Federated Contrastive Learning for Decentralized Unlabeled Medical Images
This work addresses label efficiency in medical imaging for clinical applications, but it is incremental as it adapts existing contrastive learning to a federated setting.
The paper tackles the challenge of learning from decentralized unlabeled medical images by proposing FedMoCo, a federated contrastive learning framework, which outperforms FedAvg in representation quality and reduces labeled data needs for tasks like COVID-19 detection.
A label-efficient paradigm in computer vision is based on self-supervised contrastive pre-training on unlabeled data followed by fine-tuning with a small number of labels. Making practical use of a federated computing environment in the clinical domain and learning on medical images poses specific challenges. In this work, we propose FedMoCo, a robust federated contrastive learning (FCL) framework, which makes efficient use of decentralized unlabeled medical data. FedMoCo has two novel modules: metadata transfer, an inter-node statistical data augmentation module, and self-adaptive aggregation, an aggregation module based on representational similarity analysis. To the best of our knowledge, this is the first FCL work on medical images. Our experiments show that FedMoCo can consistently outperform FedAvg, a seminal federated learning framework, in extracting meaningful representations for downstream tasks. We further show that FedMoCo can substantially reduce the amount of labeled data required in a downstream task, such as COVID-19 detection, to achieve a reasonable performance.