Archive for the 'fMRI' Category

parallelism in AFNI

Friday, October 3rd, 2003

On the AFNI message board, there is some chatter about SMP-enabled 3dDeconvolve calls. Some people aren’t getting the expected speedup. We’ve only got one 2-CPU machine in our cluster, so we’re not using the SMP features of 3dDeconvolve. What we’re finding is that our deconvolution calls come in groups of at least […]

upgrading AFNI

Friday, August 1st, 2003

There’s a new version of AFNI for August (2.55a). Okay, when isn’t there a new version of AFNI? And what is up with the version numbers? Anyway, this one may or may not fix some problems we’re having with the DICOM files produced by the 3T scanner we’re using. Misaligned functional […]

incomplete DICOM headers

Tuesday, February 25th, 2003

So last week, we discovered a new problem using data from the new scanner. Certain functional fMRI runs wouldn’t convert from DICOM to AFNI formats. AFNI would die, complaining about indecipherable SIEMENS MOSAIC info.

This refers to a DICOM header (0029 1020), which is supposed to contain information about the composition of the image […]

AFNI on Jaguar (again)

Monday, January 13th, 2003

I posted a month or two ago about my efforts to get AFNI to compile and run on Mac OSX 10.2. I should mention now that AFNI is available through fink, as the ‘afni’ package. “apt-get install afni” will do it, and so will “fink install afni”. Good news, if you’re me […]

AFNI on Jaguar

Wednesday, October 9th, 2002

I’m working on compiling AFNI for my Jaguar box. Here’s the recipe:

I used fink to install the following packages: xfree86-base-threaded, xfree86-rootless-threaded, openmotif3, and dlcompat. Installing fink on 10.2 is well-documented elsewhere. I’m using OroborOSX as my X-on-Aqua deal. Install OroborOSX after you install the xfree86 packages with fink.

I copied the […]