Gm and Wm

[rogiedodgie]

When using .nii images, should the WM and GM complexity be calculated based on segmented .nii files c1 and c2 as in the code below?
Roger

indir = '/media/chris/OCTO/ABCPROJECT_IN';
outdir = '/media/chris/GoldenBoy/FractalsFinally';
cd (indir);

% 20180313 CRM
gmf = dir('/c1.nii')
wmf = dir('/c2.nii')

for i = 1:length(gmf)

    %% load nii data, using one of three options
    % (1) Matlab 2017b+ has this built-in as 'niftiread'
    targetfile = fullfile(gmf(i).folder, gmf(i).name);
    V = niftiread(targetfile);
    bsplit = strsplit(gmf(i).name,'_');
    subjName = bsplit{2};
    
    %% select voxels to be part of region label
    % make sure this value is one of the regions included in the intensity coding of the parcellated/segmented image
    % e.g., confirm with "unique(V)"fullfil
    % if this region code doesn't exist, the toolbox will enter debug mode and say the region wasn't found
    vol = V == 1;

    %% do calcFD
    % add the calcFD path to the MATLAB environment
    %addpath('../calcFD')
    % define desired box kernel sizes
    r = 2.^[0:4]; % i.e., [1,2,4,8,16]
    % determine the 'counts' using the dilation algorithm
    n = calcFD_dilate(vol,r); 
    % perform a simple linear regression of the log2-transformed box sizes and log2-transformed counts
    c = [log2(r)' ones(length(r),1)] \ -log2(n)'; 
    % take the coefficient corresponding to the slope as the fractal dimensionality (fd) value
    fd = c(1);

    data(i).name = subjName;
    data(i).gm_fd = fd;
     fprintf('Successfully processed GM for %s with value of %d\n',subjName,fd);
    
    % (1) Matlab 2017b+ has this built-in as 'niftiread'
    targetfile = fullfile(wmf(i).folder, wmf(i).name);
    V = niftiread(targetfile);
    bsplit = strsplit(wmf(i).name,'_');
    subjName = bsplit{2};
    
    %% select voxels to be part of region label
    % make sure this value is one of the regions included in the intensity coding of the parcellated/segmented image
    % e.g., confirm with "unique(V)"fullfil
    % if this region code doesn't exist, the toolbox will enter debug mode and say the region wasn't found
    vol = V == 1;

    %% do calcFD
    % add the calcFD path to the MATLAB environment
    %addpath('../calcFD')
    % define desired box kernel sizes
    r = 2.^[0:4]; % i.e., [1,2,4,8,16]
    % determine the 'counts' using the dilation algorithm
    n = calcFD_dilate(vol,r); 
    % perform a simple linear regression of the log2-transformed box sizes and log2-transformed counts
    c = [log2(r)' ones(length(r),1)] \ -log2(n)'; 
    % take the coefficient corresponding to the slope as the fractal dimensionality (fd) value
    fd = c(1);

    data(i).wm_fd = fd;
    
    fprintf('Successfully processed WM for %s with value of %d\n',subjName,fd);

end
t= struct2table(data)
writetable(t,'subjID_GM_WM_spreadsheet.xls')
copyfile('subjID_GM_WM_spreadsheet.xls',outdir)