help > RE: denoising without MNI?
Dec 13, 2021  08:12 PM | Daniel Brennan - Icahn School of Medicine at Mount Sinai; New York University
RE: denoising without MNI?
Originally posted by Daniel Brennan:
Hello,
I am hoping to run the denoising pipeline in subject space. However, I've noticed that during the denoising step, the MNI atlas is used to mask the data. Obviously this is not ideal - I have GM,WM and CSF masks in subject space to use for any sort of tissue prior needs. How can I make it so the MNI brain is not masking the denoised data? 

I've attached the MNI brain over my GM outline. My functional data looks exactly like the MNI brain and the functional data overlap, so its clear that its being masked.

I've answered my own question. CONN defaults to an "explicit" analysis mask, which if none is provided, uses MNI brain mask as a default. Switching to implicit mask (batch.Setup.voxelmask=2;) or providing a subject mask works. 

This was a little confusing to find and a bit unexpected default behavior to use an MNI brain mask for analysis. Perhaps option 2 (implicit mask) should be default, especially since option 1 will proceed without a supplied explicit mask. just my 2c

Threaded View

TitleAuthorDate
Daniel Brennan Dec 13, 2021
RE: denoising without MNI?
Daniel Brennan Dec 13, 2021