Back to Physics: Operator-Guided Generative Paths for SMS MRI Reconstruction

2026-02-08

AI Summary

This work addresses the highly ill-posed inverse problem in Simultaneous Multi-Slice (SMS) MRI, which arises from in-plane undersampling and manifests as inter-slice interference and missing k-space data. The authors propose a physics operator-guided generative reconstruction framework that uniquely embeds the acquisition physics operator directly into the generative pathway, thereby aligning the degradation modeling with the inversion process. They introduce an Operator-Conditioned Dual-Interaction Network (OCDI-Net) to explicitly decouple target-slice and interference-slice information, and employ a two-stage chained inference strategy to jointly perform slice separation and k-space completion. Evaluated on both fastMRI brain and prospective in vivo diffusion MRI datasets, the method substantially outperforms conventional and state-of-the-art learning-based approaches, achieving higher reconstruction fidelity and effectively suppressing inter-slice leakage artifacts.

Technology Category

• Computer Vision: Generative Adversarial Networks (GANs) for Vision
• Machine Learning: Multi-instance/Multi-view Learning
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking
• Web Mining and Content Analysis: Web data generation and simulation
• Social Networks and Social Media: Generative AI / large language models and their impact on social systems

Abstract

Simultaneous multi-slice (SMS) imaging with in-plane undersampling enables highly accelerated MRI but yields a strongly coupled inverse problem with deterministic inter-slice interference and missing k-space data. Most diffusion-based reconstructions are formulated around Gaussian-noise corruption and rely on additional consistency steps to incorporate SMS physics, which can be mismatched to the operator-governed degradations in SMS acquisition. We propose an operator-guided framework that models the degradation trajectory using known acquisition operators and inverts this process via deterministic updates. Within this framework, we introduce an operator-conditional dual-stream interaction network (OCDI-Net) that explicitly disentangles target-slice content from inter-slice interference and predicts structured degradations for operator-aligned inversion, and we instantiate reconstruction as a two-stage chained inference procedure that performs SMS slice separation followed by in-plane completion. Experiments on fastMRI brain data and prospectively acquired in vivo diffusion MRI data demonstrate improved fidelity and reduced slice leakage over conventional and learning-based SMS reconstructions.

Problem

• SMS MRI
• inverse problem
• inter-slice interference
• k-space undersampling
• operator-governed degradation

Innovation

• operator-guided
• SMS MRI reconstruction
• dual-stream interaction
• structured degradation
• physics-informed generative model