search

neurofractal / analyse_OPMEG

Nic & Rob's lair of scripts to analyse OPM data, using the Fieldtrip toolbox
https://neurofractal.github.io/analyse_OPMEG/
MIT License
9 stars 1 forks source link

DSSP in SPM #30

Closed tierneytim closed 3 years ago

tierneytim commented 3 years ago

Hi Guys,

Would be good to see relative contributions of the projection on the leadffields step vs canonical correlations step. My worry is that the CCA step is not as stable at low sensor numbers (we've already seen this with it wiping out large amplitude signals). If possible could a flag be added to output the first step alone?

neurofractal commented 3 years ago

Yes good point. I guess this is what was happening with some of the earlier data when we only had 12-or-so channels

georgeoneill commented 3 years ago

Hi Tim can you have a look at the deltas and see if this is what you were proposing?

tierneytim commented 3 years ago

Yes, that should do the trick. Of course, it is rank deficient so it might do some spicy things to the beamformers but it would be good to untangle the contributions.