PISCO: Self-Supervised k-Space Regularization for Improved Neural Implicit k-Space Representations of Dynamic MRI

2025-01-16

AI Summary

To address overfitting of neural implicit k-space representations caused by severe undersampling in dynamic MRI accelerated imaging, this paper proposes a self-supervised k-space regularization method that requires no additional annotated data. The core innovation is a parallel-imaging-inspired self-consistency (PISCO) loss, which, for the first time, enforces global neighborhood constraints in k-space without paired training data. The method synergistically integrates neural implicit representation, self-supervised learning, and parallel-imaging priors to enhance reconstruction robustness. Experiments demonstrate that our approach achieves significantly superior spatiotemporal reconstruction quality over state-of-the-art methods at acceleration factors ≥54. Moreover, it exhibits strong generalization and stability across both static and dynamic MRI tasks, confirming its effectiveness under diverse acquisition scenarios and anatomical variations.

Technology Category

• Machine Learning: Unsupervised & Self-Supervised Learning
• Computer Vision: Representation Learning for Vision
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Neural implicit k-space representations (NIK) have shown promising results for dynamic magnetic resonance imaging (MRI) at high temporal resolutions. Yet, reducing acquisition time, and thereby available training data, results in severe performance drops due to overfitting. To address this, we introduce a novel self-supervised k-space loss function $mathcal{L}_mathrm{PISCO}$, applicable for regularization of NIK-based reconstructions. The proposed loss function is based on the concept of parallel imaging-inspired self-consistency (PISCO), enforcing a consistent global k-space neighborhood relationship without requiring additional data. Quantitative and qualitative evaluations on static and dynamic MR reconstructions show that integrating PISCO significantly improves NIK representations. Particularly for high acceleration factors (R$geq$54), NIK with PISCO achieves superior spatio-temporal reconstruction quality compared to state-of-the-art methods. Furthermore, an extensive analysis of the loss assumptions and stability shows PISCO's potential as versatile self-supervised k-space loss function for further applications and architectures. Code is available at: https://github.com/compai-lab/2025-pisco-spieker

Problem

• Dynamic MRI
• Data Reduction
• Image Inaccuracy

Innovation

• PISCO
• Self-supervised Learning
• Dynamic MRI