[Mristudio-users] Results using analyze format in DTIStudio

[susumu]

Hi Graeme,

 

I believe FA, ADC, and color maps are insensitive to right-left, front-back,
and up-down flip. It becomes in issue when you do fiber tracking.

 

DtiStudio support the Analyze format to save 3D images after tensor
calculation. If you save data as Analyze and read as Analyze, the
orientation should be correct, but if it is saved as Analyze and read as
Raw, you'd find that the axial image is upside-down, due to Neurology -
Radiology conventions. 

 

In DtiStudio, we do not support tensor calculation of DWIs in the Analyze
format. Also, it is difficult to confirm what would happen if DtiStudio
processes data partially processed by other software programs because it is
difficult for us to test many combinations of data processing by multiple
programs. I would suggest to stick to single software for DWI - tensor
calculation. 

 

It is indeed interesting that you saw 0.2 FA difference by going through the
Analyze format, but again, it difficult for us to follow the results that
were not processed by DtiStudio for the entire processing. If you experience
0.2 difference with and without AIR in DtiStudio, we would definitely look
into the issue. If you mean difference is between;

 

1)       raw DWI -> DtiStudio -> AIR in DtiStudio -> FA

2)       raw DWI -> FSL (eddy correction) -> Analyze -> DtiStudio -> FA

 

We first need to know 

1)       how did you read Analyze into DtiStudio?

2)       Have you compared raw DWIs and Analyze (after FSL eddy correction)
DWIs?

 

  _____  

From: mristudio-users-bounces at mristudio.org
[mailto:mristudio-users-bounces at mristudio.org] On Behalf Of Graeme Schwindt
Sent: Sunday, July 06, 2008 2:44 PM
To: Mristudio-users at mristudio.org
Subject: [Mristudio-users] Results using analyze format in DTIStudio

 

Hello,

At the Toronto workshop, it was mentioned that loading analyze format images
into dtistudio is not recommended as the images are loaded in flipped, and
so the gradient tables will not apply properly, as I understood. 

We have analyzed a couple of subjects from two separate datasets using both
the raw dicoms in dtistudio (following AIR), but also with some analyze
format images that have been preprocessed in FSL (eddy-current correction). 

Surprisingly, the FA maps and colour maps are qualitatively very similar
between the two methods, which is counterintuitive given the improper
mapping between the gradient table and orientation of the analyze imgs.
Also, the FA values are generally higher by 0.2 or so in the analyze version
of the analysis. Is it possible that the analyze format is actually being
read properly though its display is wrong? Any thoughts on the validity of
using analyze format would be appreciated.

Thanks,
Graeme



-- 
Graeme C. Schwindt, HBSc
MD/PhD Student
University of Toronto
graeme.schwindt at gmail.com
416.546.5453 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20080707/7d4c5d23/attachment.html