Timo,
The latest version of the wfu_pickatlas includes the ability to
generate statistics from each roi for a list of images (mean, std, T).
This is done using the "generate table" button at the bottom of the
pickatlas gui. This will prompt you for a list of analyze images.
Alternatively, a text file-list of analyze images can be provided (with
a .flist ending).
------------------------------------------------------------------------
----------
Joseph A. Maldjian, MD
Associate Professor, Dept. of Radiology and Biomedical Engineering
Director Advanced Neuroscience Imaging Research Core (ANSIR)
Wake Forest University School of Medicine
Office: 336 716-2815
fax: 336 716-2870
email: [log in to unmask]
website: www.ansir.wfubmc.edu/maldjian.htm
------------------------------------------------------------------------
----------
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]]
On Behalf Of Timo Lukkarinen
Sent: Friday, June 18, 2004 3:08 AM
To: [log in to unmask]
Subject: [SPM] FDR and WFU_pickatlas
Hi,
After a few hours of searching the archives of SPM mailing list, without
finding a reasonable answer to my question, I decided to ask you a few
questions.
I'm using WFU_pickatlas to specify ROIs to my group contrasts. After
selecting the ROI I want to investigate, I apply FDR correction with p <
0,05 and extent treshold of 5. I get an image with activated voxels and
after clicking the volume button, I discover that the list of peak
activations may contain values above pFDR < 0,05. Why so?
I reasoned that this happens, because FDR is applied on the roi in hand,
not for some randomly picked peaks, which are presented after p values
for the volume are requested. If this is the case, every time I see an
activation inside my ROI, it is significant, whatever the values are in
the volume table of p-values, am I correct? In this case, how can I get
a t-value or a z-score for the ROI?
To clarify my question, there's a screencapture attached to this mail.
Thanks,
Timo Lukkarinen
|