Robust Learning Protocol for Federated Tumor Segmentation Challenge
This work addresses incremental improvements in federated learning protocols for medical imaging, specifically targeting tumor segmentation in collaborative settings.
The authors tackled the challenges of data heterogeneity and communication costs in federated learning for tumor segmentation by proposing RoLePRO, a two-phase protocol combining server-side adaptive optimization and adaptive weighted aggregation, achieving improved performance on the FeTS 2022 challenge.
In this work, we devise robust and efficient learning protocols for orchestrating a Federated Learning (FL) process for the Federated Tumor Segmentation Challenge (FeTS 2022). Enabling FL for FeTS setup is challenging mainly due to data heterogeneity among collaborators and communication cost of training. To tackle these challenges, we propose Robust Learning Protocol (RoLePRO) which is a combination of server-side adaptive optimisation (e.g., server-side Adam) and judicious parameter (weights) aggregation schemes (e.g., adaptive weighted aggregation). RoLePRO takes a two-phase approach, where the first phase consists of vanilla Federated Averaging, while the second phase consists of a judicious aggregation scheme that uses a sophisticated reweighting, all in the presence of an adaptive optimisation algorithm at the server. We draw insights from extensive experimentation to tune learning rates for the two phases.