000179696 001__ 179696
000179696 005__ 20240229145551.0
000179696 0247_ $$2doi$$a10.1038/s41591-022-01768-5
000179696 0247_ $$2pmid$$apmid:35469069
000179696 0247_ $$2ISSN$$a1078-8956
000179696 0247_ $$2ISSN$$a1546-170X
000179696 0247_ $$2altmetric$$aaltmetric:127264908
000179696 037__ $$aDKFZ-2022-00845
000179696 041__ $$aEnglish
000179696 082__ $$a610
000179696 1001_ $$aSaldanha, Oliver Lester$$b0
000179696 245__ $$aSwarm learning for decentralized artificial intelligence in cancer histopathology.
000179696 260__ $$aNew York, NY$$bNature America Inc.$$c2022
000179696 3367_ $$2DRIVER$$aarticle
000179696 3367_ $$2DataCite$$aOutput Types/Journal article
000179696 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1655905740_23458
000179696 3367_ $$2BibTeX$$aARTICLE
000179696 3367_ $$2ORCID$$aJOURNAL_ARTICLE
000179696 3367_ $$00$$2EndNote$$aJournal Article
000179696 500__ $$a2022 Jun;28(6):1232-1239
000179696 520__ $$aArtificial intelligence (AI) can predict the presence of molecular alterations directly from routine histopathology slides. However, training robust AI systems requires large datasets for which data collection faces practical, ethical and legal obstacles. These obstacles could be overcome with swarm learning (SL), in which partners jointly train AI models while avoiding data transfer and monopolistic data governance. Here, we demonstrate the successful use of SL in large, multicentric datasets of gigapixel histopathology images from over 5,000 patients. We show that AI models trained using SL can predict BRAF mutational status and microsatellite instability directly from hematoxylin and eosin (H&E)-stained pathology slides of colorectal cancer. We trained AI models on three patient cohorts from Northern Ireland, Germany and the United States, and validated the prediction performance in two independent datasets from the United Kingdom. Our data show that SL-trained AI models outperform most locally trained models, and perform on par with models that are trained on the merged datasets. In addition, we show that SL-based AI models are data efficient. In the future, SL can be used to train distributed AI models for any histopathology image analysis task, eliminating the need for data transfer.
000179696 536__ $$0G:(DE-HGF)POF4-313$$a313 - Krebsrisikofaktoren und Prävention (POF4-313)$$cPOF4-313$$fPOF IV$$x0
000179696 588__ $$aDataset connected to CrossRef, PubMed, , Journals: inrepo02.dkfz.de
000179696 7001_ $$00000-0002-3597-5444$$aQuirke, Philip$$b1
000179696 7001_ $$00000-0002-0346-6709$$aWest, Nicholas P$$b2
000179696 7001_ $$00000-0002-6945-6060$$aJames, Jacqueline A$$b3
000179696 7001_ $$aLoughrey, Maurice B$$b4
000179696 7001_ $$00000-0001-9520-6228$$aGrabsch, Heike I$$b5
000179696 7001_ $$aSalto-Tellez, Manuel$$b6
000179696 7001_ $$0P:(DE-He78)9b2a61b2abe4a64ca23b6783b7c4fe63$$aAlwers, Elizabeth$$b7$$udkfz
000179696 7001_ $$00000-0002-2647-9959$$aCifci, Didem$$b8
000179696 7001_ $$aGhaffari Laleh, Narmin$$b9
000179696 7001_ $$aSeibel, Tobias$$b10
000179696 7001_ $$00000-0003-4440-574X$$aGray, Richard$$b11
000179696 7001_ $$aHutchins, Gordon G A$$b12
000179696 7001_ $$0P:(DE-He78)90d5535ff896e70eed81f4a4f6f22ae2$$aBrenner, Hermann$$b13$$udkfz
000179696 7001_ $$avan Treeck, Marko$$b14
000179696 7001_ $$0P:(DE-He78)b9e439a1aa1244925f92d547c0919349$$aYuan, Tanwei$$b15$$udkfz
000179696 7001_ $$0P:(DE-He78)1e33961c8780aca9b76d776d1fdc1ebb$$aBrinker, Titus J$$b16$$udkfz
000179696 7001_ $$0P:(DE-He78)c259d6cc99edf5c7bc7ce22c7f87c253$$aChang-Claude, Jenny$$b17$$udkfz
000179696 7001_ $$aKhader, Firas$$b18
000179696 7001_ $$aSchuppert, Andreas$$b19
000179696 7001_ $$00000-0002-6288-8821$$aLuedde, Tom$$b20
000179696 7001_ $$00000-0003-2762-8247$$aTrautwein, Christian$$b21
000179696 7001_ $$aMuti, Hannah Sophie$$b22
000179696 7001_ $$aFoersch, Sebastian$$b23
000179696 7001_ $$0P:(DE-He78)6c5d058b7552d071a7fa4c5e943fff0f$$aHoffmeister, Michael$$b24$$udkfz
000179696 7001_ $$aTruhn, Daniel$$b25
000179696 7001_ $$00000-0002-3730-5348$$aKather, Jakob Nikolas$$b26
000179696 773__ $$0PERI:(DE-600)1484517-9$$a10.1038/s41591-022-01768-5$$n6$$p1232-1239$$tNature medicine$$v28$$x1078-8956$$y2022
000179696 909CO $$ooai:inrepo02.dkfz.de:179696$$pVDB
000179696 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)9b2a61b2abe4a64ca23b6783b7c4fe63$$aDeutsches Krebsforschungszentrum$$b7$$kDKFZ
000179696 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)90d5535ff896e70eed81f4a4f6f22ae2$$aDeutsches Krebsforschungszentrum$$b13$$kDKFZ
000179696 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)b9e439a1aa1244925f92d547c0919349$$aDeutsches Krebsforschungszentrum$$b15$$kDKFZ
000179696 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)1e33961c8780aca9b76d776d1fdc1ebb$$aDeutsches Krebsforschungszentrum$$b16$$kDKFZ
000179696 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)c259d6cc99edf5c7bc7ce22c7f87c253$$aDeutsches Krebsforschungszentrum$$b17$$kDKFZ
000179696 9101_ $$0I:(DE-588b)2036810-0$$6P:(DE-He78)6c5d058b7552d071a7fa4c5e943fff0f$$aDeutsches Krebsforschungszentrum$$b24$$kDKFZ
000179696 9131_ $$0G:(DE-HGF)POF4-313$$1G:(DE-HGF)POF4-310$$2G:(DE-HGF)POF4-300$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$aDE-HGF$$bGesundheit$$lKrebsforschung$$vKrebsrisikofaktoren und Prävention$$x0
000179696 9141_ $$y2022
000179696 915__ $$0StatID:(DE-HGF)0160$$2StatID$$aDBCoverage$$bEssential Science Indicators$$d2021-02-03
000179696 915__ $$0StatID:(DE-HGF)1190$$2StatID$$aDBCoverage$$bBiological Abstracts$$d2021-02-03
000179696 915__ $$0StatID:(DE-HGF)0113$$2StatID$$aWoS$$bScience Citation Index Expanded$$d2021-02-03
000179696 915__ $$0StatID:(DE-HGF)0420$$2StatID$$aNationallizenz$$d2022-11-17$$wger
000179696 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)1050$$2StatID$$aDBCoverage$$bBIOSIS Previews$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)1110$$2StatID$$aDBCoverage$$bCurrent Contents - Clinical Medicine$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)1030$$2StatID$$aDBCoverage$$bCurrent Contents - Life Sciences$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bNAT MED : 2021$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)0600$$2StatID$$aDBCoverage$$bEbsco Academic Search$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)0030$$2StatID$$aPeer Review$$bASC$$d2022-11-17
000179696 915__ $$0StatID:(DE-HGF)9980$$2StatID$$aIF >= 80$$bNAT MED : 2021$$d2022-11-17
000179696 9201_ $$0I:(DE-He78)C070-20160331$$kC070$$lC070 Klinische Epidemiologie und Alternf.$$x0
000179696 9201_ $$0I:(DE-He78)C120-20160331$$kC120$$lPräventive Onkologie$$x1
000179696 9201_ $$0I:(DE-He78)HD01-20160331$$kHD01$$lDKTK HD zentral$$x2
000179696 9201_ $$0I:(DE-He78)C140-20160331$$kC140$$lNWG Digitale Biomarker in der Onkologie$$x3
000179696 9201_ $$0I:(DE-He78)C020-20160331$$kC020$$lC020 Epidemiologie von Krebs$$x4
000179696 980__ $$ajournal
000179696 980__ $$aVDB
000179696 980__ $$aI:(DE-He78)C070-20160331
000179696 980__ $$aI:(DE-He78)C120-20160331
000179696 980__ $$aI:(DE-He78)HD01-20160331
000179696 980__ $$aI:(DE-He78)C140-20160331
000179696 980__ $$aI:(DE-He78)C020-20160331
000179696 980__ $$aUNRESTRICTED