Subject-Specific Low-Field MRI Synthesis via a Neural Operator

2026-03-25

AI Summary

This work addresses the challenges of low-field MRI, which suffers from low signal-to-noise ratio and poor contrast, and for which existing simulation methods fail to accurately capture realistic degradation characteristics. To overcome these limitations, the authors propose H2LO, a coordinate-image decoupled neural operator framework that enables end-to-end learning to synthesize high-fidelity low-field MR images using only a small number of paired high- and low-field scans. H2LO is the first method to jointly model both high-frequency noise and structural degradation inherent in low-field MRI, significantly outperforming conventional noise-injection and image-translation approaches. The framework generates more realistic low-field images across T1-weighted and T2-weighted sequences and demonstrably enhances performance in downstream image enhancement tasks.

Technology Category

• Computer Vision: Learning & Optimization for CV
• Natural Language Processing: Code Generation / Program Synthesis from Natural Language
• Machine Learning: Neuro-Symbolic Learning

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
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs

Abstract

Low-field (LF) magnetic resonance imaging (MRI) improves accessibility and reduces costs but generally has lower signal-to-noise ratios and degraded contrast compared to high field (HF) MRI, limiting its clinical utility. Simulating LF MRI from HF MRI enables virtual evaluation of novel imaging devices and development of LF algorithms. Existing low field simulators rely on noise injection and smoothing, which fail to capture the contrast degradation seen in LF acquisitions. To this end, we introduce an end-to-end LF-MRI synthesis framework that learns HF to LF image degradation directly from a small number of paired HF-LF MRIs. Specifically, we introduce a novel HF to LF coordinate-image decoupled neural operator (H2LO) to model the underlying degradation process, and tailor it to capture high-frequency noise textures and image structure. Experimental results in T1w and T2w MRI demonstrate that H2LO produces more faithful simulated low-field images than existing parameterized noise synthesis models and popular image-to-image translation models. Furthermore, it improves performance in downstream image enhancement tasks, showcasing its potential to enhance LF MRI diagnostic capabilities.

Problem

• low-field MRI
• image synthesis
• contrast degradation
• signal-to-noise ratio
• MRI simulation

Innovation

• neural operator
• low-field MRI synthesis
• image degradation modeling
• coordinate-image decoupling
• paired MRI translation