Computationally Efficient 3D MRI Reconstruction with Adaptive MLP

TL;DR

This paper introduces Recon3DMLP, a hybrid neural network combining CNN and adaptive MLP modules with a circular shift operation, enabling efficient high-resolution 3D MRI reconstruction with improved accuracy and reduced memory usage.

Contribution

The paper proposes a novel hybrid CNN-MLP architecture with a circular shift operation for scalable, high-quality 3D MRI reconstruction, addressing limitations of existing CNN-based methods.

Findings

Recon3DMLP outperforms existing CNN models in high-resolution 3D MRI reconstruction.

The method reduces GPU memory consumption by over 50%.

Recon3DMLP effectively captures long-distance information for better image quality.

Abstract

Compared with 2D MRI, 3D MRI provides superior volumetric spatial resolution and signal-to-noise ratio. However, it is more challenging to reconstruct 3D MRI images. Current methods are mainly based on convolutional neural networks (CNN) with small kernels, which are difficult to scale up to have sufficient fitting power for 3D MRI reconstruction due to the large image size and GPU memory constraint. Furthermore, MRI reconstruction is a deconvolution problem, which demands long-distance information that is difficult to capture by CNNs with small convolution kernels. The multi-layer perceptron (MLP) can model such long-distance information, but it requires a fixed input size. In this paper, we proposed Recon3DMLP, a hybrid of CNN modules with small kernels for low-frequency reconstruction and adaptive MLP (dMLP) modules with large kernels to boost the high-frequency reconstruction, for 3D MRI reconstruction. We further utilized the circular shift operation based on MRI physics such that dMLP accepts arbitrary image size and can extract global information from the entire FOV. We also propose a GPU memory efficient data fidelity module that can reduce $>$50$\%$ memory. We compared Recon3DMLP with other CNN-based models on a high-resolution (HR) 3D MRI dataset. Recon3DMLP improves HR 3D reconstruction and outperforms several existing CNN-based models under similar GPU memory consumption, which demonstrates that Recon3DMLP is a practical solution for HR 3D MRI reconstruction.