Generalizable Learning Reconstruction for Accelerating MR Imaging via Federated Neural Architecture Search

TL;DR

This paper introduces GAutoMRI, a federated neural architecture search framework that enhances generalization, fairness, and efficiency in MR image reconstruction across heterogeneous data sources without sharing sensitive data.

Contribution

It proposes an automatic neural architecture search method combined with a fairness adjustment to improve federated MR reconstruction, addressing heterogeneity and bias issues.

Findings

GAutoMRI outperforms six state-of-the-art federated methods in accuracy and generalization.

The framework produces a lightweight model suitable for practical MR reconstruction.

Extensive experiments validate the effectiveness of the proposed approach.

Abstract

Heterogeneous data captured by different scanning devices and imaging protocols can affect the generalization performance of the deep learning magnetic resonance (MR) reconstruction model. While a centralized training model is effective in mitigating this problem, it raises concerns about privacy protection. Federated learning is a distributed training paradigm that can utilize multi-institutional data for collaborative training without sharing data. However, existing federated learning MR image reconstruction methods rely on models designed manually by experts, which are complex and computational expensive, suffering from performance degradation when facing heterogeneous data distributions. In addition, these methods give inadequate consideration to fairness issues, namely, ensuring that the model's training does not introduce bias towards any specific dataset's distribution. To this end, this paper proposes a generalizable federated neural architecture search framework for accelerating MR imaging (GAutoMRI). Specifically, automatic neural architecture search is investigated for effective and efficient neural network representation learning of MR images from different centers. Furthermore, we design a fairness adjustment approach that can enable the model to learn features fairly from inconsistent distributions of different devices and centers, and thus enforce the model generalize to the unseen center. Extensive experiments show that our proposed GAutoMRI has better performances and generalization ability compared with six state-of-the-art federated learning methods. Moreover, the GAutoMRI model is significantly more lightweight, making it an efficient choice for MR image reconstruction tasks. The code will be made available at https://github.com/ternencewu123/GAutoMRI.