[Camino-users] problem with mbalign.

[Philip A Cook]

Hi Darren,

So is the problem with the mask or the data? Camino should read a little-endian mask in .nii format. The data should be appropriately converted to Bfloat by image2voxel. If either of these things didn't happen, please let me know more details. If you can also send me the header details by doing

niftiheader -readheader <image> 

I will try to reproduce and fix the problem.

We can do this off list if you like, I will post the solution once we have it

Thanks,


Phil

On Nov 18, 2011, at 7:52 AM, Darren Gitelman wrote:

> Dear Phil
> 
> Yes and yes. There are several b=0 measurements and they are in the scheme file with the directions and b value = 0. 
> 
> I did figure out that one problem was that my images were not all big endian and unfortunately camino doesn't really seem to understand 1 file nifti files, but I fixed all that and confirmed it by making sure that the image files had proper values. The mask file also appropriately covers the brain.
> 
> Darren
> 
> 
> On Thu, Nov 17, 2011 at 10:09 AM, Philip A Cook <cookpa at mail.med.upenn.edu> wrote:
> Hi,
> 
> Does your data include b=0 measurements? Also, is your mask including non-brain regions? These kinds of errors are common if you include background voxels.
> 
> BTW, if you supply a mask, the -bgthresh option is ignored.
> 
> 
> Phil
> 
> On Nov 15, 2011, at 7:27 PM, Darren Gitelman wrote:
> 
> > Hi
> >
> > I am trying to use camino to analyze my DTI data.
> >
> > I am first doing eddy current correction.
> >
> > This is my command line:
> >
> > [drg at zonda camino]$ mbalign -datatype float -schemefile 36dwi_b1000_bvector.scheme -components 36 -datadims 260 260 65 -voxeldims .73846 .73846 2.0 -sigma 14 -inputfile dti001_ser7.Bfloat -bgmask dti001_mask.nii -eddy -outputfile dti001_ser7_eddy.Bfloat -slicecheck /data/zonda1/pddti/dti001/camino/slicecheck/dti001_slicecheck -bgthresh 20
> >
> > This gives me some initial output:
> >
> > :-) I have everything I need!
> > 36 components inside inputfile.
> > No temp directory specified. Trying to use /tmp/dti001_ser7_15.11.11_181736.
> > Successfully created /tmp/dti001_ser7_15.11.11_181736.
> > Linux system detected.
> > Disk space temporarily used during calculation is about 5694624000. Make sure space is available!
> > Fitting Tensors...
> > mask input
> >
> > After this I get a lot of output that looks like the following. I get this output whether or not I specify a bgthresh and at a variety of sigma values. Is this to be expected or do I need to fix something?
> >
> > Nov 15, 2011 6:17:42 PM imaging.DW_Scheme normalizeData
> > WARNING: Can't use normalization constant 0.0
> > Nov 15, 2011 6:17:42 PM inverters.RestoreDT_Inversion invert
> > INFO: optimizers.MarquardtMinimiserException: Singular Matrix in gaussj - 2  Fitting failed.  Outputting best DT found, which may be affected by outliers.
> > Nov 15, 2011 6:17:43 PM imaging.DW_Scheme normalizeData
> > WARNING: Can't use normalization constant 0.0
> > Nov 15, 2011 6:17:43 PM inverters.RestoreDT_Inversion invert
> > INFO: optimizers.MarquardtMinimiserException: Singular Matrix in gaussj - 2  Fitting failed.  Outputting best DT found, which may be affected by outliers.
> > Nov 15, 2011 6:17:45 PM imaging.DW_Scheme normalizeData
> > WARNING: Can't use normalization constant 0.0
> > Nov 15, 2011 6:17:45 PM inverters.RestoreDT_Inversion invert
> > INFO: optimizers.MarquardtMinimiserException: Singular Matrix in gaussj - 2  Fitting failed.  Outputting best DT found, which may be affected by outliers.
> > Nov 15, 2011 6:17:47 PM imaging.DW_Scheme normalizeData
> > WARNING: Can't use normalization constant 0.0
> > Nov 15, 2011 6:17:47 PM inverters.RestoreDT_Inversion invert
> > INFO: optimizers.MarquardtMinimiserException: Singular Matrix in gaussj - 2  Fitting failed.  Outputting best DT found, which may be affected by outliers.
> > Nov 15, 2011 6:17:57 PM inverters.RestoreDT_Inversion invert
> > INFO: Maximum iterations reached without convergence. Final tensor:
> >         1.2312659747573168E-9 -1.0053586166159994E-10 -1.4422426735710772E-11
> > -1.0053586166159994E-10 1.0873050187771824E-9 2.6988536554535517E-11
> > -1.4422426735710772E-11 2.6988536554535517E-11 9.300247616640151E-10
> >
> > --
> > Darren
> > _______________________________________________
> > Camino-users mailing list
> > Camino-users at www.nitrc.org
> > http://www.nitrc.org/mailman/listinfo/camino-users
> 
> 
> 
> 
> -- 
> Darren Gitelman, MD
> Northwestern University
> 710 N. Lake Shore Dr.
> Abbott 11th Floor
> Chicago, IL 60611
> Ph: (312) 908-8614
> Fax: (312) 908-5073
> 
>