In regards to interpretting empty room signal in general, I've
attached some examples of peculiar data I'm seeing. "Example Signal"
are just a few time steps (sampled at 250Hz, bandpass filtered) from
both a single human and empty room data (with and without SSS).
"Signal Variability" is the distribution of each signal around its own
mean.

From what I see here, putting a human in the scanner makes the signal
quieter and more regular. This was unexpected to me, and I still don't
feel that I understand it.

In regards to criticality, I've looked in empty room data for cascades
of events (where "events" are defined as activity beyond a threshold),
and analyzed their statistics. Such a measurement is perhaps most
generally well known in the context of the 1988 Bak-Tang-Wiesenfield
paper on criticality, but in neuroscience a major empirical work is
the "neuronal avalanches" of Beggs & Plenz 2003. The empty room data
appears to have the fingerprints of criticality, using these
cascade-based measures. Based on my confusion described above with the
empty room signal in general, however, I suspect I'm misinterpreting
the data.

Thanks to everyone for their thoughts!
Jeff

On 3/9/11, Matti Hamalainen <[log in to unmask]> wrote:
>
> Hi Maarten,
>
> It is, indeed, difficult to say exactly what the pure SQUID noise
> spectrum is in the very lowest frequencies from MSR noise data only.
> However, the SQUID noise is basically white down to the point where it
> hits the 1/f corner. Neuromag would be able to tell you exactly where
> that corner is. The white noise level you can find yourself by looking
> at your PSD estimate or simply from the "Manhattan skyline" plot in
> tuner.
>
> As far as the double grounding goes, it will affect the noise in the
> room only if as a result you get current flowing through the room
> walls, i.e., there is a potential difference between the two grounds.
> The double grounding does not really change the qualities of the room
> as a magnetic shield.
>
> - Matti
>
>
> On Mar 9, 2011, at 4:51 AM, Maarten van-Casteren wrote:
>
>> Hi Jeff,
>>
>> I have been having the same problem.
>> It can be difficult to interpret empty room
>> recordings. We make one every day, after
>> tuning the machine, and we have a Matlab
>> script that will give noise, spectral plots and
>> covariance matrixes for all channels, but it
>> is not easy to get useful information from
>> those.
>>
>> Basically you have two main sources of noise:
>> external noise and noise that's coming from
>> the sensors and the electronics themselves.
>> The internal noise is minimised by tuning,
>> but we have less control over external noise.
>> I wouldn't know how to tell them apart in
>> the recording, except that perhaps external
>> noise should be more affected by SSS, so
>> should be removed to a greater extend by
>> Maxfilter.
>>
>> As a test I double grounded our MSR a few
>> weeks ago. This should seriously increase
>> external noise, as the MSR is then much less
>> effective. But I really couldn't see much
>> difference between the normal empty room
>> recordings and the ones made with double
>> grounding. This surprised me, I have to admit,
>> but there might be more analyses I can still
>> do on the data. Any suggestions are very
>> welcome.
>>
>> Maarten
>>
>> From: Announcement for the Neuro MEG list [mailto:[log in to unmask]
>> ] On Behalf Of Jeff Alstott
>> Sent: 08 March 2011 19:39
>> To: [log in to unmask]
>> Subject: Re: SQUID behavior in an empty room
>>
>> I should restate, for simplicity, that my question really is:
>>
>> What does the MEG signal "mean" in an empty room setting?
>> In a setting with zero change in magnetic flux?
>> What would be some appropriate resources to draw upon to examine
>> this more?
>>
>> The answers could relate to the behavior of the SQUIDs themselves,
>> or to all the other hardware downstream of them. I'm just trying to
>> get a fuller picture of what is happening during such null cases.
>>
>> On Fri, Mar 4, 2011 at 6:15 PM, Jeff Alstott
>> <[log in to unmask]> wrote:
>> Hello all,
>>
>> I am analyzing data from a Neuromag Vectorview system for evidence
>> of critical phenomena, and empty rooms keep puzzling me. I am still
>> quite new to MEG, but my analysis of the "signal" seen from an empty
>> room, with proper shielding, shows signatures of a system at
>> criticality. This is surprising to me.
>>
>> While I have no experience with SQUIDs, an initial literature search
>> seemed to show that SQUIDs have critical dynamics when there is no
>> external magnetic flux. Is this correct? What would be some further
>> resources to understand what the MEG signal "means" in an empty room
>> setting?
>>
>> Thank you for your help!
>> Jeff Alstott
>> Cambridge, BCNI
>> US NIH, NIMH
>>
>>
>
>
>
> ---------
>
> Matti Hamalainen, Ph.D.
> Athinoula A. Martinos Center for Biomedical Imaging
> Massachusetts General Hospital
>
> [log in to unmask]
>
>
>
>
>
>
>
>
> The information in this e-mail is intended only for the person to whom it is
> addressed. If you believe this e-mail was sent to you in error and the
> e-mail
> contains patient information, please contact the Partners Compliance
> HelpLine at
> http://www.partners.org/complianceline . If the e-mail was sent to you in
> error
> but does not contain patient information, please contact the sender and
> properly
> dispose of the e-mail.
>