[Mristudio-users] DTIStudio isn't reading headers correctly
susumu mori
susumu at mri.jhu.edu
Tue Nov 24 17:26:55 EST 2009
Judy is right. I think you did two signal averaging.
For DTI, Siemens scanners do not actually average data and save them into
separate files. So if you do 2 signal averaging of 13 DWI+b0 scans, you get
26 Mosaic files. Below, I provide some related info you may be interested
in.
1) In regular MR scans, such as T1 and T2, signal averaging is done in
time-domain data (called k-space) with so-called phase-cycling. For example,
if you do 2 averaging, the actual data in the time domain looks inverted. If
one seems to have a positive k-space peak, the other has a negative peak.
These two data are acquired in a way that they are completely opposite. When
the actual averaging is done, they are subtracted, not added; a - (-a) = 2a,
suppose "a" is the signal. They do this to remove consistent noise. Suppose
the non-random consistent noise (such as leakage of certain radio signal
into the scanner), the first scan signal becomes a+b (b = noise) and the
second scan is -a+b. If you subtract these two (a+b) - (-a+b) = 2a; you can
see that the consistent noise is canceled out.
2) In DTI, there is a reason that you can't do this "time domain (k-space)
averaging". This is because of physiological motion such as pulsation. If
you do two scans and one is denoted as "a", then the second scan becomes
"a*", in which "*" means some phase inconsistency among the scans due to
motion. Note that this motion-caused phase inconsistency is due to very tiny
tissue motion. Much smaller than the pixel size, which can not be suppressed
unless the subject is dead. Because of this issue, you can not use the
phase-cycling approach; a - (-a*) is not equal to 2a. As a result, DTI scans
do not do k-space (time-domain) averaging. They calculate image first; the
time domain signal "a" is converted to image "A". Because after magnitude
image calculation, phase information is discarded, the second scan "a*" also
becomes "A". So, you can do simple signal averaging A + A = 2A, in the image
domain.
3) Now, once the time domain data "a" or "a*" are converted to image "A", we
may ask, "what the point of creating 2A by adding two images?". Please note
that once you calculate the magnitude image "A", you can not expect to
cancel out the consistent noise "b" like the time-domain averaging. Such
noise just keeps adding up. Creating 2A images, on the other hand, have
negative impacts. For example, if the subject moved between the two scans,
you can not do image registration anymore once the two scans are added. If
one of the images are corrupted and two images are added, both images get
corrupted. Therefore, it is a good idea to keep the two images stored
separately.
4) Going back to your problem, you need to create a new table like;
0: 0.000, 0.000, 0.000
1: 1.000, 0.414, -0.414
2: 1.000, -0.414, -0.414
3: 1.000, -0.414, 0.414
4: 1.000, 0.414, 0.414
5: 0.414, 0.414, 1.000
6: 0.414, 1.000, 0.414
7: 0.414, 1.000, -0.414
8: 0.414, 0.414, -1.000
9: 0.414, -0.414, -1.000
10: 0.414, -1.000, -0.414
11: 0.414, -1.000, 0.414
12: 0.414, -0.414, 1.000
13: 0.000, 0.000, 0.000
14: 1.000, 0.414, -0.414
15: 1.000, -0.414, -0.414
16: 1.000, -0.414, 0.414
17: 1.000, 0.414, 0.414
18: 0.414, 0.414, 1.000
19: 0.414, 1.000, 0.414
20: 0.414, 1.000, -0.414
21: 0.414, 0.414, -1.000
22: 0.414, -0.414, -1.000
23: 0.414, -1.000, -0.414
24: 0.414, -1.000, 0.414
25: 0.414, -0.414, 1.000
, which is a simple repeated copy of the 12-orientation table, or submit two
separate scans rather than two signal averaging, from which you will get two
separate directories, each contain 13 images. In this case, you can list the
two directories as the sources of data and use
0: 0, 0, 0
1: 1.000000, 0.414250, -0.414250
2: 1.000000, -0.414250, -0.414250
3: 1.000000, -0.414250, 0.414250
4: 1.000000, 0.414250, 0.414250
5: 0.414250, 0.414250, 1.000000
6: 0.414250, 1.000000, 0.414250
7: 0.414250, 1.000000, -0.414250
8: 0.414250, 0.414250, -1.000000
9: 0.414250, -0.414250, -1.000000
10: 0.414250, -1.000000, -0.414250
11: 0.414250, -1.000000, 0.414250
12: 0.414250, -0.414250, 1.000000
The former dataset is considered as one set of 26 DWI+b0 data. The latter
will be considered as (13 DWI+b0)x2 data. The advantage of the second
approach is that DtiStudio can provide you subtracted images of the two
sets, from which you can get idea about patient motion.
I hope it helps.
Susumu
On Tue, Nov 24, 2009 at 3:57 PM, Judy James <JJames2 at radiology.umsmed.edu>wrote:
> How many averages have you collected your data with?
> Some scanners if the mosaic option is turned on, will collect images twice
> instead of averaging it into 1 data set and this can cause 12 directions x 2
> avg. + 1(b0) x 2 avg. = 26 sets of images.
>
> Judy
>
>
> *****************************************************
> Dr. Judy Rose James
> MR Physicist/Assistant Professor
> Department of Radiology
> University of Mississippi Medical Centre
> Jackson, MS 39216
> Ph: 601-984-2585
> Pager: 601-929-2442
> email: jjames2 at radiology.umsmed.edu
> *****************************************************
> >>> Ryan <wmcgarry at gmu.edu> 11/24/09 2:42 PM >>>
> I'm encountering some trouble where DTIStudio isn't reading my headers
> correctly. I have MRI data collected with an Allegra scanner. It
> contains a b0 with 12 directions, but 26 appear in the drop-down list
> on the "Image" tab. I would greatly appreciate any ideas people may
> have. Thanks.
> -Ryan
> _______________________________________________
> Mristudio-users mailing list
> Mristudio-users at mristudio.org
> http://lists.mristudio.org/mailman/listinfo/
> Unsubscribe, send a blank email to:
> Mristudio-users-unsubscribe at mristudio.org
>
>
>
> Individuals who have received this information in error or are not
> authorized to receive it must promptly return or dispose of the information
> and notify the sender. Those individuals are hereby notified that they are
> strictly prohibited from reviewing, forwarding, printing, copying,
> distributing or using this information in any way.
>
>
>
> _______________________________________________
> Mristudio-users mailing list
> Mristudio-users at mristudio.org
> http://lists.mristudio.org/mailman/listinfo/
> Unsubscribe, send a blank email to:
> Mristudio-users-unsubscribe at mristudio.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20091124/92183793/attachment.html
More information about the Mristudio-users
mailing list