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

00:13:18
https://www.youtube.com/watch?v=wLvGABoTV-s

Summary

TLDRThe concept of a 'mechanical nose' for detecting diseases is discussed, inspired by how smoke detectors work. This device would detect diseases by analyzing specific odor profiles associated with infections, potentially allowing for 24/7 monitoring and immediate notification upon detection at a certain threshold. Lessons from the COVID-19 pandemic emphasize the importance of early detection. Current research is exploring how animals like dogs and ferrets can detect diseases through their sense of smell. Dogs have shown the ability to detect cancer, blood sugar levels, and seizures. Ferrets have been trained to identify avian influenza in ducks with high accuracy. The challenge lies in replicating the complexity of the mammalian olfactory system in a mechanical form, as it involves detecting thousands of odors across millions of olfactory receptors. The research aims to leverage what is known about animals' extraordinary sense of smell, with studies showing dogs and ferrets can detect diseases earlier than traditional methods. This might also be applicable at a large scale for stopping the spread of infectious diseases among humans and animals, like avian influenza and chronic wasting disease in deer. The end goal is to create devices that can rapidly and non-invasively detect diseases through odor, thus potentially preventing future pandemics.

Takeaways

  • 🔍 The idea is to create a 'mechanical nose' for detecting diseases using odor detection.
  • 🐾 Animals like dogs and ferrets are being studied for their ability to detect diseases by smell.
  • 🦠 Early detection of diseases through unique odor profiles could prevent pandemics.
  • 🐕 Dogs have shown they can detect diseases such as cancer through smell.
  • 🐾 Ferrets have been used to identify avian influenza in lab tests.
  • 🤖 Creating a mechanical nose is challenging due to the complexity of the olfactory system.
  • 🚫 Early detection can stop the spread of diseases among humans and animals.
  • 👃 Research in odor detection could lead to rapid, non-invasive disease tests.
  • 🔬 The project involves biologists, chemists, mechanical engineers, and physicists.
  • 📉 Timing is crucial in disease detection and prevention.

Timeline

  • 00:00:00 - 00:05:00

    The speaker introduces an innovative concept of a mechanical nose that detects diseases by identifying specific odor profiles, akin to a smoke detector identifying smoke particles. Such a device would provide an early alert for virus infections, potentially preventing the spread by promoting early isolation or medical intervention. This idea stems from observations during the COVID-19 pandemic, emphasizing the critical need for timely infection detection. The speaker mentions the vast implications for both humans and animals, citing the avian influenza virus, which devastated poultry populations. Although testing exists, it's impractical due to the scale of farms and the asymptomatic nature of wild carriers. A continuous air monitoring system could revolutionize how quickly and efficiently outbreaks are managed, making it a research focal point for many scientists across various fields.

  • 00:05:00 - 00:13:18

    The speaker shares personal anecdotes, transitioning from a music career to becoming a scientist working with animals to test for diseases. Starting with a failed idea of using dogs in hospitals to detect infections like MRSA, the speaker collaborated with research institutes to explore alternative methods. Ferrets were trained to distinguish between samples of infected and non-infected cases with high accuracy, revealing their potential as biodetectors despite challenges in practical application due to their distractibility. Dogs, however, proved effective and adaptable across various environments, successfully identifying avian influenza and chronic wasting disease in a wide range of settings. These developments highlight the powerful role animals can play in disease detection, achieving a level of accuracy and speed that could surpass traditional medical tests, thereby paving the way towards creating a mechanical solution.

Mind Map

Video Q&A

  • What is a mechanical nose?

    A conceptual device that detects diseases by identifying specific odor profiles associated with infections.

  • How could a mechanical nose help in disease prevention?

    It could provide early detection of diseases, preventing their spread by alerting individuals before symptoms appear.

  • Can animals detect diseases?

    Yes, dogs and ferrets, among other animals, can detect diseases by sensing changes in odor profiles.

  • Why is early disease detection important?

    Early detection helps prevent the spread of diseases and allows for quicker medical intervention.

  • Have animals been used in disease detection studies?

    Yes, studies have shown animals like ferrets and dogs accurately detecting avian influenza and other diseases.

  • What challenges exist in creating a mechanical nose?

    The human olfactory system is complex and difficult to replicate; understanding odor detection fully remains a challenge.

  • Why use ferrets in disease detection research?

    Ferrets can hold human gaze and follow pointers, making them more engaging for this research.

  • What diseases have been successfully detected by dogs or other animals?

    Diseases like avian influenza and chronic wasting disease in deer have been detected using such methods.

  • How effective are animals compared to traditional tests in disease detection?

    In some cases, animals have detected signs of disease sooner than traditional testing methods.

View more video summaries

Get instant access to free YouTube video summaries powered by AI!
Subtitles
en
Auto Scroll:
  • 00:00:04
    you know how smoke detectors work
  • 00:00:07
    right so imagine if we had something
  • 00:00:09
    like a smoke detector but for
  • 00:00:12
    diseases instead of testing the air for
  • 00:00:15
    smoke
  • 00:00:16
    particles this alarm would look for tiny
  • 00:00:18
    traces of an odor profile associated
  • 00:00:21
    with being infected by a virus for
  • 00:00:24
    example all day long
  • 00:00:27
    24/7 but the second Theodor profile
  • 00:00:30
    reached a certain concentration the
  • 00:00:33
    alarm would start
  • 00:00:34
    blaring immediately you would know that
  • 00:00:38
    you were sick that you should see a
  • 00:00:40
    doctor or at the very least that you
  • 00:00:42
    should isolate yourself ideally this
  • 00:00:45
    mechanical nose would sniff out a virus
  • 00:00:48
    before you showed any symptoms or spread
  • 00:00:51
    it to anyone else because if we learned
  • 00:00:54
    anything from the covid-19 pandemic it's
  • 00:00:57
    that testing matters and the sooner we
  • 00:01:00
    know you're infected the better a
  • 00:01:04
    mechanical nose like this could stop
  • 00:01:06
    pandemics in their tracks and not just
  • 00:01:09
    for humans but for animals too take the
  • 00:01:12
    high pathogen Avan influenza virus that
  • 00:01:16
    left 53 million chickens dead in the us
  • 00:01:19
    alone and caused egg prices to
  • 00:01:23
    Skyrocket it's easy to test chickens for
  • 00:01:26
    Aven influenza virus but there are tens
  • 00:01:28
    of thousands of chickens at any given
  • 00:01:31
    farm so which ones do you
  • 00:01:35
    test Wild Water foul are basically
  • 00:01:38
    asymptomatic and can transfer the virus
  • 00:01:41
    just by landing on the farm and if you
  • 00:01:44
    do have a sick chicken by the time you
  • 00:01:47
    get the results back it's probably too
  • 00:01:49
    late the whole flock is sick what we
  • 00:01:53
    need is a mechanical nose testing the
  • 00:01:55
    air
  • 00:01:56
    247 nonstop
  • 00:02:00
    that is the ultimate end goal of my
  • 00:02:02
    research well not just my research but a
  • 00:02:06
    whole field of biologists chemists
  • 00:02:09
    mechanical engineers and physicists
  • 00:02:11
    studying odor and scent
  • 00:02:14
    detection but building a mechanical nose
  • 00:02:17
    is really hard the Maman old factory
  • 00:02:20
    system is incredibly
  • 00:02:23
    complex detecting thousands of odors
  • 00:02:26
    across combinations of millions of ol
  • 00:02:29
    Factory recept C we have a lot to learn
  • 00:02:32
    before we can design this disease
  • 00:02:34
    detecting mechanical nose but I know how
  • 00:02:37
    we're going to get there using
  • 00:02:39
    everything we're learning about how dogs
  • 00:02:41
    smell disease in
  • 00:02:44
    Wildlife for decades we've known that
  • 00:02:46
    dogs can smell cancer high and low blood
  • 00:02:49
    sugar seizures and other non-infectious
  • 00:02:52
    diseases they're not just smelling the
  • 00:02:55
    disease per se but the metabolic results
  • 00:02:58
    of being infected if you have diabetes
  • 00:03:01
    for example your sweat blood and urine
  • 00:03:04
    smells
  • 00:03:06
    differently when you have high blood
  • 00:03:08
    sugar that's why doctors use the stiff
  • 00:03:11
    or even taste their patients urine to
  • 00:03:15
    test for
  • 00:03:17
    diabetes thank God for modern science
  • 00:03:20
    right but you can train a dog to detect
  • 00:03:23
    high blood sugar the same way by
  • 00:03:25
    rewarding them with their favorite treat
  • 00:03:27
    or toy anytime indicate they smell the
  • 00:03:31
    correct
  • 00:03:32
    sample they quickly learn to pick up the
  • 00:03:34
    odor that will get them the
  • 00:03:37
    reward of course this is very useful
  • 00:03:39
    information for an individual with a
  • 00:03:42
    specific disease but what about on a
  • 00:03:44
    larger scale can we train animals to
  • 00:03:48
    sniff out infectious diseases and stop
  • 00:03:51
    their spread I believe the answer is
  • 00:03:54
    yes my obsession with all things animal
  • 00:03:57
    started when I was a little kid but for
  • 00:04:00
    a long time it was just a hobby my real
  • 00:04:03
    passion was being the frontman and lead
  • 00:04:05
    vocalist of a band called Fatal
  • 00:04:08
    Attraction that toured around the New
  • 00:04:10
    York tri-state
  • 00:04:12
    area in fact this is the first time I've
  • 00:04:16
    been on a stage since
  • 00:04:21
    1991 after my best friend and drummer
  • 00:04:24
    died in a motorcycle
  • 00:04:26
    accident I can tell you that being here
  • 00:04:29
    on stage AG is nothing but
  • 00:04:32
    surreal and if I look behind me it's not
  • 00:04:34
    because I'm looking at the slides but
  • 00:04:36
    I'm looking for that all tooth grin of
  • 00:04:38
    my brother behind the red drum
  • 00:04:41
    kit the end of his life was the end of
  • 00:04:44
    my music career I'd have to find a new
  • 00:04:48
    passion after a long deep depression I
  • 00:04:51
    reemerged as a scientist and continued
  • 00:04:54
    on as a dog
  • 00:04:55
    decoy you've probably seen us on TV
  • 00:04:58
    before the guys wearing the big padded
  • 00:05:01
    suits getting repeatedly pummeled by
  • 00:05:04
    police and military dogs and
  • 00:05:06
    training yeah sounds like the perfect
  • 00:05:09
    career for a big guy like me who just
  • 00:05:11
    went through a traumatic life event
  • 00:05:13
    getting attacked repeatedly by
  • 00:05:18
    dogs but I loved
  • 00:05:21
    it I spent all day working with dogs and
  • 00:05:25
    around the same time I kept seeing MCA
  • 00:05:28
    all over the news
  • 00:05:30
    hospitals developed extensive cleaning
  • 00:05:33
    protocols and still patients got Mara I
  • 00:05:36
    thought surely dogs must be able to
  • 00:05:38
    smell Mera so my initial naive as hell
  • 00:05:42
    idea was to bring dogs around hospitals
  • 00:05:46
    to detect specific areas where Mera
  • 00:05:49
    colonies were growing kind of like a
  • 00:05:51
    drug dog at the
  • 00:05:53
    airport I just missed the small detail
  • 00:05:55
    that no surgeon was going to let me let
  • 00:05:58
    a dog into a room where they'd be
  • 00:06:00
    exposing patients with large open
  • 00:06:03
    incisions not that I would let that kind
  • 00:06:05
    of detail stop
  • 00:06:07
    me so off I went to monell chemical
  • 00:06:10
    sensus Center the leading Research
  • 00:06:12
    Institute for chemosensory Sciences
  • 00:06:15
    where I met Dr Bruce Kimbell an
  • 00:06:17
    analytical chemist and the late Dr kunio
  • 00:06:20
    yamazaki Who had successfully shown that
  • 00:06:23
    mice can detect Aven influenza in Duck
  • 00:06:26
    feal samples while I was working on
  • 00:06:29
    another post-doctoral project for Bruce
  • 00:06:32
    they were looking to do the next round
  • 00:06:34
    of Aven influenza virus studies this
  • 00:06:37
    time using a different species as the
  • 00:06:40
    biodetector my first thought was giant
  • 00:06:43
    gambian pouch Rats the hero rats you've
  • 00:06:46
    seen detecting landmines in Africa but
  • 00:06:50
    gambian pouch rats are a highly invasive
  • 00:06:53
    species in the US and there was no way
  • 00:06:56
    that they were going to let me import
  • 00:06:58
    any
  • 00:07:00
    and then I read a research paper about
  • 00:07:02
    how ferrets interact with humans like
  • 00:07:05
    dogs do yeah
  • 00:07:07
    ferrets ferrets can hold your gaze and
  • 00:07:10
    if you point your finger they'll follow
  • 00:07:12
    it to see what you're pointing at so we
  • 00:07:15
    ordered eight young ferrets and got to
  • 00:07:17
    work to figure out whether ferrets can
  • 00:07:20
    accurately detect Aven influenza virus
  • 00:07:23
    we trained them using two different
  • 00:07:25
    chemical samples one with an odor
  • 00:07:27
    similar to an infected duck and one with
  • 00:07:30
    the odor similar to a non-infected duck
  • 00:07:33
    if the feret alerted by scratching on
  • 00:07:35
    the correct box we clicked the clicker
  • 00:07:38
    to Mark's
  • 00:07:40
    behavior and then gave them a tasty
  • 00:07:42
    reward with a
  • 00:07:44
    syringe very quickly ferrets learned to
  • 00:07:46
    identify the sick duck chemical samples
  • 00:07:50
    with 90%
  • 00:07:52
    accuracy and that's great for in the lab
  • 00:07:56
    but if we wanted to use ferrets to
  • 00:07:58
    detect Avan influenza in real life they
  • 00:08:01
    wouldn't be sniffing pure chemical
  • 00:08:03
    samples but real duck poop and poop is
  • 00:08:06
    obviously more socially Salient
  • 00:08:09
    aromatically complex a lot of competing
  • 00:08:13
    compelling odors in a very tiny little
  • 00:08:16
    package we still had the feal samples
  • 00:08:18
    from the mouse experiment that had been
  • 00:08:20
    sitting in a minus 80 freezer for the
  • 00:08:22
    past 6
  • 00:08:23
    years we put five boxes in a row four
  • 00:08:28
    with negative samples and one with a
  • 00:08:30
    positive sample in a different order
  • 00:08:32
    each time so the ferrets couldn't
  • 00:08:34
    predict which box contain the positive
  • 00:08:37
    sample I remember watching each feret
  • 00:08:40
    approach the positive box smell it for a
  • 00:08:43
    bit and then start to move on to the
  • 00:08:45
    next box and my heart would sink but
  • 00:08:49
    before I could even sigh every feret
  • 00:08:53
    spun around and slam that paw down on
  • 00:08:56
    the box holding the positive sample the
  • 00:08:59
    ferrets were right 79% of the time and
  • 00:09:04
    they could detect the odor change a day
  • 00:09:07
    before and at least one day after
  • 00:09:10
    traditional medical testing the next
  • 00:09:13
    step was to transition to dogs because
  • 00:09:15
    as much as I love ferrets they just
  • 00:09:17
    aren't suited to hight trffic places I
  • 00:09:20
    mean imagine seeing a ferret on a little
  • 00:09:22
    leash at the
  • 00:09:23
    airport on an egg Farm or at a highway
  • 00:09:27
    stop sniffing livestock Crossing state
  • 00:09:30
    lines it would be
  • 00:09:32
    hilarious it would be adorable but it
  • 00:09:35
    would also be imperfect because ferrets
  • 00:09:38
    are prone to distraction and less
  • 00:09:40
    motivated than dogs especially in
  • 00:09:43
    chaotic
  • 00:09:44
    environments so just like the ferrets we
  • 00:09:47
    trained up a group of dogs and they
  • 00:09:49
    could detect Avan influenza virus it
  • 00:09:51
    didn't really matter which breed of dog
  • 00:09:54
    so long as they were physically capable
  • 00:09:56
    of doing the job and very motivated
  • 00:10:00
    these dogs could work on chicken farms
  • 00:10:02
    do a quick scan of the chickens every
  • 00:10:05
    day and sniff out bird flu before it
  • 00:10:07
    spreads to thousands of other birds
  • 00:10:11
    Hunters coming off the lake with their
  • 00:10:12
    Harvest could use these dogs to tell
  • 00:10:15
    them which Ducks were infected and which
  • 00:10:19
    weren't but what about other
  • 00:10:22
    diseases recently my team has been
  • 00:10:24
    working on chronic wasting disease in
  • 00:10:26
    white tailed deer chronic wasting
  • 00:10:29
    disease or CWD is a fatal highly
  • 00:10:33
    transmissible disease that affects cids
  • 00:10:36
    like deer elk and
  • 00:10:38
    moose if you never want to sleep again
  • 00:10:41
    feel free to look up pron disease after
  • 00:10:44
    my
  • 00:10:45
    talk basically prons cause cause
  • 00:10:48
    proteins to misfold in the body which
  • 00:10:52
    then accumulate in the nervous system
  • 00:10:55
    resulting in spongiform and sepil apathy
  • 00:10:58
    which is really just a fancy phrase for
  • 00:11:00
    huge holes in your brain literally the
  • 00:11:03
    stuff of
  • 00:11:04
    nightmares CWD spreads from infected
  • 00:11:07
    deer through bodily fluids and tissues
  • 00:11:11
    so it's not just deer to Dear contact
  • 00:11:14
    that's a problem but the environment as
  • 00:11:16
    well a healthy deer can get CWD by
  • 00:11:20
    drinking from a water source where a
  • 00:11:22
    sick deer had drank from days or even
  • 00:11:25
    weeks
  • 00:11:26
    before so using the same techniques we
  • 00:11:29
    did with the Aven influenza we trained
  • 00:11:32
    dogs to detect feal samples from CWD
  • 00:11:35
    infected deer the dogs were 94% accurate
  • 00:11:39
    in the lab and 80% accurate in the field
  • 00:11:43
    even when they were distracted by moving
  • 00:11:46
    tractors and combines that means we
  • 00:11:48
    could use labt trained dogs as a rapid
  • 00:11:51
    non-invasive tested Hunter check-in
  • 00:11:54
    stations to prevent hunters from eating
  • 00:11:57
    CWD infected deer meat
  • 00:12:00
    and we could deploy them to live deer
  • 00:12:02
    forms to alert farmers of infected deer
  • 00:12:05
    in the
  • 00:12:07
    herd This research is
  • 00:12:09
    groundbreaking our results suggest that
  • 00:12:12
    not only can dogs ferrets and rats
  • 00:12:15
    accurately detect the odor profiles of
  • 00:12:17
    infectious diseases but in some cases
  • 00:12:21
    they can do it sooner than a traditional
  • 00:12:23
    medical test and like I said earlier
  • 00:12:27
    timing is everything the faster you
  • 00:12:30
    detect the disease the sooner you can
  • 00:12:32
    eradicate
  • 00:12:34
    it in the long term the more we learn
  • 00:12:37
    about odor detection the faster we get
  • 00:12:40
    to a mechanical nose and though it may
  • 00:12:42
    be many years before we have both a
  • 00:12:45
    smoke and a virus detector in our homes
  • 00:12:48
    dogs ferrets and rats can be deployed to
  • 00:12:51
    Labs worldwide in a matter of months so
  • 00:12:54
    the next time you get a rapid
  • 00:12:56
    non-invasive pathogen test it might not
  • 00:12:59
    be a human with a pipet but a little
  • 00:13:02
    feret on a leash doing all the
  • 00:13:04
    work animals really are our best friends
  • 00:13:08
    it's not just that they love us
  • 00:13:10
    unconditionally it's that they can truly
  • 00:13:12
    Save Our Lives thank you
  • 00:13:16
    [Applause]
Tags
  • mechanical nose
  • disease detection
  • dogs
  • ferrets
  • odor detection
  • COVID-19
  • avian influenza
  • chronic wasting disease
  • early detection
  • biological research