This gives a basic idea of the process used to test equipment (displays, etc) for noise. If a device is not shielded properly it may emit deterministic noise capable of corrupting images and generating artifacts.

• Set up device (e.g. monitor inside shielded box).  The device must be powered up.  In the case of the monitor, a stimulus doesn'tneed to be running - the monitor itself must be on and receiving input from the computer.
• Set up post-head coil with round phantom. Use the phantom marked silicon oil - this one has a T2 that is most similar to a brain.
• Enter patient information as usual
• Enter patient position and specify pulse sequence parameters as follows:
  • patient position: supine; patient entry: head first; coil: posthead (or whatever)
  • plane: axial; mode: 2D; pulse seq: gradient echo (GRE)
  • TE: Min Full; TR: 50 at 3T; flip angle: 40; bandwidth: 32 (actually 31.25 but you can type 32)
    Note: Gary told Amy to use TR=100 at 1.5T but we used 50 at 3T. Ask Gary if this is a real difference or if it just doesn't matter much.
  • freq: 256; phase : 128; NEX: 1; Phase FOV: 1
  • freq dir: R/L; auto center freq: water
  • autoshim button selected
  • FOV: 20; slice thickness: 5; spacing: 0
  • start: 0; end: 0; # slices: 1
    
• Scan the phantom: select save series and then scan .
• Disable transmitter switch in machine room. Flipping it down turns it off. Write yourself a note to remind you to re-enable transmitter when you are through testing.
• With the transmitted disabled, scan the phantom again.
• Unplug/disable device being tested for noise. Scan phantom again.
• Repeat these last two steps (scan with device on and off)
• View the images. Click on the top right of the 6 icons and then click on browser (on the left). Select the series corresponding to the scans that you just did. Then select viewer (on the far right).
• Select format (bottom left) that allows you to view multiple images at once. Adjust brightness & contrast using the middle mouse button as usual. For the scan collected on the phantom with the transmitter enabled, you should see an image of the phantom. For all the others you should see noise.
• Examine images for noise. These should look like 'zippers', vertical bands of higher contrast noise. Also look for systematic artifacts, e.g., a bright spot might correspond to a periodic signal like the vertical blank in an video device.
• Select measure (bottom left). Adjust the brightness/contrast so that you can clearly see the phantom from the first scan (acquired with the transmitter turned on). Choose the circle/oval tool. Position an oval on the phantom near the bottom close to the coil.
• Copy the oval to the other images. Cntrl-c to "copy". Click left on the next image. Cntrl-v to "paste".
• At the bottom of each image in the center, you should see the mean displayed.  If there is no deterministic noise, the mean for all scans should be roughly identical. In a recent noise test at 1.5T, this number was ~12. In another recent test at 3T, this number was ~33.
• Re-enable transmitter by flipping switch up.