Show simple item record

SurvMaximin: Robust federated approach to transporting survival risk prediction models

dc.rights.licenseopenen_US
dc.contributor.authorWANG, Xuan
dc.contributor.authorZHANG, Harrison G.
dc.contributor.authorXIONG, Xin
dc.contributor.authorHONG, Chuan
dc.contributor.authorWEBER, Griffin M.
dc.contributor.authorBRAT, Gabriel A.
dc.contributor.authorBONZEL, Clara-Lea
dc.contributor.authorLUO, Yuan
dc.contributor.authorDUAN, Rui
dc.contributor.authorPALMER, Nathan P.
dc.contributor.authorHUTCH, Meghan R.
dc.contributor.authorGUTIERREZ-SACRISTAN, Alba
dc.contributor.authorBELLAZZI, Riccardo
dc.contributor.authorCHIOVATO, Luca
dc.contributor.authorCHO, Kelly
dc.contributor.authorDAGLIATI, Arianna
dc.contributor.authorESTIRI, Hossein
dc.contributor.authorGARCIA-BARRIO, Noelia
hal.structure.identifierBordeaux population health [BPH]
dc.contributor.authorGRIFFIER, Romain
dc.contributor.authorHANAUER, David A.
dc.contributor.authorHO, Yuk-Lam
dc.contributor.authorHOLMES, John H.
dc.contributor.authorKELLER, Mark S.
dc.contributor.authorKLANN MENG, Jeffrey G.
dc.contributor.authorL'YI, Sehi
dc.contributor.authorLOZANO-ZAHONERO, Sara
dc.contributor.authorMAIDLOW, Sarah E.
dc.contributor.authorMAKOUDJOU, Adeline
dc.contributor.authorMALOVINI, Alberto
dc.contributor.authorMOAL, Bertrand
dc.contributor.authorMOORE, Jason H.
dc.contributor.authorMORRIS, Michele
dc.contributor.authorMOWERY, Danielle L.
dc.contributor.authorMURPHY, Shawn N.
dc.contributor.authorNEURAZ, Antoine
dc.contributor.authorYUAN NGIAM, Kee
dc.contributor.authorOMENN, Gilbert S.
dc.contributor.authorPATEL, Lav P.
dc.contributor.authorPEDRERA-JIMENEZ, Miguel
dc.contributor.authorPRUNOTTO, Andrea
dc.contributor.authorJEBATHILAGAM SAMAYAMUTHU, Malarkodi
dc.contributor.authorSANZ VIDORRETA, Fernando J.
dc.contributor.authorSCHRIVER, Emily R.
dc.contributor.authorSCHUBERT, Petra
dc.contributor.authorSERRANO-BALAZOTE, Pablo
dc.contributor.authorSOUTH, Andrew M.
dc.contributor.authorTAN, Amelia L. M.
dc.contributor.authorTAN, Byorn W. L.
dc.contributor.authorTIBOLLO, Valentina
dc.contributor.authorTIPPMANN, Patric
dc.contributor.authorVISWESWARAN, Shyam
dc.contributor.authorXIA, Zongqi
dc.contributor.authorYUAN, William
dc.contributor.authorZOLLER, Daniela
dc.contributor.authorKOHANE, Isaac S.
dc.contributor.authorAVILLACH, Paul
dc.contributor.authorGUO, Zijian
dc.contributor.authorCAI, Tianxi
dc.date.accessioned2022-10-29T10:20:11Z
dc.date.available2022-10-29T10:20:11Z
dc.date.issued2022-08-23
dc.identifier.issn1532-0480 (Electronic) 1532-0464 (Linking)en_US
dc.identifier.urihttps://oskar-bordeaux.fr/handle/20.500.12278/170159
dc.description.abstractEnOBJECTIVE: For multi-center heterogeneous Real-World Data (RWD) with time-to-event outcomes and high-dimensional features, we propose the SurvMaximin algorithm to estimate Cox model feature coefficients for a target population by borrowing summary information from a set of health care centers without sharing patient-level information. MATERIALS AND METHODS: For each of the centers from which we want to borrow information to improve the prediction performance for the target population, a penalized Cox model is fitted to estimate feature coefficients for the center. Using estimated feature coefficients and the covariance matrix of the target population, we then obtain a SurvMaximin estimated set of feature coefficients for the target population. The target population can be an entire cohort comprised of all centers, corresponding to federated learning, or a single center, corresponding to transfer learning. RESULTS: Simulation studies and a real-world international electronic health records application study, with 15 participating health care centers across three countries (France, Germany, and the U.S.), show that the proposed SurvMaximin algorithm achieves comparable or higher accuracy compared with the estimator using only the information of the target site and other existing methods. The SurvMaximin estimator is robust to variations in sample sizes and estimated feature coefficients between centers, which amounts to significantly improved estimates for target sites with fewer observations. CONCLUSIONS: The SurvMaximin method is well suited for both federated and transfer learning in the high-dimensional survival analysis setting. SurvMaximin only requires a one-time summary information exchange from participating centers. Estimated regression vectors can be very heterogeneous. SurvMaximin provides robust Cox feature coefficient estimates without outcome information in the target population and is privacy-preserving.
dc.language.isoENen_US
dc.title.enSurvMaximin: Robust federated approach to transporting survival risk prediction models
dc.typeArticle de revueen_US
dc.identifier.doi10.1016/j.jbi.2022.104176en_US
dc.subject.halSciences du Vivant [q-bio]/Santé publique et épidémiologieen_US
dc.identifier.pubmed36007785en_US
bordeaux.journalJournal of Biomedical Informaticsen_US
bordeaux.volume134en_US
bordeaux.hal.laboratoriesBordeaux Population Health Research Center (BPH) - UMR 1219en_US
bordeaux.institutionUniversité de Bordeauxen_US
bordeaux.institutionINSERMen_US
bordeaux.teamAHEAD_BPHen_US
bordeaux.peerReviewedouien_US
bordeaux.inpressnonen_US
hal.identifierhal-03834307
hal.version1
hal.date.transferred2022-10-29T10:20:29Z
hal.exporttrue
dc.rights.ccPas de Licence CCen_US
bordeaux.COinSctx_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.jtitle=Journal%20of%20Biomedical%20Informatics&rft.date=2022-08-23&rft.volume=134&rft.eissn=1532-0480%20(Electronic)%201532-0464%20(Linking)&rft.issn=1532-0480%20(Electronic)%201532-0464%20(Linking)&rft.au=WANG,%20Xuan&ZHANG,%20Harrison%20G.&XIONG,%20Xin&HONG,%20Chuan&WEBER,%20Griffin%20M.&rft.genre=article


Files in this item

FilesSizeFormatView

There are no files associated with this item.

This item appears in the following Collection(s)

Show simple item record