[Mristudio-users] question on landmarker and tensor image

[susumu]

Hi Elizabeth,

1) First of all, the AIR-based linear transformation can take images with
matrix and pixel sizes different from the atlas and save the normalized
image in the atlas dimension. So you don't need the pre-processing, unless
your data is sagittal and the atlas is axial. AIR would be confused with
this 90-degree initial location difference. As long as the two images are
reasonably close in terms of image orientation, you can just submit the two
images to AIR directory. 
2) AIR results are sensitive to the threshold values. Please make sure using
the "landmarking" capability to see if the AIRed images are registered
reasonably well. The brain height, width, length, and overall orientations
should be well adjusted after the linear normalization. This will set all
required conditions (matrix size, pixel size, and brain locations) for the
next LDDMM. Once you set the good threshold value, that should be good for
all other images (usually) using the same protocol.
3) If you use DWI (the sum of all DWIs) or b0 to drive AIR, you don't have
to worry too much about skull stripping, because skull is well suppressed in
these images and the simple threshold used in AIR can remove it. Sometimes,
you can see some strong intensity from outside the brain such as sinus,
which interfere with AIR linear registration and may need skull stripping.
For skull stripping, you can test the skull-strip button in Landmarker or
threshold/region growing x "masking" function (inside =/-* button) in
RoiEditor.
4) When you load *.d, your *.d is still in the original dimension, while you
said you already change the image dimensions of other images (FA, DWI, b0).
Please use the original dimensions for all images (FA, DWI, b0, tensor) and
submit them to AIR all together. You get normalized FA, DWI, b0, tensor in
the atlas coordinates. AIR takes care for everything. I think you can
resample *.d using the resampling function, save the resampled *.d, and
reload with other resampled images, but it sounds too many steps....

Please let me know, if it doesn't work or you have further questions.

Susumu

-----Original Message-----
From: mristudio-users-bounces at mristudio.org
[mailto:mristudio-users-bounces at mristudio.org] On Behalf Of Elizabeth Finger
Sent: Wednesday, May 21, 2008 4:06 PM
To: mristudio-users at mristudio.org
Subject: [Mristudio-users] question on landmarker and tensor image

Hello-


I have a data set in which I would like to compare the uncinate
fasiculus in the patients vs. controls.  Ideally, I would like to try
both the automated white matter segmentation approach as well as a
manual ROI approach.  So far I have loaded the DTI data (2 runs), run
AIR to register the runs together, and done the manual quality check and
image flag file. I then reloaded the two AIR series as .rec format with
the updated gradient table into DTI mapping and did the tensor
calculation, saving the six tensors in one file (Dxx, Dyy, Dzz, etc.) as
well as the FA, eigen vector, color maps, recalculated mean B0 and DWI
separately as raw format.
My next goal is to normalize the images so that I can use the LDDMM to
use the automated white matter parcelation, to look at the statististics
for the uncinate fasiculus and some control tracts. To do this, what I
think needs to be done next (this is a question) is to choose an atlas,
load the individual subject data and tensor, and perform the AIR
transformation prior to LDDMM?  To do this, I first resampled my images
in Landmarker to fit the parameters of the atlas. I also did skull
stripping in Landmarker, with the GUI button for skull stripping, and
multiplied the resulting mask from the skull stripped B0 by the other
images.  Next, I am able to load the atlas images, and my own resampled
images (FA, B0, DWI) successfully. However, when I try to load the
tensor image using the Open.d file button, the image parameters of the
atlas come up in the pop-up, and whether I load the tensor from DTI
mapping, or a tensor which I attempted to resample, I receive an error
message "File size is smaller than expected value".

I would very much appreciate any suggestions as to the steps I have
taken and what I should do in response to this error. I saw an email
prior with the same error, but can not find how it was resolved.

Thank you,
Elizabeth Finger


----------------------------------------------------------------------------
----
This information is directed in confidence solely to the person named
above and may contain confidential and/or privileged material. This
information may not otherwise be distributed, copied or disclosed. If
you have received this e-mail in error, please notify the sender
immediately via a return e-mail and destroy original message. Thank you
for your cooperation.
_______________________________________________
Mristudio-users mailing list
Mristudio-users at mristudio.org
http://lists.mristudio.org/mailman/listinfo/mristudio-users