IUBio

An Experiment to do -- EEG Skull Cap

kenneth collins kenneth.p.collins at worldnet.att.net
Sun Dec 12 10:11:13 EST 2004


"kenneth collins" <kenneth.p.collins at worldnet.att.net> wrote in 
message 
news:D1Hud.114444$7i4.8627 at bgtnsc05-news.ops.worldnet.att.net...
| "kenneth collins" <kenneth.p.collins at worldnet.att.net> wrote in
| message
| 
news:c2psd.1030833$Gx4.115719 at bgtnsc04-news.ops.worldnet.att.net...
|| [...]
|| Variations in the technigue include
|| using it to explore the neural dyn-
|| amics that underpin "curiosity's"
|| behavioral orientation dynamics
|| as they are discussed in AoK,
|| Ap5. [Here, the problems of
|| getting "novelty" into the scan-
|| ning apparatus, and of scanning
|| regardless of orientation move-
|| ment are left to the Experimenter
|| to resolve :-] [I know it can be
|| accomplished, but I've no hands-
|| on experience with "scanning",
|| and I'd need that before I can
|| figure out how I'd do it.]
|| [...]
|
| Can LCD monitors remain func-
| tional in the 'magnetic' environ-
| ments of a scanner?
|
| If not, then perhaps some sort of
| Rub[e]-Goldberg chain of non-metalic
| mirrors could be employed to give
| subjects something to see while
| they're being scanned.
| [...]
| The "multi-Mobius-ed Klein
| bottle" stuff will just precipitate-
| out -- like in thise sets of sup-
| posed 'single-photon' increment-
| al-accumulation images that are
| invoked as 'substantiation of
| quantum mechanics'.

In this post I'm going to focus up-
on the EEG Skull Cap technique
because it's relatively-easy to ob-
tain really-spectacular results us-
ing it in pairs [or multiplets] that
make careful use of the rigorous-
ly-delimited externally-imposed
artificial restrictedness that I brief-
ly discussed in my preceding post
[linked-to above].

----- DATA ANALYSIS -----

The results are data-intense, though,
and the necessary data-analysis is
a bit demanding. But this is not an
insurmountable "obstacle" because
the data can be saved indefinitely,
so the data-analysis algorithms can,
themselves, be refined incrementally.
More on data-analysis below.

The general idea is to have two, or
more subjects, each wearing one of
the EEG Skull caps, and each hav-
ing it's independent data-collection
system. This's not as expensive as
it might seem because, in the tech-
nique I'm proposing, all that's nec-
essary, at this point in the experi-
mental design, is to simply collect
the data for subsequent analysis.
This makes the necessary hardware
much less complex than setups that
do real-'time' processing. The one
Critical thing is that all of the record-
ings that are grouped in a single tri-
al have to be sequentially-synchro-
nized in a way that is referenced in
each data subset. [This "syncroniza-
tion-reference can then be used as
a "primary key" in a database table,
enabling the data in each table to be
accessed in a completely-synchron-
ized fashion. Do this one thing and
everything else that I'll discuss in this
post just becomes flat-out doable.]

The general idea is to present the
subjects with specifically-delimited
tasks to focus upon during each
trial. The task instructions should be
written-up and stored in a "trial spec-
ifics" table [or group of tables] in the
database. [It would also be nice to
record video of the subjects' behav-
ior during each trial. If this is practi-
cable, the video recordings must al-
so be referencable via the synchro-
nization key. And while an Experi-
menter is at it, record and store as
much other data as one can -- EMG
from as much of the musculature as
is possible [here, I'm reiterating my
long-prior discussions of "Myographic
Array Diagnosis ["MAD", which I
invented in the late 1980s]. Ideally,
this would be done via body suits
studded with myographic sensors.
Build temperature, perspiration, and
any other type of sensors that can
be dreamed-up, into this body suit
[but understand that the more cum-
bersome the suit becomes for its
wearer [e.g. the more data-collection
leads that tether subjects], the more
the data collected will be shunted-off
into extraneous directions, skewing
the data that will be collected. This
can be controlled for, do useful de-
gree, via suit-design trials -- incre-
mentally cross-reference data col-
lected from body suits with and
without each of the individual encum-
berances, extracting the "background"
that's correlated to the "encumbrance",
and building it cummulatively. If this
is done, this "background" must be
updated periodically because, as
subjects gain experience, this back-
ground that's due to suit encumbrance
will alter [which will be extractable
in the data, and, when it is, will be
found to be correlated to 'just' more
TD E/I-minimization :-]

Record heart-rate, blood pressure,
respiration, eye movements, and
anything else that has any possibility
of being useful [of course, I'm going-
ahead, looking 'round the bend in the
technique's development -- early im-
plimentations should be as simple
as possible -- two EEG skull caps -- 
because what's most-important, early-
on is to concentrate on developing
the database and data-analysis al-
gorithms =with care with respect to
cleanness= of the code so that it can
be used as a "base" for all manner of
enhancements, branched designs [de-
signs that are the same, up to a point,
but which diverge in their foci at that
point]. When developing data-analysis
code, it's extremely-important to be
=Careful= about such open-ended
"cleanness".]

Remember, one of the most-useful
attributes of Computers is that they
allow incremental development, so
be sure to "backup, backup, backup",
and, while you're doing so, keep
careful version-'state' data.

Anyway, after getting a start with a
simple version of all of the above [of
course, in early versions, most of
what's above will only be "allowed-
for" in your "clean" design], go ahead
and start collecting data. At this point,
your main focus must remain on the
development and verification of your
data-handling. The most-important
thing is to assure that any bit can be
rigorously cross-correlated with any
other synchronized bit, so build all
the necessary stuff into your database
and focus upon maximizing the effic-
iency of such cross-correlation [most
data processors fail to do this initial
work adequately. Don't fail, here,
and everything that's subsequently
done becomes flat-out easy to ac-
complish. Fail here, and everything
subsequent becomes Hard-to-do].

When the above data-handling stuff
is accomplished, it's 'time' to start
collecting real data that will pass
through open-ended ["cleanly"-de-
signed] data-analysis algorithms.
Capture all the data that can be cap-
tured, not 'worrying' about what you
think is "important" to capture. All of
it is important. It's just that, since you
are =saving= the data that you collect,
you can, subsequently, analyze it to
your 'heart's content -- for decades,
and centuries :-] which is why the
essential thing is to work "cleanly" -- 
to develop code that enables the
sort of open-ended building of data-
analysis capabilities that are necessary
if all that's in the data you'll collect
is to be extracted.

----- NEUROSCIENCE -----

Everything you need, from the per-
spective of Neuroscience, has been
in AoK, all along.

In designing your trials, focus upon
delimiting the subjects' tasks to min-
imal requirements.

We're working with yoked-pairs
or multiplets] of subjects in interac-
tion. In each trial, one subject will,
be "motor-" ["active-"] phase dom-
inant and the other "sensory-" ["pas-
sive-"] phase dominant [AoK, Ap5
and Ap7].

Trials can be as 'simple' as having
two yoked subjects converse, which,
of course, would mean that active-
and passive-phase processing will
'alternate' between subjects. This
is no problem, however, because
you've developed your data-proces-
sing capabilities "cleanly" :-]

But it's best to start simple -- prob-
ably with simple somatosensory in-
teraction. Blindfold the subjects
and instruct one of them to reach-
out and touch the other. Record
from both.

What you'll see, here, will be the
generation of "probing" movements
in the "active" subject, and the gen-
eration of orienting-to-stimulus move-
ments in the "passive" subject.

It's necessary to be aware that the
subjects' performances, and, hence,
the data that's collected, will vary as
their 'familiarity' builds [as TD E/I-
minimization occurs] with respect to
their experience within the experi-
mental design. Since you're saving
=all= of the data you collect, and
since it's all 'time' synchronized, this
"habituation" is not a problem -- it
is 'just' more of what's in 'normal'
nervous system function. You want
to See it.

>From here, the methodology is in-
finitely-useful. That is, there's no
end to the ways in which it can be
used to explore nervous system
function [AoK, Ap1].

This is doubly-True because you
can not only vary the tasks submit-
ted to your subjects in infinite ways,
you can also analyze the data that's
collected in any trial in infinite ways.

But, since you've designed "cleanly",
this open-endedness is not only not-
a-problem, it's your =Opportunity=.

And as the dust settles in this exper-
imental approach, you'll end up in-
dependently Verifying what's been
in AoK all along [except that you'll
find that there are 4 "dimensions"
instead of 5 [AoK, Ap2] because
there is no such thing as "time" [what's
been referred to as "time" is =just=
energy flowing in rigorous accord
with WDB2T]].

What's discussed above is exactly
how I developed NDT, except that
I had to do it all in the ol' noggin' lab,
working with the data that had been
collected by my Colleagues in Neuro-
science, as it was deposited in the
Neuroscience stacks and standard
references.

Working in one's noggin' lab has some
powerful advantages, a main one be-
ing that noggin' labs can do millions
of iterations in parallel, billions of
'times', reiteratively, and once one
comprehends TD E/I-minimization,
all of this innate information-processing
capacity becomes flat-out Directable,
just like "language" is during 'normal'
conversation.

This's why I've taken the 'time' to
'translate' what I do in the ol' noggin'
lab into this multi-EEG-skull-cap
data-analysis methodology. If folks
do it, folks'll see, in it, an analogue
of the way NDT was developed, and,
hopefully, in that, gain a sense for
what's been in AoK all along.

k. p. collins 





More information about the Neur-sci mailing list

Send comments to us at biosci-help [At] net.bio.net