| Author |
 Topic  |
|
|
syam.gadde
BIAC Staff
USA
421 Posts |
 Posted - Dec 21 2004 : 1:24:00 PM
|
Hello all,
A new version of the QA is available, and will run automatically for those scans occuring after 1PM today, and will be used for anyone re-running the QA using bputil. The differences between the old QA and new QA include:
- a new summary section at the top that indicates how many volumes are more than 3 or 4 std. deviations from the mean (individual or across runs) as well as an SFNR number for the middle slice.
- in the last section of the QA page is an SFNR image for the middle slice and a mask used in calculating the mean SFNR.
- an XML file storing the mean intensity and center of mass values for each volume as a list of events in the file qa_events_runXXX_XX.xml. Future versions of analysis tools may use this file to detect outlier points.
If there are any problems with the new version, the old version is still available as fmriqa_generate_old.pl. You may run it manually, as described at the end of the following forum post: http://www.biac.duke.edu/forums/topic.asp?TOPIC_ID=360 |
Edited by - syam.gadde on Dec 21 2004 2:09:17 PM |
|
|
diaz
BIAC Alum
USA
212 Posts |
Posted - Jan 04 2005 : 09:21:25 AM
|
I recently tried to run QA using bputil (option 1) on several Mask.05 subjects but am having trouble. The process runs, but the output file is empty, (0 KB, and nothing is in the file). Any thoughts?
Thanks! |
Michele T. Diaz, Ph.D.
Associate Director
Brain Imaging and Analysis Center |
 |
|
|
syam.gadde
BIAC Staff
USA
421 Posts |
Posted - Jan 04 2005 : 09:38:03 AM
|
Hi Michele,
I can't find any error message in the log files (but maybe they don't get redirected there). Can you run it manually on one of those exams using /usr/local/bin/fmriqa_biac Mask.05 Func 200XXXXX_XXXXX and report any errors it produces? Thanks! |
|
|
|
mb
BIAC Alum
27 Posts |
Posted - Jan 04 2005 : 09:46:46 AM
|
Michele,
Are you maybe out of space? |
|
|
|
diaz
BIAC Alum
USA
212 Posts |
Posted - Jan 04 2005 : 10:22:15 AM
|
There is enough space available on Nernst.
I tried using the line syam provided above and had no problem. I noticed that the output is going to a different place from bputil. The bputil output went to Mask.05\Analysis\subject#\AutoFMRIQA\ and the line I used above went to Mask.05\Data\Func\sub#\QA\.
Even though there was no error, I'll put the display output from the line below.
/usr/local/bin/fmriqa_biac Mask.05 Func 20041008_43810
Output directory /home/diaz/net/nernst/data2/BIAC/Mask.05/Data/Func/20041008_43810/QA already exists! Do you want to overwrite?
(y/n)y
Increasing datasize resource limits to 2GB
fmriqa_volmeasures
run004_01... run004_02... run004_03... run004_04... run004_05... run004_06... run004_07... run004_08... run004_09... run004_10... run004_11... run004_12...
calculating sfnr
run004_01... run004_02... run004_03... run004_04... run004_05... run004_06... run004_07... run004_08... run004_09... run004_10... run004_11... run004_12...
Plotting volume means...
Plotting center of mass (X)...
Plotting center of mass (Y)...
Plotting center of mass (Z)...
calculating mean/stddev/slicevar
run004_01... run004_02... run004_03... run004_04... run004_05... run004_06... run004_07... run004_08... run004_09... run004_10... run004_11... run004_12...
calculating min/max of mean/stddev/sfnr/mask/slicevar
converting bxh to jpg (via ppm)
Done! Output is in /home/diaz/net/nernst/data2/BIAC/Mask.05/Data/Func/20041008_43810/QA/index.html
|
Michele T. Diaz, Ph.D.
Associate Director
Brain Imaging and Analysis Center |
|
|
|
jimmy.dias
BIAC Alum
USA
210 Posts |
Posted - Jan 04 2005 : 10:40:25 AM
|
Hi All,
I've installed an older version of bputil because the older version handles proper directory copy over the network properly. Unfortunately, this version of bputil however does not do the QA properly and as you noticed puts the output in the wrong directory. I'm working on merging the changes for both versions and will post here when done. Until then, please hold off on using bputil for QA. Everything else should be working properly.
Apologies for the inconvenience.
Cheers,
Jimmy |
|
|
|
gunes
BIAC Alum
45 Posts |
Posted - Feb 07 2005 : 12:19:16 PM
|
I am testing "eventstats" with our old data. I need qa_events_runXXX_XX.xml files. I encountered a problem with using new QA for a subject's data. This subject was not scanned with the correct sliced number and Chuck helped us by writing a script that would take care this problem. I ran the QA stat with 35 slices and 34 slices separately. The results:
The output with 34 slices: only run1 and run5
The output with 35 slices: I got all runs without runs'label and qa_events_runXXX_XX.xml files.
the display output:
[yucel@golgi: ~] $ fmriqa_biac ADVTN.01 Func 42791
Increasing datasize resource limits to 2GB
fmriqa_volmeasures
... ... ... ... ... ... ... ... ... ...
calculating sfnr
... ... ... ... ... ... ... ... ... ...
Plotting volume means...
Plotting center of mass (X)...
Plotting center of mass (Y)...
Plotting center of mass (Z)...
calculating mean/stddev/slicevar
... ... ... ... ... ... ... ... ... ...
calculating min/max of mean/stddev/sfnr/mask/slicevar
converting bxh to jpg (via ppm)
Done! Output is in /home/yucel/net/huxley/data/ADVTN.01/Data/Func/42791/QA/index.html
[yucel@golgi: ~] $
Any suggestions?
Thanks.
Gunes |
|
|
|
syam.gadde
BIAC Staff
USA
421 Posts |
Posted - Feb 07 2005 : 1:21:46 PM
|
There was a bug in fmriqa_generate.pl that created incorrect labels. I believe this is now fixed. Can you try fmriqa_biac again?
Also if you have moved your run directories to a lower level directory in order to separate them for some reason, you can specify this lower-level directory when running the QA. Gunes' specific example: fmriqa_biac ADVTN.01 Func 42791/34_slices fmriqa_biac ADVTN.01 Func 42791/35slices should both work if there are runXXX_XX subdirectories in the ADVTN.01/Func/42791/34_slices or ADVTN.01/Func/42791/35slices directories. The QA will then show up in the 34_slices or 35slices directory. |
|
|
|
gunes
BIAC Alum
45 Posts |
Posted - Feb 07 2005 : 1:23:26 PM
|
Thank you very much, Syam.
Gunes |
|
|
| |
Topic |
|
new QA version
