Print

---

The .mat file is optional and only if you want the final perfusion image to be transformed into MNI standard space. The registration of ASL data to structural image will be done for you by ASL_gui and you don't need to worry that the input data is 4D. However, it might be that this registration is not perfect, in which case you might like to try your own registration - the transformation from this can be applied to the perfusion image in the 'native_space' output directory from ASL_gui. That directory contains the perfusion image in the same resolution and orientation as the original data. You should not provide ASL_gui with ASL data that has already been normalised, it is expecting the original raw data.

Michael

Dear Michael, Thank you for your response. I've found an issue with the dicom conversion, like explained in another post in this list, the dcm2nii (McCausland Center) is not working properly for FSL- at least in win 8. Now using the MRIConvert I can see the ASL 3D volumes in FSL properly. The last step I guess is to use fslmerge to create a 4D volume with all the pairs (ctr/tag). Does anybody know another option to create the 4D? It seems that the T1 segmentation is done well, so there has to be a problem with the registration. And I have a doubt now: In ASL_gui we are asked to enter the .mat related to the T1 to MNIStandardBrain, but I guess the 4DASL volume has to be registered to the T1 when providing it to the ASL_gui, is that right? it has to be normalized too? It seems that FLIRT does not support 4D volumes registration, do I have to register the first ASL to the T1 and apply the .mat to the rest of the 3D volumes (100 images)? Thank you once again, Marta