CNI Scanning Instructions


Jump to: navigation, search


[edit] Scanning fMRI at the CNI

The following assumes that you are scanning fairly standard back-of-the-brain EPI and want to collect field maps, an inplane T1, T2 and multi-shot EPI (protocol 1.154 is an example of that).

[edit] Preliminaries

  • Get the subject to fill in a consent and screening form.
  • Connect the display computer to the LCD display and to the USB (the one labeled "responses" will get you both the TTL pulses and the behavioral responses through the button boxes. (Actually, you need to also check that the BNC connector in the back of the fORP control box is connected to the wire that delivers the TTL pulses. Often it is not connected because some people don't want those pulses.) See the cni wiki for more details on configuring that). The TTL pulse is a character '5' when using the 'HID NAR 12345', which is recommended. In other modes, the pulses and buttons show up as different characters (and other input-related changes happen). Set the silver forp box to the 'HID NAR 12345' mode, which can be adjusted using the dial in the front of the box:

By temporarily changing the mode, this can be used to avoid getting pulses from the manual pre-scan (see below).

  • On the scanner computer: add the participant, entering the participant's weight and using a Patient ID in the following way: 'wandell/name_of_study'. This will allow you to easily find the data on the CNI NIMS web interface. Choose the protocol you want to use.

[edit] Inside the scanner room

  1. Make sure the LCD brightness is maximal: to control the LCD: click the arrows at the top (what is the top now, when it's up-side-down) of the monitor. If there is no change in brightness press the mode button and keep trying. Unfortunately, you will get no visual feedback when you change modes.
  2. Position the subject in the coil and pad them in. Don't forget ear-plugs.
  3. Optional: Attach the physiological monitoring devices to the subject. If you do that you will want to make sure that your fMRI sequences are set to save these data ("Save Physio Data").
  4. Remove the standard mirror and attach the flex-stand mirror (which may be attached to the 16-channel coil) to the coil. Tape it to the coil, so that it doesn't vibrate during the scan. Be careful not to touch the surface of the mirror. The mirror is front-coated and you will leave finger prints, so it will be easy to track you down.
  5. Advance the bed into the bore. Use the laser to position the center of the head under the cross and landmark. Advance participant all the way in.

[edit] In the control room

1. Acquire a three-slice localizer (1 slice in each of three dimensions). This is done by highlighting the localizer in the task menu (top left) => "Save Rx" => "Scan".

2. Copy and paste the localizer back in and increase the number of slices in each dimension to 10 and the slice gap to 5 mm, to get better coverage for the placement of the slices you will use in your data acquisition. Click "Save Rx" => "Scan".

3. Acquire an ASSET Calibration (this is the calibration for the parallel imaging). Note that if there is any motion during this calibration, this will affect everything that follows, so try particularly hard not to move during this time (Actually, this applies throughout the entire session, including fieldmaps, etc. So don't move.) Double-click on this procedure in the task menu, then click in slice window to define the location of the (big) slab. Make sure the slab covers the entire head, leaving some space on all sides. Then hit "Save Rx" => "Scan".

4. Set up the slices for the high-res inplane T1). Double-click on this task and click once within one of the localizer images (probably you want to use the sagittal) to place the slices. Then, use the mouse to drag and rotate your slice perscription. Make sure that:

  • The '1' in the sagittal view indicating the side on which the first slice is located in the back of the head, meaning that slices are acquired in the posterior-to-anterior direction (note that the slices are actually acquired in interleaved (1,3,5,...) temporal order, but when reconstructed, the slices will be spit out in the usual numerical order).
  • The slices are symmetrically placed relative to the head
  • You are covering the parts of the brain you want to cover and you are not cutting out the back of the brain, or any other important bits. Use the scroll bar to go through all slices/orientations.

5. When you are happy with your prescription, hit "Save Rx" and then the down arrow next to "Scan" => "Download". But don't scan the sequence!

6. You can now copy the slices into any additional scan you prepare, by setting up the additional scan that you want to perform, and then hitting the eraser icon and the "Rx => Rx" icon in the "Graphic Rx toolbar". Note that you want to copy from the "3d" protocols (chosen through a drop-down that comes up when you click the "Rx=>Rx" icon). Always choose the high-res T1 to copy from.

7. Now would be a good time to grab a screen-shot (for future reference). Hit the print screen button and save a png. This should be somewhere on the scanner computer and also on the computer next to it, for you to grab later.

8. Now that we have setup and downloaded the slices that we care about, double click on "HO shim". Dismiss the text box that pops up and press SCAN. Set up the red ellipse to cover the part of the brain you are interested in. Be liberal, don't risk not getting all of the brain. Hit "Calculate shim". If you have a value under 8-ish, you are good to go (hit "done"). Otherwise, perhaps reposition your subject or re-define the ellipse. Repeat until the shim value is acceptable. Following this, you will want to always verify that the shim is set to "off" in all of your sequences (not "auto"!), so that you don't get additional automatic shimming occurring.

9. Now acquire all of your inplanes (T2, EPI multishot, and T1). For example: set up and scan a multi-shot inplane EPI, which can be used to assess distortion correction in your subsequent single-shot EPI scans. Copy the slices over from the T1 high-res (see above). There should be only 1 volume to be acquired (in the "multi-phase" tab, set the "phases per location" to 1 (you only want 1 TR of this)). In the "Details" tab, make sure the shim is set to "off". Click "Save Rx" => downarrow => Research => "Download" => downarrow => Research => "Display CVs". Now you need to change some of these CV vars, specifically, set rhfermr to <matrixsize> (e.g. "64", when in doubt set it to a number larger than your maximal matrix-size dimension) and rhfermw to 1. This in effect turns off the Fermi spatial smoothing that is usually applied. Once you do that, you don't need to do it again - these changes propagate through copy-pasting.

10. Now set up a 2-volume test of your actual EPI data. Hit SCAN as usual. This will trigger all the pre-scan stuff. We want to allow it to happen here once and for all. After this point, we will specifically bypass all the pre-scan stuff.

11. Set up the spiral field-map scan. Copy over slices from your inplane T1. Hit the down-arrow icon and "download". Then - hit the down arrow =>"research" and "CVs". Set the CV var 'rf1_type' = 3. Then scan the sequence by doing the Manual Prescan and then Prep scan. Note that the field of view of these is intentionally rather large, to avoid wrap-around artifacts. Set the TE to 9.091 and NEX. If need be (if your subject is in pain...), you can make the field-map shorter, by changing the NEX to 1 (instead of the 2 value that is there by default).

12. Set up the EPI scan. This is essentially the same as the multi-shot inplane EPI: Copy over the slices from your inplane T1. In the "multi-phase" tab, set the "number of phases" to the number of TRs you want to acquire. Note that you can control the triggering behavior (laptop triggers scanner or the other way around) and whether or not physio gets saved in one of the tabs. Make sure that the shim is set to "off". Once you have copied the Rx from the high-res T1, click "Save Rx" => downarrow => "Research" => "Download" => downarrow => "Research" => "display CVs". You will need to change some of these CV vars: rhfermr to <matrixsize> (e.g. "64"). and rhfermw to 1. Then, trigger the scan by doing the usual Manual Prescan and then Prep scan. You want to alternate field-map acquisitions with about 10 minutes of EPI acquisitions, finishing off with one last field-map acquisition. The procedure for each one of these is the same.

13. For the EPI scans: keep copying and pasting a previous successful sequence you had into the task menu. For each one double-click on it, or click "Setup". Then "Save Rx" => downarrow => "Research" => "Download". Then downarrow => "Manual prescan" (a window will pop-up and do the pre-scan procedure) => "Done". Down arrow => "prepscan".

  • Note that the Manual prescan does some measurements that results in TTL pulses being sent out. Make sure this doesn't confuse your stimulus computer. You can fake it out by temporarily changing the mode of the fORP box, or unplugging the cable connecting your stimulus computer to the scanner.

14. After Prep scan, you should hit the "START" button on the extra control key-board above the computer key-board, which should be lit green. Upon pressing that key, this should immediately send a TTL pulse to the display computer and start the acquisition.

15. For the field-map scans: ditto. Just keep copying and pasting, hitting "Save Rx" => downarrow => "Research" => "Download" => Manual Prescan, Prep Scan. and "START"

[edit] When you are done

  • Use a measuring tape to measure the distance of the participants eyes from the screen (eye-to-mirror + mirror-to-screen). This will allow you to calculate the actual FOV occupied by your stimulus.
  • Remove the flex-stand mirror and re-attach the standard mirror to the coil. Turn off the LCD monitor.

Click "end" at the top right of the task menu, to close out the session. You data should be on NIMS by the time you get back to the lab...

Personal tools