Open Source MRF is a Matlab based software package which enables vendor neutral, fast prototyping of magnetic resonance fingerprinting using Pulseq. It is supported by open-soure standards.
Magnetic resonance fingerprinting is a framework that allows for simultaneous quantification of tissue properties and hence is a significant tool to understand multi-vendor multi-site variability. However, a vendor-neutral, open soure implementation of MRF has not been developed to the best of our knowlegde. In this work, we develop the package to enable comparisons between two sites with two vendors.
Open Source MRF consists of five modules. Sequence design, image reconstruction, dictionary simulation, dictionary matching, and ROI analysis.
Please note: You need to install Pulseq and Michigan Image Reconstruction Toolbox (MIRT) in your system for the package to run. Visit Pulseq and MIRT for more information.
This package contains five different parts. To use the package, please pull the complete mrf repository.
- Call Spiral_Design/Function/gen_MRF_sequence_pulseq.m. The function returns parameters of the sequence that are used in other parts of the package
- A MRF sequence based on Jiang's paper [1] is implemented using Pulseq framework.
- The user should be able play the generated .seq file on a scanner with Pulseq interpreter installed. This sequence has been verified on a scanner by the author.
- Call Image_Reconstruction/Function/MRF_recon.m
- The script lets user select raw data file (.dat) from scanner and trajectory file (.mat) and reconstruct the image using sliding window method and complex channel combination [2].
- Call EPG_dict_sim/Function/gen_MRF_dict.m.
- A dictionary is simulated based on the T1/T2 ranges using Extend Phase Graph(EPG) [3] method.
- Call Dictionary_Matching/Function/MRF_Dic_Matching.m.
- The script uses vector dot product to find the best matching dictionary entry and retrieves all parameters for that entry.
- Call ROI_analysis_tool/Function/NIST_ROI_ana.m .
- The function takes in the parameter maps of ISMRM/NIST phantom and return locations of spheres in correct orders.
To test the package, a demo script can be found at mrf/demo_MRF.m. This script generates a sequence, then read a sample data, reconstruct the data. A dictionary is generated and dictionary matching is performed. Then ROI analysis is done to provide T1 and T2 values for each spheres.
In addition to the overall demo script, each module can be tested by itself. The demo script can be located under each module, which will help verify the correctness of the module.
[1] Jiang Y, Ma D, Seiberlich N, Gulani V, Griswold M. MR fingerprinting using fast imaging with steady state precession (FISP) with spiral readout. Magn Reson Med. 2015;74(6):spcone-spcone. doi:10.1002/mrm.26048
[2] Roemer P, Edelstein W, Hayes C, Souza S, Mueller O. The NMR phased array. Magn Reson Med. 1990;16(2):192-225. doi:10.1002/mrm.1910160203
[3] Weigel M. Extended phase graphs: Dephasing, RF pulses, and echoes - pure and simple. Journal of Magnetic Resonance Imaging. 2014;41(2):266-295. doi:10.1002/jmri.24619