Martin,
This script is for rotating all of the
vectors the same amount, for example if you register your DTI data to another space
and want to recalculate the quantitative DTI parameters. It would certainly be
possible to make a script that takes the rotation components of the
eddy_correct for each gradient individually and applies them to the table.
Peace,
Matt.
From:
Sent: Sunday, September 07, 2008
1:48 AM
To: [log in to unmask]
Subject: Re: [FSL] correcting dti
bvecs after registration
Hi Matt,
could you please explain me: you take one flirt matrix and you use it to rotate
components of the whole bvecs? While this could be correct for rotating
gradient component of, say 5th DTI image with respect to the reference, the
10th image may be rotated differently than the 5th one.
Thanks,
Martin
On Sat, Sep 6, 2008 at 6:26 PM, Matt Glasser <[log in to unmask]> wrote:
It is attached.
Matt.
From:
Sent: Saturday, September 06, 2008
11:18 AM
To: [log in to unmask]
Subject: Re: [FSL] correcting dti
bvecs after registration
Hi Matt,
I would like to try this script. how I can get it ?
reg, vin
On Sat,
Sep 6, 2008 at 7:40 AM, Matt Glasser <[log in to unmask]> wrote:
It is currently extra, but will be included in FSL in the
future.
Peace,
Matt.
From:
Sent: Friday, September 05, 2008
6:17 PM
To: [log in to unmask]
Subject: Re: [FSL] correcting dti
bvecs after registration
also mriconvert software(dicom--fsl nifti converstion)
will give you the desired format as Matt wrote.
I want to ask , is this script to rotate the gradients using FLIRT transform is
included in FSL ? or extra.
vin
On Fri,
Sep 5, 2008 at 9:08 PM, Matt Glasser <[log in to unmask]> wrote:
Like I said, you need to convert your bvecs file to be of the
form:
x1 x2 x3 … x12
y1 y2 y3 … y12
z1 z2 z3 … z12
The script I wrote will not work with the gradient table in
the format you have. If you want to convert the gradient table to the
correct format quickly, highlight the cells in excel, select paste special, and
transpose.
Peace,
Matt.
From:
Sent: Friday, September 05, 2008
8:33 AM
To: [log in to unmask]
Subject: Re: [FSL] correcting dti
bvecs after registration
Hi, I'm not sure if I'm correctly interpreting the
output of Tbvecs.
My bvecs file looks like this:
0.89440000 0.00000000 -0.44730000
-0.00010000 0.44730000 -0.89450000
0.44720000 0.89450000 0.00000000
0.89400000 0.44830000 0.00000000.
.....
.....
-0.44730000 0.00000000 -0.89450000
where the rows are n=12. The output I get is:
0.8951819 0 0
-0.001100128 0.4473 0
0.4456314 0.8945 0
What's the meaning of this 3x3 matrix? I expected the output to contain 12
(new) directions...I got no warning nor errors.
Thanks,
Michela
On
9/5/08, Matt Glasser <[log in to unmask]> wrote:
An FSL
formatted bvecs file contains x, y, and z rows, and a number of
columns that equals the number of directions + number of b0 images acquired.
The code won't work unless the bvecs file is formatted in this way.
Peace,
Matt.
-----Original Message-----
From:
Of Michela Pievani
Sent: Friday, September 05, 2008 5:23 AM
To: [log in to unmask]
Subject: Re: [FSL] correcting dti bvecs after registration
Hi,
I'm using your script to rotate the gradients using FLIRT transform. I have
a question about the input/output files:
Q1. Should the input file (original bvecs) be formatted as a bvecs FSL file?
If so, there's something wrong with my files: I provided a bvecs
file
containing 12 directions, but the output contains only three rows. What's
wrong in it?
Thanks!
Michela