- How to use PET Session Profile

Before you start to analyze PET scan data, you need to input data and
parameters. You also would like to save the inputted information into
a file for later retrieval.

This SESSION PROFILE window provides you with an efficient way to input
and save the data and parameters, as well as to create a corresponding
datamat.



- Session Description

It is nice to have a Session Description to describe what this experiment 
will include.



- Working Directory

To use this window, you MUST provide a Working Directory name (or you
can also call it 'PLS Data Directory'). In this directory, you will 
save the PET session file, PET datamat, and PET result file.

Like the way you prepare the data for command_line PLS analysis for PET
scan, you already have subjects' data. They should be organized in a way
such as each subject has their own directory, and inside the directory,
there are several condition files of this subject. It will be convenient
for you if you put subjects' directories immediately under your working 
directory.



- Behavior Data File

If you prepare to run Behavior PLS on those PET scan data, you MUST 
also provide the behavior data, which is not required for running Task
PLS. The number of rows of your behavior data must match the Behavior 
PLS test you are running. For example, if you run 4 conditions with 9
subjects, then, the number of rows in behavior data will be 4*9 = 36

The behavior data is prepared in ASCII format.

The row order of behavior data should be in the form of [cond1/subj1; 
cond1/subj2; ... ; cond2/subj1; cond2/subj2; ... ;]. i.e. in the order
of each subject, then in the order of each condition.

Once you click 'Edit Behavior Data' button, the 'Edit Behavior Data'
window will open. You can click the 'Browse' button to get the behavior
data text file you prepared, or you can also directly enter the number
into the text input area.

Depending on how many column in the 'Behavior Data', the 'Behavior Name'
will automatically be filled as 'behav1 behav2 ...'. You can change the
behavior name by edit them. You can use any alphanumeric or '_' charactor
for behavior name. Don't use special charactors, also don't use space,
which acts like a delimiter.



- Datamat Prefix

It is more convenient to tell the data from its file name, before open
it. We use '_PETsession.mat' to represent PET session profile; use
'_PETdatamat.mat' to represent PET datamat; use '_PETresult.mat' to 
represent PET result data. So, enter a meaningful datamat prefix here
will make yourself easy later.



- Input Conditions

Click 'Input Conditions' to enter the condition names under which you 
get those experiment data. The number of conditions you entered will be 
displayed in 'Number of Conditions:' field. 



- Select Subjects

Clicking on 'Select Subjects' will open the Subject Directory window.
From here, you can add all the subjects into the session. 

Consistent filenames: Filenames are consistent if subject initials are 
the same length, and if the same characters at the beginning 
of each filename are allocated as the subject identifier, with the 
remaining characters identifying the condition (e.g., 'SubjInit_CondName').
If you have consistent filenames, enter the length of the subject initials
in the 'Number of Characters for Subject Initial' box. '-1' will disable 
this feature.

Click 'Add ...' in the Subject Directory window. This will lead you to 
Subject Detail window to add subjects. The left listbox should list the 
subject directories, which will also be used as subject name later. (If 
they are not subject directories, browse to the correct location.)

Then, click on a directory to see if the subject files match the inputted
conditions (right window). Note: the gui initially assigns data files to 
conditions alphabetically. If the assignment is not correct, you can click
subject files to swap them.  If you have consistent filenames, you should
only need to reassign files to conditions for the first subject.

If you click DONE, the highlighted subject will be selected, and you will 
go back to the 'Subject Directory' window. However, if you hold Ctrl key 
and select additional subject directories, all (and only) the highlighted
ones will be added. Previously selected subjects will be ignored.

By default, the 'Consistent subject file name convention' checkbox is 
disabled, which will not force you to use the naming convention mentioned
above. However, you can change the value in previous window to enable 
this feature.

Once subjects are added, they can be edited/checked by clicking 'Edit ...'
in the Subject Directory window. This will lead you to 'Subject Detail' 
window, with the edited subject highlighted. You can choose another subject
from the Directory listbox, or change the condition order by clicking the 
subject file name.

After subjects are selected, you can remove any unwanted conditions, across
all subjects.  However, if you would like to add more conditions, you will
need to edit all the selected subjects, and add the new subject file(s) for
the new condition(s).



- Create Datamat

In PET, one session file can have more datamat files. Once datamat is 
created, the analysis and result will only deal with datamat file, not
session file.

The datamat is composed of rows of scanned volumes that stacked together.
It is a 2D data matrix. The column of the matrix includes all the slices
for this volume, while the row of the matrix includes all the subjects
under different conditions. To save space, only the brain voxels will be 
stored in the matrix. The non-brain voxels are defined as the voxels 
that are above a quarter of the maximum value in the datamat.

Before creating datamat, you muse have the session profile saved first.
Click the 'Create Datamat ...' will lead you to Create Datamat window.

In the Brain Region frame, you can either use a predefined brain region 
image, or enter a value in the threshold field to define the brain 
region. Any voxels with values below 1/threshold of the max. value in 
the volumes will be removed. The 'Max. standard deviation allowed:' is 
the maximum absolution standard deviation allowed in the brain voxels.
You can also select whether or not to normalize the datamat with its 
volume mean.