If users want, they should be able to denoise the optimally combined data produced by fMRIPrep in whichever space they want. The tedana pipeline would still use the native-space echo-wise data, of course, but the bad components could be regressed out of any target space's data.
If users want, they should be able to denoise the optimally combined data produced by fMRIPrep in whichever space they want. The tedana pipeline would still use the native-space echo-wise data, of course, but the bad components could be regressed out of any target space's data.