Abstract:Federated learning (FL) enables multiple sites to collaboratively train powerful deep models without compromising data privacy and security. The statistical heterogeneity (e.g., non-IID data and domain shifts) is a primary obstacle in FL, impairing the generalization performance of the global model. Weakly supervised segmentation, which uses sparsely-grained (i.e., point-, bounding box-, scribble-, block-wise) supervision, is increasingly being paid attention to due to its great potential of reducing annotation costs. However, there may exist label heterogeneity, i.e., different annotation forms across sites. In this paper, we propose a novel personalized FL framework for medical image segmentation, named FedICRA, which uniformly leverages heterogeneous weak supervision via adaptIve Contrastive Representation and Aggregation. Concretely, to facilitate personalized modeling and to avoid confusion, a channel selection based site contrastive representation module is employed to adaptively cluster intra-site embeddings and separate inter-site ones. To effectively integrate the common knowledge from the global model with the unique knowledge from each local model, an adaptive aggregation module is applied for updating and initializing local models at the element level. Additionally, a weakly supervised objective function that leverages a multiscale tree energy loss and a gated CRF loss is employed to generate more precise pseudo-labels and further boost the segmentation performance. Through extensive experiments on two distinct medical image segmentation tasks of different modalities, the proposed FedICRA demonstrates overwhelming performance over other state-of-the-art personalized FL methods. Its performance even approaches that of fully supervised training on centralized data. Our code and data are available at <a class="link-external link-https" href="https://github.com/llmir/FedICRA" rel="external noopener nofollow">this https URL</a>.

FedContrast-GPA: Heterogeneous Federated Optimization Via Local Contrastive Learning and Global Process-Aware Aggregation

FedDGP: Disentangling Global and Personal Models for Federated Learning

FedCL: Federated Contrastive Learning for Multi-center Medical Image Classification

Federated Ophthalmic Disease Diagnosis Based on Weighted Averaging and Local Optimization

Unifying and Personalizing Weakly-supervised Federated Medical Image Segmentation via Adaptive Representation and Aggregation

FedProc: Prototypical Contrastive Federated Learning on Non-IID data

FedCCL: Federated Dual-Clustered Feature Contrast Under Domain Heterogeneity

Distributed contrastive learning for medical image segmentation

FedLPPA: Learning Personalized Prompt and Aggregation for Federated Weakly-supervised Medical Image Segmentation

FedGrav: An Adaptive Federated Aggregation Algorithm for Multi-institutional Medical Image Segmentation

Improving Global Generalization and Local Personalization for Federated Learning

MpFedcon : Model-Contrastive Personalized Federated Learning with the Class Center

FedTGP: Trainable Global Prototypes with Adaptive-Margin-Enhanced Contrastive Learning for Data and Model Heterogeneity in Federated Learning

Federated Optimization in Heterogeneous Networks

FedPer++: Toward Improved Personalized Federated Learning on Heterogeneous and Imbalanced Data

GRACE: A Generalized and Personalized Federated Learning Method for Medical Imaging

MH-pFLGB: Model Heterogeneous personalized Federated Learning via Global Bypass for Medical Image Analysis

FedGS: Federated Gradient Scaling for Heterogeneous Medical Image Segmentation

Heterogeneity-Aware Federated Learning with Adaptive Client Selection and Gradient Compression.