>> Tell us, the people talk a lot about frequencies in the brain, gamma, alpha and

so forth.

We've mostly been talking so far about particular regions or

circuits in the brain.

Can you tell us a little bit about frequencies and

what that has to do with the brain's functioning and neuroscientific research?

Yeah, so one of the things that you may know is that

a nerve cell has something called an action potential.

So a nerve cell is actually in its resting state,

it's slightly negative in its valance.

Outside the nerve cell site,

there's this electrical tension that's sort of part of its basic state.

And when certain neurotransmitters hit it, channels open up that allow

charged particles to enter that changes the valance of the cell and

that creates and electrical pulse that goes down the cell and

you can measure this pulse as a little wave of electrical potential.

Well, that's a single nerve cell.

When you start adding nerve cells, those electrical pulses, they sum.

It's a process, it's called summation, it's very, very clear.

Where you get a magnified level of summed

electrical potential throughout the brain.

Now the degree to which a cell fires can be periodic and

it can move at different frequencies and

that's true of bundles of cells, as well.

And it turns out that the frequency of firing that a cell fires

has important implications for how that cell's been used and

this has been known at a descriptive level again for more than a hundred years.

People have been trying to sort of measure this phenomenon, but

we're understanding more and more about what the cells are actually doing.

So for example, a certain frequency above 15 hertz is by the way,

the number of periods within a second.

So the number of pulses within a second.

Both sort of 15 pulses per second, you're starting to,

that's associated with really processing information in a more active way.

There's a period or a range of activity between 8 and

13 hertz where we know that cells are sort of in a kind of resting state.

They're not actively processing information as a collection, so

much as remaining ready to process information.

And just staying in week through,

I think of it as nerve cells idling at a red light in traffic.

The car's on and they're ready to go, but there's a red light right now.

And that's a frequency band the 8 to 13 hertz that I've been particularly

interested in, because it seems to be implicated in a lot of psychological

processes when certain part of the way we're learning that the brain works

is that it doesn't necessarily always work by activating certain processes

as deactivating certain processes and letting other processes run free.

And one of the ways that you can measure that is looking at which parts

of the brain start showing larger amplitudes of alpha power.

Amplitude is how strong the signal is and it becomes stronger when larger

bundles of cells are engaging in it I hope that wasn't too technical.

>> No, that was very helpful.

>> Can you take us from that particular

frequency to examples of psychological experience?

Like you said, you're interested in it,

because it's associated with a number of psychological phenomena.

>> Yeah, for example, one of the things that I have done a lot of is look

at alpha power frequencies in the prefrontal cortex.

And it turns out that when people have

more alpha- >> [COUGH]

>> More alpha even in

a resting state in their left prefrontal cortex.

Pushed into the prefrontal cortex, people call the dorsal lateral or

ventral lateral prefrontal cortexes.

That suggests that their left portion of their

prefrontal cortex is less active than their right.

And individuals who tend to be in that state more often are at risk,

it turns out significant risk for depression and anxiety and

we're starting to figure out why that is.

And we've just completed a large grant of hundreds of people where

we've measured this and identified this pattern of prefrontal

activity based on alpha power asymmetries that looks like a reasonably

good marker of risk of depression and risk and probability of recovery.

We think that's partly true, because the left prefrontal cortex is other studies.

Alpha power in the left prefrontal cortex is inversely associated

with the degree that people engage problems.

So people who have less alpha power that means that their nerve cells are doing

other things than resting, they're processing information in other ways.

Those individuals who are processing more information in their left prefrontal

cortex seem to sort of be more likely to take the bull

by the horns to go after something to approach instead of to avoid.

And some people, for example, people like Richie Davidson suggest have

suggested that this is one way to understand what's going on in medication.

That one of the things that you're doing is being open to and

approaching stimuli write large.

Stimuli over any any kind, as opposed to being more wary and avoiding stimuli.

>> So how do you go about measuring brain frequencies and

what are the challenges in terms of localizing precisely where a given set

of frequency measurements might be coming from in terms of brain regions?

>> So traditionally,

the way that you measure brain frequencies as with the electroencephalogram, the EEG.

And it's a series of electrodes you place on people's scalps all over and

it gives you very imprecise readings of broad regions of activity.

Very imprecise,

because what you're actually measuring is the brain's electrical activity

through the scalp, through the skull and the skin on the scalp.

And what the skull does is it takes even very focal electrical

information and diffuses it, so there are two problems.

There are two problems with localization.

One is you have to first, just the information that you're getting,

the electrical information you're getting, you have to decompose

it into its constituent frequencies and that's a tricky task.

We use something called a Fourier transform,

it's a mathematical formula to disentangle all of the frequencies that go

into making those seemingly random squiggly lines that you get with an EEG.

But the other thing is you have an electrode here that's getting

some information that's relatively focal to that point.

An electrode here,

that's getting some information relative that's focal to that point.

And then both electrodes are receiving information that's shared between them and

shared with other electrodes all around and determining what's shared and

what's specific is really tricky.

And there's yet

another problem with localization from scalp related potentials.

Which is that, information that, electrically, is fairly focal to this

electrode, can be generated in a lot of different locations in the brain.

The way that electrical potential, it's like a storm.

It's like trying to measure where the source of the lightning is,

versus where it strikes.

And they can be very, very different.

We've gotten better at it, and we get better all the time.

But there's probably some upper limit that we've either hit already or we're

close to hitting in terms of using EEG signals to really localize information.

There's a lot of excitement about that over the last 15 years or so but

I am seeing that excitement precipitously abating because again,

we're hitting some kind of limit.

The advantage for using those kinds of neural signals at

the electrical level is that they're very precise timewise, so

using EEG you can measure brain activity down to the millisecond.

Okay?

That's really wonderful.

You can't do that with any other imaging technique.

The downside, as you've pointed out,

is that you can't be very precise about where that signal is deriving from.

>> And EEG is a relatively old technology.

>> EEG is old, yeah.

>> In scientific terms.

>> Mm-hm, yeah and it keeps getting reinvented in important ways, so

it's not something that we should think of as obsolete at all.

>> Yeah, I get that impression,

so can you tell us about the relationship between that relatively old technology,

I don't know, you can tell us whether it's a stable technology or not,

whether it's our ability to make use of it is expanding.

I'd like to know about the relationship between EEG and MRI,

which is the new technology on the block.

>> So EEG and MRI have complementary advantages and disadvantages.

FMRI is very spatially precise but the timing resolution is really pretty poor.

We're talking about, and you know,

when you're talking about the activity of nerve cells, these things happen quickly.

And so

the fact that with FMRI you're able to only measure things in terms of seconds.

That's an eternity in terms of how the brain processes information.

So, FMRI is sort of plodding

along measuring large movements of blood through the brain, and

then again making inferences about why that blood moved the way that it did.

It's not about measuring nerve cell activity at all.

EEG is measuring nerve cell activity directly, and

that's why the timing is so precise.

It's able to capture the activity of nerve cells, but of course,

the spatial resolution is very, very poor.

So we have these complimentary capabilities.

Now many people are trying very hard

to create a standard approach to measure both simultaneously.

There are certainly people who do it.

I've tried it.

I've played around with it.

But there's another problem.

And this is just wonderfully sort of sidebar, but it is wonderfully

illustrative of the kinds of headaches that scientists deal with all the time.

There are methodological conundrums.

So EEG is taking signals that are already very faint,

vanishingly small at the scalp and amplifying them 20,000 times,

just to see if we can then start decomposing the frequencies into

something that makes sense and is associated with psychological processes or

sensory processes or something in a reliable fashion.

That process alone is almost miraculous because when you take a little electrode

and you amplify any signal that it's getting 20,000 times,

you're not only getting brain signal, you get power outlets.

You get [SOUND] some one goes like that and it will show up in the electrodes.

Talking, noise all kinds of things wind up in that signal.

And filtering that out is very difficult to do.

It's much easier now,

I mean it's ridiculously easy even compared to when I started.

But that there's lots of problems with that and then you take that and

you put that into an FMRI environment where you've got an unbelievably

powerful magnet that's rattling around parts inside and very, very noisy.

And what you've got is like, if you could get Isaac Asimov to design just the worst

possible conditions under which to record EEG, that's the FMRI environment.

So trying to do both simultaneously is almost impossible.

And what people have done is try to model what happens to the EEG when

it's in a FMRI environment and use that model to remove variation in EEG signal

that looks just like the noise and then process the signal that's left over.

And there's been some progress there.

I'm quite skeptical of a lot of it, because when you partial away,

when you remove that much variance, what you're left with, you know,

it's been really the task is to really show really reliable associations between

the signal that's left and what's going on in the FMRI measurement or

in other behavioral measurements.

And that people are making progress, but it's slow.

It's hard to do both simultaneously.

>> So that's just one anecdote in what really is a technological revolution

in our ability to study the brain and

it strikes me that you're living through that.

We all are right now, but in particular,

when you began this area of research and when you we're

a graduate student that our technological ability was very different, right?

>> Yeah. >> Than it is today, right?

So, you've actually lived through the transformation that something like MRI

represents and its possibilities.

So, how does it look different, when you started versus now?

>> Well, when I started doing research, there was such a thing as FMRI and

people were starting to use it to study psychological processes and

behavioral processes.

But it was, I mean, I was just to burn a compact disk of data was still going to

a special lab in the Health Sciences Center at the University of Washington.

And signing up for time at 10:00 at night just to burn data onto a compact disc.

That's what I was doing when I started.

So, this was the idea that you could

study an interactive social process, an MRI scanner,

which is what I do now, was absurd.

>> Mm-hm.

>> Absurd.

It was almost absurd to think about doing it with EEG because a single amplifier

of EEG would take up a pretty good section of wall space in a room.

Right and take you a half an hour to an hour to set up and

calibrate when you brought the person in and then getting the EEG electrodes

all placed on their scalp and making sure the recordings were sufficient

in terms of the electrical interference called impedance at the scalp line.

It was a very large process, and the idea that you do that with two people and

you'd be interacting was just insane.

By the time I started in my faculty position here at the University of

Virginia, I had already done an MRI study, had received some training in doing MRI.

And I had done one of the first social-interactive studies in fMRI ever.

And that was a challenge, and it was very limited.

Just hand holding, just touch was challenging to implement.

At the EEG level, that big wall-sized amplifier that we used to have to use when

I was originally trained had been reduced to,

looked like a couple of little shoe boxes that sat on a desk.

And, that was pretty amazing.

It was living Star Trek as far as I was concerned.

And now, we have an amplifier at the lab with a small number of channels in a hat,

a little device that fits on people's heads that records,

again a small number of EEG channels through a bluetooth wireless connection

into an amplifier that fits on a thumb drive that sticks in I mean where is it?

It's amazing.

It's absolutely amazing.

>> So you're almost at the point of getting people out of the lab and

still being able to- >> Yes.

In fact, >> Monitor them systematically.

>> Just talking today with a colleague about doing a study of EEG and

about sort of idea generation while walking.

Because walking is an approach related activity and

we wonder what it would look like.

I was telling you earlier about this left prefrontal engagement activity.

If you can just ramp up that activity just by walking, we already know that walking

is associated with recovery from depression and anxiety symptoms.

>> So how does neuroscientific research relate to other forms of research into

human being, human functioning?

For example, people who are looking at the cardiac system,

or people who are studying more at the behavioral side.

I'm just wondering to what degree are these often silos,

where people are doing their work within their specific kinds of context and

methodologies, but not necessarily taking advantage of these other approaches?

And to what degree to often really work in teams where people are looking at

how the circulatory system is working in terms of blood?

But then also looking at the nervous system then also looking at

behavioral studies.

Well unfortunately,

people are doing their work mostly in the little silos at this point.

I think that there are a lot of sort of sociological, historical reasons for

that the way that we parse disciplines.

The way funds get distributed as a function of that parsing.

Just a lot of reasons for that.

It's not efficient as far as science goes.

It wastes a lot of time and money when we're taking that approach.

So there is a lot of excitement.

A lot of talk about teams of researchers that

utilize these tools in complimentary fashions.

And there's actually real progress in that regard as well.

But it's still the case, especially in the United States, that funding goes

very often to single labs or single individuals.

And what they have to do is construct teams of experts to do their thing.

But they're not really working across modalities or

across systems as much as they ought to be.

That's absolutely the case.

>> That's different elsewhere?

Is that different in Europe for instance?

>> It's a little more common in Europe or in Canada where the grants are smaller but

more plentiful so more people have the means to,

a larger number of people have the means to do the kinds of research that they do.

Here in the US, the model is a little bit more different.

A smaller number of people have enormous means [LAUGH].

It's sort of that way about everything in the US, isn't it?

As far as interdisciplinary teamwork,

I think other sciences do a somewhat better job of this.

So for example if a fossil is discovered in,

the Yukon territories of some significance, right away.

Within weeks, perhaps,

you could have teams in Australia that specialize in something.

Teams in, you know, North American and in Europe, all collaborating together to

understand what happened, what's the story of that fossil?

There's nothing like that I'm aware of really happening in brain science,

and I think that's too bad because we'd be a lot more efficient,

we'd make a lot more rapid progress if it were.

"Tools of Neurological Research," Jim Coan

藏传佛教禅修与现代世界：小乘佛教

与讲师见面