Re: subject-specific hrf

Posted by bob cox on June 14, 2002 at 19:53:34:

In Reply to: subject-specific hrf posted by Bob on June 14, 2002 at 18:12:13:

It is not really practicable to get a HRF from a periodic block experiment with such long blocks and so few of them. Unless you are interested only in the HRF to long stimuli (usually, HRF refers to the response to short stimuli, and linear superposition is used to synthesize the predicted response to long stimuli).

One way to proceed would be

• Assume some simple HRF to find "clearly" active voxels, and analyze the data. (e.g., AFNI 3dfim, 3dfim+, 3dDeconvolve.)
  
• Extract the time series for these active voxels, and average them (possibly normalizing them in some way first). (e.g., AFNI 3dmaskave.)
  
• Then analyze this resulting single time series to get a single HRF.
  (e.g., AFNI 3dDeconvolve with the -input1D option.)
  

bob cox

Follow Ups:

Post a Followup