TARQUIN example confirmation and Gannet comparison

[samuraimichelle.he]

Hi Dear TARQUIN members,

I just wondered whether it's possible to ask two questions about 1) TARQUIN manual example 2) Gannet result comparison. If will be really appreciated if any suggestion could be given. TARQUIN version 4.3.2

1. For the manual example, I didn't change any parameter as a default and followed the guideline carefully, however the result I've got for the philips SPAR/SDAT as water suppressed scan and unsupressed scan pair, the result(insert below this paragraph) is way off compared to the result given in the manual... I tried each ref. signal in Data.parameter and preprocessing, but it seems doesn't work... I've tested all the troubleshooting method but it doesn't work either. My guess is this result should be correct w/o changing any parameter as it's an example... but it just totally off, any suggestion will be really appreciated.

2. For the comparison between Gannet and TARQUIN:

For the MEGA-PRESS data, I confirmed it's correct for Gannet result, and the correlation between TARQUIN's GABA'S amplitude and Gannet's GABA area is about 0.3. Now I just wondered is there any way to direct confirm the GABA/Cr ratio from TARQUIN to Gannet's result? I found a previous topic of this, but I didn't quite understand sorry... and on the manual it's the "Dynamic averaging" scheme I should alter to get this ratio... but it seems doesn't work again...

Thank you so much!

Best,

Xuanzi

[Martin Wilson]

Dear Xuanzi,

1) I don't think the result is as far off as you think, just a scaling issue with the y-axis. If you go to "Plot" > "Show raw data" you'll get something comparable. 

2) To get GABA/Cr you need to perform two fits. One where you get GABA from the edited data, and one where you get TCr from the non-edited data. Then you divide GABA from fit 1 by Cr from fit 2. To get the non-edited data for Cr fitting you need to select the option to average just the odd scans or even scans (depends on your protocol).

Hope that makes sense!

Martin

--
You received this message because you are subscribed to the Google Groups "TARQUIN users' group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tarquin_users_g...@googlegroups.com.
To post to this group, send email to tarquin_u...@googlegroups.com.
Visit this group at http://groups.google.com/group/tarquin_users_group.
For more options, visit https://groups.google.com/d/optout.

[Xuanzi He]

Hi Martin,

Thank you so much for this quick reply and it was quite helpful! For the example I've got identical result as it showed on manual via ur guideline, thank you.

I'm sure you must be very busy, but I just wondered whether I could ask some following questions... thanks a lot!

Basically, I got 3 pair of Philips SPAR/SDAT data for a signal subject(Monkey) as following:

 

I treated the first one (GABA excitation) as MEGA-PRESS data, while the later two as water-supress/water-unsupress pair for PRESS analysis.

1) For MEGA-PRESS data, I followed the tutorial and ur sugguestion(really appreciated that). The protocol I found via using "no average scan" and it showed as following:

As the unedited data are odd here, I used "average odd only" for caculating Cr. value in MEGA-PRESS mode while leaving other settings to be default. This one has Cr. value.

While when I altered all the parameters following the tutorial's MEGA-PRESS part, the Cr value dissapeared:

The normal GABA mega-press(fit2) is as following:

Now I just wondered whether I should use the Cr value from default setting since it showed up while the expected Cr value is not in the parameter-altered version...but another concern is whether the GABA value from fit2(which altered parameters) is comarable to the Cr value from fit1(default value)...

2) For the ratio, I just wondered whether amplitude ratio is comparable to area ratio...

3) I just wondered whether the sum of GABA_A and GABA_B amplitude  in TARQUIN is comparable to the GABA area in Gannet...

4) sorry this is my last question...

For water supressed and water unsupressed SPAR/SDAT(the later two set of file), I used PRESS default for processing, but the result seems quite strange to me as it seems NAA didn't show a large amplitude compared to other variables like GABA... I just wondered whether this normal... sorry I'm trying to figure this out but just have no clue....

Thank you so  much for your time and any suggestions or reply!

Best,

Xuanzi

[Martin Wilson]

Hi Xuanzi,

1) Yes you should be able to calculate ratio's between fit1 and fit2.

2) Yes amplitude and area are directly proportional (unless there is de-phasing where it get a bit less intuitive).

3) Yes use the sum of GABA_A and GABA_B as your final measure. It should correlate with GANNET however there may be a consistent scale factor difference.

4) I'm a bit confused about this also, if you can send me the data I'll take a closer look.

You should also be aware that your MEGA-PRESS data has a very large lipid component and this may bias your results. The cause is most likely poor voxel placement (over scalp) or perhaps monkey-motion :)

Martin

[Xuanzi He]

Hi Martin,

Thank you so much for the quick and very helpful reply! I've attached the files I showed on the previous post. I hope besides the last 4th question, I could clarify some confusions...

For the part of getting Cr. value, I found if I set the "internal basis set" to be "1H MEGA-PRESS GABA" then the result just don't have any result of Cr, that makes sense because Cr should come from unedited(water unsupressed) sequence. Then I just confused that (1) should I leave this variable to "1H brain"...?

(2) "pulse sequence" setting in "data parameter" should be "PRESS" or 'MEGA-PRESS'...? this odd scan sequences come from MEGA-PRESS sequence so my guess is it should be set to M-P...?

(3) Also any other parameter setting for getting Cr value should I pay attention...

(4) The protocol I got from MEGA-PRESS is:

by using NAA as map numerator (the default setting). I just wanna clarify the meaning of color here. I think blue color means "ON" as an editing pulse is applied to GABA spins at 1.9 ppm in order to selectively refocus the evolution of J-coupling to the GABA spins at 3 ppm, while the various color grids means "OFF", the inversion pulse is applied elsewhere so that the J-coupling evolves freely throughout the echo time.... Please correct me if I got this wrong....

Sorry for asking tons of questions, but I really appreciate your help!

Thanks!

Best,

Xuanzi

[Martin Wilson]

Hi,

1) I think you may be a bit confused here, Cr should come from unedited water *suppressed* (ie the odd or even scans).

2) I don't think MEGA-PRESS does anything different to PRESS, this is a bit of a hang-up from a previous version.

3) As long as you do the same for the edited data and non-edited data (apart from basis set) you should be ok.

4) That sounds correct to me.

Martin