Hi @Francois
When using CAT12 for segmentation, I noticed a warning message. I wonder if this is a bug?
By default (in windows), the brainstorm uses the TPM files under the path of *.brainstorm\defaults* but the strcmp used in CAT12 is under SPM toolbox which is: *\brainstorm\plugins\spm12\spm12\tmp* this makes the mismatch triggers the warning message (code below).
isSPMtpm = strcmp(job.extopts.species,'human') && ...
( strcmp(job.opts.tpm , fullfile(spm('dir'),'tpm','TPM.nii') ) || ...
strcmp(job.opts.tpm , fullfile(spm('dir'),'tpm','TPM.nii,1') ) );
[ppt,fft] = spm_fileparts(job.opts.tpm{1});
isLONGtpm = strcmp(fft(1:min(numel(fft),7)),'longTPM');
if exist('Ybg','var') && job.extopts.setCOM ~= 120 % setCOM == 120 - useCOM,useMaffreg,noMask
if (length(ff)>4 && strcmp(ff(1:5),'navg_')) || ...
(isfield(job,'useprior') && ~isempty(job.useprior)) || ...
(isfield(job.extopts,'new_release') && job.extopts.new_release)
cat_io_cprintf('g8',' Use new longitudinal background setting. ');
% new minimal masking approach in longitudinal processing to avoid backgound peak erros and for future releases
Ymsk = cat_vol_morph( ~Ybg ,'dd',10,vx_vol) & ... % remove voxels far from head
~( Ybg & rand(size(Ybg))>0.5) & ... % have a noisy corona
~( cat_vol_grad( Ysrc , vx_vol)==0 & Ysrc==0 ); % remove voxel that are 0 and have no gradient
else
% RD20220103: old cross-sectional setting with small correction for own TPMs
if isSPMtpm || isLONGtpm
Ymsk = ~Ybg; % old default - mask background
else
cat_io_addwarning([mfilename ':noSPMTPM-noBGmasking'],...
show original
Thank you for your time
Best regards
Hi Ken,
Thank you for the detailed information.TPM.nii file if it is available: 3a38bdb
1 Like
