OPM MEG three axis measurements

Dear Dr. Cohen,

I hope you are fine and in good health,

I would like to compare between SQUID-MEG and OPM-MEG, we are trying to show that OPM sensors record more brain sources than SQUID,
and the simulation data should be helpful,

The main idea is with OPM we are dealing with three-axis measurement, and as you know MEG is most sensitive to tangential sources, while EEG ‘sees’ both components,

Your course is very interesting, and I will try to use it but with OPM we have 3 axis measurements, I saw your lesson about the simulation dipole when you discuss the lead field matrix you said it is on 3 dimensions x y z…either you average or you used one direction…so on our case how can I deal with it?

Another challenge is that I found some data in mne python with .fif file, and I want to study it to convert it to your file format.

Best regards

Hi Saeed. The leadfield model I typically use is actually 4D, with the 4th dimension being the cardinal orientations of the dipoles at each XYZ location. I then either compute the orientation that is normal to the cortical surface, or simply take one of the directions. That’s done for convenience in teaching so that we don’t need to worry about that additional layer of complexity.

But for your case, it sounds like this would be important to take into consideration. I guess you’d need to compute the three directions that are normal and tangential to the curvature of the cortex, to compute the sensitivities to each orientation. Also make sure that you are creating your own leadfield for your MEG setup; the leadfield I use is for 64-channel EEG.

As for importing fif files, you can either export the data from Python into .mat format, or you can use fieldtrip to import the .fif files directly.