hi lei,
please see below.
On Tue, 12 Sep 2006 Lei.Zhang@uniklinik-freiburg.de wrote:
Hi, Jonathan,
I'm glad to know you via Nick's email and I want you could give me some guidances with my retinotopic analysis. I have scanned the retinotopic runs with eccen and polar, and the parameters were as following. Eccen run: TR 3, the first 8 TRs will discard, then follow Ncycles 7, 48 sec per cycle, all scan 24+336 sec. Polar run: TR 3, the first 8 TRs will discard, then follow Ncycles 6, 60 sec per cycle, all scan 24+360 sec.
one thing to note here: you mentioned that you have two rotating wedges presented simultaneously, and that the full stimulus rotates once per 60 seconds. each position in the visual field would then be stimulated *twice* per 60 seconds, so, if i am understanding correctly, it sounds like your 'ncycles' for the polar angle maps should therefore be 12.
Now I can view my final processed retinotopic results, but unfortunately it looks not ideal. This is my first time to try Freesurfer, so I was confused by some processing steps, and I want to get your kind help.
- in mkanalysis-sess.new command, -ncycles option, I don't know how to
fill it because my eccen run was 7 and polar run was 12 cycles (sorry, here should be 6 cycles).
doug attached to his previous email a new version of "mkanalysis-sess" in which he has added the ability to specify the 'ncycles' parameter value on a per-run basis. as he explained, the command no longer requires specification of 'ncycles' on the command line. this parameter value can now be specified within each paradigm file contained in the run directories.
so you may produce, e.g., an 'rtopy_mixedncycles.par' file that, in the eccentricity map runs, looks like...
stimtype eccen direction neg ncycles 7
...and, in the polar angle map runs, looks like...
stimtype polar direction neg ncycles 12
then just specify '-paradigm rtopy_mixedncycles.par' in your "mkanalysis-sess" command and leave out the '-ncycles' parameter. to take advantage of this new feature, you should use the updated version of "mkanalysis-sess" that doug provided in his email attachment.
- also in mkanalysis-sess.new, can I use -nskip? I want to discard the
first 8 time points (8 TRs) in eccen and polar runs.
yup. just append '-nskip 8' to your "mkanalysis-sess" command on the command line.
- what can we find in sliceview-sess (intermediate results)? I mean
that the activations should be found in the occipital lobe if the processing is OK?
"sliceview-sess" is great for checking the significance maps calculated from the phase encoding analysis---which i believe are stored in those first 3 images of the analyzed phase encoding volume that is loaded---prior to computing the functional-structural registration. note that you can separately view the significance maps of the 'eccen' and 'polar' analyses within the functional volumes.
however, to see that the phases are mapped out correctly you will probably need to view the 'eccen' and 'polar' phases along the cortical surface. these last two steps (involving "paint-sess" and "surf-sess") are described in the "FsFastIndividualRetinotopyAnalysis" workflow on the FreeSurfer wiki.
- can I process the eccen and polar runs separately? because I don't
know if ther will be interferences with different cycle time in the analysis, my eccen run was 48 sec per cycles and polar was 60 sec per cycle.
sure. again, now that doug has extended "mkanalysis-sess" you can easily process both the eccentricity and polar angle maps at once even if they were acquired with different stimulus presentation frequencies. internally, the analysis processes each separately when it computes the corresponding phase maps, so there will be no interference.
hope that helps. good luck!
-jon