Deep Plug-and-Play Prior for Parallel MRI Reconstruction

TL;DR

This paper introduces a deep plug-and-play prior framework for parallel MRI reconstruction that leverages deep neural networks as denoisers, significantly improving image quality at higher acceleration factors and reducing scan time.

Contribution

It presents a novel deep learning-based plug-and-play approach for MRI reconstruction that outperforms traditional methods like GRAPPA in quality and speed.

Findings

Higher quality images at high acceleration factors

Outperforms clinical gold standard GRAPPA method

Enables faster MRI scans with maintained image quality

Abstract

Fast data acquisition in Magnetic Resonance Imaging (MRI) is vastly in demand and scan time directly depends on the number of acquired k-space samples. Conventional MRI reconstruction methods for fast MRI acquisition mostly relied on different regularizers which represent analytical models of sparsity. However, recent data-driven methods based on deep learning has resulted in promising improvements in image reconstruction algorithms. In this paper, we propose a deep plug-and-play prior framework for parallel MRI reconstruction problems which utilize a deep neural network (DNN) as an advanced denoiser within an iterative method. This, in turn, enables rapid acquisition of MR images with improved image quality. The proposed method was compared with the reconstructions using the clinical gold standard GRAPPA method. Our results with undersampled data demonstrate that our method can deliver considerably higher quality images at high acceleration factors in comparison to clinical gold standard method for MRI reconstructions. Our proposed reconstruction enables an increase in acceleration factor, and a reduction in acquisition time while maintaining high image quality.