[Mristudio-users] saving fibers - flipping?

[Hangyi Jiang]

hi, Miguel,

if your file is in Analyze, you should open/load it as Analyze, not Philips REC (athough the .img looks like .rec file) .  their coordinate system are different.  Analyze uses left-hand system, REC is right-handed.

in DtiStudio -> DTI Mapping,  it does't accept Analzye file because it only takes the original DW-Image, not pre-processed images (which may contain scaling factors, and affact the tenosr calculation results).

you may try:  Dtistudio->MriView3D-> load Analyze files  -> resave all images as on "RAW" file and then
                    Dtistudio->DtiMapping-> load "RAW" file as Philips REC format..... OK


basically, when you loaded the Analyze file, the program will adjust its coordinate system accordingly, and re-construct the images correctly.  then, resaved "RAW" file is in same format as that of Philips REC.  in this way, you can processing the "RAW" file as REC format.

regards,

hangyi









________________________________
From: mristudio-users-bounces at mristudio.org [mristudio-users-bounces at mristudio.org] On Behalf Of Miguel Burgaleta [miguel.burgaleta at uam.es]
Sent: Friday, November 19, 2010 5:16 AM
To: DTI Studio, ROI Editor, DiffeoMap Questions/Support
Subject: Re: [Mristudio-users] saving fibers - flipping?


Hi Susumu,

Not sure if that solves my problem... Also, Hangyi's answer was:

please beware that the Analyze file is in Format 7.5.  it is left-handed  system.
if you loaded it as Nifti-Analyze,  the program may loaded as Nifti format, which is right-hand system.

Let me give you more info.

I processed my DWI in FSL for motion and eddy currents correction. Saved the images as 4D with .img extension, and opened them in DTIStudio as REC (the only option that worked out... when trying to open as Analyze, it complained about the header. None of the other format options worked). Did the fiber tracking and saved the fiber as Analyze.

Please see attached a capture of a small bundle (but strategically located in the brain to track potential rotations) as it was reproduced in DTIStudio, and then opened with fslview. First, in DTIStudio, right seems to be right. Second, it looks like, when importing into FSL, the fiber is flipped along the AP and SI axes, and perhaps LR as well...

This looks like a real mess...Any suggestions? Perhaps start from the scratch?? ;)
Thanks,
Miguel