[Mristudio-users] Gradient Table for Philips

[susumu mori]

Hi Chunxiao and Alex,

I would be cautious about using the gradient information in PAR file if you
are using oblique-angle imaging.
For example, suppose you apply [1, 0, 0] gradient, which is along X (right -
left). If you rotate the FOV about the Z axes, there are two ways to deliver
[1, 0, 0]. In the first option, regardless of the rotation angle, the [1, 0,
0] is always the physical gradient X. In the second option, [1, 0, 0]
rotates with FOV such that [1, 0, 0] gradient is always along the X axis of
the image, meaning the combination of physical X and Y gradients (e.g. [0.9,
0.4, 0]).

Unless you are using "Overplus = on", the gradient table
is automatically rotated to make sure [1, 0, 0] always means the X axis of
the image, not the physical gradient. In this case, if the first line of the
"vendor-supplied" table is [1, 0, 0], you can use the information as is. The
problem is, if you read the gradient table from PAR file, I am not sure if
it is stored in the image axes ([1, 0, 0]) or physical axes ([0.9, 0.4. 0]).
If the latter is the case, you can't use the information as is.

I wonder if Jonathan or Bennett know the detail of the PAR file gradient
table.

Susumu

On Thu, Dec 2, 2010 at 12:54 PM, Dresner, Alex <alex.dresner at philips.com>wrote:

>  Dear Andrew,
>
> Dr. Mori’s message was more complete than mine was, so I’ve copied his
> response below.  In your .par file, you are correct that those three columns
> are your diffusion directions.  And though it says there are 33 directions,
> the last direction is 0.00, 0.00, 0.00 for the calculated DWI which should
> be replaced with 100, 100, 100 in your table for DTI Studio.  And the
> direction before that (next-to-last in the .par file) is the b0 image,
> because it is labeled with direction 0.00, 0.00, 0.00 and b-value = 1 (the
> column to the left of the gradient number, so those lines read … 1 33 0 0
> 0.000 0.000 0.000 …).
>
>
>
> As far as checking the results to see if you have entered the correct
> gradient table, there’s no specific method.  The easiest thing to do is to
> compare the FA color map to that of the sample dataset and check obvious
> structures like the corpus callosum (should be red) and the optical fibers
> (should be green).  Again I will quote Professor Mori: (DTI Studio mailing
> list, 6/10/10)
>
> There are several ways to notice wrong gradient tables;
>
> 1) Totally wrong table: If the table is totally wrong, the color map
> doesn't make sense. So check color map first.
> 2) X/Y/Z permutation: If X/Y/Z is miss signed, the color of the color maps
> doesn't look right. For example, the corpus callosum should look red, not
> green
> 3) +/- sign error: Perform tractography. If sign is wrong, it doesn't make
> any sense.
> 4) More subtle errors: Look at fitting residual (error) map. There is a
> file called "DtiStudioUpdate.doc" at
> https://www.mristudio.org/wiki/changelog/dtistudio-3.0.2. This explains
> how to get the fitting error maps.
> 5) Oblique angle: The attached file explains in detail. To make sure
> everything is right, you can scan a person twice. One in a regular position
> and one in severe oblique. If everything is working, you should get
> anatomically comparable tractography results.
>
>  Regards,
>
> Alex
>
>
>
> Dear Alex,
>
>
>
> Thank you very much for your prompt response.
>
>
>
>  The attached is my .par file. Do you mean that  the "#  diffusion (ap, fh,
> rl)     (3*float)" , for example  -0.001   -0.076   -0.997, is a diffusion
> direction? Since there are 33 directions in my .PAR file and are labeled
> from 1 to 33, should I change their label to 0,1,..32?  In addition , should
> I add a extra line (33) with 100, 100, 100 in the end of the gradient table?
>
>
>
> By the way, is there any way based on the result I can check whether I use
> the right gradient table?
>
>
>
> Again, thank you for your time and help!
>
>
>
> Chunxiao(Andrew)
>
> * *
>
>
>
> *From:* mristudio-users-bounces at mristudio.org [mailto:
> mristudio-users-bounces at mristudio.org] *On Behalf Of *susumu mori
> *Sent:* Wednesday, November 24, 2010 6:27 AM
> *To:* DTI Studio, ROI Editor, DiffeoMap Questions/Support
> *Subject:* Re: [Mristudio-users] Gradient Table for Philips
>
>
>
> 1) Please read your REC file using Mri3DView. Use the image dimensions you
> get from the PAR file.
>
> 2) After the images are read, please go through the images using the
> pull-down menu in the right column of the viewing window. Check if all the
> images (b0 and DWIs) look fine.
>
> 3) Find out how many 3D images are loaded. If you used Philips "High"
> gradient orientation, you should have 34 images loaded; one b0, 32 DWIs, and
> one calculated DWIs (I think this is the sum of all DWIs).
>
> 4) Make sure where is b0. It could be the first or the second from the
> last. The calculated DWI is always at the end.
>
> 5) Close the window and start DtiMapping.
>
> 6) Cut&Paste your gradient table to the appropriate section. You have to
> make sure several things here;
>
> a: The position of b0; First or the second from the last
>
> b: The number of rows; It has to be 34. Sometimes you find a table with
> only 33 lines ignoring the last calculated image. In this case, you get the
> error message, complaining that the file size doesn't match. Because the
> last image is not used anyway, you can ignore the message and proceed. The
> last image is thrown away. Or, you can increase the number of row to 34 by
> adding 33: 100, 100, 100, in which [100, 100, 100] send the message to
> DtiStudio to remove this image.
>
> c: Please make sure if you used "gradient overplus ON" or "OFF", because
> you have to use different tables.
>
> 7) If you still get the file size error message, something is wrong.
>
> On Tue, Nov 23, 2010 at 5:21 PM, Dresner, Alex <alex.dresner at philips.com>
> wrote:
>
> Dear Chunxiao,
>
> Your gradient table is also in the .PAR file.  You should be able to read
> the .PAR file with any text reader such as WordPad, NotePad or Word; lower
> in the file is a table of parameters per-image.  On the right-hand-side of
> that table are the diffusion directions in three columns.
>
> Regards,
>
> Alex
>
> ____________________
>
> M. Alex Dresner, Ph.D.
>
> MR Clinical Scientist
>
> alex.dresner at philips.com
>
> ____________________
>
> *From:* mristudio-users-bounces at mristudio.org [mailto:
> mristudio-users-bounces at mristudio.org] *On Behalf Of *Chunxiao Zhou
> *Sent:* Tuesday, November 23, 2010 5:07 PM
> *To:* mristudio-users at mristudio.org
> *Subject:* [Mristudio-users] Gradient Table for Philips
>
>
>
> Hello,
>
> When I do tensor calculation for my own data , there is a warning message
> "File size is bigger than DW-image size, continue anyway?". I get all image
> size parameters from my .par file ( I use Philip .REC .PAR data). I guess
> the reason is that I use the default ParRec gradient table from your
> GradientTable file. How could I get my gradient table from my file?
>
> Thank you.
>
> Chunxiao(Andrew)
>
> _______________________________________________
> 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
>
>
>
> ------------------------------
> The information contained in this message may be confidential and legally
> protected under applicable law. The message is intended solely for the
> addressee(s). If you are not the intended recipient, you are hereby notified
> that any use, forwarding, dissemination, or reproduction of this message is
> strictly prohibited and may be unlawful. If you are not the intended
> recipient, please contact the sender by return e-mail and destroy all copies
> of the original message.
>
> _______________________________________________
> 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/20101204/0275d4dd/attachment-0001.html