So they get greater pay.

I'll try not to roll around too much.

So I have to make a disclaimer
I was a little worried when

Paul asked me to speak.

I'm very much a system person.

My background and training can
take me to the national Brown.

But the work that I do shows how
you can integrate Microsystems' and

microelectronics with
Ultimately at some point.

Nanotechnology.

But those of you who have done
some now technology work.

I think you may experience that although
the the sort of the information and

energy exchanges on the national level but
the moment you try

to interface with nanotechnology
you suddenly jump to micro scale.

So a lot of my work is
at the micro scale and

basically what I do the bulk
of what I do is I use

electrical charge to
overcome a Larson either

the auditory system or
the balance vestibular system.

So a lot of times people think of the
vestibular system as primarily balance but

we'll talk a little bit about that.

So as

like as I speak I hope that I can
underscore these themes about you know.

At a at the first level maybe what
relates to the nanotechnology aspect is

how do we better couple electrical
signals through the neural interface

and ultimately that's
the exchange of ions.

And coming from a system perspective
you know how do we do that well and

that is reducing size of the micro
system reducing power with

the ultimate goal of improving a patient
benefit from the system or the device.

So.

I put this slide up because we're going to
spend a lot of time here during my talk.

We are going to be talking
about the other Tory system.

Which I guess I'll be my own pointer.

But you know if you come through
the outer ear to the middle ear

you get our atory system.

And coupled closely to the auditory
system is the vestibular system.

And we don't often think about the
vestibular system and less we're dizzy.

We lose balance.

So it's a system that works
all the time for us and

the moment it doesn't work
we realize it's there.

There's a lot of similarities in
the way that information is transferred

in each of these systems.

So start with the other Tory system and

branch into the vestibular system and
says we have such a good mix here.

Please feel free to start me and ask
any questions as I go along because I'm

probably going to speed up
pretty quickly to get through.

A lot of the topics today.

So he's let me let me provide
a little bit of reference here.

So we all know that sound is
made up of pressure waves and we

saw that early on and you're seeing this
now as I keep scraping this mike around.

And so we have our outer ear which is
designed really well to collect that

sound information and provided to
the middle ear we have that your drone we

have these beautiful bones here.

The smallest bones in our body and
these are so

impressive because what they do is they
amplify and mechanically filter sound so

that you have a gain of twenty B. from
the eardrum until you have the cochlear

the Coakley as a fluid filled to so
fluid vibrations serve

to couple two neurons really and
how does that happen.

That occurs through what's called
the organ of Corti here so

imagine that we've taken
a wedge out of the Coakley or.

These are repeats of that quiet
old into us and there's a sensory

structure here called the organ of Corti
and it's designed with hair cells.

These are the electrical fusing
mechanical to electrical transducers and

it's the motion of fluid that sets this

membrane to move up and down it's called
the basil membrane and the motion of that

membrane as couple through the hair cells
so as the membrane moves up and down.

This top of the stereo cilia band.

The bending of the stereo Sylvia issues
a series of electro chemical reactions

effectively that serve to activate
nerve fibers here in yellow.

So mechanical energy is directly
coupled into neurons firing.

It's a kind of an interesting structure
because we have three rows of

outer her cells and
we have one row of inner hair cells.

So these outer hair cells serve as
amplifiers they can actually elongate and

shorten based on the acoustic input
coming in and that's important for

hearing preservation which we may
have a few minutes to talk about.

So in a normal you know for
all of us here this is working just fine.

This is great.

So I've talked about how that information
is transferred the next pieces.

We hear pitch.

How do we hear different pitches and

that's because that Basil or
membrane structure.

Has a variable stiffness so
it's a mechanical filter.

And it's more sensitive to high
frequencies as you come into the base.

It's close structure.

And as you go up to the top.

It's more sensitive to lower frequencies.

So that is the way that we
resolve pitch in the Coakley.

So what happens.

What breaks down.

You know what is sensory neural
hearing loss which is how I

started this discussion.

I've shown here the hair cells.

Damaged hair of cells are pretty obvious
to pick out how does that happen.

You can be born that way you can even
we can be born with without her cells.

If your mother had rubella mumps
measles while she was carrying you.

You could be born with her cell loss.

Also there are N.

about extent are still prescribed
today that can cause hair cell loss.

So the electrical mechanical to electrical
transduction mechanism has failed.

Fortunately we can overcome
that through electrical

stimulation by directly
stimulating auditory nerve fibers.

So let me get a cochlear prosthesis.

This is the most successful
neural prosthesis to date.

And over.

Actually that and then them are needs
to be updated over one hundred one

hundred eighty thousand
individuals worldwide have these.

And through electrical stimulation
we can convey a perception of sound

in Europe they're implanting
children as early as six

months the reason being we
all have a formative period.

And I don't know if any of I see a very
diverse multicultural group here and and

a lot of people say well you need to
teach your child a foreign language and

you really do if you want your child
to pick up a foreign language they

need to hear it during their first nine No
nine months because that's when the brain

is kind of speaking non-variable.

Technical I'm not a neuroscientist but
there's a formative period during which

their brain is going to dedicate
that space to language.

So there is a big push to
implant children very young so

they have that auditory input so
they can build those connections.

I have talked to
an audiologist locally and

they've implanted someone as old as nine.

So there's a wide range of
people who are implanted.

So people who do very well they can get
one hundred percent of speech in quiet.

And you all know if
you've gone to a party.

It's really hard to hear
the person next to you.

So speech and
noise is a very active area of research.

How do we pursue or
perceive speech in noise and

there's a lot of signal processing work
out there and I'm going to talk more about

the neural interface work but really it's
the you really need to work on these from

multiple avenues to give people
the best perception of sound.

So from a system level.

That's why I made that disclaimer because
I tend to start from the system level.

The basic components of
a Coke layer prosthesis

are the instrumentation
to collect the sound.

Through of microphone and you know
number two there is a speech processor.

It's an amazing piece of hardware there's
actually four digital signal processing

cores in a cochlear implant and that also

it's externally the bat is powered by
a battery the same kind of batteries.

One would buy for a hearing aid and
across the skin the commands

are transmitted to an implanted receiver
and stimulator number five which then

takes which really sets up the signals
that stimulate auditory nerve fibers.

So what you see there is number six.

And that's the implanted
electrode array and

that's really where you start to get to
these low small scale interfaces between

the electrons using between
the electrical hardware and the neurons.

So.

There that should help a little bit.

So let's focus on this structure
a little bit more because this is where

this is where I'm going to really get
into the micro systems aspect of it.

So this is a cochlear electrode array
the black bear arms are the contacts where

where really the the electrical
signals are sent out of so

you can see if you recall that frequency
the placemat being will go from twenty

twenty kill herds down to twenty hertz but
we only have sixteen

to twenty two contacts in and
then planted electrode array.

That's all the space that we have
to convey the richness of sound.

So it's made of bundles of wire and
incased in silicone and

it still hand and I had the opportunity
just recently to go out and

talk to a cochlear
implant manufacturer and

they're really interested in how can we
use micro technologies to reduce the cost.

Because if you can back
fabricate something like this

if you can reduce the time maybe
these are all made by hand.

It's amazing it's really an art but

if you can reduce that cost forget even
providing enhanced sensation of sound.

That's kind of where my work went but

if you can just reduce the cost of that
because this cost forty thousand dollars.

And so
we've got some are emerging economies.

It would be great to be able
to have people use these.

So from a performance perspective you
know if we can somehow better recruit

the neurons so what I've shown here is
kind of an abstraction of the hair cells.

So it's all spiky things
are the stereo sillier and

the red illustrates where
we're activating neurons.

And if we can focus the fields and.

Activate a restricted neural population.

You can think just from the selectivity
of a specific city point of view.

We're going to better provide a sensation
a sound we can better restrict

we can better restrict the fields then we
can provide a better resolution is sound.

So that's where my research took me

the other thing is that neural
population changes over time.

I know in me the way I
can hear now is different

than the way I heard when I was five.

So if you have a system that can adapt
with more sites better selectively

you can couple to that neural
interface as it's changing

also talked a little bit
about electric fields.

If you have more sites you can play
games by waving the current and

we do that we do simulations in my lab
on how do you restrict that field.

How do you better activate those neurons
and you can just think you know in

the limit you would want to be able to
activate whatever frequencies you need and

that comes back to the hardware.

So this

represents the first Coakley or
electrode array the first high density so

we put more sight two to three times
more sites in a commercial array but

this is all micro fabricated.

So the same way we make
integrated circuits.

We made electro to race.

And not to trivialize it but
it's a fancy cable.

It's a really fancy cable that enables you
to better coupled to the neural tissue.

It's it was made for a guinea pig
model a very short insertion will to

talk more about why we only went into the
first turn of the guinea pig the guinea

pig actually has three and a half
turns to very tall conical Coakley and

in the guinea pig model
a very very thin material.

Very very thin.

So this is from my work at
University of Michigan and

the interesting thing about the.

This technology is that we
used Boron dope silicon.

Just one obligatory process
flow slide because I.

I do marker technologies
of fabric ation So

those of you are interested
in verification.

Really the exciting piece
in this is if you don't

silicon with Boron it serves as an F.
stop.

So instead of dicing out your devices.

You can put it in a very aggressive
etch so you can release them and

they just sort of float
around in a tray and

essentially it's made up of layering
on top of the silicon you layer

dielectrics you layer insulators
metals insulator and more metals.

So it's a fairly stable technology
Michigan has been doing this for

years but this is the first time we've
scared to go into a structure as

aggressively as the Coakley or
for the coiling.

And that presents some
challenges when you think of

an integrated circuit that's whack and
hard.

The cocoa here is spiraling and
soft in a very aggressive trajectory.

So if you're going to scale down in size

an important issue is what how do you
couple to the neural tissue once again.

And that's through your surface for
exchanging charge and

so we had to go to a different material

commercial arrays are made out of platinum
very stable we scaled down the size but

we went to radium oxide because you
can grow an oxide on the electrode.

And that really enhances the kind
of the space for exchanging charge

because that's what it's all about through
these electron neural interface devices.

But from a systems person perspective
if you're a circuit designer and

you're designing a current stimulator
the amount of charge you put in has to be

the amount of charge you put out
because what we're doing is.

It's very nice but nine.

Charge exchange is that
the electrode interface

the moment you get very
aggressive you start to build up.

Well first of all you can start to release
some of your metal into the tissue

you change the PH locally.

So there are limits on how much
charge you can put in tissue.

And you have to we followed that for
our device design here.

I'll go through this really briefly but

what I wanted to give you a perspective
of is OK You know you can go down from

the micro scale nano micro and
you can start to integrate circuitry.

So you can envision building
implantable devices.

So once you understand how you
can best couple to the tissue.

You can keep layer naan
technologies that exist.

So what we did is we built our old
integrated circuit for stimulation.

We coupled it with his me.

We coupled it with the electrode array and
of course this is a research device those

of you who you know have
to try to get F.D.A.

approval you wouldn't put up
a picture like that which

that is going to be implanted
into the tissue are insulated.

But that's to show how.

You how we made the connection
between all our

traces that needed to go to
electrodes with the circuit.

OK it's a very small device
it's about two point three Boyd

the the circuit here is two point three
by two point three millimeters not

an aggressive technology by any sense in
terms of feature size point five micron.

But it did the job.

So we were able to generate by
phase it current pulses that's

really the take home message from
the slide with the circuitry.

I want talked too much about testing but

I always get these questions about well
you made a device for a guinea pig.

How's a guinea pig going to tell.

So you that it that it can here and

through making really through
understanding the animal model

the abstraction that we make is the same
way that you have kind of a test board.

And the animal.

Becomes a test board you stimulate
at one end and we record.

So we can record the response to

electrical stimulation
at another location.

Further up in the auditorium system
called the inferior click yes.

That also has a frequency
to place mapping.

So in summary the way we do
the experiments is we first provide

an acoustic input and

calibrate through recording electrode
away ripping up neurons firing.

Ripping up the polarizations
as voltage spikes.

And we're recording you so course
there's some nice amplification going on

on these signals but then we can count

spikes that we can safely stimulate
here acoustically recording here.

So we can then do our own frequency
to placemat being in this

fine structure called
the inferior could careless.

Sort of just those of you who would
like to see where that is the knee

mid-brain there.

The in very quickly listeners pointed out
those kind of those two greenish knobs

in there so we put a recording array
in there to measure the signals.

It also has a nice frequency to make place
mapping it's like an onion the further you

stick in the recording array
the higher the frequency yes.

And he's so sorry for

my not an artist here but I wanted to
show you know you did this we did this.

What did we know at a first pass
when you're doing these experiments

you want to know that if I'm putting
in charge into the tissue number one.

Am I really can I record a response.

Number two are is the level of
current that we need to put

in similar to a commercial device so

this is towards benchmarking I
wouldn't call a special marking.

But it's towards benchmarking and what we
found is that the levels were very similar

two types of stimulation
briefly one is monopole or

simulate in the Coakley and
your return electrode is outside.

A broader field.

Another type is bipolar your sourcing and
sinking in the cold air.

You're restricting the fields.

It always takes more current.

So what we did is we stepped out
through the electrode sites and

made them further and
further apart for the bipolar and

you can see as we go further and further
apart it takes less less current So

we're approaching monopole or so this
is kind of to what your appetite about

the different games you can play with
the electrodes sites and focus fields.

So I'll kind of skip through this part but

say you know do these experiments first
we do this acoustic input then we

cut that that link the acoustic
link by damaging the hair cells by

actually applying one of the antibiotics
that I said damages her cells and

we validate that the system does not
respond to an acoustic input and

then we come in with our
electrical signal and measure.

So I will get through this
pretty briefly and say are we

validated that we can stimulate in the
Coakley or with a thin film based array.

The problem was silicon.

It's a great material but it's very
brittle and so we've moved on from silicon

here the work we're doing at Georgia Tech
is we're using a poly image based You're

a very a very flexible device people
views poly image in the body.

For a long time now.

So there's enough history there
to support it as a material and

instead of so it's very thin and it lacks
integrity to go on go in there on its own.

So we're using commercial commercial
insertion devices from Coakley or

implant manufacture and
were a couple ing the thin film

array with the commercial device and
we've done studies in.

Human Cadavers looking at how well

do these two pieces stay together
during the insertion and after.

So this is work that we're doing
with Medical College of Georgia and

it's proven to be a really
really exciting research topic.

We're on our second phase of the.

The cadaver trials and
we'll be doing animal studies in the fall.

So.

Let me try to launch into some
of the way out there work.

So I've talked about how you can use
electrical So signals to overcome a loss.

Now we're thinking about how
to use electrical send signals

to dampen the response of
the auditory system and

this came about through a proposal or
some area of research with DARPA.

And that is you know how do you
prove preserve soldiers from

impact noise in the battlefield.

And what I've shown down at
the bottom there in that profile is

you know what that signature of
the pressure wave looks like.

And the the directive is if we can
detect if we can make the outer

hair cells less responsive to the signal
we can detect it and beat it and

we can electrically if we can
detect it and beat that signal

before it gets to the inner hair cells
as you see the outer hair cells.

We can make those in her hair.

So she can make those outer
herself less responsive

to the electrical signal to the acoustic
input by hyper polarizing them.

And so that's a really new area of
research that we're moving into to use

electrical signals for auditory
prevention and the motivation is that

yes there are devices for
soldiers to wear in the battlefield.

But if you if you talk to the military
the soldiers don't like to wear them

they're uncomfortable
there's a payload and

I think as humans we tend
to feel like if someone.

Our senses are compromised.

We're not comfortable I mean think about
how well you know if it's all quiet and

you're in the forest think about
how you can hear if a twig snaps.

And so we're trying to figure out
ways that we can make soldiers

more natural in a way in the battlefield.

There's also the other
side of maybe you can make

the auditory system more responsive.

There may be situations where you'd
want to amplify the ability to hear.

So let me launch into some of the work

I'm still in the ear but now in the
balance system in the vestibular system.

And those three loops there
are semi-circular canals fluid

filled again fluid vibrations
are the key fluid vibrate in a fluid

motion in those canals
to flex their cells.

And that's the coupling
that we're looking at.

There's also a linear acceleration sensors

in the kind of the voters portion
there where you can see some of the.

The vestibular nerve fibers coming out.

That's actually the cochlear vestibular
nerve coming together there so

we do have linear motion sensor slows
the hair so mapping is a very confusing.

So it's a lot harder to try
to work in that regime so

we pick the simpler stuff and want to look
working with angular rotation sensors so

why is this so important.

I didn't really know this until
I started researching this but

many many of us will see our
clinician because we're dizzy.

Not all and not every one of us is
going to have a vestibular dysfunction.

But it is very common
in the elderly a large

amount of patients have vestibular
dysfunction I work at Emory and

every you know every Friday and
there are these clinics meetings and

you'd be surprised about the number of
people who have a vestibular deficit.

So the reason why it's very scary I
think in the elderly population is that.

If you know Lou.

Person and I don't even know what elderly
means anymore because we're you know

sixty five's not that far away from me.

If a person falls.

The likelihood of them
passing is it's very wild and

it's very strongly connected and
so for people to have command

of motion and to feel secure when
they move around is very important.

So I work in two different
areas in this regime.

So the good news is I work
with a rehab therapist.

You can train the body.

Depending on the level of
your vestibular deficit.

You can train the body to overcome that.

And if individuals are dedicated and

they do their exercises they actually
regain they regain a lot of their function

especially if the vestibular deficit
is only on one side a unilateral case.

Now the problem becomes tough.

When.

It's bilateral.

Because you can think about the body is so

smart of part of us is missing
the other part seems to work it out but

in the cases where there is
biological vestibular dysfunction.

There is research being done at
a number of very prestigious labs

on of us to real or prosthesis.

And I'm doing some of that
research here as well.

These are for people of no other
therapeutic benefit because really and

truly you don't really go around
implanting people you know with

a lot of research.

So let's talk about the of
some of the good news part and

that is just talk for a bit about how
would you train your vestibular system one

class of exercises that I'm
going to be working with at

Emory are patients doing these
exercises is gaze stabilization.

So even athletes do these exercises
you hold something in front of you and

you move your head and
you're trying to induce retinal slip.

But not so much that it makes you this.

But you're trying to go fast enough.

So if I gave you those instructions and
told.

If you to go home and do this.

Ten minutes a day two times a day.

And those are the instructions.

I think you may have trouble.

I have trouble.

So you know where do I see the
Microsystems' can make an impact in this

and I'm sure in the first generation of
this device it monitors your head motion

and the idea is a clinician
sets these targets for you.

They can tell you how fast you
should be moving your head.

They can monitor how long
you're moving your head.

And that way when you come back in the
individual come back to the clinic they

can start to correlate how well
an individual is doing with how much they

actually did their exercises because I
don't know every time I go to the dentist

they ask me do you floss and I say yes and
I'm a floss like twice the whole year.

So what patients report when they come
into the clinic is very different than

what they do and so now we can start to
monitor that the reason I'm showing this

slide is because all the all
the electronics are visible.

Do X. the stronger skull.

Data Capture power supply.

This is all integrated into the Kathy Now
you can see it the patient just takes

the cap home.

Nobody can see the circuitry at all.

And I've recently got some and
I funding to pursue this at Emory and

we also have you know this
is just a nice matlab.

This is a Matlab file that
spitting out what a patient did.

Over time of course and they were doing
you know some nice periodic motion there

but we can capture how you know how
much they do it how fast they do it and

you can even you know take
this data in between visits.

So you can monitor patient over the week.

And the patient doesn't have to come into
the clinic because when the patients come

in with clean these these
visits are longer and hour so

we're trying to do is see how
Microsystems can improve rehabilitation.

So now into the kind of
the back maybe linking toward.

The neural interface here.

Great looking to the neurons
here is a vestibular prosthesis.

So we capture angular head
rotation with some sensors.

We can then use that rate to control the
stimulation rate of vestibular neurons.

The reason why this works is because that

our peripheral vestibular
system what it does is that

it's senses Excel aeration
it integrates it once.

It sends an angular velocity signal
up through the nervous system.

A nervous system actually
integrates it one more time and

it controls the motion of our eyes.

So people have a similar deficit
if they move their head and

try to focus on something.

That I just can't follow because
they're not getting the right signal.

And so that's our one of you know that's
what that's why we can easily tested

in animal models is that
we apply the signal and

then we look at the motion of the eyes.

Actually there's a group at
University of Washington and

they're building a device that controls
it's a vestibular pacemaker and using

their pacemaker the animals moving around
but the neurons are telling the animal.

No motion is happening.

So without talking too long
about that we're developing

an integrated micro system to do that and
we have.

You know really this is this is probably
maybe one of the most interesting slides

is that the way our body codes angular
rotation is through a firing rate and

it's a sigmoid all relationship you can
program listen to a chip very easily and

implements that through.

An electrode array and
simulate vestibular nerve fibers.

So the one chala.

In this area that I'm
working on is the sensor.

So this is the circuitry.

But the sensor.

We're using commercial gyroscopes and

you know if you've got your i Phone
you have a gyroscope in there.

So these devices are becoming
readily available.

They're shrinking in size.

They're not shrinking in power.

So if you think in a room
of an implantable device.

In a Coakley or implant ninety percent
of the system power is to the tissue

in a vestibular in prime and they
are working hard to reduce the power but

it's still at about eighty percent
of the powers in the sensor.

And we're trying to overcome that through
mimicking the actual vestibular system so

this is a model of the semicircular canal
just one of them think of a horizontal

canal and we're building a pressure
sensor out of a very flexible material to

as the sensor to flex as a sensor moves
that gets coupled to a firing rate.

So we can detect motion through the
pressure sensor it's purely passive and

control the firing rate of
a similar nerve fibers.

So let's briefly some work
that we're doing there and

just a summary slide
here to show you why I

feel that the kind of the work that
we do in my lab is very integrated.

So I guess we're called you know
it's a Biosystems interface lab but

really to get the job done.

It involves all of those pieces for
our work.

So with that being said if if we
have a few minutes for a question.

So thank you.

Yes yes.

So the question is of early on in cochlear
implants there's a fibrous tissue

growth so the body reacts.

And the final phase of
the reaction is virus and

capsule ation because
Coco implants stimuli

from the electrode location
they tend to burn it off.

That's what I've been told.

But that also means that you can't
pull a cochlear implant out.

So you can't upgrade the electrode
ray you can upgrade the processor.

But not the electorate or
re there they are doing work by coding

putting coatings on Coakley or implants.

It's not quite.

You know at the at the stage where
they're doing more than animal studies

to encourage the neurons so
this is where the nanotechnology comes in

to encourage neurons to grow toward
the implant because the implant and

it's different than a quarter Bill
stimulating application then you're in it.

You're in a channel.

And the neurons are not there the you
have connective tissue and bone.

So if you can somehow encourage
the neurons you improve that interface.

Thank you.

Yes.

There's a couple of reasons.

Actually some of the experiments
we did weren't cat.

And it depends on
the expertise of the lab.

And animals.

It costs a lot to have them get in pigs or

trainable they do studies where they
train the guinea pig to step on a.

It all and I get some juice in so
my kids are very

smart animals in a way that you can
do behavioral studies with them.

Thank you.

All right thank you.