Photos: James Duncan Davidson

When Jack Andraka was 15 years old, he didn’t know what a pancreas was. Now, this teenager has created a test for the early detection of pancreatic cancer that, while still in the preliminary stages, looks promising. So how did he become an health innovator?

Andraka tells the story during Session 6 of TED2013.

“Have you ever experienced a moment in your life that was so painful and confusing, you just want to learn everything you can to make sense of it all?” he asks.

For him, that moment came when a family friend, who’d been like an uncle to him, passed away from pancreatic cancer. In Andraka’s Googling, he discovered startling statistics about this kind of cancer — that in 85% of cases, pancreatic cancer is diagnosed late when a person only has a 2% chance of survival. As Andraka explains on the stage, this is because the same (very expensive) pancreatic cancer test has been used for decades, and is only given if a doctor already suspects you have the disease.

“It’s a 60-year-old technique — that’s older than my dad,” says Andraka.

Andraka set out to develop a new test for pancreatic cancer that’s inexpensive, rapid, simple, sensitive, selective and minimally invasive. He began by looking for a protein in the bloodstream that would be a biomarker for pancreatic cancer — one that would be found in all cases, even in the earliest stages. The problem: there were 8,000 possible proteins. When Andraka was “close to losing sanity on the 4,000 protein,” he finally found one that could work — mesothelin.

But then he found a whole new problem — how would he go about detecting it?

“My inspiration came from the most unlikely place for innovation — high school biology class, that absolute stifler of innovation,” says Andraka, to big laughs from the audience.

While studying carbon nanotubes, Andraka had a flash of insight — that he could lace antibodies to these nanotubes so that they would react to mesothelin. This gave him the idea to make his cancer sensor out of paper. While he swears that doing this was “as easy as making chocolate chip cookies,” he realized that he needed to find a lab in which to do his work. “I can’t really do cancer research on my kitchen countertop,” says Andraka. “My mom doesn’t like that.”

Andraka wrote to 200 scientists asking for space in their lab. He received 199 rejections. And even at the one lab at Johns Hopkins University where a professor was willing to entertain his theory, he was bombarded with questions from grad students trying to sink his procedure. Andraka realized that his method did indeed have blank spots.

“Over the course of the next months, I painstakingly filled all those holes,” he says.

In the end, Andraka has created a paper censor that costs 3 cents — about 26,000 times less expensive than the current pancreatic test. The test takes five minutes. And it appears to have close to 100% accuracy, potentially allowing pancreatic cancer to be detected in its early stages, when a person has a much better prognosis. This accomplishment not only made Andraka the winner of the Intel International Science Fair – it has the potential to save many lives.

Even better, Andraka thinks it could potentially be used to test for ovarian and lung cancer too. And by switching out the protein the test reacts to, it could — down the road — be used for diseases as varied as heart disease and HIV/AIDS.

“Thorough this journey, I’ve learned an important lesson — that anything is possible with the internet,” says Andraka. “You don’t have to be a professor with multiple degrees to have your idea work.”

Read the TED Blog’s Q&A with Andraka.

In yesterday’s TEDTalk, Max Little explained how a 30-second phone call could soon be a diagnostic test for Parkinson’s disease, thanks to 7,500 people who’ve placed calls from around the world to ensure the technology’s accuracy. (See how you can get involved at ParkinsonsVoice.org.) This brings up an interesting question — what other medical tests could we see getting cheaper, easier and more accessible in the coming years? Below, find seven exciting possibilities.

Sam Gambhir: Molecular spies that detect cancer and Alzheimer’s early
In this bold talk, Dr. Sam Gambhir — a professor of cancer research at Stanford University School of Medicine — calls cancer a “solvable problem.” Rather than diagnosing diseases like cancer and Alzheimer’s in their late stages, Gambhir explains how “molecular spies” can be used to detect the very early stages of disease, allowing for “rogue cells”  to be destroyed. (Filmed at TEDxStanford.)

George Whitesides: A lab the size of a postage stamp
Traditional lab tests for diseases are far too expensive to be used in the areas that need them most — the developing world. In this talk, chemist George Whitesides explains his research in developing diagnostic tests on tiny pieces of paper, which can be manufactured at close to zero cost and shipped very cheaply. (Filmed at TEDxBoston 2009.)

Deborah Rhodes: A tool that finds 3x more breast tumors
Dr. Deborah Rhodes and her team developed a tumor detection tool that’s three times as effective as traditional mammograms for women with dense breast tissue. The life-saving implications are stunning. So why haven’t we heard of it? Rhodes reveals the web of politics that have kept the test from going mainstream. (Filmed at TEDWomen 2010.)

Daniel Kraft: Medicine’s future? There’s an app for that
Physician and scientist Daniel Kraft — the future-med chair at Singularity University — doesn’t understand why the grand majority of a person’s healthcare costs are spent in the last 20 years of life. In this talk, Kraft walks us through the myriad iPhone apps that exist for diagnostic testing, from apps that detect STDs to ones that monitor blood sugar, and imagines what we could see in coming years. (Filmed at TEDxMaastricht.)

Joe DeRisi: Solving medical mysteries
There are zillions of viruses out there, with large numbers often categorized under the same name. In this talk, biochemist Joe DeRisi explains his lab’s work making DNA chips, which allow for the easy diagnosis of viruses — down to specific strains — by their DNA sequence. Not only will this help in the diagnosis of viruses, but also in treatment.  (Filmed at TED2006.)

Jack Andraka: A test for pancreatic cancer
15-year-old Jack Andraka was shocked by the statistic that only 5.5% of those diagnosed with pancreatic cancer survive past five years, because diagnosis usually happens so late. In this spirited talk, Jack explains his cheap and easy test for the early detection of pancreatic cancer, using carbon nanotubes. (Read our Q&A with Jack, whose talk was filmed at TED@NewYork, part of the TED2013 Talent Search.)

Bonus: Read about Brittany Wagner, winner of Google’s second annual International Science Fair, who created an artificial neural network that allows a computer to diagnose breast cancer, making the least invasive breast cancer test — fine needle aspirates — far, far more accurate.

In today’s TEDTalk, Max Little tells you about a test — now in trials — that could help diagnose Parkinson’s disease with a 30-second phone call. But the first challenge: He and his team needed to record 10,000 voices from across the world to make sure that the technique is accurate and scientifically valid.

In late June, when he announced the trial, 600 people called in during the first eight hours. Now, just over a month later, they’ve got 75% of the phone calls they need for their data set. And you can help: Visit ParkinsonsVoice.org to be part of the data for the test trial.

How does the test work? Parkinson’s affects muscle function, including the muscles in the vocal cords, and people with the disease generally display characteristic properties in the voice. Little’s insight: An algorithm should be able to detect those vocal properties over the phone.

The first application is obvious — enabling cheaper, earlier, more accessible diagnosis of an incurable disease that affects 6.3 million people worldwide.

Reached by email, Little explained, “Among poorer communities who have little in the way of health care resources, the awareness of Parkinson’s is low, and many people go undiagnosed and untreated.” Using a call-in system they “can very rapidly detect those who appear to be at high risk and concentrate scarce clinical resources on those picked out by voice-based algorithms.”

In the future, Little and his team are exploring further applications.

“Another tantalizing possibility is that vocal disturbances might be among the first symptoms of the disease,” said Little, quickly adding that the evidence for that is currently low, but that the creation of a massive database allows them to explore the question.

Finally, could a call-in system be used for other diseases?

It’s possible, says Little, who is putting together a research program to test patients with ALS (Lou Gehrig’s disease, or motor neuron disease), MS (multiple sclerosis), MSA (multiple system atrophy), ET (essential tremor) and other diseases.

As always with research, the team won’t know if the program works until they try it, but it’s an exciting new direction for diagnosis.