English-Video.net comment policy

The comment field is common to all languages

Let's write in your language and use "Google Translate" together

Please refer to informative community guidelines on TED.com

TED2018

Dina Katabi: A new way to monitor vital signs (that can see through walls)

Filmed:
404,506 views

At MIT, Dina Katabi and her team are working on a bold new way to monitor patients' vital signs in a hospital (or even at home), without wearables or bulky, beeping devices. Bonus: it can see through walls. In a mind-blowing talk and demo, Katabi previews a system that captures the reflections of signals like Wi-Fi as they bounce off people, creating a reliable record of vitals for healthcare workers and patients. And in a brief Q&A with TED curator Helen Walters, Katabi discusses safeguards being put in place to prevent people from using this tech to monitor somebody without their consent.

- Technologist
Dina Katabi investigates how AI can make wireless devices sense human motion and vital signs. Full bio

When I was a kid,
00:13
I was, like many of you in this room,
very much fascinated by Star Wars,
00:14
and what fascinated me the most
is this notion of the Force,
00:19
this energy that connects
all people and all objects
00:23
and allows you to feel people
that you can't even see.
00:26
And I remember many nights,
I would be sitting at home,
00:30
just, like, concentrating and focusing,
trying to feel the Force,
00:33
and I didn't feel anything, don't worry.
00:38
(Laughter)
00:41
And later in life, I became a scientist.
00:43
I joined the MIT faculty
and started working on wireless signals.
00:45
These are things like Wi-Fi
or cellular systems,
00:51
and I did a lot of work in that domain.
00:54
But then, again, this Force thing
kept nagging me,
00:57
and at some point, I was just like,
01:02
"Wait a minute, these wireless signals --
they are like the Force."
01:04
So if you think about it,
01:08
wireless signals,
they travel through space,
01:09
they go through obstacles
and walls and occlusions,
01:12
and some of them,
01:16
they reflect off our bodies,
because our bodies are full of water,
01:17
and some of these minute reflections,
01:21
they come back.
01:23
And if, just if, I had a device that can
just sense these minute reflections,
01:25
then I would be able to feel people
that I cannot see.
01:32
So I started working with my students
on building such a device,
01:36
and I want to show you
some of our early results.
01:40
So here, you see my student standing,
01:44
and here is our device.
01:46
And we are going to put the device
in the other office, behind the wall,
01:48
and we are going
to monitor him as he moves.
01:54
This red dot is tracking him
using wireless signals.
01:57
And as you can see, the red dot
is tracking his movements very accurately,
02:02
purely based on how his body interacts
with the surrounding wireless signals.
02:07
Pretty accurate, isn't it?
02:13
He has no wearables, nothing.
02:16
(Applause)
02:18
Now you might be wondering,
02:21
how is it possible
that we can sense people
02:23
and track them, without
any wearables, through walls,
02:26
and the easiest analogy
to think about is radar.
02:30
I'm sure many of you
have seen this picture.
02:33
You transmit a wireless signal to the sky,
02:35
it reflects off some airplane,
comes back to you,
02:37
and you start detecting these airplanes.
02:40
But if it were just radar,
02:44
then we would have this 50 years ago.
02:46
So it's not just radar.
02:49
There are two key differences.
02:51
So the first difference, of course --
02:54
you can't, like radar, just blast
wireless power at somebody.
02:56
You're going to fry them
like if they were in a microwave.
02:59
Don't do that.
03:02
So it means that you have to be able
to deal with very weak signals,
03:04
and that means that your device
has to be very sensitive.
03:09
The second difference is that,
unlike the sky, where it's empty --
03:14
if you are lucky, there is one airplane
that you can catch there.
03:18
Like, look at the room
03:21
and look how many objects
and people there are.
03:22
So in indoor environments, the signal
not only reflects off the person,
03:25
if reflects off the person,
off the floor, the ceiling,
03:29
off other people around,
03:33
and you get very complex reflections
03:35
where the same signal reflects
off me and then off you,
03:37
and then off the ceiling,
then off the floor.
03:40
And you have to make sense of that mess.
03:42
But we were lucky.
03:47
We were coming at the right time.
03:49
So two things helped us.
03:53
The first thing is radiotechnologies
have evolved a lot,
03:55
and over the last decade,
03:59
radio technology
became much more powerful,
04:01
so we were able to build
very sensitive radios
04:03
that can sense weak and minute RF signals.
04:07
The second thing: machine learning.
04:11
So you keep hearing about machine learning
04:14
and there was a revolution
of machine learning recently,
04:16
in deep learning,
04:19
and that allowed us to build
machine-learning models
04:20
that can understand wireless signals
and interpret them
04:24
so they would know what happened
in the environment.
04:28
So if you think of it,
the radio is like the ear of our device
04:31
and the machine learning
is like the brain,
04:35
and together, they have
a very powerful device.
04:38
So what else can we sense about people
using wireless signals?
04:44
Sleep.
04:50
Sleep, actually, is something
very dear to my heart,
04:51
because my sleep is a disaster.
04:54
(Laughter)
04:56
So one thing is when you start working
on some physiological signal
04:57
and you discover that yours sucks.
05:00
(Laughter)
05:03
So you can see why we can capture sleep,
05:07
because the person walks and the device
sees him as he walks to bed,
05:09
when he stops tossing around in bed,
05:12
when he steps out of bed,
05:15
and that measure of sleep
is what people call actigraphy.
05:16
It's based on motion.
05:20
But it turned out
that we can actually get sleep
05:23
at a much more important level.
05:26
We can understand
the change in the brain waves
05:28
that occur during sleep.
05:32
So, many of you probably know
that as we go to sleep,
05:35
our brainwaves change
and we enter different stages:
05:38
awake, light sleep, deep sleep
and REM, or rapid eye movement.
05:41
These stages are of course
related to sleep disorders,
05:46
but they are also related
to various diseases.
05:50
So for example, disturbances in REM
are associated with depression.
05:53
Disturbances in deep sleep
are associated with Alzheimer's.
05:59
So if you want to get sleep staging,
06:04
today, you will send the person
to the hospital,
06:07
they put all of these
electrodes on their head,
06:09
and they ask them to sleep like that.
06:12
(Laughter)
06:13
It's not really a happy experience.
06:16
So what if I tell you
that I can do the same thing
06:20
but without any of these electrodes
on the person's body?
06:23
So here is our device,
06:27
transmitting very low power
wireless signal,
06:29
analyzes the reflections using AI
06:32
and spits out the sleep stages
throughout the night.
06:34
So we know, for example,
when this person is dreaming.
06:38
Not just that ...
06:44
we can even get your breathing
while you are sitting like that,
06:46
and without touching you.
06:49
So he is sitting and reading
06:51
and this is his inhales, exhales.
06:53
We asked him to hold his breath,
06:55
and you see the signal
staying at a steady level
06:58
because he exhaled.
07:00
He did not inhale.
07:02
And I want to zoom in on the signal.
07:05
And this is the same signal as before.
07:07
These are the inhales,
07:09
these are the exhales.
07:11
And you see these blips on the signal?
07:13
These are not noise.
07:15
They are his heartbeats.
07:17
And you can see them beat by beat.
07:19
So I want to stop here for a moment
and show you a live demo.
07:23
Zach is going to help me with the demo,
07:27
and we're going to use the device
to monitor Zach's breathing.
07:29
So this white box
that you see here is the device,
07:35
and Zach is turning it on ...
07:39
and I see that he breathes well.
07:43
So we're going to do exactly what we did
in the video with the other guy,
07:46
so the wireless signal is going through,
07:50
it's touching Zach's body,
07:53
and it's reflecting back to the device,
07:55
and we want to monitor his breathing,
his inhale-exhale motion.
07:57
So we see the inhales, exhales --
08:00
so see, these ups and downs
are Zach breathing.
08:03
Inhaling, exhaling.
08:10
(Applause)
08:12
So, he can breathe.
08:17
(Laughter)
08:20
Zach, can you hold your breath, please?
08:21
OK, so now he's holding his breath,
08:25
so you see the signal stays
at a steady level,
08:28
and these are his heartbeats.
08:30
Beat, beat, beat, beat, beat.
08:32
(Applause)
08:35
OK, Zach, you can breathe again.
08:36
(Laughter)
08:38
We don't want accidents here.
08:40
(Laughter)
08:42
OK, thank you.
08:43
(Applause)
08:45
So as you can see, we have this device
08:51
that can monitor so many
physiological signals for you,
08:54
and what is really interesting
about this device
08:58
is that it does all this
without any wearables,
09:01
without asking the person
to change his behavior
09:03
or to wear anything
or charge anything special.
09:06
And that got doctors very excited,
09:09
because doctors,
09:12
they always want to know
more information about their patients,
09:13
particularly at home,
09:16
and this is particularly true
in chronic diseases,
09:18
like pulmonary diseases, like COPD,
09:20
or heart failure or Alzheimer's
and even depression.
09:24
All of these chronic diseases
are very important.
09:29
In fact -- perhaps you know --
09:31
two-thirds of the cost
of health care in the US
09:34
is due to chronic diseases.
09:37
But what is really interesting
about chronic diseases
09:39
is that when the person, for example,
09:42
has a problem that leads
to the hospital and the emergency room,
09:43
this problem doesn't happen overnight.
09:48
Actually, things happen gradually.
09:51
So if we can monitor
chronic disease patients in their home,
09:53
we can detect changes in their breathing,
heartbeat, mobility, sleep --
09:56
and we can detect emergencies
before they occur
10:01
and have the doctor intervene earlier
10:05
so that we can avoid hospitalization.
10:07
And indeed, today we are working
with multiple doctors
10:10
in different disease categories.
10:14
So I'm really excited
10:16
because we have deployed the device
with many patients.
10:17
We have deployed the device
with patients that have COPD,
10:20
which is a pulmonary disease,
10:24
patients that have Alzheimer's,
10:25
patients that have depression and anxiety
10:27
and people that have Parkinson's.
10:29
And we are working with the doctors
on improving their life,
10:32
understanding the disease better.
10:35
So when I started, I told you
10:38
that I'm really fascinated with Star Wars
and the Force in Star Wars,
10:41
and indeed, I'm still
very much fascinated,
10:46
even now, as a grown-up, with Star Wars,
10:50
waiting for the next movie.
10:52
But I'm very fascinated now and excited
10:54
about this new Force of wireless signals,
10:59
and the potential of changing
health care with this new force.
11:03
One of the patients with whom
we deployed is actually my aunt.
11:07
She has heart failure,
11:10
and I'm sure many of you guys
in the audience
11:12
have parents, grandparents,
loved ones who have chronic diseases.
11:16
So I want you to imagine with me a future
11:21
where in every home
that has a chronic disease patient,
11:24
there is a device like this device
sitting in the background
11:26
and just monitoring passively
11:30
sleep, breathing, the health
of this chronic disease patient,
11:32
and before an emergency occurs,
11:37
it would detect the degradation
in the physiological signal
11:39
and alert the doctor
11:42
so that we can avoid hospitalization.
11:43
This can change health care
as we know it today,
11:46
improve how we understand
chronic diseases
11:49
and also save many lives.
11:52
Thank you.
11:54
(Applause)
11:56
Helen Walters: Dina, thank you so much.
12:01
Thank you too, Zach.
12:03
So glad you're breathing.
12:04
So Dina, this is amazing.
12:05
The positive applications are incredible.
12:08
What is the framework, though,
like the ethical framework around this?
12:11
What are you doing to prevent
this technology from being used
12:15
for other, perhaps less positive
types of applications?
12:18
Dina Katabi: Yeah, this is
a very important question, of course,
12:22
like, what about misuse,
12:25
or what about, I guess you could say,
about the Dark Side of the Force?
12:26
HW: Right, right.
12:30
(Laughter)
12:31
DK: So we actually have technologies
12:33
that prevent people
from trying to use this device
12:36
to monitor somebody without their consent.
12:41
Because the device understands space,
12:43
it will ask you to prove,
by doing certain movements,
12:45
that you have access to the space
12:49
and you are the person
who you are asking the device to monitor.
12:51
So technology-wise,
12:54
we have technology
that we integrate to prevent misuse,
12:56
but also, I think there is a role
for policy, like everything else,
13:00
and hopefully, with the two of them,
we can control any misuse.
13:03
HW: Amazing. Thank you so much.
13:09
DK: Thank you.
13:10
(Applause)
13:12

▲Back to top

About the speaker:

Dina Katabi - Technologist
Dina Katabi investigates how AI can make wireless devices sense human motion and vital signs.

Why you should listen

Dina Katabi designs new wireless devices that use machine learning to sense people through walls and occlusions. Her devices look like a Wi-Fi box. They transmit a low-power wireless signal and capture its reflections as it bounces off people and objects. They analyze those reflections to learn how people walk, measure their gait and detect elderly falls. The device can also measure a person's breathing, heart rate and sleep quality using wireless signals, without any sensor on the person's body. Katabi is working with medical doctors to use her technology to detect health emergencies and provide a better understanding of chronic diseases such as Alzheimer's and Parkinson's. 

Katabi is the Andrew & Erna Viterbi Professor of Electrical Engineering and Computer Science at MIT. She is also the director of the MIT's Center for Wireless Networks and Mobile Computing, a member of the National Academy of Engineering and a recipient of the MacArthur Fellowship. Her research has been recognized by the ACM Prize in Computing, the ACM Grace Murray Hopper Award, the SIGCOMM Test-of-Time Award, the IEEE William R. Bennett prize, the Faculty Research Innovation Fellowship, a Sloan Fellowship and multiple best paper awards. Several startups have been spun out of her lab, such as PiCharging and Emerald.

More profile about the speaker
Dina Katabi | Speaker | TED.com