Cross-institutional HER2 assessment via a computer-aided system using federated learning and stain composition augmentation
C.-H. Yang, Y.-A. Chen, S.-Y. Chang, Y.-H. Hsieh, Y.-L. Hung, Y.-W. Lin, Y.-H. Lee, C.-H. Lin, Y.-C. Lin, Y.-S. Lu, Y.-Y. Lin
This paper develops a HER2 assessment system using federated learning and stain composition augmentation, achieving over 90% accuracy across two institutions with improved generalizability.
bioRxiv · Jan 17, 2024 · PreprintPathology