Hi there,
I am new to fNIRS processing and Homer 3, and I'm wondering how best to include short channels in my processing pipeline. Our study has very long trials of 240-300 seconds (free play in children) so we are looking to use block averaging rather than a GLM. Is there a way to incorporate short channels beyond using them as regressors in the GLM? Any help/advice is much appreciated!
Just use the GLM. It basically is the same thing as doing a block average while doing the short separation regression at the same time. I guess we never wrote the code to let you do SSR first and then block average because we had no reason to do that. I'd be interested to hear if ever some has a reason to do it sequentially rather than simultaneously.
Thanks David! As we have very long trials of natural play, is there a way to figure out which basis function of the GLM would be most suitable? I don't think we'll find a canonical response across the 4 minutes.
English