Yes, you can average the parameter estimates and derive an average HRF. You can also fit that with a double gamma afterwards if you prefer, although it isn't really a question of things being valid or not - they are just different modelling decisions
and it depends on how well the models fit. You just don't want the fit to suffer too much for the sake or niceness or simplicity of the model, though there is always some tradeoff.
Hello,
We're trying to use FLOBs to optimize the HRF for the rat fMRI data. According our understanding, the individual HRF can be calculated by pe1*h1+pe2*h2+pe3*h3, where pe is parameter estimate and h is FLOBS basis functions. Could we average these HRF between
subjects to get an HRF model generalized for the population? If we're interested in the double gamma function specific for our subject population, is it valid to fit the averaged HRF with double gamma function?
Thank you,
Pei
