[MUSIC PLAYING]

DR. JOHN WHYTE: Welcome,

everyone.

I'm Dr. John Whyte,

the chief medical officer

at WebMD, and you're watching

Your Health on Tech.

What if I could diagnose

your health conditions just

by listening to your voice?

Sound far fetched?

It may not be joining me to help

explain what we're learning

about the relationship

between our voices

and our health, is Dr. Yael

Bensoussan.

She's the director

of the University of South

Florida health voices center.

Dr. Bensoussan Thanks

for joining me.

DR. YAEL BENSOUSSAN: Hi Thank

you so much for the invitation.

DR. JOHN WHYTE: You've been

studying how the voice

and our health are related.

And you're working on a study

that I find fascinating.

And you said to the doctors,

close your eyes when a patient

comes in.

And just by listening

to their voice,

do you have an idea

about the diagnosis?

DR. YAEL BENSOUSSAN: This is

a very big project.

My co-principal investigator

is Dr. Olivier Elemento

at Cornell, and we have a team

of 10 institutions with 45 to 50

investigators on that project.

So really a large scale project.

And what's interesting is

that we find it, we know

and we've known for a very, very

long time that voice can give

biomarkers of certain diseases.

And you know that because when

your friend calls you

for example,

you can tell if he's feeling

sad, you can tell if he sounds

sick.

A lot of times we hear "You

don't sound like you're well."

And that's just from the sound

of the voice.

The interesting part is when we

talk to doctors to find out what

disease to study, that's really

when we ask them well close

your eyes, listen to the voice,

can you tell?

And that's when we ask experts

about the disease

we think that could be

a diagnosed with voices.

Some of them are intuitive.

What I do is voice disorders.

So obviously for voice

disorders, when I close my eyes

and I listen to voices

I hear different types of voice

disorders

like cancers of the voice box,

like neurological conditions

called spasmodic dysphonia

for example.

But there's other ones that are

less intuitive.

So for example, mycopia.

I spoke to one

of the oncologists at Cornell,

and he said, well, I think I can

hear when people have brain

metastases from lung cancer

when they come to see me.

And those are less intuitive.

And obviously it's always nicer

to say, well just by the voice.

But really in terms

of the science,

we're going into multimodal data

integration

when we speak

about artificial intelligence

in the future.

What's going to be really

important is not

single modality, so not just

the voice but really taking

the voice with other integration

of other modalities.

It can become,

people say it could become

like another heart rate,

or another blood pressure.

DR. JOHN WHYTE: What's

the pathophysiology of a voice

disorder

that's signaling perhaps

there's something else going on

in the body?

DR. YAEL BENSOUSSAN: So that's

a great question.

So I just to make it clear we're

not talking about specifically

diagnoses pancreatic cancer

with voice here.

What's interesting

is that we looked back

at the science,

and looked at what was proven

or what was in the terms

of pilot studies and diseases.

And we found really five

categories of diseases.

So when you think about voice

disorders that's intuitive.

When your vocal chords change,

that's the first thing that

changes.

If there's a disease

on your vocal cords,

then the sound of your voice

will change.

That's the first thing that

changes.

The science is easier.

When we talk

about neurological disorders

for example like--

DR. JOHN WHYTE: Parkinson's.

DR. YAEL BENSOUSSAN: Parkinson's

DR. JOHN WHYTE: Parkinson's is

an example.

DR. YAEL BENSOUSSAN: ALS, people

have

neurodegenerative disorders.

These are people who have

changes in speech and in voice.

When we go back

to neurological disorders,

we know for example somebody's

having a stroke that what you

say changes.

Sometimes they tell you on TV

if your speech is slurred,

go to the doctor.

So those are things that we've

known for a long time.

This is not new Parkinson's

same thing for people who know

that they work with Parkinson's

disease patients,

they can recognize the voice

of a Parkinson's patient.

DR. JOHN WHYTE: But you're

collecting 30,000 voices

to analyze.

DR. YAEL BENSOUSSAN: That's

the goal over 4 years.

And that's really the most

important thing.

Artificial intelligence

is a great technology.

We know that the technology is

there.

Machine learning is very

advanced.

Algorithms are very advanced.

What we're missing is really

the data.

And that's the most important

because we know that to train

these algorithms,

if you don't have accurate data

in a large quantity

then your algorithms are just

not generalizable.

That's really the goal

of our big project

and that's why so many people

are involved.

And it's important for us

to keep it

within academic institution,

is to get the quality

and the accuracy

of the diagnosis

as well as the quantity.

DR. JOHN WHYTE: Well I just talk

into my smartphone.

And it'll analyze my voice is

that what's coming?

DR. YAEL BENSOUSSAN: Well that's

the goal.

The goal is really to make

something a technology that's

accessible.

And a lot of startups

are looking at this already

and are making you give donate

your voice through your phone.

What's different

here is that, first of all,

we need to make the experiments

to make sure that the quality is

good enough through a phone.

And maybe for different type

of diseases the quality you need

is different.

Maybe for Parkinson's,

or for Alzheimer's, you can give

your voice through your phone

because it's really the content

that's important-- the content

of the speech.

But maybe for diseases

like cough that can have

different etiologies or causes,

maybe you need a better

microphone, maybe you need

a digital stethoscope to have

the accuracy that you need

to diagnose this.

So this is really what our teams

are working on this year is

looking at what

is the minimal quality we need,

but obviously the goal

for public health

is to have something accessible

not in a doctor's office

with high technology,

but really

to remote populations.

DR. JOHN WHYTE: Well you say

this technology ultimately could

be a bridge to provide access

to those

in rural and underserved

communities.

Everybody says that though

for every technology

that we talk about.

Do you think this one will be

different if we know that it

works?

DR. YAEL BENSOUSSAN: The goal

is not to replace doctors.

The goal is not to replace

experts.

Every categories of diseases

that we are studying,

there are experts in the field.

If you go to the expert then

you'll get a diagnosis.

That's probably as good

or better than the technology.

What's important

is to provide the same expertise

to people that don't necessarily

have it.

So when we think about family

doctors, you go to your family

doctor your voice changed,

often they don't have the tools

to look into your voice box.

They don't have the tools to do

a CT scan, they don't have

in their offices,

and that's why the voice can

help

and you would be speaking

through your phone, for example,

and the app could say,

well, there's a really high

chance this is a voice box

cancer based on all the data

that we've seen

and that we've analyzed

with our algorithms.

This patient should be sent

to an ENT fast.

I see people in my practice

unfortunately that come really

late with very advanced voice

box cancers.

They were told it was a cold

or laryngitis or reflux,

and not because they're

bad physicians just because they

don't have the tools

and it takes times to see

an expert.

That's really the goal is really

the screening earlier.

Not necessarily diagnosis where

you will not

won't have to go to the doctor.

That's not what we're trying

to develop.

We're trying to help physician

with better screening tools.

DR. JOHN WHYTE: What I love

about this Dr. Bensoussan is

you're talking about the power

of voice.

Yet as physicians,

or known to interrupt patients

after just a few minutes.

We tend to do most

of the talking.

And we don't really listen

to them.

Do you think this might re

sensitize clinicians actually

listening to their patients?

DR. YAEL BENSOUSSAN:

In my clinic

I often make the patient repeat

the story for me,

then I have my students come in

and I say listen to what they're

going to say,

and listen to the cues

in their voice.

So this in my practice

is very intuitive.

Again I'm really hoping.

I think in medicine we really

need to listen to patients.

I think listen to what they say,

listen to how they say it.

And I always teach

my medical student

that a good clinical history is

worth a million CT scans.

DR. JOHN WHYTE: What tips

if any, do you have for viewers

that even now could maybe

be on the lookout

for certain changes

in voice of loved ones,

of spouses, of children,

of elderly parents that might

trigger the idea, "Hey

maybe you might need to see

a doctor?"

DR. YAEL BENSOUSSAN:

Well I would say,

listen to your friends

and families.

I often see people in my office

who say, well my wife told me

that my voice changed so I'm

here.

Sometimes people listen

to a change in their voice

or with some diseases that are

more neurological, people don't

necessarily realize.

So they'll tell me,

well my friends are telling me

that I speak differently,

that my speech is slurred.

So when people give you

that feedback, that's really

the time to come and see

your physician.

I hear, my voice is tired you

know I see a lot of people

and they're like, "Well my voice

is not as strong."

The voice can tell a lot

about your health in general.

And it's not necessarily.

It doesn't mean

that your vocal cords are

injured it could mean

a lot about your whole body

system.

DR. JOHN WHYTE: The power

of voice relating

to our overall health

fascinating research.

Dr. Bensoussan I want to thank

you for all the work

that you're doing here

and sharing it with us today.

We look forward to following it.

DR. YAEL BENSOUSSAN: Thank you

so much.

[MUSIC PLAYING]