Great, thanks Michael. That makes much more sense!
------------------------------
Date: Thu, 17 Jan 2013 14:01:34 -0600
From: Michael Harms <[log in to unmask]>
Subject: Re: averaging functional datasets
Hi Jodie,
None of those 3 :)
The standard way of doing this is to run FEAT individually for each BOLD
run, and then combine them in a "second level" analysis using a Higher
Level, Fixed Effects analysis in FEAT. See the FEAT documentation.
cheers,
-MH
--
Michael Harms, Ph.D.
-----------------------------------------------------------
Conte Center for the Neuroscience of Mental Disorders
Washington University School of Medicine
Department of Psychiatry, Box 8134
660 South Euclid Ave. Tel: 314-747-6173
St. Louis, MO 63110 Email: [log in to unmask]
From: Jodie Davies-Thompson <[log in to unmask]>
Reply-To: FSL - FMRIB's Software Library <[log in to unmask]>
Date: Thursday, January 17, 2013 1:51 PM
To: FSL - FMRIB's Software Library <[log in to unmask]>
Subject: [FSL] averaging functional datasets
Dear FSL-experts,
I have two identical functional data sets of an individual subject
(collected one directly after the other) that I would like to average to
increase SNR. However, I am unsure on the best point in the analysis to
average them. I can see three possibilities:
1. Average the raw data files using fslmaths. However, it doesn¹t seem
to make much sense to average two datasets that have not been motion
corrected especially if the subject moved in between scans.
2. Run FEAT individually for both, average the two filteredfunc
outputs, and then run FEAT again using this single filteredfunc file.
Presumably this would mean not including motion correction or slice timing
the 2nd time around.
3. Run FEAT individually for both, and average the two zstat files.
Is there a standard way of doing this?
Cheers,
Jodie
-
---
Jodie Davies-Thompson, Postdoctoral Fellow
Department of Ophthalmology & Visual Sciences
UBC/VGH Eye Care Centre
2550 Willow Street
Vancouver, BC, V5Z 3N9
Canada
Tel: 604-875-4111 <Tel:604-875-411> ext 69003
|