Hi,
the size of 4D files I'm merge is around 1000MB and the data type is
int16. Using fs -l on the resulting file produces
hanne@cortex:~$ ls -l data.nii.gz
-rw-r--r-- 1 hanne hanne 12387464423 2008-09-10 16:03 data.nii.gz
The merging works for two 4D datasets of this size, not more. I'll try
to merge data with fewer voxels, and see how it goes.
Hanne Antila
[log in to unmask]
0500-563674
----- Original Message -----
From: Mark Jenkinson <[log in to unmask]>
Date: Wednesday, September 10, 2008 12:00 pm
Subject: Re: [FSL] fslmerge and bad image/header file
To: [log in to unmask]
> Hi,
>
> That's really strange - although the image you are trying to
> make is *very* large (3.4GB for 32-bit int/float datatype or
> 6.8GB for 64-bit doubles).
>
> Can you send us the results of ls -l on the resulting file (or files).
> If fslhd is failing to open it then this is a very bad sign.
>
> Another useful test would be to try and merge slightly smaller
> images to find out when it stops working and hence what the
> critical size is.
>
> All the best,
> Mark
>
>
>
> On 10 Sep 2008, at 08:45, Hanne Antila wrote:
>
> > Hi,
> >
> > By "completing ok" I mean that there are no error messages and the
> > result of fslmerge is about the right size if one thinks it's size
> > should be the sum of the sizes of the files beeing merged.
> >
> > The dimensions of 4D files are x=182 y=218 z=182 and the amount of
> > images (t) is 128. These dimensions are the same for all subjects
> and> indeed, I'm trying to merge the 4D files from different
> subjects into
> > one file.
> >
> > Using fslhd to the result of fslmerge produces an error message:
> >
> > ** ERROR (nifti_image_read): failed to find header file for 'data'
> > ** ERROR: nifti_image_open(data): bad header info
> > Error: failed to open file data
> > ERROR: Could not open file
> >
> > Naturally, the files I'm trying to merge are quite large. Could
> this
> > be
> > a memory problem? And if so, how this could be solved? I'm
> running a
> > 64
> > bit linux machine.
> >
> > Hanne Antila
> > [log in to unmask]
> > 0500-563674
> >
> > ----- Original Message -----
> > From: Markus Gschwind <[log in to unmask]>
> > Date: Tuesday, September 9, 2008 6:53 pm
> > Subject: Re: [FSL] fslmerge and bad image/header file
> > To: [log in to unmask]
> >
> >> Hi!
> >>
> >> Are you trying to concatenate the 4D files from different subjects
> >> into ONE file?
> >>
> >> Why are you thinking that the process completed OK?
> >>
> >> Are the diemensions between the different subjects the same?
> (fslhd)>>
> >> What are the dimensions of your 4D+4D+4D file? (fslhd)?
> >>
> >> All the best,
> >> Markus
> >>
> >> 2008/9/9 Hanne Antila <[log in to unmask]>:
> >>> Dear experts,
> >>>
> >>> I'm trying to merge 4D diffusion datasets from different subjects
> >> using> fslmerge, but I have been running into some problems. The
> >> process seems to
> >>> be completed ok and no error messages emerge. However, when I try
> >> to open
> >>> the result of fslmerge in fslview, the program annouces
> >> following: "Missing
> >>> header/image file" and "Failed to load base image. Command line
> >> processing> aborted"
> >>>
> >>> Using bedpostx_datacheck on the same file the error message is
> >>>
> >>> ** ERROR (nifti_image_read): failed to find header file for
> './data'>>> ** ERROR: nifti_image_open(./data): bad header info
> >>> Error: failed to open file ./data
> >>> ERROR: Could not open file
> >>>
> >>> I use the following syntax
> >>>
> >>> fslmerge -t data data_1 data_2 data_3 data_4 ... data_12
> >>>
> >>> When merging only two datasets, though, everything works ok.
> >>>
> >>> Does anyone know what could be wrong and how can I fix this? All
> >> help will
> >>> be appreciated!
> >>>
> >>
> >>
> >>
> >> --
> >> Dr. med. Markus Gschwind, M.D.
> >> Laboratory for Neurology and Imaging of Cognition
> >> Dept of Neurosciences
> >> University Medical Center (CMU)
> >> 1 Michel-Servet - 1211 GENEVA - CH
> >>
> >> Tel 0041 (0) 22 379 5324
> >> Fax 0041 (0) 22 379 5402
> >> email: [log in to unmask]
> >> http://labnic.unige.ch
> >>
> >
>
|