This repository provides Matlab codes to recover accelerated phase contrast MRI data using ReVEAL. Codes are provided to recover planar images with 1-3 velocity endoding dimensions, and 4D flow data. Two example data sets are provided, one for planar imaging and one for 4D flow.
To regularize the ill-posed inverse problem, we present a novel inversion algorithm that applies regularization based on structure unique to phase contrast MRI data. Adopting an empirical Bayes approach, spatial and temporal redundancies are exploited via sparsity in the wavelet domain, and the voxel-wise magnitudeand phase structure across encodings is captured in a conditional mixture prior that applies regularizing constraints based on the presence of flow. From the Bayesian model, a factor graph is developed and an iterative inversion algorithm is derived using message passing computation. The loopy portions of the graph are greatly accelerated using generalized approximate message passing (GAMP). The technique has been validated using prospectively accelerated mecanical flow phantom and in vivo data. To install the codes, simply download the ReVEAL repository and add the folder and its subfolders to the Matlab path. To run the example scripts, you will need to download the datasets provided below. Make sure the datasets are included in the Matlab path as well.
Planar phase contrast dataset
4D flow dataset
1) A Bayesian model for highly accelerated phase-contrast MRI
2) A Bayesian Approach for 4D Flow Imaging of Aortic Valve in a Single Breath-hold
3) Quantification of aortic stenosis diagnostic parameters: comparison of fast 3 direction and 1 direction phase contrast CMR and transthoracic echocardiography
4) A bayesian method for accelerated magnetic resonance elastography of the liver
To generate additional VISTA sampling patterns, download the VISTA repository.
Codes for 1-4 dimensional non-decimated wavelet transforms are provided at Wavelets
This repository also uses the GAMPLAB software package available at GAMPLAB
.