Hi - the SIENA/X web-page reporting gives you images showing the various stages of the processing - you need to look carefully through this to find out at what stage the registration has not worked.   To increase the robustness of such analyses it is always wise/safe to make sure the input data is oriented as with the MNI152, and also potentially use fslroi to remove large amounts of unnecessary neck slices.

Steve.



On 16 Apr 2013, at 10:52, Patricia Pires <[log in to unmask]> wrote:

Dear all,

I wrote this a few days ago, unfortunately I still have not solved the problem. Any help would be appreciated.

Yours sincerely,

Patricia


---------- Forwarded message ----------
From: SUBSCRIBE FSL Patricia Pires <[log in to unmask]>
Date: 2013/4/12
Subject: [FSL] Sienax
To: [log in to unmask]


Dear FSLexperts,

i am using Sienax for a cross-sectional volume comparison (GM and WM) between two groups.
I am using crop T13D NIfTI images and run sienax:

sienax co_image.nii -B "-f 0.22" -d

I've noticed that some images are well segmented or it is easy to modify them just moving "-f" option, however other images are very bad segmented because it was used a quite strange image. I attach both examples, respectively.

How is it possible if i proceed equally for all the analyses?

Any help would be very appreciated.


<I2std.png><I2std.png>


---------------------------------------------------------------------------
Stephen M. Smith, Professor of Biomedical Engineering
Associate Director,  Oxford University FMRIB Centre

FMRIB, JR Hospital, Headington, Oxford  OX3 9DU, UK
+44 (0) 1865 222726  (fax 222717)  2176
[log in to unmask]    http://www.fmrib.ox.ac.uk/~steve
---------------------------------------------------------------------------