Print

Print


Dear Elin,

You seem to be right - bugger!!

Thanks for finding that. I'll email about patches shortly.

Tim



On Wed, 25 May 2005, Elin Lundström wrote:

> Hello,
> 
> I am using FSL with the Linux Mandrake operative
> system. Earlier I had problems with FLIRT so Mark
> Jenkinson made a new version of FLIRT which I
> downloaded. Now when I run FEEDS (with my new fsl
> version), I get some errors. I put the run output at
> the end of this mail and the result directory at
> http://www.smile.ki.se/results.tar.gz. I have run
> eddycorrect on several brains and I am a bit worried
> that the FLIRT error will have affected the results.
> 
> I also have another problem: I have calculated FA and
> MD values on a couple of pixels manually, starting
> from eddy current corrected brains. Unfortunately I
> did not get the same values as DTIFit. I have looked
> at the dtifit.cc to see the formulae you are using. I
> saw one formula saying:
> 
> float numer=(1.5*(Dd(1)-mDd)*(Dd(1)-mDd)... and so on
> 
> If I am not mistaken it seems like there is a
> parenthesis missing (in order to multiply all
> eigenvalues with 1.5). Could you please give me the
> exact formulae that you have used to calculate MD and
> FA so that I can compare it with my manually
> calculated ones?
> 
> Thank you for your time!
> 
> Best regards,
> Elin
> 
> 
> 
> [elin@freja feeds]$ time ./RUN all
> 
> FSL Evaluation and Example Data Suite v3.2
> 
> start time = Wed May 25 13:45:37 CEST 2005
> hostname = freja.hs.sll.se
> os = Linux freja.hs.sll.se 2.4.25eva_patch #1 SMP Mon
> Apr 5 16:52:08 CEST 2004 i686 unknown unknown
> GNU/Linux
> 
> 
> /bin/rm -rf /home/elin/test_fsl/feeds/results ; mkdir
> /home/elin/test_fsl/feeds/results
> 
> 
> Starting PRELUDE & FUGUE at Wed May 25 13:45:37 CEST
> 2005
> % error = 0.0
> % error = 0.0
> 
> Starting SUSAN at Wed May 25 13:45:41 CEST 2005
> % error = 0.0
> 
> Starting SIENAX (including testing BET and FLIRT and
> FAST) at Wed May 25 13:47:39 CEST 2005
> checking error on BET:
> % error = 0.0
> checking error on FLIRT:
> % error = 1085.46
> Warning - test failed!
> checking error on FAST:
> checking error on single-image binary segmentation:
> % error = 2.07
> Warning - test failed!
> checking error on partial volume images:
> % error = 1.1
> Warning - test failed!
> % error = 0.44
> % error = 0.28
> checking error on SIENAX volume outputs:
> % error = 0.17
> % error = 0.02
> % error = 0.57
> % error = 0.46
> % error = 1.03
> Warning - test failed!
> 
> Starting BET2 at Wed May 25 14:03:19 CEST 2005
> checking error on T1 brain extraction:
> % error = 0.0
> checking error on skull and scalp surfaces:
> % error = 0.09
> % error = 0.16
> % error = 0.32
> 
> Starting FEAT at Wed May 25 14:30:59 CEST 2005
> checking error on filtered functional data:
> % error = 0.19
> checking error on raw Z stat images:
> % error = 0.33
> % error = 0.17
> % error = 0.3
> checking error on thresholded Z stat images:
> % error = 0.29
> % error = 0.18
> % error = 0.35
> checking error on position of largest cluster of
> Talairached zfstat1:
> % error = 0.0
> % error = 0.0
> % error = 0.1
> % error = 0.07
> % error = 0.1
> % error = 0.19
> 
> Starting MELODIC at Wed May 25 14:49:36 CEST 2005
> % error = 0.0
> 
> Starting FDT (bedpost) at Wed May 25 15:09:08 CEST
> 2005
> checking error on bedpost output:
> % error = 0.25
> % error = 0.02
> % error = 0.07
> % error = 0.07
> % error = 0.1
> % error = 0.22
> % error = 0.22
> 
> Warning - not all tests passed
> 
> end time = Wed May 25 15:50:14 CEST 2005
> 
> 6818.41user 269.82system 2:04:37elapsed 94%CPU
> (0avgtext+0avgdata 0maxresident)k
> 0inputs+0outputs (645543major+28231604minor)pagefaults
> 0swaps
> [4]+  Done                    emacs dtifit.cc  (wd:
> ~/fsl/src/fdt)
> (wd now: ~/test_fsl/feeds)
> [elin@freja feeds]$
> 

-- 
-------------------------------------------------------------------------------
Tim Behrens
Centre for Functional MRI of the Brain
The John Radcliffe Hospital
Headley Way Oxford OX3 9DU
Oxford University
Work 01865 222782
Mobile 07980 884537
-------------------------------------------------------------------------------