Combining planar gradiometers vs Remapping channels

[anapesq]

Hello all,

I see that the issue of how to combined planar gradiometers in mne has come up a couple of times in the past in this mailing list. I am trying to solve the same issue and have some questions about the previous suggestion of remapping channel types using the as_type function
https://mne.tools/dev/auto_examples/preprocessing/virtual_evoked.html

Questions:

• Does the computation of the interpolation that is applied by the as_type function take into account the different orientations of the sensors (and consequently the magnetic fields they capture)?
• When visualizing combined magnetometers and gradiometers as magnetometers is it safe to interpret the output as the gradient of the magnetic field at the triplet sensor location (mag1,grad2,grad3)?
• Does the interpolation in the as_type function take into account that mag and grad have different magnitudes?

Many thanks!

Ana P

[agramfort (Alexandre Gramfort)]

@anapesq

• Does the computation of the interpolation that is applied by the as_type function take into account the different orientations of the sensors (and consequently the magnetic fields they capture)?

yes it’s using the sensor geometry

@anapesq

• When visualizing combined magnetometers and gradiometers as magnetometers is it safe to interpret the output as the gradient of the magnetic field at the triplet sensor location (mag1,grad2,grad3)?

when you visualize as magnetometers it’s no grad. It’s an estimate of the raw field assuming the gradients are also observed. Hence you see dipole patterns of B field.

@anapesq

• Does the interpolation in the as_type function take into account that mag and grad have different magnitudes?

of course. It’s using a true physics model of the field.

Alex

[anapesq]

@agramfort Thank you!