MAP4SL - Multiband Artefact Probe for Slice Leakage

beta version

MAP4SL is a Matlab tool which is designed to identify the effect of an artefact in MRI data on other areas depending on the slice leakage in MR multiband EPI functional images.

This tool calculates the possible artefact positions of a given seed region depending on the multislice acceleration factor (and GRAPPA if used). Then it creates a sphere or disk around this positions and extracts and plots the time courses of all voxels in these regions. Additionally, two independent control regions are extracted to compare with. Further this tool does all by all voxel correlations between the voxels of the seed area and all voxels of other possible artefact and control areas. The given seed region is always labeled as A and the expected artefact positions in other slices are B,C,D,... depending on the multislice acceleration factor. If GRAPPA was used, each expected artefact positions has a second region in the same slice where the artefact can occur depending on the parallel imaging. They are labels A_g, B_g, etc.

USAGE

You can use MAP4SL as a function giving the following inputs via a script in the following way: MAP4SL(cfg,filename,outputfolder) You can also start/run MAP4SL directly. In this case, when no input is given, MAP4SL will open three input dialogues to select the data file and outputfolder and to specify the input values directly.

INPUT

• filename - '<yourpath2BIDSsubjectfolder\niifilename>' The data needs to be in the BIDS format (at least nii.gz and a corresponding json file is needed)

• cfg is a configuration structure with the following fields: cfg.seed Coordinate of the seed voxel (e.g. [69,83,14]); cfg.sphere_radius Radius of a sphere or a disk around the estimated artefact positions cfg.disk 'yes' if want a disk or no if you want a sphere around the estimated artefact positions cfg.corr_thresh Threshold for thresholded images of all correlation coefficients rho
  cfg.saveprefix Prefix of the outpuf files cfg.flipLR Flip left and right of seed position ("yes" or "no") cfg.Shift_FOVdevX CAIPI shift of your sequence in form of FieldOfView divided by this number (e.g. normally 3)

• outputfolder The folder in which the output files will be saved;

OUTPUT

MAP4SL provided various output files

• masks in nifti format containing: -- possible artefact regions (including seed): single voxel 'prefix_artefact_mask_single_voxel.nii' and sphere/disk 'prefix_artefact_mask_conv.nii' -- control regions single voxel 'prefix_control_mask_single_voxel.nii' and sphere/disk 'prefix_control_mask_conv.nii' -- only seed single voxel 'prefix_seed_mask_single_voxel.nii' and sphere/disk 'prefix_seed_mask_conv.nii'
• text file with the a table of all artefact and control positions and their labels in the images.
• Bitmaps showing all expected artefact (and projected control) positions in 2D and 3D
• MP4 movie showing all slices containing expected artefact positions over time.
• Bitmap of time courses of the seed and all possible artefact regions
• Bitmap of time courses of both control regions
• Bitmaps of correlations coefficients rho between seed and artefact and control positions(*_rho.bmp)
• Bitmaps of all absolute correlations coefficients (*_abs_rho.bmp)
• Bitmaps of all thresholded absolute correlations coefficients (*_corr_abs_rho_thresholded.bmp)
• Matlab .m file containing all correlation values (rho and p)
• Matlab .m file containing all time series (*_data_correlations.mat)

Example script:

cfg.seed = [31,85,22];
cfg.sphere_radius = 6;
cfg.disk = 'yes';
cfg.corr_thresh = 0.6;
cfg.saveprefix = 'TEST';
cfg.flipLR = 'no';
cfg.Shift_FOVdevX = 3;
inputfilename = 'D:\BIDSfolder\s001\func\sub-01_task-X_bold.nii.gz';
outputfolder='D:\MAP4SL_output';

MAP4SL(cfg, inputfilename, outputfolder)

DEPENDECIES

MAP4SL needs the toolbox "Tools for nifti and analyze image" for Matlab in the MATLAB path. The toolbox can be downloaded from here: https://uk.mathworks.com/matlabcentral/fileexchange/8797-tools-for-nifti-and-analyze-image

License

MAP4SL by Michael Lindner is licensed under CC BY 4.0 This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY;

Author:
Michael Lindner
University of Reading, 2018
School of Psychology and Clinical Language Sciences
Center for Integrative Neuroscience and Neurodynamics