Federated learning enables big data for rare cancer boundary detection

• Verfasst von: Pati, Sarthak [VerfasserIn]
• Verfasst von: Vollmuth, Philipp [VerfasserIn]
• Verfasst von: Brugnara, Gianluca [VerfasserIn]
• Verfasst von: Sahm, Felix [VerfasserIn]
• Verfasst von: Maier-Hein, Klaus H. [VerfasserIn]
• Verfasst von: Bendszus, Martin [VerfasserIn]
• Verfasst von: Wick, Wolfgang [VerfasserIn]
• Titel: Federated learning enables big data for rare cancer boundary detection
• Verf.angabe: Sarthak Pati, Ujjwal Baid, Brandon Edwards, Micah Sheller, Shih-Han Wang, G. Anthony Reina, Patrick Foley, Alexey Gruzdev, Deepthi Karkada, Christos Davatzikos, Chiharu Sako, Satyam Ghodasara, Michel Bilello, Suyash Mohan, Philipp Vollmuth, Gianluca Brugnara, Chandrakanth J. Preetha, Felix Sahm, Klaus Maier-Hein, Maximilian Zenk, Martin Bendszus, Wolfgang Wick [und 257 weitere Personen]
• E-Jahr: 2022
• Jahr: 05 December 2022
• Umfang: 17 S.
• Fussnoten: Gesehen am 18.07.2023
• Titel Quelle: Enthalten in: Nature Communications
• Ort Quelle: [London] : Nature Publishing Group UK, 2010
• Jahr Quelle: 2022
• Band/Heft Quelle: 13(2022) vom: Dez., Artikel-ID 7346, Seite 1-17
• ISSN Quelle: 2041-1723
• DOI: doi:10.1038/s41467-022-33407-5
• Datenträger: Online-Ressource
• Sprache: eng
• Sach-SW: Biomedical engineering
• Sach-SW: CNS cancer
• Sach-SW: Computer science
• Sach-SW: Medical imaging
• Sach-SW: Medical research
• K10plus-PPN: 185291808X

Abstract

Although machine learning (ML) has shown promise across disciplines, out-of-sample generalizability is concerning. This is currently addressed by sharing multi-site data, but such centralization is challenging/infeasible to scale due to various limitations. Federated ML (FL) provides an alternative paradigm for accurate and generalizable ML, by only sharing numerical model updates. Here we present the largest FL study to-date, involving data from 71 sites across 6 continents, to generate an automatic tumor boundary detector for the rare disease of glioblastoma, reporting the largest such dataset in the literature (n = 6, 314). We demonstrate a 33% delineation improvement for the surgically targetable tumor, and 23% for the complete tumor extent, over a publicly trained model. We anticipate our study to: 1) enable more healthcare studies informed by large diverse data, ensuring meaningful results for rare diseases and underrepresented populations, 2) facilitate further analyses for glioblastoma by releasing our consensus model, and 3) demonstrate the FL effectiveness at such scale and task-complexity as a paradigm shift for multi-site collaborations, alleviating the need for data-sharing.