|
Hi Nina,
I generally add motion regressors to my analysis. It seems to help a lot.
Normally I import the data strait from the .txt files that SPM realign
produces.
However other options exist. Firstly, a clever fellow named Damien Fair has
shown it may be better to calculate the Pythagorean transform for each
motion to see the true total movement.
Example:
http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0088297
You can also use the first derivative of the motion transforms (scan1-scan2,
scan2-scan3, etc). This measures scan-to-scan motion, which may be more
informative.
I don't usually bother too much with those, but have in the past from time
to time and they may work a bit better, but it can also be circumstantial.
Like so many things in fMRI there isn't a right or wrong choice here. I
often find a given choice will have positive impacts in some regions and
negative impacts in others, in terms of looking things like t-values.
Best,
Colin Hawco, PhD
Neuranalysis Consulting
Neuroimaging analysis and consultation
www.neuranalysis.com
[log in to unmask]
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On
Behalf Of Nina Goossens
Sent: December-15-15 4:33 AM
To: [log in to unmask]
Subject: Re: [SPM] define contrasts
Dear Colin,
I am working on a first-level analysis myself and really liked your comment
on the fact that you better not specifically model the OFF conditions.
I added to my model the movement parameters estimated in the motion
correction preprocessing steps as multiple regressors. What is your opinion
on this? Do you think this is a good idea or would you leave it out?
Kind regards,
Nina
Nina Goossens
PhD researcher
KU Leuven - Department of Rehabilitation Sciences
Tervuursevest 101 box 1501
3001 Leuven
tel. +32 16 37 90 40
[log in to unmask]
________________________________________
Van: SPM (Statistical Parametric Mapping) [[log in to unmask]] namens
[log in to unmask] [[log in to unmask]]
Verzonden: maandag 14 december 2015 23:20
Aan: [log in to unmask]
Onderwerp: Re: [SPM] define contrasts
Firstly, it is not generally considered necessary, or even wise, to
specifically model the OFF conditions. If you contrast this with any other
condition, you'll see a lot of stuff like negative findings in default and
positive activity in visual regions or general cognitive control areas, and
unless you have a really specific reason to do so, not much that would be
very interesting or interpretable. If you leave the OFF condition out of
your model, it is still IMPLICITLY modeled as 'everything else'.
Your post provides very minimal info, but presumably you need to hit the
"Define New Contrast" button. Then enter your contrast. For example Positive
Vs. Negative t-contrast would be 0 -1 1 0 assuming the even order below. If
you are attempting an F-test then you need to enter a valid F-contrast,
which is structured differently.
Again I think you should really consider spending some more time going
through the tutorials as most of these issues is covered in there
: )
Colin Hawco, PhD
Neuranalysis Consulting
Neuroimaging analysis and consultation
www.neuranalysis.com
[log in to unmask]
-----Original Message-----
From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On
Behalf Of Bel Rach
Sent: December-14-15 5:01 PM
To: [log in to unmask]
Subject: [SPM] define contrasts
dear SPMers,
when i am trying to define contrasts in spm contrast manager gui it says "no
valid contrast defined"
I HAVE
4 conditions (negative / positive / neutral / rest or off)
- OFF: onsets at 0, 36, 72, 108, 144, 180, 216, 252, 270, 288, 306 with a
duration of 18 seconds
- Positive: 18, 162, 306 with a duration of 18 seconds
- Negative: 54, 198, 234 with a duration of 18 seconds
- Neutre: 90, 126, 270 with a duration of 18 seconds.
help please
|
