[Mristudio-users] Is better to do 3 DTI series or just one well done?

Sat Dec 12 12:21:40 EST 2009

I'm using a GE too, and only 33 images (for the dirrections) are created.
My doubt is wich one is better, use the 3NEX ou make 3 series, but as you said before, I will keep using the 3 NEX and hope for patients cooperation.
Thank you very much.

--- Em sáb, 12/12/09, susumu mori <susumu at mri.jhu.edu> escreveu:

De: susumu mori <susumu at mri.jhu.edu>
Assunto: Re: [Mristudio-users] Is better to do 3 DTI series or just one well done?
Para: "DTI Studio, ROI Editor, Landmarker Questions/Support" <mristudio-users at mristudio.org>
Data: Sábado, 12 de Dezembro de 2009, 0:14

Do you mean you are doing 33-orientation x 3 signal averaging for 5 mm slice? That is equivalent to 99 DWIs. With such as thick slice and the 99 DWI equivalent, you should have enough SNR even with 1.5T.

3 NEX and 3 series are equivalent in terms of SNR. The only difference is whether the scanner average the data for you or you get them saved separately. Usually, if you do 3 separate series, MR scanner can not combine them and does not offer an FA map based on the all 3 series. So, for clinical scans, you have no choice but using 3 NEX.

If you are using Siemens, 3NEX and 3 series are completely identical because they actually save the 3 NEX scans separately and use all of them for tensor calculation. For example, if you do 3NEX of 33-orientaiton, you can find 99 Mosaic files. They don't actually average 3 scans. So it's the same as 3 series.

I'm not sure how GE works. I believe if you do 3NEX of 33-orientaiton, you may get only 33 images, in which the 3 scans are already averaged. The disadvantage of this approach is, if the subject moved or one of the images are corrupted, you can not correct them once 3 images are averaged. These issues can be corrected to some degree if you are using off-line programs like DtiStudio or FSL. However, if you rely on FA map generation by the scanner, I guess you'd better use 3NEX.

On Fri, Dec 11, 2009 at 11:05 AM, Lucas Lessa <lucas_lessa at yahoo.com.br> wrote:

Once again thank you Dr Mori.
and what do you think about increase the SNR in GE 1.5T scanners?
We only do one serie with 33 directions, but we have to use slices with 5mm thickness, and NEX 3. Is it better to do 3 series like in Phillips or push the parameters?

--- Em qui, 10/12/09, susumu mori <susumu at mri.jhu.edu> escreveu:

De: susumu mori <susumu at mri.jhu.edu>
Assunto: Re: [Mristudio-users] Is better to do 3 DTI series or just one well done?
Para: "DTI Studio, ROI Editor, Landmarker Questions/Support" <mristudio-users at mristudio.org>

Data: Quinta-feira, 10 de Dezembro de 2009, 16:59

I wouldn't push the resolution that much. I even hesitate to use 128x128, unless we can
 use SENSE factor of 4 so that the echotrain is 128/4 =32

On Thu, Dec 10, 2009 at 1:33 PM, Lucas Lessa <lucas_lessa at yahoo.com.br> wrote:

Dr. Susumu Mori you helped me a lot.
I'm using 3 repetions of a 33 dir DTI in a Philips 1.5T (the voxel is about 2.2mm.) when I want to do an exam for a research, for the real life I use just one repetion.
Indeed the SNR is much better with 3 repetions, but just in DTI Studio, because Philip's software don't handle the 3 series like one.
What do you think in increase my resolution to 256x256 in this scanner?
Will the exam take so long? Can I use just 2 repetitions?
Thank you.

Lucas Lessa.

--- Em qui, 10/12/09, susumu mori <susumu at mri.jhu.edu> escreveu:

De: susumu mori <susumu at mri.jhu.edu>

Assunto: Re: [Mristudio-users] Is better to do 3 DTI series or just one well done?

Para: "DTI Studio, ROI Editor, Landmarker Questions/Support" <mristudio-users at mristudio.org>
Data: Quinta-feira, 10 de Dezembro de 2009, 15:22

If you are using a 1.5 T scanner, I think 2.5 mm isotropic resolution is a good compromise between scanning time and SNR. With a 30-orientation scheme, you may want to do at least two repetition (total 60 DWIs) to get a decent SNR. If you want to increase the resolution, you need more dwis and a longer scanning time.

What is important is the total number of images, in this case 60. So, if you are using a 15-orientation scheme, you need 4 repetitions. A 30-orientation scheme should take about 4-5min in Siemens and Philips. GE may take 7-8 min in some versions because of gradient overheating precaution (TR can't be too short). So, two repetitions (total 60 dwis) should take 8-10 min in Siemens/Philips. This is certainly a long scan, especially for clinical. If time is important, you can do one repetition (total 30 dwis), scarifying the SNR. In this case, you definitely don't want to increase the resolution, which would make the
 SNR worse.

For 3T, 30 dwis can give you a decent SNR with 2.5mm resolution. Or if you stick with a 60 dwis protocol, you can use 2.2 mm resolution. I have seen nice 2.0 mm resolution with the latest head coils. For 2.2 mm, you need to use 112x112 matrix. Please note that larger matrix leads to more distortion. Because 3T has larger distortion than 1.5 T, using a larger matrix in 3T is not a good idea in general. 

Hope it helps.

Susumu

On Thu, Dec 10, 2009 at 9:47 AM, Lucas Lessa <lucas_lessa at yahoo.com.br> wrote:

Good morning,

As I said in aonther post, I'm in the middle of a DTI study that, obsviously, uses DTI studio to process the data.

The other researcher brought (a think from John Hopkins) a model that uses 3 DTI series, just to increase the ratio between sign/noise.
The exams are taking so long and I would like to increase the resolution from 112x110 to 256x256, and maybe, 512x512.

Is this a good idea? Or it's better to continue making the exams with the 3 DTI series?
And wich is the bennefit of doing 3 DTI series (the vendor software only uses one at a time)?
One more time thank you.

Lucas Lessa
São Paulo/Brazil

Veja quais são os assuntos do momento no Yahoo! + Buscados: Top 10 - Celebridades - Música - Esportes

_______________________________________________
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

-----Anexo incorporado-----

_______________________________________________
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

      Veja quais são os assuntos do momento no Yahoo! + Buscados: Top 10 - Celebridades - Música - Esportes

_______________________________________________

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

-----Anexo incorporado-----

_______________________________________________
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

      Veja quais são os assuntos do momento no Yahoo! + Buscados: Top 10 - Celebridades - Música - Esportes

_______________________________________________

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

-----Anexo incorporado-----

_______________________________________________
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

      ____________________________________________________________________________________
Veja quais são os assuntos do momento no Yahoo! +Buscados
http://br.maisbuscados.yahoo.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.mristudio.org/pipermail/mristudio-users/attachments/20091212/2a713feb/attachment-0001.html 