Greetings,
I am having trouble please help. So I have been following the Dartmouth
guide for DTI data. I have a few issues that have popped up. I have posted
about them before. After running bedpostX I am missing the mean and dyads
files. I was informed in my thread "Missing mean and dyads after
"successful" bedpostx?" that I do not need the mean files for probtrackX.
When I run a probtrackX I run out of memory st9::bad alloc. I was running in
32bit mode and would only get to 1_2 before it would crash. I was instructed
to switch to 64 bit and increase the swap. So I did now with 2 gigs of ram
and an 8.2 gig swap. I can get to 2_3 before I crash. Watching the system
monitor it shows that my ram is fully utilized and with ever iteration 1_1,
1_2, 1_3, 2_1... it uses about a gig of swap space until it runs out and
crashes.
I don't view my data as anything unique. All of my scans are from a philips
scanner. My control group (the data I am currently working with) is older
2005 and was version 3 Par files. I used PRIDE to convert to version 4 and
then combined my directions into an interav- par rec file. I then used cmd
line program to coregister my dti data to my anatomy data and converted it
to analyze (this process splits the data back into individual directions). I
then used MRIcron to convert to NifTI and then followed the Dartmouth guide
and FSL.
Please any suggestions would be high appreciated. Even if I have to start
all over from my original version 3 files please let me know of any ideas. I
am completely at a loss of what I can do.
Thank you.
Sincerely,
-James Sheehan
|