Summary of Point Transformer with Federated Learning for Predicting Breast Cancer HER2 Status from Hematoxylin and Eosin-Stained Whole Slide Images
This work addresses breast cancer diagnosis by improving HER2 prediction from standard histopathology images, potentially aiding clinicians in faster and cheaper treatment decisions, but it is incremental as it builds on existing federated learning and transformer methods.
The study tackled predicting HER2 status from HE-stained whole slide images to reduce costs and speed up treatment decisions, achieving state-of-the-art performance across four sites with 2687 WSIs and strong generalization to two unseen sites with 229 WSIs.
This study introduces a federated learning-based approach to predict HER2 status from hematoxylin and eosin (HE)-stained whole slide images (WSIs), reducing costs and speeding up treatment decisions. To address label imbalance and feature representation challenges in multisite datasets, a point transformer is proposed, incorporating dynamic label distribution, an auxiliary classifier, and farthest cosine sampling. Extensive experiments demonstrate state-of-the-art performance across four sites (2687 WSIs) and strong generalization to two unseen sites (229 WSIs).