Subject: Instructions for how to see possible-RF artifact

Date: Fri, 16 Oct 1998 12:34:12 +0900
To: Torborg, Paul J (MED) Paul.Torborg@amermsx.med.ge.com
From: Michael Zeineh mmz@ucla.edu
Subject: Re: paul's e-mail
Cc: ralphlee@ucla.edu
Mime-Version: 1.0

Note, I've saved this info in "/ppq/zeineh/rftest/instructions"
Instructions for how to see possible-RF artifact (maybe Ralph you can post this to the howto pages)

1. Scan a EPI gradient echo time series on a phantom, at least 1 slice but preferrably at least 6. (I think Basic Function 2, series 7, scan #2)
2. After reconstruction, log into getz (or porkpie on the instascan)
3. cd to a working directory, i.e. /ppq/zeineh/rftest, or a directory allocated to you.
4. locate the location and name of your EPI scan in #1.
eg. /apd4/anmr_scans/image/geservice_50480_00005_00005_00001_001.img
5. making sure you are in the working directory specified in #3, type the following command to start the file conversion procedure epi2bshort -i
/apd4/anmr_scans/image/geservice_50480_00005_00005_00001_001.img
6. type "ls" Look at the new files created, you will use this in # 7, eg. the new files will be named A5.s1.bshort thru A5.s6.bshort if there were six slices
7. then type "mdisp A5.s*.bshort -o run1 -A" run1.bshort will be the output file, which will then be displayed.
8. Next, "stdev -i run1.bshort -o run1.stdev" which will generate the variance map and put it in run1.stdev.bfloat
9. "xds run1.stdev.bfloat &" will display the variance map. For step 10-11 make sure the mouse pointer is in this new window.
10. To see if the RF artifact is present, go to the 4th image in the variance map. You can do this by hitting "+" 3 times while in the window, and the # in the bottom left corner will change from "00" to "03".
11. The whole window should be bright white because the scaling is wrong. Use and hold down the left mouse button to highlight a big (yellow on a color monitor) box that will cover at least 1/4 of the window. Then, hit "w" to rescale, and "shift-c" to remove the box so you can see the image.
12. The artifact looks like a bright horizontal line (there may be 2) within a slice. It is usually only a couple of pixels wide. If you see these lines, then the artifact is there. Move the mouse to the x,y location of the brightest point on the line. At the lower left, you will be given the x,y coordinates, and the variance value. Make sure you are on the point of highest variance (I've gotten around 2000 on a phantom) and record this x,y location.
13. If you now move the mouse to the window with the time-series up (which appeared after completion of step 7), type "g". A window will appear which will display the time series of the current mouse location. If you move the mouse to the same x,y coordinates specified in step 12, you will see a much higher variance in the timecourse. I've seen signal changes of 10-20%. If you click the mouse button, it will highlight the point in yellow. Then hit "s", and it will spit out the numbers for the timeseries of that pixel on the screen, from which you can easily approximate the % change.

Michael Zeineh
From original by: MMZ on date10/16/98
Edited&posted to how2 page 10/16 , R. Lee