Federated brain tumor segmentation: an extensive benchmark

Abstract

Recently, federated learning has raised increasing interest in the medical image analysis field due to its ability to aggregate multi-center data with privacy-preserving properties. A large amount of federated training schemes have been published, which we categorize into global (one final model), personalized (one model per institution) or hybrid (one model per cluster of institutions) methods. However, their applicability on the recently published Federated Brain Tumor Segmentation 2022 dataset has not been explored yet. We propose an extensive benchmark of federated learning algorithms from all three classes on this task. While standard FedAvg already performs very well, we show that some methods from each category can bring a slight performance improvement and potentially limit the final model(s) bias toward the predominant data distribution of the federation. Moreover, we provide a deeper understanding of the behaviour of federated learning on this task through alternative ways of distributing the pooled dataset among institutions, namely an Independent and Identical Distributed (IID) setup, and a limited data setup.

References in corpus (28)

• Automated Design of Deep Learning Methods for Biomedical Image Segmentation
• Federated Learning with Non-IID Data
• The Liver Tumor Segmentation Benchmark (LiTS)
• On the Convergence of FedAvg on Non-IID Data
• SCAFFOLD: Stochastic Controlled Averaging for Federated Learning
• Tackling the Objective Inconsistency Problem in Heterogeneous Federated Optimization
• Federated Learning with Personalization Layers
• Ensemble Distillation for Robust Model Fusion in Federated Learning
• MONAI: An open-source framework for deep learning in healthcare
• Federated Learning Enables Big Data for Rare Cancer Boundary Detection
• Fair Resource Allocation in Federated Learning
• Three Approaches for Personalization with Applications to Federated Learning
• Survey on Federated Learning Threats: concepts, taxonomy on attacks and defences, experimental study and challenges
• The reliability of a deep learning model in clinical out-of-distribution MRI data: a multicohort study
• Overcoming Forgetting in Federated Learning on Non-IID Data
• Do Gradient Inversion Attacks Make Federated Learning Unsafe?
• Detect and Correct Bias in Multi-Site Neuroimaging Datasets
• Federated Multi-Task Learning under a Mixture of Distributions
• The Brain Tumor Segmentation (BraTS) Challenge 2023: Focus on Pediatrics (CBTN-CONNECT-DIPGR-ASNR-MICCAI BraTS-PEDs)
• The Brain Tumor Segmentation (BraTS) Challenge 2023: Glioma Segmentation in Sub-Saharan Africa Patient Population (BraTS-Africa)
• Federated Learning with Partial Model Personalization
• Local Adaptivity in Federated Learning: Convergence and Consistency
• Personalized Federated Learning through Local Memorization
• A Unified Analysis of Federated Learning with Arbitrary Client Participation
• Accelerated Federated Learning with Decoupled Adaptive Optimization
• What Do We Mean by Generalization in Federated Learning?
• FedExP: Speeding Up Federated Averaging via Extrapolation
• CD${}^2$-pFed: Cyclic Distillation-guided Channel Decoupling for Model Personalization in Federated Learning

Cited by in corpus (1)

• Decentralized Personalization for Federated Medical Image Segmentation via Gossip Contrastive Mutual Learning