight tool to provide fine-grained annotations from coarsely annotated pathology sets. Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing only coarse and approximate annotations is a much easier task, less costly, and it alleviates pathologists’ workload. In this paper, we study the problem of refining these approximate annotations in digital pathology to obtain more accurate ones. Some previous works have explored obtaining machine learning models from these inaccurate annotations, but few of them tackle the refinement problem where the mislabeled regions should be explicitly identified and corrected, and all of them require a – often very large – number of training samples. We present a method, named Label Cleaning Multiple Instance Learning (LC-MIL), to refine coarse annotations on a single WSI without the need for external training data. Patches cropped from a WSI with inaccurate labels are processed jointly within a multiple instance learning framework, mitigating their impact on the predictive model and refining the segmentation. Our experiments on a heterogeneous WSI set with breast cancer lymph node metastasis, liver cancer, and colorectal cancer samples show that LC-MIL significantly refines the coarse annotations, out-performing state-of-the-art alternatives, even while learning from a single slide. Moreover, we demonstrate how real annotations drawn by pathologists can be efficiently refined and improved by the proposed approach. All these results demonstrate that LC-MIL is a promising, lightweight tool to provide fine-grained annotations from coarsely annotated pathology sets.
Author	Wang, Zhenzhen Wangsiricharoen, Sintawat James, Aaron W. Popel, Aleksander S. Saoud, Carla Sulam, Jeremias
Author_xml	– sequence: 1 givenname: Zhenzhen orcidid: 0000-0002-0195-4362 surname: Wang fullname: Wang, Zhenzhen email: zwang218@jhu.edu organization: Department of Biomedical Engineering, Mathematical Institute of Data Science, Johns Hopkins University, Baltimore, MD, USA – sequence: 2 givenname: Carla surname: Saoud fullname: Saoud, Carla email: csaoud1@jhmi.edu organization: Department of Pathology, Johns Hopkins Medicine, Baltimore, MD, USA – sequence: 3 givenname: Sintawat surname: Wangsiricharoen fullname: Wangsiricharoen, Sintawat email: swangsi1@jhmi.edu organization: Department of Pathology, Johns Hopkins Medicine, Baltimore, MD, USA – sequence: 4 givenname: Aaron W. surname: James fullname: James, Aaron W. email: awjames@jhmi.edu organization: Department of Pathology, Johns Hopkins Medicine, Baltimore, MD, USA – sequence: 5 givenname: Aleksander S. surname: Popel fullname: Popel, Aleksander S. email: apopel@jhu.edu organization: Department of Biomedical Engineering, Johns Hopkins University, Baltimore, MD, USA – sequence: 6 givenname: Jeremias orcidid: 0000-0003-0946-1957 surname: Sulam fullname: Sulam, Jeremias email: jsulam1@jhu.edu organization: Department of Biomedical Engineering, Mathematical Institute of Data Science, Johns Hopkins University, Baltimore, MD, USA
BackLink	https://www.ncbi.nlm.nih.gov/pubmed/36037454$$D View this record in MEDLINE/PubMed
BookMark	eNp9kc2LFDEQxYOsuLOjd0GQBi9eekw6H93xICyDHwOzCO6K3kJ1pno2SyaZ7XQL_vemnXHQPXhJIPV7L1X1LshZiAEJec7ogjGq39xcrRYVraoFz2ct9SMyY1I2ZSXF9zMyy29NSamqzslFSneUMiGpfkLOuaK8FlLMiF1Di75YeoTgwra4Gv3g9h6LVUgDBIvFGqGfSm-LL9i539AyQp-wuAwhDjC4GFIRQ3GdS1n47TZ6LK-922STHWwxPSWPO_AJnx3vOfn64f3N8lO5_vxxtbxcl1YIMZR1KxR2VcesAsF103Kx4VrABjQIUExboVvetp3EmjUbLlHxFmurrJR1B8Dn5N3Bdz-2O9xYDEMP3ux7t4P-p4ngzL-V4G7NNv4wuqnktJI5eX006OP9iGkwO5cseg8B45hMVdNMalarjL56gN7FsQ95vEwJyaqaN5Phy787OrXyZ_8ZoAfA9jGlHrsTwqiZIjY5YjNFbI4RZ4l6ILHukEKeyfn_CV8chA4RT__oRumm0fwX7P-z4w
CODEN	ITMID4
CitedBy_id	crossref_primary_10_1002_jor_25693 crossref_primary_10_1016_j_inffus_2025_103027 crossref_primary_10_1016_j_patter_2025_101178 crossref_primary_10_1016_j_compbiomed_2025_110293 crossref_primary_10_3390_diagnostics14040388 crossref_primary_10_3390_jlpea15030054 crossref_primary_10_1016_j_bspc_2025_107599 crossref_primary_10_1002_aisy_202300885 crossref_primary_10_1016_j_jpi_2024_100367 crossref_primary_10_1109_TMI_2024_3351213 crossref_primary_10_1186_s12859_024_06007_x crossref_primary_10_1016_j_jpi_2024_100363 crossref_primary_10_1109_TMI_2024_3485120 crossref_primary_10_1007_s10489_023_05004_6 crossref_primary_10_1016_j_compmedimag_2024_102337 crossref_primary_10_1016_j_pt_2024_05_005 crossref_primary_10_1109_RBME_2024_3357877
Cites_doi	10.1016/j.media.2019.03.009 10.1007/978-3-030-78191-0_52 10.1109/WACV.2018.00138 10.3389/fonc.2018.00649 10.1109/ISBI.2018.8363599 10.1016/j.media.2020.101813 10.1016/j.canlet.2014.10.021 10.3389/fmed.2019.00310 10.1109/TIP.2017.2689998 10.1093/gigascience/giy065 10.1016/j.neucom.2020.03.127 10.1038/s41551-020-00682-w 10.1007/s10916-009-9299-0 10.1007/978-3-030-33391-1_24 10.1111/j.1365-2559.2011.04154.x 10.1038/s41591-019-0508-1 10.1609/aaai.v31i1.10894 10.1038/modpathol.2012.235 10.1097/PAP.0000000000000273 10.1016/j.media.2020.101854 10.1109/ICCV.2019.01078 10.1371/journal.pone.0070221 10.3389/fphys.2020.583333 10.1038/modpathol.3800496 10.1109/TBME.2017.2690863 10.1186/1471-2342-6-14 10.1109/TSMC.1979.4310076 10.1016/j.patcog.2017.08.026 10.1016/j.media.2014.01.010 10.1016/j.ajpath.2019.05.007 10.1126/science.aaa6204 10.3389/fonc.2018.00627 10.1016/S0004-3702(96)00034-3 10.1109/CBMI.2015.7153607 10.1007/978-3-030-32239-7_60 10.1109/TSMC.1972.4309137 10.1109/RBME.2009.2034865 10.1109/CVPR.2009.5206848 10.1109/ACCESS.2021.3086020 10.1109/TMI.2017.2724070 10.1109/ISBI.2018.8363518 10.1109/CVPR.2018.00571
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2022
DBID	97E ESBDL RIA RIE AAYXX CITATION CGR CUY CVF ECM EIF NPM 7QF 7QO 7QQ 7SC 7SE 7SP 7SR 7TA 7TB 7U5 8BQ 8FD F28 FR3 H8D JG9 JQ2 KR7 L7M L~C L~D NAPCQ P64 7X8 5PM
DOI	10.1109/TMI.2022.3202759
DatabaseName	IEEE Xplore (IEEE) IEEE Xplore Open Access Journals IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Electronic Library (IEL) CrossRef Medline MEDLINE MEDLINE (Ovid) MEDLINE MEDLINE PubMed Aluminium Industry Abstracts Biotechnology Research Abstracts Ceramic Abstracts Computer and Information Systems Abstracts Corrosion Abstracts Electronics & Communications Abstracts Engineered Materials Abstracts Materials Business File Mechanical & Transportation Engineering Abstracts Solid State and Superconductivity Abstracts METADEX Technology Research Database ANTE: Abstracts in New Technology & Engineering Engineering Research Database Aerospace Database Materials Research Database ProQuest Computer Science Collection Civil Engineering Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional Nursing & Allied Health Premium Biotechnology and BioEngineering Abstracts MEDLINE - Academic PubMed Central (Full Participant titles)
DatabaseTitle	CrossRef MEDLINE Medline Complete MEDLINE with Full Text PubMed MEDLINE (Ovid) Materials Research Database Civil Engineering Abstracts Aluminium Industry Abstracts Technology Research Database Computer and Information Systems Abstracts – Academic Mechanical & Transportation Engineering Abstracts Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Ceramic Abstracts Materials Business File METADEX Biotechnology and BioEngineering Abstracts Computer and Information Systems Abstracts Professional Aerospace Database Nursing & Allied Health Premium Engineered Materials Abstracts Biotechnology Research Abstracts Solid State and Superconductivity Abstracts Engineering Research Database Corrosion Abstracts Advanced Technologies Database with Aerospace ANTE: Abstracts in New Technology & Engineering MEDLINE - Academic
DatabaseTitleList	Materials Research Database MEDLINE - Academic MEDLINE
Database_xml	– sequence: 1 dbid: NPM name: PubMed url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed sourceTypes: Index Database – sequence: 2 dbid: RIE name: IEEE Electronic Library (IEL) url: https://ieeexplore.ieee.org/ sourceTypes: Publisher – sequence: 3 dbid: 7X8 name: MEDLINE - Academic url: https://search.proquest.com/medline sourceTypes: Aggregation Database
DeliveryMethod	fulltext_linktorsrc
Discipline	Medicine Engineering
EISSN	1558-254X
EndPage	3968
ExternalDocumentID	PMC9825360 36037454 10_1109_TMI_2022_3202759 9869889
Genre	orig-research Research Support, Non-U.S. Gov't Journal Article Research Support, N.I.H., Extramural
GrantInformation_xml	– fundername: CISCO Research grantid: CG2686384 – fundername: NIH grantid: R01CA138264; U01CA212007 funderid: 10.13039/100000002 – fundername: NCI NIH HHS grantid: U01 CA212007 – fundername: NCI NIH HHS grantid: R01 CA138264
GroupedDBID	--- -DZ -~X .GJ 0R~ 29I 4.4 53G 5GY 5RE 5VS 6IK 97E AAJGR AARMG AASAJ AAWTH ABAZT ABQJQ ABVLG ACGFO ACGFS ACIWK ACNCT ACPRK AENEX AETIX AFRAH AGQYO AGSQL AHBIQ AI. AIBXA AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 DU5 EBS EJD ESBDL F5P HZ~ H~9 IBMZZ ICLAB IFIPE IFJZH IPLJI JAVBF LAI M43 MS~ O9- OCL P2P PQQKQ RIA RIE RNS RXW TAE TN5 VH1 AAYXX CITATION AAYOK CGR CUY CVF ECM EIF NPM RIG 7QF 7QO 7QQ 7SC 7SE 7SP 7SR 7TA 7TB 7U5 8BQ 8FD F28 FR3 H8D JG9 JQ2 KR7 L7M L~C L~D NAPCQ P64 7X8 5PM
ID	FETCH-LOGICAL-c444t-7b46ef2f1c6a4398b34d394ada9a4a619c49b3bbf5e718d35e63be7c6c557faa3
IEDL.DBID	RIE
ISICitedReferencesCount	15
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000907324600039&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	0278-0062 1558-254X
IngestDate	Tue Sep 30 17:16:20 EDT 2025 Sun Nov 09 14:05:58 EST 2025 Mon Jun 30 03:43:22 EDT 2025 Thu Apr 03 07:11:19 EDT 2025 Sat Nov 29 05:14:10 EST 2025 Tue Nov 18 21:57:15 EST 2025 Wed Aug 27 02:15:09 EDT 2025
IsDoiOpenAccess	true
IsOpenAccess	true
IsPeerReviewed	false
IsScholarly	true
Issue	12
Language	English
License	https://creativecommons.org/licenses/by/4.0/legalcode This work is licensed under a Creative Commons Attribution 4.0 License. For more information, see https://creativecommons.org/licenses/by/4.0
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c444t-7b46ef2f1c6a4398b34d394ada9a4a619c49b3bbf5e718d35e63be7c6c557faa3
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23
ORCID	0000-0003-0946-1957 0000-0002-0195-4362
OpenAccessLink	https://ieeexplore.ieee.org/document/9869889
PMID	36037454
PQID	2745127380
PQPubID	85460
PageCount	17
ParticipantIDs	proquest_journals_2745127380 pubmed_primary_36037454 crossref_primary_10_1109_TMI_2022_3202759 pubmedcentral_primary_oai_pubmedcentral_nih_gov_9825360 ieee_primary_9869889 proquest_miscellaneous_2708259176 crossref_citationtrail_10_1109_TMI_2022_3202759
PublicationCentury	2000
PublicationDate	2022-12-01
PublicationDateYYYYMMDD	2022-12-01
PublicationDate_xml	– month: 12 year: 2022 text: 2022-12-01 day: 01
PublicationDecade	2020
PublicationPlace	United States
PublicationPlace_xml	– name: United States – name: New York
PublicationTitle	IEEE transactions on medical imaging
PublicationTitleAbbrev	TMI
PublicationTitleAlternate	IEEE Trans Med Imaging
PublicationYear	2022
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref57 ref56 ref12 ref15 ref58 ref14 ref53 ref52 ref11 ref10 ref17 ref16 ref19 mukhoti (ref54) 2020 ref18 rister (ref29) 2018 simonyan (ref61) 2015 northcutt (ref33) 2017 ref51 sukhbaatar (ref24) 2014 ref50 ref46 ref48 courtiol (ref49) 2018 ren (ref41) 2018 ref42 ref44 ref43 goldberger (ref25) 2017 ref8 ref7 ref9 ref4 ref3 ref6 ilse (ref47) 2018 ronneberger (ref13) 2015 ref5 cheng (ref23) 2020 ref40 wang (ref36) 2019 ref34 ref31 ref30 köhler (ref35) 0 ref32 ref2 bahri (ref45) 2020 ref1 ref39 shu (ref37) 2019 simonyan (ref59) 2015 ref26 ref20 northcutt (ref55) 2017 ref22 ref21 ref28 ref27 guo (ref38) 2017 ref60
References_xml	– year: 2015 ident: ref61 article-title: Very deep convolutional networks for large-scale image recognition publication-title: Proc ICLR – ident: ref19 doi: 10.1016/j.media.2019.03.009 – start-page: 4334 year: 2018 ident: ref41 article-title: Learning to reweight examples for robust deep learning publication-title: Proc ICML – year: 0 ident: ref35 article-title: Uncertainty based detection and relabeling of noisy image labels – ident: ref32 doi: 10.1007/978-3-030-78191-0_52 – ident: ref34 doi: 10.1109/WACV.2018.00138 – ident: ref6 doi: 10.3389/fonc.2018.00649 – ident: ref28 doi: 10.1109/ISBI.2018.8363599 – start-page: 234 year: 2015 ident: ref13 article-title: U-Net: Convolutional networks for biomedical image segmentation publication-title: Proc MICCAI – ident: ref22 doi: 10.1016/j.media.2020.101813 – start-page: 2127 year: 2018 ident: ref47 article-title: Attention-based deep multiple instance learning publication-title: Proc ICML – ident: ref3 doi: 10.1016/j.canlet.2014.10.021 – ident: ref2 doi: 10.3389/fmed.2019.00310 – ident: ref30 doi: 10.1109/TIP.2017.2689998 – ident: ref56 doi: 10.1093/gigascience/giy065 – ident: ref31 doi: 10.1016/j.neucom.2020.03.127 – ident: ref21 doi: 10.1038/s41551-020-00682-w – ident: ref10 doi: 10.1007/s10916-009-9299-0 – year: 2017 ident: ref55 article-title: Learning with confident examples: Rank pruning for robust classification with noisy labels publication-title: arXiv 1705 01936 – ident: ref43 doi: 10.1007/978-3-030-33391-1_24 – ident: ref17 doi: 10.1111/j.1365-2559.2011.04154.x – ident: ref20 doi: 10.1038/s41591-019-0508-1 – start-page: 540 year: 2020 ident: ref45 article-title: Deep k-NN for noisy labels publication-title: Proc ICML – ident: ref27 doi: 10.1609/aaai.v31i1.10894 – year: 2019 ident: ref36 article-title: Emphasis regularisation by gradient rescaling for training deep neural networks with noisy labels publication-title: arXiv 1905 11233 – ident: ref18 doi: 10.1038/modpathol.2012.235 – ident: ref1 doi: 10.1097/PAP.0000000000000273 – ident: ref57 doi: 10.1016/j.media.2020.101854 – ident: ref52 doi: 10.1109/ICCV.2019.01078 – ident: ref11 doi: 10.1371/journal.pone.0070221 – ident: ref5 doi: 10.3389/fphys.2020.583333 – year: 2018 ident: ref29 article-title: CT organ segmentation using GPU data augmentation, unsupervised labels and IOU loss publication-title: arXiv 1811 11226 – ident: ref16 doi: 10.1038/modpathol.3800496 – year: 2014 ident: ref24 article-title: Training convolutional networks with noisy labels publication-title: arXiv 1406 2080 – ident: ref12 doi: 10.1109/TBME.2017.2690863 – start-page: 1321 year: 2017 ident: ref38 article-title: On calibration of modern neural networks publication-title: Proc Int Conf Mach Learn – ident: ref9 doi: 10.1186/1471-2342-6-14 – year: 2020 ident: ref54 article-title: Calibrating deep neural networks using focal loss publication-title: arXiv 2002 09437 – ident: ref58 doi: 10.1109/TSMC.1979.4310076 – ident: ref50 doi: 10.1016/j.patcog.2017.08.026 – ident: ref48 doi: 10.1016/j.media.2014.01.010 – ident: ref15 doi: 10.1016/j.ajpath.2019.05.007 – start-page: 117 year: 2020 ident: ref23 article-title: Self-similarity student for partial label histopathology image segmentation publication-title: Proc ECCV – year: 2017 ident: ref25 article-title: Training deep neural-networks using a noise adaptation layer publication-title: Proc ICLR – year: 2015 ident: ref59 article-title: Very deep convolutional networks for large-scale image recognition publication-title: Proc ICLR – ident: ref4 doi: 10.1126/science.aaa6204 – ident: ref7 doi: 10.3389/fonc.2018.00627 – ident: ref46 doi: 10.1016/S0004-3702(96)00034-3 – ident: ref39 doi: 10.1109/CBMI.2015.7153607 – ident: ref42 doi: 10.1007/978-3-030-32239-7_60 – ident: ref44 doi: 10.1109/TSMC.1972.4309137 – start-page: 1917 year: 2019 ident: ref37 article-title: Meta-weight-net: Learning an explicit mapping for sample weighting publication-title: Proc NeurIPS – ident: ref8 doi: 10.1109/RBME.2009.2034865 – ident: ref60 doi: 10.1109/CVPR.2009.5206848 – ident: ref53 doi: 10.1016/j.patcog.2017.08.026 – start-page: 1321 year: 2017 ident: ref33 article-title: Learning with confident examples: Rank pruning for robust classification with noisy labels publication-title: Proc UAI – ident: ref14 doi: 10.1109/ACCESS.2021.3086020 – ident: ref51 doi: 10.1109/TMI.2017.2724070 – year: 2018 ident: ref49 article-title: Classification and disease localization in histopathology using only global labels: A weakly-supervised approach publication-title: arXiv 1802 02212 – ident: ref26 doi: 10.1109/ISBI.2018.8363518 – ident: ref40 doi: 10.1109/CVPR.2018.00571
SSID	ssj0014509
Score	2.5347319
Snippet	Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine...
SourceID	pubmedcentral proquest pubmed crossref ieee
SourceType	Open Access Repository Aggregation Database Index Database Enrichment Source Publisher
StartPage	3952
SubjectTerms	Algorithms Annotations Breast cancer Breast Neoplasms - pathology Cancer Cleaning coarse annotations Colorectal carcinoma Female Humans Image segmentation Impact prediction label cleaning Learning algorithms Liver cancer Lymph nodes Lymphatic Metastasis Machine Learning Medical imaging Medical research Metastases multiple instance learning Pathology Prediction models Predictive models Refining Training Tumors Whole-slide image segmentation
Title	Label Cleaning Multiple Instance Learning: Refining Coarse Annotations on Single Whole-Slide Images
URI	https://ieeexplore.ieee.org/document/9869889 https://www.ncbi.nlm.nih.gov/pubmed/36037454 https://www.proquest.com/docview/2745127380 https://www.proquest.com/docview/2708259176 https://pubmed.ncbi.nlm.nih.gov/PMC9825360
Volume	41
WOSCitedRecordID	wos000907324600039&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: PRVIEE databaseName: IEEE Electronic Library (IEL) customDbUrl: eissn: 1558-254X dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014509 issn: 0278-0062 databaseCode: RIE dateStart: 19820101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3fa9RAEB7aImIfrLZao7Ws4Itgekl2sz98K0eLBa-IrXhvYXcz0YMzkd6df7-zSS70ShH6FtjZZcPMZOfLzH4D8F5Zk1Q-TWPUzgWAkpJLWRfbTFte2TSvdNk2m1CXl3o6NV-34ONwFwYR2-IzPAmPbS6_bPwq_CobGS2N1mYbtpWS3V2tIWMg8q6cIwuMsYnM1inJxIyuJxcEBLPsJPQKV3kgCuUy8K7kYuM0atur3Bdp3i2YvHUCne89bO_P4GkfabLTzjSewxbW-7B7i39wHx5P-sz6Afgv1uGcjedow58SNukLDdlFGz96ZD0T689P7BtWbVsJNm4IFiM7reumy-gvWFOzKxqiiT9C5934aj4raZHf9NlavIDv52fX489x34Ah9kKIZayckFhlVeqlpcBFOy5KboQtrbHCEvTywjjuXJUjHXElz1Fyh8pLn-eqspa_hJ26qfEVMEJdxuWcwA5FPCka48rUorVeuipDpyIYrRVR-J6dPDTJmBctSklMQVosghaLXosRfBhm_OmYOf4jexA0Msj1yojgaK3ronfdRUEwnYIgxXUSwbthmJwuZFJsjc0qyARkTUhXRnDYmcaw9tq0IlAbRjMIBELvzZF69qsl9ja0KE1_ff9u38CT8E5dLc0R7CxvVvgWHvm_y9ni5ph8YqqPW5_4B6myCFA
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3fb9MwED6NgRg88GMbkDHASLwgkTWJ7STe21QxraKtECtib5HtnKFSSdDa8vdzTtJoRRMSb5FsnxzdXXxf7vwdwLtMq8jZOA4xN8YDlJhcSptQJ7nmTsfS5WXTbCKbTvOrK_V5Bz70d2EQsSk-wxP_2OTyy9qu_a-ygcpTlefqDtyVgsS2t7X6nIGQbUFH4jljozTZJCUjNZhNRgQFk-TEdwvPpKcK5alnXpFi6zxqGqzcFmv-XTJ54ww6f_x_u38Cj7pYk521xvEUdrDah4c3GAj34f6ky60fgB1rgws2XKD2_0rYpCs1ZKMmgrTIOi7W76fsC7qmsQQb1gSMkZ1VVd3m9JesrtglDdHCb773bni5mJck5Cd9uJaH8PX842x4EXYtGEIrhFiFmREpusTFNtUUuuSGi5IroUuttNAEvqxQhhvjJNIhV3KJKTeY2dRKmTmt-TPYreoKXwAj3KWM5AR3KOaJUSlTxhq1tqlxCZosgMFGEYXt-Ml9m4xF0eCUSBWkxcJrsei0GMD7fsWvlpvjH3MPvEb6eZ0yAjje6LronHdZEFCnMCjjeRTA236Y3M7nUnSF9drP8diasG4awPPWNHrZG9MKINsymn6Cp_TeHqnmPxpqb0VCafnR7bt9A3sXs8m4GI-mn17CA_9-bWXNMeyurtf4Cu7Z36v58vp14xl_AEvPCq8
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=Label+Cleaning+Multiple+Instance+Learning%3A+Refining+Coarse+Annotations+on+Single+Whole-Slide+Images&rft.jtitle=IEEE+transactions+on+medical+imaging&rft.au=Wang%2C+Zhenzhen&rft.au=Saoud%2C+Carla&rft.au=Wangsiricharoen%2C+Sintawat&rft.au=James%2C+Aaron+W.&rft.date=2022-12-01&rft.issn=0278-0062&rft.eissn=1558-254X&rft.volume=41&rft.issue=12&rft.spage=3952&rft.epage=3968&rft_id=info:doi/10.1109%2FTMI.2022.3202759&rft.externalDBID=n%2Fa&rft.externalDocID=10_1109_TMI_2022_3202759
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0278-0062&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0278-0062&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0278-0062&client=summon