[Brains-users] Re: SV: Re:

[Eugene Zeien]

Reinstalling brains2 may help.  Cygwin support is rather limited, we have one person
who works on it part-time.  I found a previous version, which will probably have
different issues...

http://www.psychiatry.uiowa.edu/users/gene/brains2-cygwin-beta-3-21-2004.tar.gz

I know this one doesn't support nifti.  I doubt if the one you acquired does either.
We just got NIFTI support working properly in our Linux version this month.
/opt/brains2/lib/libconnectBrainsItkTcl.so does not exist on my computer either :-)
I do not know why BRAINS2 still wants to use that...  The absence of that library
does not seem to have any effect.

Olof Lindberg wrote:
> Dear mr. Zeien,
>   sorry for testing your patience! The more I try the more stubid I feel. I get the following messages when I´m open the image I at least can open:
>   Warning: Brains2 <-> Itk Conversion library /usr/local/brains2/lib/libconnectBrainsItkTcl does not exist
>    
>   I have tryied to open nifti files from FSL, analyze files from MRIcro and brains2 will not accept either of them - even if I´m perfectly able to open them in the other programs.
>    
>   For the moment I´m thinking of trying to do the installation al over again since som much seems to go wrong. The question I would like to ask you is this. In the manual it says that I should download : Brains2_cygwin.tar.gz. I can´t find this file however the only one that has cygwin in the name is called something like: Brains2_cygwin_DEBUG_2005.....
>    
>   Is this maybe the wrong file? If so could you please give me the name of the right file. Ofcourse if you have any other suggestions I be happy to listen
>    
>   Yours sincerely 
>   Olof Lindberg
> 
> Eugene Zeien <eugene-zeien at uiowa.edu> skrev:
>   Hmm. Brains2 should be able to read the DICOM images.
> However, brains2 will fail to read those if there are multiple
> sequences in the same directory.
> 
> 
> I have used medcon to process DICOM images before,
> (read the medcon man page to determine which axes flips you need)
> 
> medcon -c anlz -spm -fv -qc --files ${petid}_Oblique_[0-9]*
> 
> Which generates a separate .img and .hdr for each slice.
> To combine those into a single analyze image,
> 
> cat `ls -1r m*Oblique*.img` > ${petid}_Oblique.img
> Keep one .hdr. Any of the header files will suffice,
> cp `ls -1 m*Oblique*.hdr | head -1` ${petid}_Oblique.hdr
> 
> Then, we have to edit some of the .hdr internal values,
> since all the .hdr files think there's only 1 slice.
> Change the 128\n128 to match the slice dimensions of your
> images. Commonly, this is 256\n256 or 512\n512. Finally,
> replace $#obl with the correct number of slices for your image set.
> 
> printf "INT\n${petid}_Oblique\n4\n128\n128\n$#obl\n1\n\n\n\n\n\n\n\n0\nY\n\n" |\
> /usr/local/brains2/bin/AnalyzeHeaderEditor ${petid}_Oblique.hdr >& /dev/null
> 
> No guarantee this will work for you, but I still use this sequence of actions
> to convert some of our DICOM images to Analyze format with medcon.
> 
> Olof Lindberg wrote:
> 
>>Dear mr Zeien,
>>thanks again for the help in my effort trying to start Brains2.
>>I become a bit confused. I looked at the different suggetions you gave me and as far as I can tell I have a working program here. I made another effort converting an image to dicom - and suddenly I could at least open the image. I still get a problem messages and I am not able to do much with the image since the program do not seem to completely accept it. However I have done som checking - If I write strict analyze at cygwin prompt I can see that the program recognize the command since I get a lot of different options.
>>
>>The question is if I do something wrong when I convert to analyze. I use medcon: medcon -c anlz -f stack3d filename.
>>Is something else requiered by brains?
>>
>>yours sincerely
>>Olof Lindberg
>>
> 
> 
>   
> 
>