脑影像数据分析流程-优快云博客

%--brain mask with the brain tissue

mask_name = 'C:\Users\Administrator\Desktop\workspace\preprocessed\masks\within_brain_mask.nii';
M   = load_untouch_nii( mask_name ); % load mask NIFTI

mask     = double(M.img>0);          % get 3d v
 
%--brain functional 4d data

data_4d = 'C:\Users\Administrator\Desktop\workspace\preprocessed\4d\func_3d.nii';

% data_4d = 'C:\Users\Administrator\Desktop\phycaa_workspace\phycaa_plus_2104_03_27\_PHYCAA_step1+2.nii';
 
V   = load_untouch_nii( data_4d );
 
%--transform 4d array to 2d array, using brain_mask
within_brain_voxels = nifti_to_mat(V,M);
 
nt_matrix = within_brain_voxels;
 
[V_c S_c temp] = svd( nt_matrix' * nt_matrix );
% PC-space representation
Q_c = V_c * sqrt( S_c );
 
offSet=1;

pcs=4;
 
% estimate temporal autocorrelation maximized "sources"

Q1 = Q_c( 1:(end-offSet) , 1:pcs ); % un-offset

Q2 = Q_c( (offSet+1):end , 1:pcs ); % offsetted timeseries
 
% canonical correlations on time-lagged data
[A,B,R,U,V,stats] = canoncorr(Q1,Q2);
 
% getting stable "average" autocorrelated timeseries

a=[U(1,:)]; b=[U(2:end,:) + V(1:end-1,:)./2]; c=[V(end,:)];
 
tset = [a;b;c];