Swarm learning for decentralized artificial intelligence in cancer histopathology

A decentralized, privacy-preserving machine learning framework used to train a clinically relevant AI system identifies actionable molecular alterations in patients with colorectal cancer by use of routine histopathology slides collected in real-world settings. 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.

Automated summary

This study demonstrates the importance of using a decentralized, privacy-preserving machine learning framework for histopathology image analysis. By utilizing swarm learning, AI models can be trained on routine histopathology slides collected in real-world settings to predict molecular alterations in patients. The results show that AI models trained with swarm learning outperform most locally trained models and perform on par with models trained on merged datasets.