Cargando…

Swarm learning for decentralized artificial intelligence in cancer histopathology

Artificial 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...

Descripción completa

Detalles Bibliográficos
Autores principales: Saldanha, Oliver Lester, Quirke, Philip, West, Nicholas P., James, Jacqueline A., Loughrey, Maurice B., Grabsch, Heike I., Salto-Tellez, Manuel, Alwers, Elizabeth, Cifci, Didem, Ghaffari Laleh, Narmin, Seibel, Tobias, Gray, Richard, Hutchins, Gordon G. A., Brenner, Hermann, van Treeck, Marko, Yuan, Tanwei, Brinker, Titus J., Chang-Claude, Jenny, Khader, Firas, Schuppert, Andreas, Luedde, Tom, Trautwein, Christian, Muti, Hannah Sophie, Foersch, Sebastian, Hoffmeister, Michael, Truhn, Daniel, Kather, Jakob Nikolas
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Nature Publishing Group US 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9205774/
https://www.ncbi.nlm.nih.gov/pubmed/35469069
http://dx.doi.org/10.1038/s41591-022-01768-5
Descripción
Sumario:Artificial 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.