Soon after starting the MEG electronics, as they warm up, there is sometimes large noise from the system itself. We regularly see this during the empty-room noise recordings, and have to wait a few minutes and redo the noise recording. As such, please always plan to start the electronics 5-10 minutes before you collect the empty-room recording and if you do see the artefact, wait 2-3 minutes for it to disappear. As always, keep me informed of anything new or unusual, like if this startup noise lasts more than a few minutes.
Please be aware this issue can sometimes be worse, either taking much longer to appear after starting the system, or even lasting longer, up to an hour. I’m still working on minimizing it, but for now we can keep the system on, e.g. overnight for morning bookings.
Importantly, it seems to need to be in acquisition preview mode to finish warming up, not just the electronics turned on. So even if we keep the system on, it’s best for now to still put it in preview for say 30 minutes before recording, to be safe.
Please note that since about last week, this warm-up artefact tends to happen about 45 minutes after putting the system in preview. It is therefore important to start it up ahead of your booking start time when possible. I’ll try to help with that, but please also specifically ask me if you know you can’t be there ahead of time.
The warm-up is slower probably due to changing weather and temperature in the lab. The good news is that the MEG electronics are scheduled to be replaced in December (date to be confirmed), which will solve this issue.
While the electronics should now be replaced at the end of September, the warm-up artefact has lately sometimes taken over 1 h to appear after turning on the system, and 10-15 minutes to disappear after. Because of this, please keep the electronics on between scans during the day, and if there’s a scan early the next day, keep it on overnight. When I’m present, I’ll take care of logging the next person in and put the acquisition software back in “preview” mode. If I’m not present, try to come in an hour ahead if possible to log in and start the software.
Additionally, it seems that when the system is on for an extended period of time, the artefact may come back and linger in the high frequencies. We’ve seen it above 400 Hz. This is not too problematic in most cases, but turning off the electronics may get rid of it again. It’s not clear what’s the best approach, but I’d suggest that if the system has been on for many hours, to restart it, keeping it off between 10 s to 1 min, before you start your session and collect the empty-room recording.
Finally, the good news is that if the artefact appears during a recording, it can be cleaned pretty easily, e.g. with SSP, with 2 or 3 components. It’s easier to identify the components when it’s in the high frequencies, away from strong brain activity. The pattern seems stable even when the artefact sweeps across frequencies, but when that’s the case, a relatively narrow time window should be used to be able to focus on an appropriate frequency range with a strong artefact peak. See here for figures.