After running do_mritotal, do_mritopet and do_pettopet, the system
should alert you (the one who submitted the job (see big_command.txt)
that a series of "verification" images have been generated and stored
in the directory 'verify'.

You should view these images to check that the data have been properly
registered (mri from native to MNI-Tal space in the case of
do_mritotal; mri from its native space to native pet space in the case
of do_mritopet; and from all pet scans' (except the first scan) native
space to native space of the first pet scan in the case of
do_pettopet)

To look at these images, open a window on the local machine
(e.g. grumio, quince, cordelia, etc.) and go to the verify directory
for your study.  In the case of nling13, for the DAT subjects, this
would be:

$> cd /data/aging/nling/mystudy5/verify

The system produces images in the 'rgb' file format.  You can look at
them using the program display (*not* to be confused with the BIC-MNI
product Display that can only view 'mnc' files).

The command structure is this:

$> display filename.rgb 

Once the image is displayed on your screen, you click once with the
left mouse button (while the cursor is over the image) to activate
the menu.  If you click and hold the right mouse button, you have
access to the "files" menu.


In the case of do_mritotal verify images, you are looking for
something like this:

dat_mrital_1.rgb
dat_mrital_2.rgb
dat_mrital_3.rgb

Compare the subject's volume to the average volume (displayed on the
extreme right-hand side of the panel).  You are looking for the
following characteristics:

Subject's slice is at same level/cut as average
Same orientation
Brain generally fits inside reference box.

In the case of do_mritopet verify images, you are looking for
something like this:

boysenberry_robert_cor.rgb
boysenberry_robert_sag.rgb
boysenberry_robert_tra.rgb

You are looking to see that the pet activation falls inside the brain
and follows the shape of the brain.  You shouldn't see activation in
the ventricles, skull, etc.

In the case of do_pettopet verify images, you are looking for
something like this:

boysenberry_robert_petpet.rgb

This image is a large mosaic of columns of representative slices from
each pet scan: a) two motion-correction attempts made for each run
that is matched to the target scan and and one column representing the
target scan (usually the first run).  In the first of the two columns
for each scan you may see a lot of brain characteristics (red, green
and blue defining the shape of the brain, skull, etc.) - that's okay
as long as the second column is rather indistinct - ie. green and
fuzzy.

Note that the first three rows of the mosaic are transverse (or axial)
slices, the second two rows are sagittal slices, but they're upside
down and the last two rows are coronal slices, also upside down.
Don't ask me why this is so - ask the folks who wrote the programs.