Dear Bruce, Kevin and others,
Thank you for your fast reply. Sorry, to tell you Kevin, that the new tcl-script does not work either, the same error occurs, using the alpha-version (October release). In the last mail I was meaning the colour of the surfaces in tksurfer (e. i. red-green and grey). In the moment I am still frustrated by converting the SPM-maps to the freesurfer format. Doug has suggested a solution in a mail last October. He proposed to convert the Cor T1 volume to analyze, than co-register it with the func images (I used the mean image of the epis, because I did not acquire an T1 data set in the functional space), create an register.dat file by spmmat2register and than convert the spm statistical output (do one have to change in spm_write the format from uint8 to float?) with mri_convert. Also every step lokk fine, by overlaying the statistical map to the T1 image, the functional activation appears totally warped i. e. you cant foresee any activation pattern (I already change the format from little endian to big endian because the analysis were made in windows) Do I have to swap them?. For this reason I change the strategy, in state of mri_convert I used spmbfloat. Also the slices increased from 68 to 95 (standard spm output 79*95*68, 2*2*2mm, Origin 40, 57, 26) the overlay looked correct. By manipulating the restiger.dat file I was able to change the translation parameters slightly so that the statistical map fitted the anatomical one (it's a pity that one can not load bfloat-format in the tkregister). The problem is that the voxel of the overlay are a little bit too large. I tried to resize them by changing the values of the diagonal line in the register.dat file. As you can expect, it does not work proper. Does anybody have a solution for overlaying the spm data on the freesurfer data or see any mistakes in the described way? Thank you again, Felix