So my name is Kate I'm a student.

Right now and radical school so it's
a little bit of a change from what you've

been listening to it's really great
to be back on Georgia Tech campus.

I graduated here with my Ph D.

in two thousand and eight and
every time I come back.

I just recognize what a unique environment
that we have here as far as opportunities

like the one I'm going to explain to you
to really take the best opportunities

in fabrication technology and
apply it to real medical needs.

So.

Last time I was really here was
in two thousand and eight and

since then I've been
learning about medicine and

I wanted to talk a little bit about that
perspective because I thought well what

can I bring to this sort of forum and
these are technologies

that we can see in the embryo
are across town on a daily basis so.

I don't know.

So the top.

Would say.

So this is.

This is something that
they do on a daily basis.

It's called deep brain stimulation and so

the electrodes that they're using go
into different parts of the brain and

they treat things like Parkinson's
disease and depression and

tremor other things that you can see
in the Emory O.R. on a regular basis.

This is a spinal cord stimulator and

this is of particular interest to me
because they place these on to dorsal

column of the spinal cord to control
people with intractable chronic pain which

is an if you have it is awful and
this works very well for them.

And this is actually where I'm heading
in my own research in anesthesiology and

pain management and in the embryo are you
can see implantation almost on a daily

basis per for all nerve stimulators for
bladder control.

So all of these technologies
are being implemented and there is so

much additional work that they're looking.

To really be done to optimize it.

What these patients need is to have
these things once they're implanted

not get rejected by their own
bodies which is the big deal.

Compatibility is a humongous deal and
something that now technology can help

with immensely and there's a lot of great
research being done here on Georgia Tech's

campus concerning that as well as over at
Emory they also want to make sure that if

something does happen and the brain and
the spinal cord in the profile nerfs

are all dynamic systems things change you
want to be able to have the exact ability

of your interfacing system to adapt to
that and so one way that we do that is we

have multiple electrodes and that way if
one set of electrode modalities don't

work you can change without having to
go back in and having another surgery.

Another way that we're trying to
avoid having more than one surgery or

minimal surgeries for these patients who
are benefiting from the sort of technology

is making sure that once they're in they

are minimizing the amount of power that
they're using or if they're delivering

drugs they're really focally delivering
the drugs to where you need them to be.

And so that all has to do with
optimizing your system to make sure that

you're really precisely targeting
the tissue that you want and

that way it can stay in for longer.

So that's sort of the motivation from the
clinical and that I spend more time with.

Lately.

As far as my research is concerned I'm
very interested in the spinal cord.

And so I'm going to orient you all
just briefly to the cord just to give

you an appreciation of that's not just
a set of wires that transmit from brain to

muscle and then from sensory back up to
bring But before we get into that I just

wanted to orient you to a basic structure
of the spinal cord where this would be our

front end of this would be our back end so
when I say dorsal It's like dorsal fin.

And ventral is this way and

that's going to be important as far as
what we try to target with the technology.

That we developed here at Georgia Tech.

And Emory.

So.

On the inside of the spinal
cord are the cells and

these cells do an incredible
amount of cord nation for you.

Whenever you're moving your muscles
in any sort of way that's coordinated

you're not actively thinking about it.

Typically it's something where
you send the command to walk and

then your muscles know what to do.

That's all poured unaided by your
spinal cord as well as if you come

across something and you have to adjust
what you're stepping on or if you trip.

That's also not something that you
need to tell your body what to do your

spinal cord coordinates your muscles for
you.

And so a lot of these cells integrate
that a lot of the cells integrate the all

the different sensory modalities that
you're feeling to tell you how to adapt.

And all this information travels up and
down your cord some more locally

those are called intern or ons and
some that come from the brain and

some that come from the periphery and
go back up.

Typically the things
that travel up sensory or

consider travelling up the door
some the dorsal spinal cord and

a lot of the descending stuff that
goes out is through the front.

OK.

And that's that's a generalization.

So once again just bringing home
the point the spinal cord is more

than bundles of a standing and descending
axons which are the wires of the neurons.

There's an incredible amount of very
sophisticated interaction that goes on

within the spinal cord itself.

And that's really interesting to us
because many many people suffer from

issues where they have interruptions
in their ability to either receive or

transmit information through
their own spinal cord and so

what we're really interested in doing
is accessing the circuits that remain

in the spinal cord and
are accessible after injury.

So back in the early one thousand nine
hundred likes to do a lot more experiments

on cats than they do now.

I like cats but they're at.

As experimental tools as well as pets but.

What happens is that if you
put a cat on a treadmill

independent of descending control if you
turn the treadmill on they will walk.

So all of the and so we we've known for
over one hundred years at this point

that all of the circuitry all of
the neurons that we need in order to walk.

Are below descending control and so
that's to me incredibly fascinating and

to this day it's debated how
exactly our spinal cord does this.

Is it a modification of what
sort of sensory inputs we get or

is it something that's inherent
in the spinal cord and and

that idea is called central
pattern generation or C.P.G..

So the entered court once again is what
we're interested in clinically in this

case because there's so
much needed circuitry left in the cord and

we want to be able to access that
using our microelectronic race.

And so this became our target for

a clue to a project that integrated
what was going on at Georgia Tech and

they're really interested
in how these neurons work.

Learning how to mimic them
learning how to access them.

And what's going on over at Emory
in the physiology Department and

also in the hospital or we want to be able
to figure out what we can access from

the surface of the cord in
a relatively noninvasive way

in order to restore function to people
who have lost it due to injury.

So very roughly if this is
the injury what happens

is everything that's above that that
used to tell you how to walk that input.

No longer is descending and what happens
as those axons degenerate and so they're

no longer travelling down and what's
fascinating about that is that means that

what's acceptable from the surface
of the spinal cord is now different.

And that's important because
the things that we can access

are things that help us tap back into that
circuitry that coordinates are walking.

So this was our target and so
this was when Sean Hodgman over at Emory.

Our city department there's a ology came
to a talk like this where Steve De worth

using the neural lab over
here got to talking about

wouldn't it be cool if we could figure
out how to wrap in a ray of electrodes

around the entire surface of the injured
spinal cord stimulate it and see what sort

of output we can get as far as the
remaining neurons that coordinate motion.

So.

And once again that's why it's so nice to
be back in an environment like this where

things like that can happen
I think it's very unique.

So this became my Ph D. project.

So one thing to keep in mind is
these are these axons that transmit

information both up and down.

But also between neurons that live
only in the spinal cord they travel in

discrete functional bundles and
we know this based on many many years of

really amazing research that of has been
done in electrophysiology labs to look

around the world and so
they've been able to identify

what it is that all of these neurons
do and where their axons travel.

And so
the idea would be is if you could wrap

an array of electrodes all
around the spinal cord.

You could pull these sort of strings like
a marionette and get coordinated motion.

So people have been doing this
successfully multiple levels of control.

So just to review quickly.

So here's where your brain would be.

Here's the lower part of the brain and
here's what gets severed

in a spinal cord injury so descending
control which is in sort of this beige

collar has been cut off and here's your
spinal cord and brain stem circuits.

So what we're dealing with
are these local circuit neurons and

your motor neuron pools which go out
to tell your muscles what to do.

So interest bottleneck is stimulation and

this is done primarily by
the University of Alberta in Canada.

Taps into these bundles like
these gray matter neurons

that are within your spinal cord that
have been shown to produce coordinated

motion that is sensitive to
sensory feedback and this works.

Remarkably well.

And they've done incredible work
mapping out where exactly they need to

insert the spine by far
like microelectronics.

Into the brain matter in order to do it.

And so this is something that
you'll probably if you haven't

read about already hear more about if
they translated into human studies.

Another approach which for me I'm
most familiar with Reggie Edgerton's

work out new C.L.A.
does the same thing except it targets

from the surface of the cord and
the epidural dorsal pull of

an epidural just means that it's
around the outside of the hard coding.

Of membrane around the spinal cord.

And so what's great about this is we
put things in the epidural space and

outside the epidural space and
human beings all day every day and so

it's something that
clinicians know how to do.

It's relative noninvasive But what they do
is that that wire and that electrode is

close enough to stimulate sensory inputs
as they come into the spinal cord and if

you do that in just the right pattern you
can get walking motion walking movements.

And so they've been extremely successful
with that approach especially when they

combine it if the person has inputs on
a treadmill and also some pharmacologic.

Stimulus.

So but what we wanted to look at was
stimulation of the remaining tracks

themselves which was to get a little bit
closer than dorsal column stimulation.

But not as close as sticking
it into the gray matter.

We wanted to see if we could find a happy
medium just rest in the ray of electrodes

on the surface of the spinal cord and
see what happened.

So there are some advantages to
that there are inherent more

recording pathways which we.

Touched upon and there is evidence
these axons travel in distant bundles.

So I won't talk too much about the ventral
out of the Nicholas all that means

is if this is ventral and
this is lateral this is where.

If I'm a spinal chord.

This is where the white matter tracked
is that we're most interested in

an electric physiological studies
this area has been proven

in isolated spinal cord to
produce motor output and so

if you look at the motor routes that
do flex certain extensor motion.

Those light up in a rhythmic pattern when
you stimulate that part of the cord on and

so on so very promising area but
up challenge for

people who are trying to do this
clinically because how are you going to

get your wires all the way around to the
front of the spinal cord because we access

the spinal cord through the back
as clinicians and that's a very

difficult thing to do without damaging the
spinal cord given current technologies.

So I'm going to skip over this and this
was what I was talking about as far as

these rhythmic output patterns
on flex turn extensor.

And this is this is a spinal cord in
a bath and so we put it in a three respond

a fluid like bath and
we can keep it alive and do things to it.

So accessibility issues.

So this was our design goal and

this was achieved right here in
Georgia Tech in the am I R C.

And so what we wanted to do is
create an array of micro electrodes

that can access all around the surface
of the cord and then selectively

activate the white matter tracks that we
want in order to produce this motion.

So our hypothesis was that simulation
would be able to vote this walking

like motion so hind limb associated
motor output in an injured spinal cord.

So here were the three steps
design something evaluate it and

then use it toward a clinically.

Eleven And so what I'm about to talk
about was done here at the M I R C and

it could not have been done without
two other people who worked with me

on this project and the first one was
Rick Julie who's at U.C. San Diego now and

the other one was the young girl who might
he probably knows more of you than I do at

this point but he just graduated on to
the I think the Langer Lab at MIT and.

He really pioneered this and took off with
it but the material that we wanted to use

was poly dimethyl Psylocke Singh or
Pedia mass and you can see here

the difference between what would happen
if we were to put poly him in which is.

It's flexible but not conformable
versus P.D.I. mess around a tube.

That's about the size of our experimental
model spinal cord in the rat and

as you can see there's a lot of
bunch bunching in critical in here.

And so
you lose good contact with the surface but

also those points where
crinkles are things that

if there were a spinal cord in there you
would do significant damage to it and.

That's just not something that is going
to be conducive to a therapeutic.

Device and so our goal became to pattern a
bunch of little electrodes on to that and

to be able to stimulate through them.

So here's just a idea of mechanical
impedance matching challenge that

we had with trying to figure
out how to get closer to

the sort of the softness of the spinal
cord versus what's being used.

Typically to make a race and so P.M.S.
gets us much closer so.

This was what ended up being what
worked for us in the clean room.

And.

If anybody wants to talk about
this more particular I can point

you toward a really.

Pioneered this and also further iterations
of it but essentially we were able to

figure out how to pattern gold onto
a P.D.F. sandwich and then you.

In the reactive.

Holes in the surface and later on and so

this was our initial version of it and
you can see it bends quite nicely around.

A very small wire and
the aperture size was sixty microns.

And this is what Leon really
did here was he was able to

really improve that to make conical Well
features which help us to further isolate

that area and so here's here's the area
that would actually stimulate the surface

of the cord right here the rest of that
is just connections to the outside world.

So and you can see that this lift off
method was much more efficient for

us as far as fabrication.

But the electrode size was
was a little bit bigger which

actually ended up not being that important
to us as far as it worked just as well.

So we were able to do that and then it
was on to going over the Emory physiology

department to see if this actually did
something to the to the spinal cord that

we could use so stimulus selectively
became our goal of what to evaluate.

And why is that important to us.

Well if we can stimulate a focal region of
the spinal cord surface we're really not

doing any good by getting closer than the
epidural stimulators that are being used.

We wanted to really see if we had
an advantage as far as can we control

the system better by getting
closer to it because there's

this balance of invasiveness with
precision of how much control we have.

And so we had to show that this was
actually going to give an advantage.

And once again it's been shown very
thoroughly electrophysiological

experiments using very fine micro
electrodes that you can produce

very specific control of corrugated motor
systems by stimulating the surface of

the CT so obviously there's an advantage
of using a conformable M.B.A.

over a bunch of tiny little sharp
electrodes on the surface of the cord if

you want to put this in a human being and
have good.

Salt's So this was our in vitro set up
where we wrapped the spinal cord and

this actually went all the way around with
little tabs here and then we stimulated

the surface of the dorsal column to start
out with now the reason that we started

with the dorsal call and versus the
central lateral finnicky lists which was

our clinical target was the dorsal column
has a more homogenous type of accidental

fibers that run through and so it would be
easier for us to see what sort of spread.

We have when we activate through
a more homogenous set of axons So

that's what we did.

And then we what we did was here
this is actually one electrode and

we moved that in incremented fashion
across the cord and as we stimulated.

Between two electrodes and
bipolar stimulation we would compare it to

a control rigid stimulating micro
electrode which is the standard for

all of these studies that
I've been alluding to.

Of what they found they
can do as far as evoking

you know coordinated motor
output in the isolated cord.

So it doesn't look as pretty a person but

this is what it looked like
under the microscope and so

if you can imagine we're stimulating
between these two electrodes and

we're also stimulating and this is
a bi polar rigid tungsten electrode.

And then we record somewhere down here.

And this isn't a bath and so this is
what we found we found that the M.B.A.

is actually you can even see this here.

This is a much bigger surface area that
we're delivering current through and

this is by phase it current charges
that we're sending through bipolar

in both cases this is a teeny teeny tiny.

Exposed I mean almost you can even see
with under the microscope which part of

that was actually not insulated very teeny
tiny Bi-Polar stimulating electrode and

so they may require more current for
threshold but if you factored in.

How much larger that surface area.

Was it did require less charge density
which is good because charge density is

another way that you can
really damage neurons.

So we want to minimize that.

And then when we looked at what
happened with axons that were activated

in a lateral fashion in both cases
we were very pleased to find

that if you compared the EMEA with
the rigid electrode the die off.

Was essentially the same.

So this is very bless you.

Very exciting to us because it
meant that we could use the N.E.A.

in a similar fashion to what had been done
with the original electrode in decades

of electrophysiological experiments with
stimulating the surface of the cord.

So we wanted to make
sure that we could get

selectively as far as if we stimulate
between these two pairs of electrodes

versus these two pairs of electrodes
can we differentiate between the two is

what's being activated on the surface of
the cord and we Michel studies could.

So if we record here and
we're stimulating in blue we

would see a peak in the response here and
then with red a peak in the response here.

And so once again and this becomes very
important because if we implant an array

of electrodes one of the reasons that we
would do that is we need to be able to

change as the spinal chord changes as
the body changes to a different location.

That's something that we need to have
in any system that goes in the body to

interact with neural systems and
to have that control and

to be able to do that is incredibly
important with any interfacing technology.

We did find that there were limitations to
this and this actually was important for

us to see because it goes along.

What we know physiologically about the
axons in the spinal cord as we went more

lateral in the dorsal column.

We know that we come
across a lower threshold.

A lower threshold bundle of fibers and so

what happens is is it
takes a lot more current.

To stimulate certain types of fibers and
other types of fibers in the spinal cord

and so even if you're right on top of
these fibers that have a high threshold.

You're still going to activate the ones
over here that are much more willing to be

activated and we saw that as we went
more lateral with our electrodes and

so there are limitations this
goes along with what we know.

So.

So we found that we can also stimulate
some motor output which is important and

we wrapped it around a rat.

I'm not going to show you a picture
of that because it's gross.

But.

We are able to from the dorsal
surface of the cord and a juvenile rat

rat this array around the entire
spinal cord and bring it out the top.

And so it is flexible enough to
access everything that's below this.

So the conclusions that worked.

We were really happy.

It was selective and then we went
on to use selective stimulation.

On the V.L.F. and we ended up doing this.

Both with the biker electrode array as
well as the stimulating electrode and

this part I'm going to sort of gloss
over it's more about the physiology but

we went ahead and used a model for
spinal cord injury and looked at how

that changed what was on the surface of
the spinal cord as far as those tracks and

so we did a combination of imaging as
well as electrophysiology to show that it

does make a very large difference
whether you're dealing with an intact or

critically injured cord as far as what
we're accessing from the surface.

So that's a bunch of science in
electrophysiology which I'll be happy

to share with anybody who
wants to know about it later.

And then next steps in future directions.

What we're trying to do is show that this
sort of technology is something that can

be precise and can be noninvasive at the
same time it is a very good balance of.

Invasiveness and precision.

And actually surface stimulation is

really taking off in other
areas of interfacing in.

Well as far as retinal prosthetics
as well as electrocardiograms.

So this is really and that really goes
into broader application does it anyway.

Including recording which I
didn't really get into at all.

And so here's just one picture this is the
surface of a brain and here they are on

the surface of the brain using a very
similar material to stimulate and

that can actually be a really
nice balance between an E.K.G.

and deep brain stimulation and so
it's all about what you're trying to do.

But we have multiple points of access and
we need technology to adapt to that.

So these are the labs that
made this all happen.

We have some guest members Here's a Q.
Tansey who is an M.V.P. just over at Emory

in this is talk about her and her spinal
micro stimulation This is the Hotman lab.

And this is the de worth group who
are right across the way and without them.

None of this would have happened and
publications and other stuff so that's it.

Thanks.

The white matter from severe damage.

So what happens is and
this is actually something that we.

You've heard the term it's
not a flaw it's a feature.

So there's always going to be some
white matter on the surface because

the majority of what goes on in the spinal
cord isn't what generates after an injury.

So if you think about the descending
axons being sort of a cortex inside that

cortex are all those wires that connect
all the neurons that communicate within

the cord which they've shown is actually
the vast majority of the axons and so

you're still very much
have this architecture.

It's just the ones closest to the surface
are ones that are communicating for

example.

How your arm swing and how your
legs swing when you're walking and

that's the sort of stuff
that we're trying to target.

Let's go on a tour.

It certainly can.

That's not something that this particular
project went into but it has been done.

Thanks Bill.