Print

Print


Dear Steve,

There are several points to consider .  1. Using uncorrected p values will likely give you false positive clusters, emphasising the need for good control of type I error, and restricting uncorrected stats to exploratory review of your data.  2. If you have task-correlated eye movements, this may be picked up.  I cant tell what kind of first level model (block-vs event) or task (inducing eye movement or not) you had, or whether events were synchronised to volume acquisition, but these factors can affect the likelihood of ocular pseudo-“activations” due to eye movements.   3.   It suggests your brain masking is not being restrictive to the brain–did you specify masking, and if so how?  This has implications too for the FWE correction, if your volume is larger than intended, or includes secondary volumes like the eyeballs.

Bw

James

 

 

 

From: SPM (Statistical Parametric Mapping) [mailto:[log in to unmask]] On Behalf Of Steve Masson
Sent: 17 August 2011 23:29
To: [log in to unmask]
Subject: [SPM] Activation in an eye

 

Dear SPM users,

 

I observe activation in an eye (see attachment). I have a limited experience in fMRI data, so you're help will be really appreciated.

 

Is it common?

 

What is the possible cause?

 

What can be done to avoid this?

 

All thoughts and ideas will be welcomed!

 

Regards,

 

Steve