Homec4science
Diffusion Slepians (master) Browse

Slepi/Code/Utilities/DPARSF_V2.3_130615/Subfunctions/y_RegressOutImgCovariates.m
22f22bf83060master

Last Change
Raw File

y_RegressOutImgCovariates.m
View Options

function y_RegressOutImgCovariates(DataDir,CovariablesDef,Postfix,MaskFilename, ImgCovModel)
% FORMAT y_RegressOutImgCovariates(DataDir,CovariablesDef,Postfix,MaskFilename, ImgCovModel)
% Input:
% DataDir - where the 3d+time dataset stay, and there should be 3d EPI functional image files. It must not contain / or \ at the end.
% CovariablesDef - A struct which defines the coviarbles.
% CovariablesDef.polort - The order of the polynomial like 3dfim+.
% 0: constant
% 1: constant + linear trend
% 2: constant + linear trend + quadratic trend.
% 3: constant + linear trend + quadratic trend + cubic trend. ...
% CovariablesDef.ort_file - The filename of the text file which contains the covaribles.
% CovariablesDef.CovImgDir - Directory Cell: the image directory you want to regress out
% CovariablesDef.CovMat - Covariable Matrix. Time points by Cov number matrix
% CovariablesDef.CovMask - Covariable Masks. Cells of Mask file names.
% Postfix - Post fix of the resulting data directory. e.g. '_Covremoved'
% MaskFilename - The mask file for regression. Empty means perform regression on all the brain voxels.
% ImgCovModel - The model for the image covariates defined in CovariablesDef.CovImgDir. E.g., used for the voxel-specific 12 head motion regression model
% 1 (default): Use the current time point. e.g., Txi, Tyi, Tzi
% 2: Use the current time point and the previous time point. e.g., Txi, Tyi, Tzi, Txi-1, Tyi-1, Tzi-1
% 3: Use the current time point and their squares. e.g., Txi, Tyi, Tzi, Txi^2, Tyi^2, Tzi^2
% 4: Use the current time point, the previous time point and their squares. e.g., Txi, Tyi, Tzi, Txi-1, Tyi-1, Tzi-1 and their squares (total 12 items). Like the Friston autoregressive model (Friston, K.J., Williams, S., Howard, R., Frackowiak, R.S., Turner, R., 1996. Movement-related effects in fMRI time-series. Magn Reson Med 35, 346-355.)
% Output:
% *.nii - data removed the effect of covariables.
%___________________________________________________________________________
% Written by YAN Chao-Gan 111209.
% The Nathan Kline Institute for Psychiatric Research, 140 Old Orangeburg Road, Orangeburg, NY 10962, USA
% Child Mind Institute, 445 Park Avenue, New York, NY 10022, USA
% The Phyllis Green and Randolph Cowen Institute for Pediatric Neuroscience, New York University Child Study Center, New York, NY 10016, USA
% ycg.yan@gmail.com
if ~exist('MaskFilename','var')
MaskFilename='';
end
if ~exist('ImgCovModel','var')
ImgCovModel=1;
end
if ~isfield(CovariablesDef,'CovImgDir')
CovariablesDef.CovImgDir = {};
end
[AllVolume,VoxelSize,theImgFileList, Header,nVolumn] =rest_to4d(DataDir);
[nDim1,nDim2,nDim3,nDim4]=size(AllVolume);
if ~isempty(CovariablesDef.CovImgDir)
Cov5DVolume=zeros(nDim1,nDim2,nDim3,nDim4,length(CovariablesDef.CovImgDir));
for i=1:length(CovariablesDef.CovImgDir);
fprintf('\n\tRead the Image covariates: No. %d...\n',i);
[CovTempVolume] =rest_to4d(CovariablesDef.CovImgDir{i});
Cov5DVolume(:,:,:,:,i) = CovTempVolume;
end
end
clear CovTempVolume
if ~isempty(MaskFilename)
[MaskData,MaskVox,MaskHead]=rest_readfile(MaskFilename);
else
MaskData=ones(nDim1,nDim2,nDim3);
end
theCovariables=[];
%Add polynomial in the baseline model according to 3dfim+.pdf
if isfield(CovariablesDef,'polort')
thePolOrt=[];
if CovariablesDef.polort>=0,
thePolOrt =(1:nDim4)';
thePolOrt =repmat(thePolOrt, [1, (1+CovariablesDef.polort)]);
for x=1:(CovariablesDef.polort+1),
thePolOrt(:, x) =thePolOrt(:, x).^(x-1) ;
end
end
theCovariables =[theCovariables,thePolOrt];
end
if isfield(CovariablesDef,'ort_file')
if exist(CovariablesDef.ort_file, 'file')==2,
theCovariablesFromFile =load(CovariablesDef.ort_file);
theCovariables =[theCovariables,theCovariablesFromFile];
else
warning(sprintf('\n\nCovariables definition text file "%s" doesn''t exist, please check! This covariables will not be regressed out this time.', CovariablesDef.ort_file));
end
end
if isfield(CovariablesDef,'CovMat')
theCovariables=[theCovariables,CovariablesDef.CovMat];
end
if isfield(CovariablesDef,'CovMask')
for iMask=1:length(CovariablesDef.CovMask)
[iMaskData,iMaskVox,iMaskHead]=rest_readfile(CovariablesDef.CovMask{iMask});
TempTC=reshape(AllVolume,[],nDim4);
TempTC = mean(TempTC(find(iMaskData),:))';
theCovariables=[theCovariables,TempTC];
end
end
MaskData = any(AllVolume,4) .* MaskData; % skip the voxels with all zeros
VolumeAfterRemoveCov=zeros(nDim1,nDim2,nDim3,nDim4);
fprintf('\n\tRegressing Out Covariates...\n');
for i=1:nDim1
fprintf('.');
for j=1:nDim2
for k=1:nDim3
if MaskData(i,j,k)
DependentVariable=squeeze(AllVolume(i,j,k,:));
if ~isempty(CovariablesDef.CovImgDir)
if ImgCovModel==1
ImgCovTemp = squeeze(Cov5DVolume(i,j,k,:,:));
elseif ImgCovModel==2
Q1 = squeeze(Cov5DVolume(i,j,k,:,:));
ImgCovTemp = [Q1, [zeros(1,size(Q1,2));Q1(1:end-1,:)]];
elseif ImgCovModel==3
Q1 = squeeze(Cov5DVolume(i,j,k,:,:));
ImgCovTemp = [Q1, Q1.^2];
elseif ImgCovModel==4
Q1 = squeeze(Cov5DVolume(i,j,k,:,:));
ImgCovTemp = [Q1, [zeros(1,size(Q1,2));Q1(1:end-1,:)], Q1.^2, [zeros(1,size(Q1,2));Q1(1:end-1,:)].^2];
end
AlltheCovariables=[theCovariables,ImgCovTemp];
else
AlltheCovariables=theCovariables;
end
[b,r] = y_regress_ss(DependentVariable,AlltheCovariables);
VolumeAfterRemoveCov(i,j,k,:)=r;
end
end
end
end
VolumeAfterRemoveCov(isnan(VolumeAfterRemoveCov))=0;
if strcmp(DataDir(end),filesep)==1,
DataDir=DataDir(1:end-1);
end
OutputDir =sprintf('%s%s',DataDir,Postfix);
ans=rmdir(OutputDir, 's');
[theParentDir,theOutputDirName]=fileparts(OutputDir);
mkdir(theParentDir,theOutputDirName);
Header_Out = Header;
Header_Out.pinfo = [1;0;0];
Header_Out.dt =[16,0];
y_Write4DNIfTI(VolumeAfterRemoveCov,Header_Out,[OutputDir,filesep,'CovRegressed_4DVolume.nii']);
fprintf('\n\tRegressing Out Covariates finished.\n');
c4science · Help