Next in line is our rapid
fire presentations and

I know that several of our presenters
there are all of the in the room I saw

them earlier today so
we will start to visit.

Jennifer.

To both.

Hold on once again
the new go back here but

before I turn over the mike to Jennifer.

I was told to remind you that you
can actually use social media to

share some of the ideas from this from
this particularly one so if Jennifer is

this the my musing things make sure to to
be about them so if you were the know if.

You didn't Good afternoon
I'm Jennifer DuBose and

here's a picture of me at the end there
in this very room at a blue sky retreat

it was very appropriate that I that I use
this photo so I am a research faculty

here at Georgia Tech and the College of
Design and I am the associate director

of the lab called the some great design
lab which is part of the College of Design

the director is standing there taking
a picture right now Craig Zimring and

we are very well represented in the poster
presentations I'm going to actually kind

of give you a taste of some of those
posters to hopefully entice you to go and

talk to our percenters So the mission
of our lab is to transform the process

of health care through the design of
the built environment and how can we use

the design of the built environment
integrate that with the care process and

come up with better solutions that
help improve the delivery of care so

that that's the vision that we have and
we've had the opportunity to work

with lots of people in this room around
campus including I asked why and

Pienaar and so it's it's great to
be at a university where we have so

many diverse skill sets
that we can use and so

architecture is just one of those
that we use in our talk in.

So these are the 4 basic things that that
we do or how we can organize our work so

the 1st thing I'd say is we analyze
problems which is probably a pretty common

thing to do here Georgia Tech but we spend
a lot of time reading the literature or

writing literature of views analyzing
things coming up with new frameworks and

ways to to think about how
the problem of healthcare and

the delivery of health care can be
improved through our built environment and

through doing that better so we do a lot
of publishing and one of the big papers

that cited in the field is from 2008 that
I worked on with Craig when I 1st started

working on evidence based as I'm We also
conduct research we have a lab in tech

square about 5000 square foot lab for
now until they build a new high rise there

where we do some really interesting work
and you should come and check it out we

have a few pictures and mentor that
we try to speed up learning and

innovation so how can we find out what
the better ideas are in the better

ways of doing things faster before
people start pouring concrete

before it gets built in place and then
how can we disseminate those lessons and

get that out there so how can we
help these this process speed up so

we're not doing we're not building
based on what was great 5 years ago but

what's going to be great by
the time the building opens up and

then we also develop guidelines guidance
documents so we try to actually get that

out of the hands of people who are going
to use it so we don't want to work

with the military health system we're
hopefully going to be starting a project

with the V.A. so we can take these
lessons and turn it into checklists and

guidance documents so
that it gets repeated again again and.

So I'm going to tell you about
a couple of the projects that we did

we have been working with the C.D.C.
The C.D.C.

funded a project with Emory
ourselves in Georgia state and

we've been working on how to
improve the doffing process for

treating a bowl of patients so this is the
part that's the most dangerous when you've

been dealing treating in a ball a patient
high stress situation you're wearing this

kind of onerous personal protective
equipment and you've most likely got

a BOLO on the outside of your suit you've
got out now take that off safely without

contaminating yourself so we actually
went to this is into the field and

we observed 41 training exercises where
the health care workers did a simulated

patient care task they were infected
with the a harmless facsimile

of a bola and then we watched as
they went through this process and

we looked at 1st specifically where
they were interacting the environment

where the environment seemed to
be impeding their progress or

seemed to be in a insufficient for
steering them away from risky behavior so

we've been really careful to not say
where we see people making mistakes but

where we see that they're doing things
that could put them at risk and

how can we design the environment to
kind of scaffold their activity so

they don't do that so we have
a poster on that this afternoon and

I'll actually be presenting well
on this in another thing in

Phoenix at our big national
conference in a couple of days.

So this is our lab and we built
a bio containment unit there our lab

a room we're not certified to actually
treat contaminated patients but

so that that's our lab that's our
bio containment unit lab space and

we built out what we did is learning from
those 41 observations we did of people

doffing their space we came up with
what we thought was a better design and

we tested specific interventions and
were able to find and

make modifications of the this is
that rapid prototyping right and

tested it with students until we thought
we had it right then we brought healthcare

workers back in and lo and behold we
actually improve their performance and

reduced risky behaviors so that's really
exciting we're just now processing

the results on that but we are excited to
see that the built environment can in fact

help people be safer and
hopefully that will translate into fewer

self contaminations infections so
we have a poster on that as well we have

our 3rd I think and
final post for post is on

is on the work that we've done around team
rooms we've done a lot of work looking

at the layout of outpatient clinics and

how that layout can improve
the collaboration of the Care Partners so

the doctor the nurses the med
techs will they work better

will is their collaboration affected
by the layout of the space so

these are some images of the floor plan
to show you we really are architects

are we really have architects that
work for us I'm not marketing but so

these are space and text diagrams that
I can explain to you when you come and

see her at her poster but
that show kind of the visibility and

how visibility to exam rooms from team
rooms can affect that collaboration

in your sense your ability to control
the space right so can the our IN

be aware of what's going on the team
in the clinic from their team room so

we're looking specifically at how
the design is going to help that happen.

And this is my last example I'm going
to give you don't have a poster on this

because this is just emerging but
very very exciting work that we're doing.

The funding of this next week but

a collaboration we're doing again
with Emory funding a large donor or

a donor has as promised a large gift
to build a mild cognitive impairment

program to help people with mild cognitive
impairment which is can be the beginning

stages of Alzheimer's it doesn't always
lead to all timers but how can we help

through the design of a better space and
space in their homes and how the space in

the homes is lead to a better experience
and we're also running an innovation

accelerator through that that I'm going to
have the great pleasure to to operate and

will be giving out seed grants on
an annual basis 3 seed grants to work on

things at the intersection of technology
the built environment and therapeutics and

this is just an example of something
that happened just up the street at

the Aware Home couple weeks ago really
exciting the gentleman the older gentleman

in each of these pictures has mild
cognitive impairment and they were working

with a team of industrial design students
to develop a teaching kitchen or

a coaching kitchen where they could cook
and actually the gentleman on the left.

Is not a cook he never cooked and here
he is with mild cognitive impairment he

prefers prepared a beautiful meal who is
so proud so this is what we're considering

empowerment it's not that he's being
cured he's not necessarily treated but

he feels empowered so we're hoping to do
a lot more of these really exciting and

Gage's with people through working with
students in classes to come up with design

solutions so that's just a sample of
the things that that we have going on

you find these people in the room Gabby is
here in the back she'll be doing one of

the posters and you can talk to her about
how your students can earn $10.00 gift

cards to participate in research
in our lab ask her for more OK.

So I know.

You like the OP So I think I'm good OK.

Fantastic.

So you can get a slight appear fantastic
I am Danny Hughes can you all hear me I

think I can hear the echo that is
coming to the speakers I know they

control the volume on these things OK I'm
a professor in the school of economics and

director of Georgia Tech's
health economics and

analytics lab which I'll talk about today
I just joined the faculty this August so

I've been here about you know a couple
months which is why you'll know I'm using

a generic Georgia Tech slide we have yet
to develop our own templates and slides of

course we're still building up but this is
actually leveraged off of an awful lot of

research over the past several years that
I've been doing prior to arriving and

just want to spend some time today talking
about this very exciting initiative from

my perspective at least and how we can
use this to leverage different kinds of

interactions of different kinds of
research across campus so the health

economics analytics lab is really a 5
year 3000000 dollars partnership with

the Harvey Lehman Health Policy Institute
based out of our Reston Virginia for

full disclosure I served as executive
director of that institute so

we can talk about what that means
at another date at the same time

that institute is really dedicated
to examining health policy problems

which is a little different from the sort
of narrow clinical sort of problems

outcomes research problems that
many of you are thinking about but

at the same time it turns out that we've
discovered in our research over the past

56 years that it's becoming increasingly
important to have access to the same kinds

of systems thinking the same kinds
of analytical capabilities and

tools that you would otherwise leverage
in some of these other areas specifically

we're earmarked for the Center here to
examine things such as large scale data

analytics machine learning and artificial
intelligence which again buzzwords

everyone is using and everyone is playing
with it every facet of health but

there is one important
distinction with our center and

that's what we're doing this with
a policy focus so the traditional study

that you might think of using these tools
are let's evaluate clinical pathways and

see how we can optimize the various
patient treatments or but tention look for

ways to optimize the turnaround
times in a high.

Spittal suite look for those types of
efficiencies that's not what we're doing

we're looking really at the systems level
the policy level how when you move from

value based games to volume based payments
how does that fundamentally change

research utilization patient outcomes and
fundamentally even the profitability

sustainability of the organizations
participating this on the provider and

payer side so leveraging large scale
data models to better understand issues

studies are working on would be things
such as Accountable Care Organizations

are providers choosing the horizontally
integrate or vertically integrate

with many different specialists and how
does that change again their ability to

make savings targets change outcomes alter
research utilization and we do this using

large scale databases what we've brought
in addition to funding this lab are really

extensive databases of large scale
medical claims type from C.M.S.

for example we have access to a 13
year panel of every medical claim

associated with 5 percent of the Medicare
population pretty large more so than

that we have a partnership without them
inside for the next 5 years we have access

to every medical claim associated with
every United Health Care beneficiary and

every large self insured
corporation that utilizes

to manage their claims that
covers 30000000 American lives so

when I say large data I mean very
large data large panels terabytes and

terabytes of patient level data that we
can delve into and in fact with their

Optima partnership we're now one of only 6
organizations that have access to the full

database and so that's something very
exciting that we bring to the table in

terms of how can we leverage center of
course partnerships of age and ha and

other sort of large scale
databases available but the sort

of promise of what we can do when we're
looking at literally millions of claims

are the sort of prescriptive
predictive and descriptive analytics

that everyone's talked about but
in slightly different ways so for

example current projects we're working
on again some of these come from and

out who are in from previous research but
we're starting to.

The linkages and
working with working with probably 4 or

5 different folks here in Georgia Tech
now not counting the grad students we've

picked up to examine problems such as the
all timers foundation is funding a very

large clinical trial right now called
ideas designed to come up with yet

one more can we definitively test for all
timers understatement groups in the path

ology world trying to do the same thing
but suppose we have that test could you

give that test to not everyone with
dementia or cognitive disorders and

of having all high numbers one of
the markers you use to determine the right

guidelines for who should actually have
that test Well that's where we leverage

machine learning artificial intelligence
deep learning within millions of years

excuse me millions of patients
over such small panel and

try to understand can we come up with nice
sure it's acceptable in bed either at

the health system or at the payer level
that flag these patients and say maybe

these patients should have that test and
then you develop the sort of guidelines

that you don't really have the kind of
guidelines that were formulated for

say mammography or other sorts of
screening that's just one kind of project

we're using We're working carefully
with I assume when you say yes

to try to develop the right type
of algorithms we would have

other sort of projects to look at standard
sort of evaluations whether recent study

examining whether at the annual wellness
visit that C.M.S. started authorising for

patients in 2011 did that make
a difference in the preventive care

utilization why because that visit
is not a physical it was designed

specifically to try to convince patients
to meet with their providers and

discuss exactly what preventive care
they need and hopefully prevent

downstream cost of course it's a voluntary
piece was that of interest well if you

follow politics at all you know that many
specific procedures components that C.M.S.

puts on the table are always up to date.

Or really up on the table to be removed
and this is one of those procedures said

literally on the table excited
perhaps to be of low value

we perform are starting to recognize folks
who actually participate in those and

Wallace visits increased preventive
care utilization roughly.

3040 percent across a broad
range of preventive services and

from a modified a colonoscopy
to Zambia pap smears etc and so

that's very important information because
once folks are thinking about potentially

taking that off the table what happens to
the potential downstream cost savings we

have from the preventive care so again
looking at a very policy oriented approach

to how we want to use that leverage
its data to try to make a difference

at a systems level than the health care
system what do we provide with healed or

again this is the importance of
having this presentation here

today with this group many of
you partners here on campus and

away from campus we just opened
our doors so we're looking for

collaborators we're looking for
post docs we're looking for G.R.

rays I'm still staffing up the center
we're still getting our office space of

a side of that show you a nice slide but
we're going to be set up a lease for

the next year and a half or so in the camp
and they have a sham building and

we'll see if we move into code or
not that's still apparently

on the table with everyone in
this room I think right and so

this hopefully make of all of what we have
in terms of dollars opportunities for

students and faculty member and
data to collaborate on research

at this time I should also know we have
extensive collaboration is with clinical

partners I'm actually an adjunct professor
department of Radiology never University

have deep relationships their wife worked
with folks for about I don't know 67 years

but we also have extensive
partnerships Mass General Harvard I'm

actually drew my hat as executive
director of Neiman Institute funding

a similar sort of health analytics
lab with NORTH Well health and

Hofstra University that's actually
dedicated to clinical patient outcomes

research and actually sent and
I graduate student last summer

there to directly work on the comparison
for 2 different stroke pathways so

it's more traditional sort of comparative
outcomes or good research where we're

using their system wide data that
determines if a stroke patient arrives and

has a C.T.A. instead of an M.R.I.
for which the guidelines are unclear which

imaging approach you should
have better guide treatment.

Doesn't fundamentally change patient
outcomes readmission functional

mortality and even the timing of when they
have that image from when they arrive and

so that somewhere else where we also
operate outside of the space but

again still thinking very much
in terms of the systems level of

how do we approach this in addition
we see that as a foundation for

additional research we've actually already
are a partner for peace 30 center at

University of Washington which is funding
musculoskeletal research will be performed

grants looking at dedicated clinical areas
and it's something that can leverage our

retrospective data and databases to
hopefully form that sort of you know

why is this an important study when
you're putting forth there are one and P.

30 grand for something a bit more focused
in clinical and at the same time we

have existing international capabilities
a collaboration some we're working with

folks at Taiwan for example where we're
comparing our data with their data to

examine again an important policy question
what's maybe something that's key but

the current medical debate that
we hear in the news Medicare for

All right is that makes sense
does it not make sense what

time one has Medicare for
all so collaborating with folks

in academia Seneca work comparing their
patient panel of 5000000 folks which is

roughly 10 percent of the time when these
population maybe a bit more to all of our

databases and try to look at very precise
clinical areas such as downstream

the Margraf utilization and other
clinical areas to see if the Medicare for

All is fundamentally changing how patients
and providers are utilizing resources and

then ultimately proximate outcomes like
read missions mortality and etc so

we can better evaluate what Medicare for

all might look like because they also
have a fee for service national payer

system that would be proposed if
we chose Medicare for all here so

again this is ways when you can leverage
large scale databases of the kind

that are traditionally thought about for
more of an engineering standpoint into

what we think of as higher level
policy oriented research Thank you.

I.

Know no continue the question.

I don't disagree at all I mean the one
I've done a little bit of research in age

our space I'll defer to folks who spend
a lot more time there before the mentally

a trickier concept because I mean they're
60 vendors it's a competitive market and

without a world where we impose national
standards are going to let the market

generate standards and I'm not going away
and even as an economist whether that's

a good or bad component and then you've
got a market being with like epic which

basically dictates what they want to do
and position sort of have to rely upon

that it's very tricky to say I mean we can
design and give an anecdote for that so

when I was at the University of Oklahoma
several years ago we had a grant to

evaluate health information exchanges they
were trying to set up in the state they

were beacon Grant winner from the Office
of National Coordinator of health I.T. and

the folks in the state thought well we
could come up with common standards or

have a better idea let's
have competition rule and

we'll set up 7 independent
health information exchanges and

that way physicians can decide to use
the health information exchange they want

that conforms with Christy's of their
E.M.R. systems and then come up with

a reduced set a common standard for them
all the communique which of course meant

literally patient name you know not enough
to actually make any meaningful decisions

the problem is you know you can model
that in engineering framework and we can

discuss it but that's more of a more of
a political question what is to growth or

what is the reach the government should
have in terms of the finding standards and

even the Office of National Court for
health I.T.

is not that very old but
let's let the users to find standards and

hopefully we'll have a very quick
convergence from V.H.S. the Betamax and

the something that we love Sadly I don't
think we're going to see that in health

I.T. anytime soon is that an important
component Absolutely I mean by only real

research other than doing evaluation of
that program several years ago was really

looking at the degree to which health
information sharing made a difference and

that the paper earlier
this year I had examining

whether menfolk shared laboratory test
just something as simple as laboratory

test or imaging data between institutions.

There are Taliban rates.

It's kind of funny right not funny but
kind of perverse for

Taliban rates increased when they were
sharing information which one hospitals

in a different network so how is sharing
additional data leading to an increase in

mortality rates and that's when you get
into what's actually being shared what is

the quality information being shared
are they simply not using it and

that's why we see an increase from control
for different variants so I don't disagree

at all it's an important part of the
system but important part of a system that

has more part of political solution
part of an incentivized market solution

because it's the real solution to either
have incentivized mandated standards that

we fill a comprehensive knowing that we
have a lot of hospitals and you know that

are already heavily invested in specific
systems you have a lot of vendors who

are financially motivated to keep
proprietary aspects of that sharing in

place them acts a miser market share or do
we choose to reward natural monopolies or

provide otherwise financial incentive for
them to consolidate because ultimately if

they're sharing standards it's probably
not as much you know as many ways

to differentiate your product and so you
see a world where you actually have almost

like it on you know in the Internet space
we're going to have fewer providers but

much more even conform to
products as opposed to a lot of

providers providing highly
differentiated products and so

again that's what's the right question
there and is it right when you move into

that world as well because this brings it
to another system level policy thing I've

spent some time on and then I'll give
someone else's a moment sorry P.R..

But we will Let's catch up
after this actually though.

Is it on well OK I almost feel like I
probably should not use a microphone at

all but Anyways let's do it so

I also use the template Georgia Tech
slides I don't have the excuse of being

new here anymore although let's
just pretend like that's why.

Anyways my lab is is really interested in
measuring and understanding parameters

about the body using noninvasive sensing
so sensors that we can place for

example in wearable devices outside
of the body and learn things about.

Health that could actually be relevant for
making decisions so I'll 1st maybe I'll

start with kind of a question to you guys
since we have so many different people

representing the different aspects
of health systems here how many of

you think that current commercially
available wearable devices are providing

useful benefit from a clinical
standpoint for patient management.

I would say for me I would not have raised
my hand and I think part of the problem is

what we're measuring with commercially
available devices is not really

that relevant from a clinical standpoint
and most of the devices are not F.D.A.

approved or not going through proper
sort of validation channels so.

I'm an engineer by training I'm
an electrical engineer by training in fact

I came from a lab or my advisor was
an M.D. Ph D. and so he was really

instilling in us from the start this
concept of clinical relevance and

so everything we do in my lab here
although we're in Electrical and

Computer Engineering is focused on health
related applications and very much of

the scientific level we're interested
in better understanding what we're

measuring from the body and by doing so
how to better exploit those signals that

remeasuring to learn things that actually
matter from a clinical standpoint so

one of the projects I've been fascinated
and this is a rapid fire so you know how

can you get any in-depth information about
how these things work now but if you have

any questions and you want to chat with me
or if you want to check out of papers if

anyone actually does that then that's of
course those are the real options so one

of the things I've been fascinated with
since I was a grad student actually was

this concept that every time your heart
beats there are mechanical vibrations and

movements of your body that occur as
a result of the heartbeat of course if you

put your hand on your chest you
kind of already know that but

many people don't know for example if you
stand on a weighing scale your body weight

actually fluctuates by about point one
pounds every time your heart beats and

that's because of the fact that the center
of mass of blood in your arterial tree is

moving slightly up and down.

And my lab has been measuring and learning
more about that signal is a signal called

the ballistic cardiogram it
was 1st discovered in 880 S.

it was 1st measured using these tables and
giant beds in about the 1930 S.

and forty's so in my Ph D. work I showed
that a ordinary electronic weighing

scale could basically be hacked and
used to measure these signals reliably and

I started discovering that things like
changes in cardiac output which are useful

parameters especially for managing and
understanding heart failure patients that

US could be measured with the scale my
lab here in Georgia Tech dows working on

wearable measurements of these vibrations
using Celeron meters on the chest so

some of the work that we did
recently showed that for

example you could detect essentially the
compensated versus the compensated status

of heart failure patients using some of
these measurements especially before and

after some exercise tests like a 6 minute
walk test in particular and that you could

track heart failure patients in
the hospital from admin to discharge and

notice that through the signal
there were substantial changes

significant changes from a statistical
standpoint that preceded other

things that are more
commonly measured but.

Heart rate related parameters as well as
sort of walking distance what I do there.

Whatever that button was should never be
pressed we've been doing some work for

the past 4 or 5 years now through this
large collaborative network funded through

and I H. focused on cutlass blood pressure
measurement again we're using these

vibrations signals to understand when they
are valve is actually opening when blood

is 1st pushed into the aorta and then to
use that and some timing specifically how

long it takes the pulse wave to travel
through the arterial tree from approximate

location to a distal location to detect
blood pressure without the need for

a cough we're still dealing with
many problems in this project but

we've shown some promising results so
far with the wrist worn as well as again

a scale based prototype and
this is ongoing work in the lab.

Separate from this cardiovascular
measurement work over the past few years

totally by accident I started getting
excited about the sounds emitted from

the joints after finishing as a grad
student at Stanford I started working at

this pro audio company where
we built microphones that had

much better high frequency response
on this one that I'm using now.

And so.

I had learned how to basically build
miniature microphones for Broadway singers

for rock musicians and also for
basically hiding and clothing for for

T.V. shows and that kind of thing and so
when I 1st started here I started thinking

about why don't we put these around
the knee which is why we do that right but

maybe to be able to understand
some structural abnormalities or.

To see if somebody for example rehabbing
a knee injury might be at some points

ready to move from one particular sort
of activity to another during rehab.

Since then because of this cool Georgia
Tech Children's Healthcare of Atlanta

partnership we got connected
my lab got connected with some

pediatric rheumatologist there and

realized that maybe these signals
that we're measuring for athletes.

After their injuries to see if
we can track their rehab status

could also be useful for
patients with juvenile arthritis

where it's a heterogenous condition it's
very difficult to understand which type of

medication is going to work well for which
patient and where you have to wait for

months and you symptoms and pain and these
sorts of subjective criteria to determine

if a particular medication is working or
not so we're really interested in seeing

if the sounds from the joints
have particular signatures and

patterns that are representative of the
state of patients with arthritis to see

if the joint is inflamed for example does
that change the sounds that we measure and

to use that maybe as a physiological
biomarker that could help with the process

we have also been doing
some more basic studies so

again we do sort of full spectrum from.

Boring sort of engineering work all
the way to clinical work and part of that

is really trying to understand these
signals better at a fundamental level so

we do some work here with cadaver limbs.

That are use the same kind
of limbs that are used for

orthopedic surgeons in their training so

fresh frozen rim limbs that preserve the
mechanical properties of the joints and

the tissues and with these limbs Actually
I have a student who's an M.D. Ph D.

student again with this through this
collaboration with Emory who can induce

a minister perform different
surgeries on the limb to see exactly

how again the signatures of these sounds
change with those surgeries we've had some

really exciting results with this that
have shown that some of the same sorts of

algorithms used for example for analyzing
acoustic emissions from structures

in civil engineering can also be used
to analyze the structural integrity

of knee joints that's been exciting and
just one other project to briefly mention

we have this Defense Department funded
project which is focused on trying to come

up with a noninvasive and
hopefully more effective than

pharmacological approaches therapy for
patients with P.T.S.D.

This is again in collaboration with Emory
psychiatry radiology and cardiology and

we've had some really promising results
using this noninvasive peripheral nerve

stimulation device at the same time as
a patient receives a recall of their

traumatic event and

it's essentially this interesting
sort of neurophysiology behind that

that essentially blunts the sympathetic
fight or flight response to that.

To that normal sort of stimulus
that would normally elicit that.

And so I have a fantastic team
here of I guess now 19 Ph D.

students working on this work and
research engineer and

funded through writing of of course
sources that are generous and.

In this Georgia Tech collaborative
environment of course we've got a lot of

great folks that we work with so
thanks for time.

And.

So trying so hard to stay on time but
I forgot that I have time for one question

I was at a panel once where please say
your name and one sentence about your

research turned into 20
minutes per person so Rick.

Lazio is the woman.

I get out there and
I won my I mean see how Lee and

then in our side I was I was
after working with Pienaar my.

For my Ph D. degree in the US engineering
and I just graduated last year and

joined C.D.C. as a prevention
effectiveness fellow at D..

So this is where does this concern
prevention actually I mean the division of

a shiv is prevention and our team is
called prevention modeling I think in our

mix team where we actually develop
mathematical modeling for the entire

division to estimate is cost effective
I mean we perform cost of found in the.

Models and
tests out that different interventions.

So I mean obviously I'm only a small
part of the C.D.C. and only at.

All of my projects are it is
that there are other parts.

Of the C.D.C. that some work can they be

willing to cooperate with anyone
sitting here if you're interested.

OK So the primary go of our team is
to apply a quantitative science to

HIV testing and
use a charity that you in this and

that we focus our work as many
of the cost effectiveness

of prevention efforts such as the testing
here and gauge and retention.

And one of the major breakthrough.

Was the.

Viral therapy which is in 1906 and
we have been using this

drug to suppress the viral load for

people who are in fact there was actually
V. and we usually hear us think this will

be treating patients but
actually having people under.

Under Control is actually a very effective
prevention policy because it prevents

the transmission of HIV virus
in recent years we also

were investigating the effectiveness of
pre expose prophylaxis this is basically.

Having people not infect that
was to take to take the drugs so

that they won't not be getting infected.

We also.

Analyze some of the behavior interventions
where people will tend to be

less risky behaviors.

So to kind of figure out
what will be the best

way to prevent HIV We divided us
our role best medical models that

are mostly in the simulation area so

one of them although we used to estimate
the transmission rate was a simulation

model we called the progression and
transmission of HIV or path.

So basically it tracks a disease
progression treatment and

transmission of HIV at a individual level

it also simulates the sexual and
injection drug use partnership

from between people who have a shot at
the people who don't have to estimate the.

The transmission.

That now mix in the United States and

we usually divide people into some
populations based on their population risk

group whether they are men who have sex
with mad people who inject drugs or

heterosexuals and
also we incorporate the HIV.

Treatment options each of
the individual have and

we divide them into age group year
is a years we have seen because

of they showed the drug are getting better
and better people are leaving a lot

with a longer life expectancy so
there is an increase.

In researching the different
interventions for

people who are with age
more than 65 years old.

Using this Asian base model we were also
able to recognize some of the transmission

networks features and
clustering in our national hiv

surveillance systems where we're see
a lot of the infections are happening

at the same place and with a crossing in
fact so you will see a huge cluster of

100 people that were mostly in fact that
in the last 6 months to one year or so

we were able so what we're trying to do
now is to use our civilization model to

replicate that feature and
test out cluster based interventions

in and hopefully we use that
when there are there is a break.

In the future so another kind of
model that we are using is more

population base so basically we
divide every we divide people with.

Different compartments based on their age
group risk level or transmission group

sex and race and
this in this city and each of

the compartments will have a number of
people flowing into the compartment and

number of people flowing out of the
compartment and where where you are using

a system of differential you clearly
been solved to to to monitor

what will be the dynamic transmission
maybe 20 years later 30 years later.

So the 2 projects that we're
using this common though for are.

Assessing the effects of reaching national

strategy goes I'm not
sure that goes in place.

And we're also analyzing the cost of this
of different interventions including

increasing testing frequency increasing
and here is to HIV treatment and

increasing the coverage
of prevention drugs.

And then finally we are using
all of these models to estimate.

And some of the time is.

As a federal agency we allocate fundings
to state and local health departments and

we want to have and
she used to fund the A wisely so

we devised a bake sale for
the state the local health departments.

Each of the intervention programs that we
recommended and they didn't put to this

model are related to each other in
their visual problem programs and

clinical data specific to the states and

we put a budget for each of the states and
years a Opposition model

to help the states to solve for
their individual states parameters and

Octane a optimal funding strategy and
where we're trying to test it in-house and

C.D.C. and hopefully push it out to use
force the department 2 years later.

So obviously we are still in a very early
stage of developing these models and

we are partnering away as
university professors and

Research Institute is outside of
the C.D.C. and sometimes the team will

give funding out to do some of the math
and models that we so we welcome any

collaboration efforts with our team and
we look forward to it and that's all for.

Any questions.

Yes so the transmission has
a maggot that it's both

a fracture to these and
also a chronic disease I think.

For influenza we don't really
require partnership so

so in my in my own idea I think
it should be modeling is a little

bit harder than influenza you some
sense because you have to capture does

network effect whereas
the euphemism you can have

you can chest meant to everyone contact
with so I guess modifying the contact

network could result in the different
model in simulation that Gopi.

That will work out for influenza.

And more question.

Thank you.

My next speaker is Mark.

I am so ready for.

This one.

So thank you very much for
the imitation to tell you a little bit

about the work that my group is doing in
terms of developing low cost equipment

free low volume and quantitative
diagnostics that's a big old mouthful but

if I can convey to you kind of why this
big part of our lab is important and

what we hope to be able to
accomplish in the near future so

when you think about medical diagnostic
tests it's quite possible you might think

of some things like the following
potentially big needles blood draws train

lab personnel that have to go and do this
complicated test potentially on some

very expensive fancy lab equipment
that not everybody can afford and

then you wind up waiting days to
get results when you finally do

get the results eventually you'll get
attention a very big bill for them right

so that's kind of what I think of when I
think of medical medical diagnostic tests

and our goal in the research in our group
is to try and turn a bunch of that on its

head so we're trying to develop diagnostic
tests that can use drops of blood so

that you can maybe get it from you know
fingerstick and that can have really

significant ease of use right to the point
that almost like a blood glucose monitor

that home user could potentially be able
to execute some of these tests but perhaps

different than a blood glucose monitor we
want to things that don't actually require

analytical equipment because there
are a lot of cases where even relatively

inexpensive analytically equipment might
be out of reach for a lot of different

people we want to get visible readouts
in a very short period of time and

that's kind of where we come to this idea
of no analytical equipment is that if we

can get visible readouts that we might
have something that moves from like this

blood glucose monitor to something that's
like a PH strip where you can just look at

individual colors and you can have
an idea quantitatively of what the what

the reading is for that individual target
the is that you're trying to measure and

then ultimately low costs right so
to go from tens to hundreds of dollars to

tens to maybe hundreds of cents so
that's what we're trying to do and

kind of the initial motivation
we got pulled into this

area of work by the bill Melinda Gates
Foundation we got a little secret and

we're whom scuse me I'm
recovering from something.

OK Art So our original motivation we got
pulled into this was looking at vitamin

and mineral deficiencies in the developing
world and so I'll talk to you a little bit

about zinc deficiency and so here we have
a map that is really a of very basic

kind of 3rd order general estimate
of the prevalence of zinc deficiency

across the world and so as you can see 1st
of all there are different areas that have

different prevalences you'll see that
the areas that have the biggest problems

are typically the developing world and if
you look specifically at zinc deficiency

it kills $100000.00 children under
5 every year across the world but

the problem is that the agencies that are
going to try and fight this problem they

don't have the David they actually need
in order to attempt to make this fight so

I told you these are kind of like
3rd order very rough estimates and

they don't have the kind of resolution
they don't have the kind of accuracy

than agency that like
the USA ID could go in and

say well I believe I know that in this
country that specifically one sub

region is where we need to
devote our efforts right in for

these agencies that have do a very limited
resources they need this information or

to allocate those resources effectively.

And the reason why they don't have this
information is these diagnostic tests

are just really too expensive and
too logistically difficult to perform

at the scale and in the locations
where they're most needed so

I've already highlighted that the
developing world is where most of these

problems in nutritional deficiencies
exist and in the developing world

you not only will typically not
have access to a fully staffed and

equipped clinical lab often you might not
even have access to electricity at or

near the point of care and
so it's important for

you to develop extremely low cost
equipment free electricity free approaches

to take these to take these measurements
now that's one motivation but

health disparities exist at home to night
so across the United States based on your

zip code you will have a different
adjusted estimated life expectancy and

this is perhaps due to an equal
access to health care but

perhaps also when people do have access
to health care and even quality of health

care because somebody who is going to
a rural physician's office where they

might not have the equipment necessary
to do one test or whether when they take

the test that they might have
to send it very far away and

pay a lot of money that they otherwise
don't have in order to get the kind of

test readouts that they need and so

my question kind of my motivator here is
that if we had more of this diagnostic

from ation could we do a much better job
in these low risk areas if we could revive

them with low cost things they could do
in these potentially remoter under-served

offices could we do helping these folks
and so our vision to do this is what

I say is cheap easy equipment
free biosensors and

by biosensors I mean that what we need
to start with for the nuts and bolts and

guts of our work are these microbes these
bacteria that are otherwise harmless and

the 1st thing that we do these
bacteria is we bust them open and

we take away their cell membrane and
we use the stuff that's

inside which is really valuable and
can do lots of interesting things and so

we take those insides and
then we add some genetic program to it.

Right so we can take different pieces of
D.N.A. and we can add them to this now

in vitro mixture and that mixture it can
then be programmed to perform different

functions to sense different things and
to make different outputs and so

what we then do is we can put those
that combination of ours so free life

plus that D.N.A. program on to these
different spots and so this could be on

a piece of paper this could be different
tubes that are all nearby each other and

then the vision would be that you can take
your blood sample in each of these spots

and what you wind up with is really
a customized calibration for

your specific blood sample where
just like that ph paper that I point

to earlier you're to have these different
calibration spots where you can just

look at that bottom spot and immediately
tell what the quantitative value is for

the thing that you're trying to measure
and we've done this right so this isn't

pure fancy This is an actual zinc in vitro
cell for a transcription translation

diagnostic that we've now developed where
you can see these are the calibration

spots that we're all run in
the same specific blood sample and

then you see that these are individual
measurements where we spiked in

different levels of zinc and
you see that the colors do indeed match up

exactly how they're supposed to be and
so we've got this working for zinc and

of course our goal is to get this working
for more things that we can then deploy

across the world and hopefully across the
country and I want to point out that one

of the valuable things here so I've
already said this is a quick completely

equipment free measuring zinc and

using micro leader volumes of blood we
actually receive spend any of the each of

these with anywhere from 2
to 8 micro liters of blood.

And it is a generalizable
approach is really important so

Remember earlier I talked about
those different those D.N.A.

programs that we're putting into
the cell free life say so those D.N.A.

programs we can change out what's in them
we can choose different sensors are that

are inside them and so as long as you have
access to those interesting unique and

useful different proteins transcription
factors or other nucleic acids

that you can make something in theory
to try to detect arbitrary power.

There's And then also importantly is
something called Matrix effects and

this is kind of maybe a higher order
thing that maybe the analytical chemists

among us might really care about but
these are actually really big issues so

fundamentally my blood does not
look the same as your blood and

actually to be really specific the my
blood does not look the same as one of my

graduate students but she spun it down and
her blood Sara was really black bright

yellow and mine was not at
all bright yellow right and

that's due to differences in the bill
the ribbon the holes but the point being

each of our blood major cities ultimately
are fundamentally different and

that can change the output So where then
trying to measure zinc all of those other

things can change the perceived measure
of zinc that comes out of your diagnostic

that comes out of your measurement and
something that's super simple and

equipment free like ours is particularly
susceptible to this problem but

you remember earlier what I described we
actually take the same blood sample and

run it in all both our calibration
spots and our test spot so

we actually remove that matrix
variability and this makes it then

much more generalizable so it'll work for
potentially antibodies blood sample so

this is a big goal in our group is to
develop these quick meant free low cost

blood tests and so if you have any
interest of kind really buried the science

here if you want to anything about
the science please come find me after or

if you have visions for
things that you'd like to measure or

maybe how this could be used and please
also do come for me because I book to hear

you say of course thank you to the people
that actually do the work Monica and

Daniel were the people that have
really driven this from nothingness

into something that we hope will
actually grow and impact the world so

thank you very much for your attention and
maybe a time for a question on a.

With with work I mean
it's not just like yeah.

So you mean in terms of things that
you wanted to act OK So funny thing.

Yeah so so
the funny thing is that the R.N.A.

problem is actually more solved than some
of the things that we're trying to address

we've been addressing small
molecules the important part is then

if you want something that's quantitative
like if you're looking for viral load

that's where what we've come up with this
quantitative aspect is really important.

Here I know exactly who you are sharing.

I've been dying to talk to somebody how to

gas a so that's an amazing and
insightful question so the question was

specifically for zinc What do you do when
you actually get this information and so

that's kind of a difference maybe
between what we would do with zinc and

maybe with our in a biomarker
that Cherry just asked about so

for zinc actually the way that this
is addresses on a population scale so

nutritional deficiencies are typically
endemic to a region not a person and so

in USA ID or somebody like that swoops in
to a region that actually treat a whole

town a whole region with a supplement
rather than individual people so

what they're actually looking for is can
they take a statistical subsample of

an area get that information and then act
based on that so for then it's not going

to be a one to one although getting those
results quickly is extremely useful for

them so
they act on an epidemiological level.

But I can envision other
applications of this that are more

in the clinic where you could
get that one to one you know.

Well thanks so much for

time THANK YOU THANK YOU YOU THANK.

Once again to all of our fire presenters
this was amazing most of this

information even though I made
sure to take was new to me so and

hopefully you all enjoyed it as
much as I did so you may not is for

other Fife repertoire presenters
vergers or take faculty but

they all came from different
units different colleges so

that's by design one of the goals of
this event is to expose the audience

to all the great work that's happening
in Georgia take each of the professors

video of it into the Ask think we talked
about their students their team so we have

several students in the room and really
our students are truly amazing we couldn't

be luckier to have them here at Georgia
Tech they really really help progress

are he certainly impacting society's
to be are grateful to have them year.

One of the rapid fire
presenters was an along and

we actually we'll have have more along in
the audience and we will have some of them

in our panel discussion later so
when they graduate our students

go out to the world to do in mazing
things and they also come back and.

Continue their commitment to
Georgia Tech to our education and

some of our other activities so
we are truly grateful to them for keeping

this connection and for continuing to
support us and all of our activities here.