in the category of Medicine and Health Sciences,

(Cheers) (Applause)

the grand prize winner, Jack Thomas Andraka.

(Cheers) (Applause)

(Trumpets) (Applause)

And, now, ladies and gentlemen,

(Applause)

it is my honor to present to you the top award winners

of 2012 Intel International Science and Engineering Fair (Applause)

(Laughter)

Jack Andraka: So there I am.

No, that was the entire talk, bye.

We have been talking a lot about this thing called cancer.

It's really near and dear to all of us.

But, imagine a world without cancer.

What if there was a sensor that was like a diabetes test strip

that could detect cancer for less than 3 cents in 5 minutes.

Imagine a 168 times faster, over 26,000 times less expensive,

and over 400 times more sensitive.

Well, I created such a strip.

Basically, I was motivated to do this

because I had a close family member that passed due to pancreatic cancer.

It's a really devastating disease. It's the death sentence basically.

You can see that, after 5 years,

only 5.5 percent of people who are initially diagnosed, will survive.

That is a staggering statistics.

Why as a society that is so technologically advanced --

should this statistic be acceptable to us.

So, then, I got really interested and I started doing a bit of research.

What I found is that there are these certain proteins.

They are found in your blood at higher concentrations

when you have pancreatic cancer.

I was looking at one, called mesothelin.

It's highly overexpressed by pancreatic cancer as well as [at] the early stages.

Where, if you can find it in those stages, then it has close to a 100% survival rate

rather than 5.5%.

But, also a main problem with our current diagnostic test

for pancreatic cancer -- it's grossly inaccurate.

It diagnoses like, for example, inflamation of your pancreas.

That to me is unacceptable.

A 70% sensitivity and less than that for specificity --

[Which] means it gives tons of false positives and false negatives.

So then, through these really cool things called carbon nanotubes --

They are the superheros of material science.

There are these atom-thick tube of carbon and they have these fantastic properties.

So, I am super interested in that.

I began formulating an idea in unlikely place, biology class.

We were learning about antibodies which are basically lock and key molecules

that bind specifically to a certain protein.

In this case, the cancer biomarker, mesothelin.

Then, I had an idea. I was reading this paper

in biology class -- about carbon nanotubes --

I knew about this protein, mesothelin, and we were learning about antibodies.

It is kind of like a connect-the-dots puzzle where you form a draft

where in this case it is like a sensor.

So, here is the idea I came up with.

You have a network of the single walled carbon nanotubes and antibodies.

What happens is -- when you put a blood sample

that contains that target protein in it on this network,

that protein will go into the network and form a larger molecule with that antibody.

This causes an electrical change in the sensor.

I can actually measure this with a 50 dollar ohmmeter from Home Depot.

And it is really easy to produce.

All you do is to create uniform dispersion out of these.

You mix it up with these carbon nanotubes and this antibody.

You just dip stick that is like a piece of filter paper.

It's really, really simple.

Then, you dry it and you use it for a certain time.

That's not that hard.

You just measure it with this 50 dollar ohmmeter.

It requires less than a sixth of the drop of blood.

It is a tiny prick to your finger.

What you have to do for detecting a disease such as pancreatic cancer [is] --

when a protein is overexpressed you have to set a cutoff level.

If it's above this, then, it's at an abnormally high level.

If't is below that, then, it's kind of normal and you are OK.

So, I choose mine as 10 nano grams per mL.

A nano gram is a tiny amount.

Think it of as one billionth.

So, pretty small.

Then, essentially what I did is, I had to test whether my sensor was actualy working.

I took about 100 patients

who had pancreatic cancer ranging in its stage.

You can see they are all above this 20 ng.

That's well above the 10 ng per mL cutoff.

So, I had a 100 percent sensitivity.

Them, I had negative cancer samples, healthy patients.

People who did not have pancreatic cancer, maybe some other condition.

Those are all below 10 ng for mL.

That means that is not giving any false positives or any false negatives.

That means a 100% accuracy as [its] diagnosis.

This has the potential to save thusands of lives,

reshaping how we think of cancer diagnostics.

In addition to this, what happens is this -- also I found we could look at

how effective a chemotherapeutic treatment or a regimen

as well as how effective your surgery is.

You can see it here that

for each of different chemotherapeutic drugs -- I tested five on mouse models --

they all have different levels of mesothelin.

So, you could effectively see how effective your cancer drug treatment is.

You could look at drug resistance, for example.

With this, I have created a sensor that can detect pancreatic cancer.

It is very simple. It is very rapid -- taking 5 minutes for 3 cents.

It's non invasible. It's sensitive and selective.

But in addition, compared to the current gold standard of pancreatic cancer detection,

it is a 168 times faster,

over 26,000 times less expensive and over 400 times more sensitive.

That is not actually, including -- that you need specific practice

like a university training in order to conduct that gold standard,

as well as it requires this expensive bulky machine to read this thing called ELISA.

Whilst mine, it's a 50 dollar ohmmeter from Home Depot.

Also, the protein I am detecting, called mesothelin,

it is overexpressed in both ovarian and lung cancer.

Through this sensor, [that] is a generic sensor for those two cancers,

but also implications are that certain protein is found in 40 different types of cancer.

This sensor, is detecting nearly every form of cancer.

This has limitless applications.

What happens is you just change this antibody in the sensor,

you can detect anything, any protein.

That means other forms of cancer, other diseases

such as E Coli, Rotavirus, Salmonella, and those HIV, AIDS.

All those cancers and diseases are plague us.

Thusands, millions, billions of lives can be saved with this technology.

Also, it can look at how effective our treatment is.

So, this will reshape how we think of medicine.

We will now have a simple diabetic test strip

that can detect your disease in 5 minutes for 3 cents,

but also tell you how to optimally cure the disease and treat it.

And, so -- Oh, two slides are missing.

My final conclusion, what I am leaving you with,

is that I am not the only one who can do this.

Anyone can.

It just took me 200 emails and a bunch of hardwork to do it,

but that is a different story.

But, you just have to ask yourself.

Imagine a world -- or what if --

you can become like me and hopefully have a very happy running up to a stage.

Thank you.

(Applause)