Model-based Dynamic 3D MRI Reconstructions using Neural Fields and Tensor Product Expansions

TL;DR

This paper presents a discretization-free, neural field-based framework for dynamic 2D and 3D MRI reconstruction that is memory-efficient and outperforms existing methods in highly undersampled scenarios.

Contribution

It introduces a novel tensor product neural field approach for continuous representation of MRI data, enabling scalable high-dimensional optimization.

Findings

Outperforms state-of-the-art methods in dynamic MRI reconstruction.

Preserves structure and motion under aggressive undersampling (e.g., acceleration factor 16).

Memory-efficient and scalable to high-dimensional data.

Abstract

Conventional MRI reconstruction methods treat images and coil sensitivities as discrete objects, leading to high memory demands and limited structural awareness that hamper effective regularization. These limitations hinder accurate reconstruction in highly undersampled scenarios, such as dynamic 3D cardiac magnetic resonance (CMR). We introduce a discretization-free, memory-efficient, model-based framework for dynamic 2D and 3D MRI reconstruction from highly undersampled data. We represent magnetization and coil sensitivities as continuous objects -- differentiable functions -- using tensor products of univariate neural fields. This tensor product structure enables scalable optimization in high-dimensional spatiotemporal settings. Our method outperforms state-of-the-art model-based reconstructions in dynamic 2D and 3D MR settings, preserving structure and motion even under aggressive undersampling (e.g., acceleration factor 16).