you know how smoke detectors work, right? So imagine if we had something like a smoke detector but for diseases. Instead of testing the air for smoke particles, this alarm would look for tiny traces of an odor profile associated with being infected by a virus. For example, all day long, 24/7, but the second the odor profile reached a certain concentration, the alarm would start blaring immediately. You would know that you were sick, that you should see a doctor, or at the very least, that you should isolate yourself. Ideally, this mechanical nose would sniff out a virus before you showed any symptoms or spread it to anyone else. Because if we learned anything from the COVID-19 pandemic, it's that testing matters, and the sooner we know you're infected, the better. A mechanical nose like this could stop pandemics in their tracks, and not just for humans, but for animals too. Take the high-pathogen avian influenza virus that left 53 million chickens dead in the US alone and caused egg prices to skyrocket. It's easy to test chickens for

Aven influenza virus, but there are tens of thousands of chickens at any given farm, so which ones do you test? Wild waterfowl are basically asymptomatic and can transfer the virus just by landing on the farm, and if you do have a sick chicken, by the time you get the results back, it's probably too late; the whole flock is sick. What we need is a mechanical nose testing the air 24/7 nonstop. That is the ultimate end goal of my research, well, not just my research but a whole field of biologists, chemists, mechanical engineers, and physicists studying odor and scent detection. But building a mechanical nose is really hard. The mammalian olfactory system is incredibly complex, detecting thousands of odors across combinations of millions of olfactory receptors. We have a lot to learn before we can design this disease-detecting mechanical nose, but I know how we're going to get there, using everything we're learning about how dogs smell disease. In wildlife, for decades, we've known that dogs can smell cancer, high and low blood

sugar seizures and other non-infectious diseases. They're not just smelling the disease per se, but the metabolic results of being infected. If you have diabetes, for example, your sweat, blood, and urine smells differently when you have high blood sugar. That's why doctors use the stiff or even taste their patients' urine to test for diabetes. Thank God for modern science, right? But you can train a dog to detect high blood sugar the same way, by rewarding them with their favorite treat or toy anytime they indicate they smell the correct sample. They quickly learn to pick up the odor that will get them the reward. Of course, this is very useful information for an individual with a specific disease, but what about on a larger scale? Can we train animals to sniff out infectious diseases and stop their spread? I believe the answer is yes. My obsession with all things animal started when I was a little kid, but for a long time, it was just a hobby. My real passion was being the frontman and lead vocalist of a band called Fatal.

Attraction that toured around the New York tri-state area. In fact, this is the first time I've been on a stage since 1991, after my best friend and drummer died in a motorcycle accident. I can tell you that being here on stage AG is nothing but surreal, and if I look behind me, it's not because I'm looking at the slides, but I'm looking for that all-tooth grin of my brother behind the red drum kit. The end of his life was the end of my music career. I'd have to find a new passion. After a long, deep depression, I reemerged as a scientist and continued on as a dog decoy. You've probably seen us on TV before, the guys wearing the big padded suits, getting repeatedly pummeled by police and military dogs in training. Yeah, sounds like the perfect career for a big guy like me who just went through a traumatic life event, getting attacked repeatedly by dogs. But I loved it. I spent all day working with dogs, and around the same time, I kept seeing MCA.

all over the news, hospitals developed extensive cleaning protocols, and still, patients got MRSA. I thought surely dogs must be able to smell MRSA, so my initial naive-as-hell idea was to bring dogs around hospitals to detect specific areas where MRSA colonies were growing, kind of like a drug dog at the airport. I just missed the small detail that no surgeon was going to let me let a dog into a room where they'd be exposing patients with large open incisions. Not that I would let that kind of detail stop me, so off I went to Monell Chemical Senses Center, the leading research institute for chemosensory sciences, where I met Dr. Bruce Kimball, an analytical chemist, and the late Dr. Kunio Yamazaki, who had successfully shown that mice can detect avian influenza in duck fecal samples. While I was working on another post-doctoral project for Bruce, they were looking to do the next round of avian influenza virus studies, this time using a different species as the

biodetector. My first thought was giant Gambian pouch rats, the hero rats you've seen detecting landmines in Africa. But Gambian pouch rats are a highly invasive species in the US, and there was no way that they were going to let me import any. And then I read a research paper about how ferrets interact with humans like dogs do. Yeah, ferrets. Ferrets can hold your gaze, and if you point your finger, they'll follow it to see what you're pointing at. So, we ordered eight young ferrets and got to work to figure out whether ferrets can accurately detect avian influenza virus. We trained them using two different chemical samples: one with an odor similar to an infected duck and one with the odor similar to a non-infected duck. If the ferret alerted by scratching on the correct box, we clicked the clicker to mark the behavior and then gave them a tasty reward with a syringe. Very quickly, ferrets learned to identify the sick duck chemical samples with 90%.

accuracy and that's great for in the lab, but if we wanted to use ferrets to detect Avan influenza in real life, they wouldn't be sniffing pure chemical samples but real duck poop. And poop is obviously more socially salient, aromatically complex, a lot of competing compelling odors in a very tiny little package. We still had the fecal samples from the mouse experiment that had been sitting in a minus 80 freezer for the past 6 years. We put five boxes in a row, four with negative samples and one with a positive sample, in a different order each time so the ferrets couldn't predict which box contained the positive sample. I remember watching each ferret approach the positive box, smell it for a bit, and then start to move on to the next box, and my heart would sink. But before I could even sigh, every ferret spun around and slammed that paw down on the box holding the positive sample. The ferrets were right 79% of the time, and they could detect the odor change a day

Before and at least one day after traditional medical testing, the next step was to transition to dogs. Because as much as I love ferrets, they just aren't suited to high-traffic places. I mean, imagine seeing a ferret on a little leash at the airport, on an egg farm, or at a highway stop sniffing livestock, crossing state lines. It would be hilarious, it would be adorable, but it would also be imperfect because ferrets are prone to distraction and less motivated than dogs, especially in chaotic environments. So just like the ferrets, we trained up a group of dogs, and they could detect avian influenza virus. It didn't really matter which breed of dog, so long as they were physically capable of doing the job and very motivated. These dogs could work on chicken farms, do a quick scan of the chickens every day, and sniff out bird flu before it spreads to thousands of other birds. Hunters coming off the lake with their harvest could use these dogs to tell them which ducks were infected and which weren't. But what about other

diseases. Recently, my team has been working on chronic wasting disease in white-tailed deer. Chronic wasting disease, or CWD, is a fatal, highly transmissible disease that affects cervids like deer, elk, and moose. If you never want to sleep again, feel free to look up prion disease after my talk. Basically, prions cause proteins to misfold in the body, which then accumulate in the nervous system, resulting in spongiform encephalopathy, which is really just a fancy phrase for huge holes in your brain, literally the stuff of nightmares. CWD spreads from infected deer through bodily fluids and tissues, so it's not just deer-to-deer contact that's a problem, but the environment as well. A healthy deer can get CWD by drinking from a water source where a sick deer had drank from days or even weeks before. So, using the same techniques we did with the avian influenza, we trained dogs to detect fecal samples from CWD-infected deer. The dogs were 94% accurate in the lab and 80% accurate in the field.

even when they were distracted by moving tractors and combines. That means we could use lab-trained dogs as a rapid non-invasive tested hunter check-in stations to prevent hunters from eating CWD-infected deer meat. And we could deploy them to live deer farms to alert farmers of infected deer in the herd. This research is groundbreaking. Our results suggest that not only can dogs, ferrets, and rats accurately detect the odor profiles of infectious diseases, but in some cases, they can do it sooner than a traditional medical test. And like I said earlier, timing is everything. The faster you detect the disease, the sooner you can eradicate it. In the long term, the more we learn about odor detection, the faster we get to a mechanical nose. And though it may be many years before we have both a smoke and a virus detector in our homes, dogs, ferrets, and rats can be deployed to labs worldwide in a matter of months. So the next time you get a rapid non-invasive pathogen test, it might not be a human with a pipette, but a little

ferret on a leash doing all the work. Animals really are our best friends. It's not just that they love us unconditionally; it's that they can truly save our lives. Thank you. [Applause]

Can Dogs Detect the Next Pandemic Before It Begins? | Glen J. Golden | TED

The TED Talks channel features the best talks and performances from the TED Conference, where the world's leading thinkers and doers give the talk of their lives in 18 minutes (or less). Look for talks on Technology, Entertainment and Design -- plus science, business, global issues, the arts and more. You're welcome to link to or embed these videos, forward them to others and share these ideas with people you know.

TED's videos may be used for non-commercial purposes under a Creative Commons License, Attribution–Non Commercial–No Derivatives (or the CC BY – NC – ND 4.0 International) and in accordance with our TED Talks Usage Policy: https://www.ted.com/about/our-organization/our-policies-terms/ted-talks-usage-policy. For more information on using TED for commercial purposes (e.g. employee learning, in a film or online course), please submit a Media Request at https://media-requests.ted.com