[Mristudio-users] program crashes

Tue May 5 02:05:13 EDT 2009

Hi Susumu:

Thank you for answering my question regarding the gradient table. However the program crashes every time when I try to do a calculation (DTI mapping). I've tried to calculate Tensor, Color map, ADC map, Mean, but I always get this "The mriView3D MFC application has encountered a problem and needs to close".

I am a new user to DTI studio, so there are probably a lot of mistakes I've made on my end, but could you please tell me what are the likely causes for the crash of my program? Your time and help is much appreciated.

Regards

Eric Q

Date: Fri, 1 May 2009 09:32:56 -0400
From: susumu at mri.jhu.edu
To: mristudio-users at mristudio.org
Subject: Re: [Mristudio-users] Gradient table

Hi Liang,

1)      I
am not sure which coordinates (space) the ultimate gradient table was relative
to, such as frequency- and phase-encoding of image, horizontal and vertical
direction of image, or magnets' x/y/z coordinates. Based on your reply, I think
that it might be the last one.

When you present vectors, color maps, or performing
fiber tracking, X, Y, and Z need to be defined by the image matrix. This is the
most natural way. The frequency/phase/slice and image x/y/z have permutation
relationship, like frequency – image x / phase – image y / slice –
image z or frequency – image y / phase – image x / slice – image
z.

The relationship between magnet x / y / z and image x
/ y / z could have oblique relationship if oblique scan is used, and therefore
could be very complicated.

The gradient table, ultimately, needs to adopt the
image x / y / z system. Now most vendor system is adopting a way to deliver
gradient application in the frequency/phase/slice system, moving away from
gradient x/y/z. This means, when oblique scan is employed, the scanner
dynamically recalculates the gradient table in the f/p/s coordinates. If the
scanner doesn’t do this (like old Siemens system), we have to recalculate
the table based on the oblique angles by ourselves. The former is more fool-proof.

2) In my *.PAR file, there was last 3 columns representing diffusion directions
but obviously different between subjects. I don't know what diffusion were
recorded in this header file.

I don’t have much experience with the new PAR
file, but if you are using oblique scans, it is not surprising the each patient
have different gradient table in the “magnet” coordinate. Suppose
we want to apply [1, 0, 0] in the first line of the gradient table, which is the
“X” orientation. If we use an oblique scan, magnet X and image X
are no longer the same. As I mentioned above, in the end, what we want to know
is gradient orientation in the image coordinate. So we have two choices. 

Choice 1: deliver magnet [1, 0, 0], meaning pure X
gradient application. This X gradient could be something like [0.9, 0.15, 0] in
the image coordinate. We need to do [1, 0, 0] x ImageRotation to get new
gradient table to do tensor calculation.

Choice 2: deliver magnet [0.9, -0.15, 0] (I just made
up these numbers) so that the applied gradient aligns the image X axis. In this
case, we can use gradient table [1, 0, 0] for tensor calculation.

I hope this makes your questions answered.

Susumu

From:
mristudio-users-bounces at mristudio.org
[mailto:mristudio-users-bounces at mristudio.org] On Behalf Of liang wang

Sent: Saturday, April 18, 2009
2:42 PM

To: DTI Studio, ROI Editor,
Landmarker Questions/Support

Subject: Re: [Mristudio-users]
Gradient table

Dear Susumu,

I also have a similar question about gradient tables in Philips scanner.

My data were collected using following protocols:

1) Gradient table: Medium

2) Gradient overplus: no

3) Oblique plane: yes, and different angles between subjects.

Looking through your answer, I know that this set can result in the same
gradient table input into DTIstudio for all subject. However, 1) I am not sure
which coordinates (space) the ultimate gradient table was relative to, such as
frequency- and phase-encoding of image, horizontal and vertical direction of
image, or magnets' x/y/z coordinates. Based on your reply, I think that it
might be the last one.

2) In my *.PAR file, there was last 3 columns representing diffusion directions
but obviously different between subjects. I don't know what diffusion were
recorded in this header file.

Many thanks

Liang

2009/4/16 susumu <susumu at mri.jhu.edu>

Hi Eric,

For Philips scanner there are several types of gradient
tables you can specify in the scanner. Please check your protocol for the
following points;

1)       Gradient table: High (32-orientation)/ Medium (15-orientation) /
Low (6-orientation)

2)       Overplus (ON / OFF)

3)       Oblique plane (X/Y/Z angles)

The first option has 3 choices and the second option has 2
choices. Therefore, there are 3 x 2 = 6 gradient tables.

If the second option (Overplus) is NO, the third option has
no effect. If the second option is YES, the gradient table has to be rotated
based on the oblique angles.

To simplify the DTI processing, we recommend Overplus = NO.

We have the basic 6 tables, which can be used as is if you
use Overplus = OFF or Oblique angles = 0/0/0 degree.

Please let us know which options you are using.

Susumu

From: mristudio-users-bounces at mristudio.org
[mailto:mristudio-users-bounces at mristudio.org]
On Behalf Of eric Q

Sent: Thursday, April 16, 2009
9:44 PM

To: mristudio-users at mristudio.org

Subject: Re:
[Mristudio-users] Gradient table

Hello Dear DTIstudio users:

I am new to DTIstudio, I was hoping someone could give me some much needed
advises for beginners.

I was trying to obtain the FA values for my brain phantom (PHILIP 3T scanner).
When I clicked on "DTI mapping", it prompted me to enter the
"gradient table", how do I calculate the "gradient table"?

Thanks in advanced

Best regards

Eric

Date: Thu, 9 Apr 2009 11:04:09
-0400

From: susumu at mri.jhu.edu

To: mristudio-users at mristudio.org

Subject: Re: [Mristudio-users] MriView3D application close

Hi Hugh,

When you got this error, please suspect one of the following
two reasons;

1) Memory shortage:

Even if you have many GB of memories, 32-bit Windows system
can allocate only 1 - 1.5 GB to each program. To check if you have the memory
problem, please restrict the slices to be calculated. In the
initial data parameter window, there is a section called "Slices to
be processed". Here you can specify only a portion of slices to be
processed. Please limit the slices to, for example, 0-9 and see how it goes. If
you can calculate tensors without errors, then you have the memory problem. The
solution is; a) upgrade to the 64-bit system or b) devide the calculation to 2
batches (e.g. slice 0-29 for the first batch and 30-59 for the second batch),
calculate all maps (e.g. FA), and later combine them.

2) Mismatch between the gradient table and the number of
images:

For example, if you specify 12-orientation table and 30
volume images, then you get the error. After loading the data and before you go
to the Dti Map tab, make sure that you have 1 b0 + 64 DWIs in the pull-down
menu in the right column of the Image tab. If you have the correct number of
the images, then you might have an error in your gradient table. Small mistakes
such as missing "," are often difficult to detect. The best way is;
a) load the data using DtiMapping under the "File" button on the top
left, b) close the window, c) open the image paremater window again by clicking
"DtiMapping" in the "File" button again, d) the window still
remembers the gradient table you entered last time. This remembered window
tells how DtiStudio interpreted your gradient table. If there is any syntax
errors, you may find them much easier. For example, if you enter, "4: 1
0.5 0", instead of "4: 1, 0.5, 0", then you would find, "4:
1, 0, 0".

If any of these do not solve the issue, please let us know.

Susumu

From: mristudio-users-bounces at mristudio.org
[mailto:mristudio-users-bounces at mristudio.org]
On Behalf Of Hugh Wang

Sent: Wednesday, April 08, 2009
6:30 PM

To: mristudio-users at mristudio.org

Subject: [Mristudio-users]
MriView3D application close

Hi, 

I was using the dtistudio (ver. 2.4)  to look at the Siemens DICOM
diffusion images (1baseline,64 gradients). After successfully loading the
images, I click the "dti map" option at lower right corner. Whatever
calculation I choose, it always gives me an error: 

"MriView3D MFC Application has encountered a problem and needs to
close.  We are sorry for the inconvenience."

I tried several PCs, and the problem keeped the same. I would appreciate that
if there's anyone having any idea on how to solve it. Thanks.

Best,

Hugh

Find
car news, reviews and more Looking to change your car this year?

_______________________________________________

Mristudio-users mailing list

Mristudio-users at mristudio.org

http://lists.mristudio.org/mailman/listinfo/mristudio-users

-- 

Liang Wang, PhD

Postdoctoral Fellow

Woodward Lab

Department of Psychiatry

University of British Columbia

BC Mental Health & Addiction Services

938 West 28th Avenue

Vancouver BC
 V5Z 4H4

Telephone: 1-604-875-2000 (ext. 4735)

Fax: 1-604-875-3871

Email: wanglbit at gmail.com

_________________________________________________________________
Looking for a fresh way to share your photos? Check out the new Windows Live Messenger
http://windowslive.ninemsn.com.au/article.aspx?id=792335
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20090505/97a6dd37/attachment-0001.html 