Previous public version: JHUKKI_QSMToolbox_v3.0 is available @ https://godzilla.kennedykrieger.org/QSM/ with some example 3T data and a basic manual
JHU/KKI_QSM_Toolbox is a MATLAB software with GUI for doing Quantitative Susceptibility Mapping (QSM) processing from MR phase data acquired with gradient echo sequence (GRE). Command line version without GUI was added after version 3.3.
1.This is a MATLAB based toolbox, need MATLAB installation 2.This toolbox also uses processing routine from FSL, e.g. bet, thus need to have FSL installed.
- To start, open MATLAB and go to the folder where this toolbox is located.
- (a) run QSM.m for the GUI (b) To use the command line version, run QSM_cluster(ParamSetFile.m, LogFile.txt) where ParamsSetFile.m can be edited to select QSM reconstruction parameters LogFile.txt is the log file saved.check the ParamsSetting_cluster.m for a ParamSetFile template check makeTable.m for a Table making template
Supported input file format a. PAR/REC (Philips v4.2 and above)
b. DICOM (It may work for Philips/GE/Siemens/Canon data, but not fully tested. For DICOM files, the GRE Magnitude and Phase data should be sorted first into a seperate folder. This can be done using 3rd party DICOM packages, e.g. Horos)
c. EnhancedDICOM (Philips)
d. .mat (MATLAB, customized, check "readerwrapper.m" under /QSM_Utilily/fileIO/)
e. Bruker 2dseq data (not fully tested)
f. NIFTI files, e.g. converted using dcm2niix, echo combined, with name "_GRE_mag.nii.gz/_GRE_phase.nii.gz", see "GRE_preparation_2nifti" and "readerwrapper.m" under /QSM_Utilily/fileIO/
Output file format a. internal/final results are saved in MATLAB native format (.mat)
b. some results are also saved with a copy of NIFTI format (.nii.gz)
For detailed use of the Toolbox, read the manual (v3.0, will update later)
*_PhaseUnwrapped_echo2-5[Path/Lap/NLFPath].mat: Unwrapped phase using [path/laplacian/nonlinear fitting+path] method, echo 2-5 (user-defined) was selected for QSM processing
*GREMag#.nii.gz : GRE Magnitude Image of selected echo for brain masking
*GREMag#_brain.nii.gz : GRE Magnitude Image of brain only, using FSL BET
*GREMag#_brain_mask.nii.gz : brain mask from FSL BET
*_brain_mask_echo#_r#_t##.mat : final brain mask used for QSM
*[SMV/PDF/LBVSMV/iRSHARP]_[avg/LSlope/NLSlope]_echo2-5.mat: Frequency map (in Hz) or local field map generated using [VSHARP/PDF/LBV+VSHARP/iRSHARP] method, taking [echo averaging/linear fitting/nonlinear fitting], with echo2-5(user-defined). Have corresponding .nii.gz file too.
*R2star.mat/R2star.nii.gz : R2 star map
*chi_[iLSQR/TKD/iTKD/MEDI/SFCR/SFCR+0/nSFCR/nSFCR+0/FANSI/NDI/TFI]: QSM map calculated using different methods. Have corresponding .nii.gz file.
2019-09-03: Add back in MS-SFCR+0 for ARIC processing
2021-05-01: Added initial phase offset removal and weighted echo averaging
2021-06-01: Added FANSI and nSFCR with L1-nonlinear data-fidelity
2021-06-29: Added cluster version without GUI
2021-10-23: Added mcpc-3Ds/ASPIRE for coil-combination
2023-09-12: Added support for NIFTI input files
This Matlab package is available publicly, in the hope that it will be useful, but without any warranty and without even the implied warranty of merchantability or fitness for a particular purpose. It is designed to help researchers in the field of Magnetic Resonance Imaging who may have interests in doing QSM. It is not intended for use in a clinical setting.
Major GUI and pipeline Authors: Jiri van Bergen and Xu Li Affiliation: Radiology @ JHU & Kirby Center @ Kennedy Krieger Institute (KKI) Contact via xuli@mri.jhu.edu
Other contributors include: Joseph Gillen, Craig Jones, Jonathan Farrell, Issel Lim and Jiadi Xu (Kirby Center @ KKI) Lijun Bao, Jinsheng Fang (Xiamen University)
Some Codes are from or modifed from other publicly availabe codes including
- MEDI Toolbox: http://weill.cornell.edu/mri/pages/qsm.html
- Berkin Bilgic Software: http://martinos.org/~berkin/software.html
- FANSI Toolbox: https://gitlab.com/cmilovic/FANSI-toolbox
- SEPIA Toolbox: https://github.com/kschan0214/sepia