On The Role of K-Space Acquisition in MRI Reconstruction Domain-Generalization

2025-12-06

AI Summary

Existing k-space sampling strategies are typically optimized for a single dataset or imaging modality and suffer substantial performance degradation when generalized across different scanners and acquisition protocols. Method: We propose a domain-generalizable, learnable k-space sampling framework wherein the sampling trajectory itself is treated as an optimization variable. During training, we introduce stochastic trajectory perturbations to explicitly model k-space distribution shifts across imaging devices and conditions. Contribution/Results: To our knowledge, this is the first work to treat trajectory design as a critical degree of freedom for enhancing the domain robustness of reconstruction models—without requiring any target-domain data. Experiments demonstrate that our method consistently outperforms fixed and deterministic sampling baselines in cross-scanner and cross-protocol settings, achieving both superior reconstruction fidelity and enhanced stability under domain shifts. The framework thus unifies high robustness with high-fidelity MRI reconstruction.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Transfer, Domain Adaptation, Multi-Task Learning
• Constraint Satisfaction and Optimization: Constraint Learning and Acquisition

Application Category

• Search and Retrieval-Augmented AI: Vertical and domain-specific search
• Web Mining and Content Analysis: Robustness and generalizability of Web mining methods
• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs

Abstract

Recent work has established learned k-space acquisition patterns as a promising direction for improving reconstruction quality in accelerated Magnetic Resonance Imaging (MRI). Despite encouraging results, most existing research focuses on acquisition patterns optimized for a single dataset or modality, with limited consideration of their transferability across imaging domains. In this work, we demonstrate that the benefits of learned k-space sampling can extend beyond the training domain, enabling superior reconstruction performance under domain shifts. Our study presents two main contributions. First, through systematic evaluation across datasets and acquisition paradigms, we show that models trained with learned sampling patterns exhibitimproved generalization under cross-domain settings. Second, we propose a novel method that enhances domain robustness by introducing acquisition uncertainty during training-stochastically perturbing k-space trajectories to simulate variability across scanners and imaging conditions. Our results highlight the importance of treating kspace trajectory design not merely as an acceleration mechanism, but as an active degree of freedom for improving domain generalization in MRI reconstruction.

Problem

• Learned k-space acquisition patterns improve MRI reconstruction quality
• Existing methods lack transferability across imaging domains
• Proposes method to enhance domain robustness in MRI reconstruction

Innovation

• Learned k-space sampling patterns enhance cross-domain reconstruction
• Stochastic perturbation of trajectories simulates scanner variability
• Acquisition uncertainty training improves MRI domain generalization