Back to PLS Helpstructural MRI session set up
Posted on 03/16/10 13:26:23
Number of posts: 4
I just familiarized myself with the PLS tool for structural MRI analysis, but got stuck when trying to set up the sMRI session profile. Using the “Input Conditions” function, I indicated the conditions using a format like *con1.nii and *con2.nii. The number of conditions of 2 is correctly indicated in the Session Information window. When moving to the next step of selecting the subjects (there were about 4 subjects per condition in my test set), the dialog window “Select Subjects” displays only those subjects of condition 1 but not condition 2. I tried different ways of coding the file names but that would not resolve the problem.
I would appreciate any comments.
Thank you,
Michael
Replies:
Untitled Post
I'm Online
Posted on 03/16/10 13:46:27
Number of posts: 291
When you select the subjects, the dialog window
“Select Subjects” should displays neither condition 1 nor
condition 2.The condition part should be taken away.
For example, assume you have:
s1_con1.nii for subj1 cond1
s1_con2.nii for subj1 cond2
s2_con1.nii for subj2 cond1
s2_con2.nii for subj2 cond2
... ...
The dialog box should looks like s1_*.*, s2_*.*.
If not, please check the file names, and make it consistent (like the example above).
If problem persistent, please post all your file names here, and I will try them for you.
Jimmy
For example, assume you have:
s1_con1.nii for subj1 cond1
s1_con2.nii for subj1 cond2
s2_con1.nii for subj2 cond1
s2_con2.nii for subj2 cond2
... ...
The dialog box should looks like s1_*.*, s2_*.*.
If not, please check the file names, and make it consistent (like the example above).
If problem persistent, please post all your file names here, and I will try them for you.
Jimmy
Untitled Post
Posted on 03/16/10 14:09:29
Number of posts: 4
Dear Jimmy,
thank you for your answer. For the current MRI data set, I have a between-subject group design, so condition designates groups,with each subject belonging only to one of two groups (e.g. Alzheimer dementia vs healhty controls).
The data names are coded
s1_ad.nii for subject 1 belonging to the AD group
s2_ad.nii for subject 2 belonging to the AD group
s3_hc.nii for subject 3 belonging to the HC group
s4_hc.nii for subject 4 belonging to the HC group
I coded the conditions,i.e. groups, of AD and HC as *ad.nii and *hc.nii respectivly. The subject selection dialogue box will show s1*.* and s2*.* but not subject 3 and 4. Does the input window expect a within-subject repeated measures design only? Is there a way to circumvent this?
Thank you,
Michael
Untitled Post
I'm Online
Posted on 03/16/10 14:20:03
Number of posts: 291
Here're the problems:
Structural PLS is designed to analyze different part of brain structural area (we call it condition here), rather than the regular PLS. i.e. You have to divide the brain into several structural parts. e.g. Gray Matter, White Matter, CSF, etc. for each subject. So you should have files like: s1_gm.nii, s1_wm.nii, s2_gm.nii, s2_wm.nii, etc.; It will never like: s1_gm.nii, s2_wm.nii, s3_gm.nii, s4_wm.nii.
In addition, Group and Task (or condition, or brain structural here) are 2 completely different concept. You should not consider "Group" concept until you create all the subject datamats and start to run PLS analysis.
Jimmy
Structural PLS is designed to analyze different part of brain structural area (we call it condition here), rather than the regular PLS. i.e. You have to divide the brain into several structural parts. e.g. Gray Matter, White Matter, CSF, etc. for each subject. So you should have files like: s1_gm.nii, s1_wm.nii, s2_gm.nii, s2_wm.nii, etc.; It will never like: s1_gm.nii, s2_wm.nii, s3_gm.nii, s4_wm.nii.
In addition, Group and Task (or condition, or brain structural here) are 2 completely different concept. You should not consider "Group" concept until you create all the subject datamats and start to run PLS analysis.
Jimmy
Login to reply to this topic.
