search

firasm / CEST

Analysis of studies
2 stars 0 forks source link

Mouse Brain Experiment #26

Closed firasm closed 9 years ago

firasm commented 9 years ago

Hey Andrew,

Just checking in to solidify plans for tomorrow’s experiment.

Unfortunately I have a lecture from 11-12, so I can be there right after 12 PM - would it be possible you to get started without me so we have the maximum amount of time with you around ?

I can also be there from 10:15 - 10:50 to help set up if you need it. Scott will be there and can help if you if you need it.

Here’s what I was thinking we should do:

Firas

andrewcyung commented 9 years ago

Sure, that’s fine with me. However, can you send me parameters I should be using (or set up a protocol) with the new turboflash cest sequence? I have never used it.

-andrew

From: firasm [mailto:notifications@github.com] Sent: Thursday, March 26, 2015 1:07 PM To: firasm/analysis Subject: [analysis] Mouse Brain Experiment (#26)

Hey Andrew,

Just checking in to solidify plans for tomorrow’s experiment.

Unfortunately I have a lecture from 11-12, so I can be there right after 12 PM - would it be possible you to get started without me so we have the maximum amount of time with you around ?

I can also be there from 10:15 - 10:50 to help set up if you need it. Scott will be there and can help if you if you need it.

Here’s what I was thinking we should do:

Firas

— Reply to this email directly or view it on GitHubhttps://github.com/firasm/analysis/issues/26.

firasm commented 9 years ago

I will come in tomorrow morning before 11 and set up a protocol with all the parameters set!

Firas

On Thu, Mar 26, 2015 at 2:04 PM, andrewcyung notifications@github.com wrote:

Sure, that’s fine with me. However, can you send me parameters I should be using (or set up a protocol) with the new turboflash cest sequence? I have never used it. -andrew From: firasm [mailto:notifications@github.com] Sent: Thursday, March 26, 2015 1:07 PM To: firasm/analysis Subject: [analysis] Mouse Brain Experiment (#26) Hey Andrew, Just checking in to solidify plans for tomorrow’s experiment. Unfortunately I have a lecture from 11-12, so I can be there right after 12 PM - would it be possible you to get started without me so we have the maximum amount of time with you around ? I can also be there from 10:15 - 10:50 to help set up if you need it. Scott will be there and can help if you if you need it. Here’s what I was thinking we should do:

  • Localize and prescribe a slice (If we could find another place in the body that doesn't move, I would be willing to try that as well)
  • CEST with -100ppm, 0 ppm, 100ppm
  • Shim
  • CEST with -100ppm, 0 ppm, 100ppm
  • T1 map
  • T2 map
  • Coarse CEST spectrum acquisition (also doubles as an SNR check)

  • full CEST spectrum Firas — Reply to this email directly or view it on GitHubhttps://github.com/firasm/analysis/issues/26.

    Reply to this email directly or view it on GitHub: https://github.com/firasm/analysis/issues/26#issuecomment-86716620

firasm commented 9 years ago

The iPython notebook called "2015-03-27 - Mouse Brain experiment" contains further details and plots that could be helpful later in deconstructing this imaging session, but below is a quick summary of what we did and what went wrong.

First the T1 map (LL):

unknown

On the bright side (ha!), we got really nice shimming in the animal (much better than in phantoms)

unknown-2

Scan 37 - Full Z Spectrum, 64x64, 10,000 ms intersatfreq delay, AccFactor 8, Pulse Length 500

unknown-1

Scan 39 - Full Z Spectrum, 64x64, 10,000 ms intersatfreq delay, AccFactor 4, Pulse Length 500

unknown-5

Scan 40 - Full Z Spectrum, 64x64, 10,000 ms intersatfreq delay, AccFactor 4, Pulse Length 1000

unknown-6

I guess we kind of see a but of a bump at ~ 3 ppm.

Next Steps: Try to average the signal by correcting B0 shifts and then see if we can improve the CEST SNR (different from image SNR). Work will continue on this in issue #29.

DrSAR commented 9 years ago

yeah - you know, looking at it I find that SNR not too bad at all. The pixel averaging is something to try but I would be surprised if suddenly peaks were to rise from the depth.

firasm commented 9 years ago

Before someone jumps and says that Scan 40 looks much smoother, that's more likely a result of a larger step size (51 total offset freqs in scan 40 vs. 92 in scan 39). I guess the "same as above" is not quite accurate.

firasm commented 9 years ago

The SNR suggestion was Piotr's, but I'd tend to agree with you - there doesn't seem to be anything there for us to see.

DrSAR commented 9 years ago

needs to be tried anyway. It's quite reasonable. It also is a good way to confirm the B0 map.

andrewcyung commented 9 years ago

I guess I was too eager to take advantage of the turboflash acceleration, so i chose voxel sizes that were 2-3 times smaller. I should have gone back to the original resolution and worked up from there...

I do agree with you that SNR is not a problem, and yet when I squint at the data, I can see Piotr's point.

What was the TR of the FLASH readout? Was it quite small? The original CEST FLASH had long TR (500 ms) mainly because the saturation pulse was 490 ms and repeated for every TR, but it could have had the additional effect of decreased T1 weighting of the signal readout, and therefore higher image SNR.

I should really go back and measure the image SNR on the previous brain data that showed some success.

I was also thinking we need to get a better handle on how many saturation pulses the tissue has experienced by the time k-space centre is traversed, and to study the time to steady state using Bloch simulations. Firas tells me those simulations are just starting to be worked on? Need a hand with this?

-andrew

On 3/28/2015 9:54 PM, DrSAR wrote:

needs to be tried anyway. It's quite reasonable. It also is a good way to confirm the B0 map.

— Reply to this email directly or view it on GitHub https://github.com/firasm/CEST/issues/26#issuecomment-87358633.

DrSAR commented 9 years ago

"those simulations are just starting to be worked on" That;s an interesting way of putting it, Firas. There is, I find, a very nice python notebook that you should peruse. It shouldn't need too much work to help answer all those questions. Andrew, can I get you an account on pfeifer so you can play with it? I think I should really set up a public notebook server so we can all have python fun together.

DrSAR commented 9 years ago

I have uploaded that same worksheet from pfeifer to a public notebook server. I can make anyone collaborator but you would have to first register with cloud.sagemath

In the meantime, there is a read-only (do not mess with the code) version viewable at https://cloud.sagemath.com/projects/f9b8ad28-ab2f-4165-bb25-c80d57aaf6cc/files/Bloch-simulation-with-exchange.html