Search form

Donate Today


Ultra Low-Cost Medical Diagnostics in a Tiny Box | Paul Yager

Using paper-based technologies, Paul Yager and his team are developing diagnostic kits to test for many common and dangerous health conditions.

Full talk transcript

I’m here to tell you some things about how we would like to change medicine. But first I want to say what the idea was and how I try to explain things to people.

I've been working for about 20 years trying to shrink the hardware that allows us to detect pathogens — bacteria, viruses — to make it as small as possible.  In about 2008, my colleagues Elaine Fu, Barry Lutz, and I said, “Wait a minute, we can switch and make things in paper and throw away the instruments completely and move things to the home."  But then you've got this idea and you're in a university and you've got to raise some money, and you have to tell people a story.  And my way of telling stories involves pictures, and often movies.   So I ran to my graphics program and we made this picture of  something we thought we could do.

So imagine you had a swab and you've had it in someone's nose, so you can see that it’s green.  And then you stick it into a tube and you twirl it into the tube to get the stuff off the swab.  You push it to activate it and then liquids start moving around and they go onto paper, they run up to the paper, and minutes later you've got dots on a piece of paper.   You pick up your cell phone over there and you take a picture  and that picture sends  your medical data any place on the planet.  That's the core idea: If you forget everything else and you go to sleep now. it’s fine.

I've been working for about 20 years trying to shrink the hardware that allows us to detect pathogens.

But I want to tell you two little things about me and how I got to be the way I am.  And my problem is that I didn't grow up in this planet, really — I grew up in the future.  Specifically,  I grew up in New York City around the corner from the Leroy Street branch of the New York City Public Library, which had in this beautiful nineteenth century building a whole bunch of science fiction books.  You know, everybody from Jules Verne through Larry Niven, Lasni, Bruner, Delaney, wonderful writers who told stories.  Convincing stories about times in the future, where a lot of  the problems we have didn’t exist. 

In the end, when I was 12 years old, this thing called Star Trek happened.   And, you know, with a 12-year-old's hormones it's a pretty exciting moment in a kid's life, and  there it was on TV like it was real, right?  So I've been splitting infinitives pretty much continuously for about the last, 40, 50 years.   One of the problems, however, of having that in your head is this cognitive dissonance problem. Which is that you've grown up in one place and you live here, this weird place.  You know a lot of people put the bumper sticker on their cars and just, you know, said, "Drop out, this is just not for me."

I've been fortunate enough to be able to go through an educational process where I  now hang out with bio-engineers who are wonderful, bright people.  What stuns them is that my bookshelf is mostly my science fiction collection — to let them know that I'm actually up there, not down on this planet, and we’re  just trying to get from here to there.  So let me tell you what we're going to try to do in the near future to help us get there.

We all know that when I was a kid and went to the computing center and it was a big centralized place and you brought data cards, and today we've all got these things on our hips, which have far more computing power than the best computing centers back in the '70s.   It's an amazing revolution and it's changed the world.

What we haven't quite done is fix the hospital.  The hospital is still the equivalent to the computing center of the 1970s where we go and we do everything there.  What we would like to do is use that sort of hardware and bring the medical testing to wherever you are when you are there.  We think that could help a lot. 

Now, what is it you have to do in order to cure sick people?  Well, the first thing you have to do is find out what's wrong with them, and you can do that in a major hospital, or ideally you could do so when the patient’s at home.  That home could be here or a hundred miles north of Monrovia — these are seemingly different problems, but we want them to be done the same way; we think the same technology could be used in both places, and it could be quite liberating, and yes, we're going to use cell phone technology as part of the tool.

So, if you want to treat someone with an infectious disease, the first thing you've really got to know is what is wrong with them.   And today and for a long time, you first look at symptoms.  Coughs, spots, temperature. But those things are shared by lot of different diseases, so you really can't do that.  And what we do today is we try to then look at history and environment, and part of the problem is if you live in the tropics and you have a fever, the first thing they will say is, "Well, you've got malaria," and there are often people treated for malaria when that isn’t what they have.  They have something else or multiple things, and we need to know that.

So the state of the art here in Seattle is you look for the presence of chemical fingerprints of the suspected disease.  And it could be antibody responses, things your body produces against these things, or it could be looking for bits of the pathogen itself, a virus or bacteria  This is a flu virus, we could be looking at the proteins, or we could be looking at the DNA or RNA inside, and either one works. We’re actually pursuing both of those.

What we're trying to do: If we can’t bring the person to the lab, bring the lab to the person.

What we're addressing is that today if you go down the street here to the UW Hospital, they’re going to have huge machines that work incredibly efficiently and very effectively; however, you and I would not want to have them at home — I don’t know how big your home is, but mine won’t fit that.  And if your home or your laboratory is not like the University, but it's like this developing world laboratory, a little ways out of Delhi, you don't have the infrastructure to do this under any circumstances, let alone the power, the way to support it, the service contract.  So what we're trying to do: If we can’t bring the person to the lab, bring the lab to the person.

In the U.S. we’re concerned about the AMA and the FDA, and we want to keep the doctor in the loop.   We don't want to say, “Hey I'm going to become a doctor even though I don't have a degree and give me the tools to do that.”  We’re saying, how can the doctor-patient relationship be extended back to the home or wherever the patient is?  And that, I think, is a really potentially successful approach.

Now, if your home happens to be someplace like this, and this is on the road from Gwalior, India to the airport a couple years back, you really don't have any infrastructure.  There's no power, no running water; there's nothing, so you really want this technology to be very simple, very robust.  And there's a roadmap for how to do this: The World Health Organization has what they call their ASSURED criteria: It's an acronym for Affordable, Sensitive, Specific — those are medical terms about the test — but User-friendly, we all understand.  So something new kids could use.  Rapid and robust, meaning I could store it in the glove compartment of my car or in one of those homes in Gwalior.  Or it should be equipment-free — that means no power cord, ideally something you could carry around with you.

The trickiest part of course is “D" — it’s the delivered,  and that's the part where a faculty member has to become an entrepreneur, but that's a whole other lecture, and we’ll go to that one some other day.  Next year, maybe.  So, we have this idea which is let's focus on the user experience, and in my research group, I’m constantly beating people back and saying no.  

No user activity, no pipetting, no big robots, none of that stuff that you’d use in a regular laboratory has happened. We shrunk all that and put it in the box.

Fortunately, Lisa Lafleur and Samantha Burns did this movie, which is what we thought the user experience should be.  So you take something out of a pouch in which it has been stored for up to a year at room temperature, and in that pouch there will be a small box, and maybe in this case a nasal swab.  So she's going to take out a nasal swab and she's going to pretend to stick it in her nose —no noses were harmed in the making of this movie — and then twirl it around a few times, and then after she's twirled it a few times, she's done.  And she can then at this point get up, walk away, do stuff for a bit. No user activity, no pipetting, no big robots, none of that stuff that you’d use in a regular laboratory has happened. We shrunk all that and put it in the box.

Then, when the timer goes off, your cell phone dings, you go take a picture with your cell phone of the output of the device — in our case we’re using a picture of an image that comes at the end of the thing, at which point the data is present on your phone and can go anywhere on the planet.  You can analyze it on your phone, you could send it to your doctor's office, you could send it to your health care provider anywhere.  That's the model.

What we've put inside the box is paper-based technology, and you've probably seen certain types of paper-based technology if you've ever had a pregnancy test in your hand.  That's what we're talking about; it's very simple, it doesn't use any pumps, it uses wicking of water through the paper to drive the fluids around. They’re called lateral flow testers, and they’re good, they’re inexpensive, and they're just however a little bit dumb.  They don't do complicated things and they don't do all the complicated hand-pipetting step. 

So, we said, what if we can actually do some of that complexity stuff, and we call it programming in paper.  It's not really computer programming, it’s setting up the paper so it's going to do interesting things.  In this case, we put food dye in pads; the rising  water  wets them,  the water level drops as the water gets sucked into those things; there's a wick on the left side that just sucks extra water in and increases the humidity, and if you watch over here, imagine that's where we're going to do some chemical testing, where we're going to have some captured molecules and grab proteins and nucleic acids.

So this is just what happens — this is approximately real-time: Notice the liquids are dropping as they get sucked into the paper. First you have yellow liquid going across that spot, then blue liquid, now the red liquid is going to come up;  the other ones have been shut off because they’ve fallen out of the water.   And then eventually the rest of it will run out,  and you’ll actually get white stuff. So now run four different liquids across the spot with no human hands whatsoever. 

That movie was enough to get us a relatively large amount of federal funding from the National Institutes of Health — a good story and a good picture and something that is simple enough that you could imagine a commercial product is really what it takes these days.  So that's what we're pursuing in our lab.

One of the things we're doing is we're looking at proteins.   And this is the flu virus — again, we're looking at the proteins on the outside of them.  We’re fortunate to be working with David Baker, who is a fabulous protein designer and he's been working on proteins that stick to the flu virus that are smaller and much cheaper to make than antibodies.  And then we’re coupling those to paper and using those to grab hold of the whole virus or pieces of the virus, in order to detect it.   And that's great — there are lots of copies of this protein so it's relatively easy to do and relatively fast.  

The idea is that in one version of this, the entire thing could be burned, and there would be no waste whatsoever — it would go to ash.

We've also — and watch the switching here — made a lot of toys with paper that allow us to do validating and the kinds of things robots would do. In this case the liquid is turning on another liquid, and they’re actuators, and the actuators are actually bits of sponge.  And all of this stuff has no metal parts, no electronics or things of that sort. The idea is that in one version of this the entire thing could be burned and there would be no waste whatsoever — it would go to ash.

Now the bigger challenge of the two was to go after nucleic acids, because today if you want to find nucleic acids you want a big instrument.  And they’re really similar in many ways to the protein-based approach in terms of what the box would look like, but nucleic acids can be detected down to single copy, and there's no way we can do that with proteins at this point in our system.  This gives you enormous sensitivity, so you can work with people who just have a very small amount of bacteria or virus in their sample.  You get a higher part count because it’s a little more complicated, and that’s why we get a lot of money to work on the problem.   

So we have this device we called the MAD NAAT — I envision a small angry gnat when I think of this, but you've got to have an acronym if you’re going to go for DARPA money.  Swab transfer is what you start with,  then you lyse things — that means breaking open the cells,  busting them and getting out the nucleic acid.  It could be as few as a single copy, we’d  be happy with 10 to the 4th. We then have to make lots of copies.  So you amplify, and we have a method that we use for amplifying these things and making lots and lots of copies — it takes 20 to 30 minutes to do it, but when you're done you can have 10 to the 12th copies; so many copies that you can use the same lateral flow technology used for the pregnancy test and see them easily on a piece of paper, and hence take the picture with your camera.

So this is a demo we did back in July to prove that we weren't crazy. This is Lisa La Fleur, the director of the previous movie, actually inserting the swab — she’s sort of chief designer for these things. And she's putting in the swab, running it, and then the timer will start going a little faster as this thing goes through.  So you swab for 10 seconds, start the device when the lid is closed.  This thing is sitting on a table there are no other moving parts outside this thing all the parts are inside the box, and you open valves, move things in the application zones, and a variety of things happen.  Application for 30 minutes in this case to make sure we get lots of copies. Finally more valves open up, and at the end of it when you're done then you actually run flow and you can probably just barely see it's a little bit of liquid going up into that white window next to the QR code, which is actually the detection event.  At which point, out comes the cell phone — we’re actually putting the cell phone on top of the device. We've got a piece of code running and it'll take a picture, and the picture shows up over there, and afterwards you can run this through a piece of software that we loaded into the phones, so the phone can actually analyze that you had a sample there and how much.

What would you do with this yourselves?  Why is this relevant — the "So what?" question?  Well, you can imagine if you wonder how long it takes to get an appointment with your primary care physician, you could do the test now rather than waiting for that.   You may have been on a trip, and the question is what did you bring back from that trip, whether it was West Africa or someplace else.   Employers don't really want you at work when you're sick, because you're going to bring it to everybody else and shut down the production line,  so that's another issue that can be a big deal.  Sometimes you don't know whether what you've got is a serious thing or something trivial, and you can test it at home.

There are lots of things that we don't necessarily want to tell everybody about, but you want to know, and you want to know as soon as possible.

You might be able to pick stuff up from the pharmacy to do that, but if you are thinking maybe you need to go to the ER to have it checked out and it’s 50 miles away and it’s at night, and there's ice on the roads, that could be the problem.  This is a big one.  This is one that there are lots of things that we don't necessarily want to tell everybody about, but you want to know, and you want to know as soon as possible. So this is one actually that might be the primary market.

If you're on antibiotics for that or something else and the question is are you really getting better fast enough, was it the right antibiotic, did the test get it right, you can check things and repeatedly test to make sure that things are getting better.  Sometimes your kid may have a problem and the question is do you send them to school or not, and you could say, "Nope, it's not this" or "It's over."

Finally, one that I think is really important here and elsewhere is the HIV question. You might want to know whether that person you had sex with was HIV-positive. And if the answer could be proven to be yes because you had a swab test within the next 72 hours, you can get post-exposure prophylaxis and not spend a life on anti-retrovirals.  So that's actually a broader, interesting one that has some significance here in this country, but a huge significance the developing world.

So, lots of benefits for more testing if you did it right, assuming you didn't have a lot of false positives and people medicating themselves for things they shouldn't be doing.  

How do we do this?  Well, one possibility is your doctor prescribes the test and you take it home, you go stop off at the pharmacy.  Alternately, you go direct to the pharmacy and say, “I want one of those, one of those and one of these.”  And maybe the pharmacist  suggests, “Ah, you have young kids in the home, get one of these too.”

Finally there's the Amazon route.   A lot of people are buying their meds at Amazon these days — there happens to be an HIV test that you can buy online at Amazon, not an ad for Amazon, but there are lots of issues about the business model.  And this is going to take lawyers, ethicists, and a lot of doctors dealing with how we're going to do this well.

Now, finall,y and it's hard not to think about this, but what about Ebola?  And there’s a beautiful online picture of the Ebola virus.  A beautiful and incredibly deadly pathogen.  Could we use this cheap, small, fast technology to actually address something like an outbreak?   Whether it's here or someplace else?   And the answer is, “Oh, my God, yes!”  What we're doing is we’ve brought relatively sophisticated medical testing to Africa, but we bring it with hardware, laboratory equipment, and people have to gown up in all the personal protective equipment to allow them to work in the presence of the pathogen.

This person over here from the Navy is simply eliminating the deadliest of the blood sample so that it could be taken to the next person, still wearing protective gear to run the tests.  And that sample often has to go a long distance.   And the problem is in places like the developing world, often the roads are terrible, so you can't get the sample from point A to point B, or the patient to go there.  We have a really serious problem with this, and the result is you take people you think might have Ebola and you warehouse them with people who may also have a fever but have malaria, or had the flu.  And all those people get the chance of catching Ebola. So you actually kill more people because you're testing poorly.  It's a really big issue for outbreaks like this.

We think we could address that with this kind of technology by sticking your finger in a box, getting a blood sample, getting a reading, either 20 minutes for protein or 15 minutes total for nucleic acid, eliminate the warehousing and go directly from “Gee, I think this person sick” to  “You have Ebola, and you don't.” That would be relatively straightforward to do.  The problem we have is we're not actually ready to get it out there.  And when we talk to people at the Gates Foundation, they said, “Can you get it out there in three months?” We took a deep breath and said we're not quite ready yet. So this is a really important and interesting problem for us, and we would like to get this done as quickly as we could.

If you pull back a little bit, you could use it for things like farmers. Farmers could check the health of their animals or their crops right where they are, without having to go through long-term testing.  A really big thing in the developing world where there’s virtually no infrastructure for this.  Finally you could also screen individuals for whether they should go on to that boat or onto that plane, or cross that border, or enter that hospital.

You can pull it right back to the fact that we're all in the same boat, and we all have the same issues.

We think this is increasing access to healthcare, this is a central feature of what I care about in the laboratory, whether it iss here in the United States or in the developing world. In this real 21st century we’re in, we do it partially because we love the technology and we love proving that we understand it — we’re engineers by avocation.  On the other hand, we also want to do it because we think that distribution is fair, it's the right thing to do for people. It’s something that we need to do to help the world.  And finally, it makes us all healthier and safer.  You can pull it right back to the fact that we're all in the same boat, and we all have the same issues.

I want to thank my group — they do wonderful work. This is all done by people in my research group — only a fraction of the people are shown in the picture.  It's a great group. Thank you for your attention, and finally, live long and prosper.

Paul Yager


Speaker Bio

Paul Yager received his A.B. in Biochemistry from Princeton in 1975, and a Ph.D. in Chemistry from the University of Oregon in 1980, specializing in vibrational spectroscopy of biomolecules. After an NRC Fellowship at the Naval Research Laboratory from 1980–1982, he joined the NRL staff as a research chemist. In 1987, he moved to Seattle and the Center (now Department) of Bioengineering at the University of Washington as an Associate Professor, advancing to Professor in 1995, and served as Chair of the department from 2007–2013. Read more