you can’t really pick up any major
magazine or newspaper today without
seeing articles about Big Data
artificial intelligence machine learning
and there’s a lot of fear factor going
on right there’s a lot of fear that our
elections are going to be impacted that
our livelihoods are being threatened
that our economy is under attack in fact
it really seems like some of these
articles are proposing that data is the
next gold or the next oil and in fact
interning says that if data is the next
oil then artificial intelligence is the
next engine it really seems like big
data is threatening our privacy and
artificial intelligence is ready to take
our jobs today I’m going to try to
reassure you about artificial
intelligence big data and machine
learning and maybe convince you this is
the most amazing time to be an innovator
in this space it’s not about fear it’s
about the opportunity right now and it’s
an amazing opportunity at that but first
let’s make sure everyone is level set it
on exactly what do these terms mean
because they’re often used
interchangeably the first term is
artificial intelligence and that’s maybe
the more grander of the terms artificial
intelligence just means mimicking
aspects of human intelligence it might
be driving a car translating a language
it might be recognizing faces in a photo
one kind of artificial intelligence is
machine learning where we’re gonna teach
that computer to do those skills using
data now it might be obvious of course
we’re gonna use data to teach a computer
what else would you use but you could
also use rules you could go gather a
roomful of experts and ask them what the
rules are to learn something but machine
learning is about using the data in fact
are two kinds of machine learning
they’re supervised and unsupervised
supervised means that we know the right
answer we know who’s in this picture and
the computer can get it right or wrong
and we can measure accuracy sensitivity
specificity we can come up with the
score how good is the computer
unsupervised means we really don’t know
what’s going on in this data computer
help us find something find some
patterns some interesting way to
organize the data there’s no right or
wrong there in unsupervised and then one
kind of machine learning is deep
learning and deep learning means to use
different neurons to kind of look like
the human brain try to mimic the human
neural architecture to do the machine
learning it crudely mimics that at least
it did at the beginning now now these
are the terms that we use in this field
and it’s amazing they still seem
sophisticated but part of why I’m on
stage is to demystify these concepts
because in the end these are software
tools that you borrow and put into your
software as you’re writing some code in
fact you can go to Amazon and buy
dummies books on these subjects right
now I just checked they’re about
nineteen dollars each
right even dummies can learn about this
I guess now if these terms still seem
sophisticated to you let me introduce
you to the most basic form of machine
learning almost all of us are taught in
eighth or ninth or 10th grade the
concept that we can fit a line to a
bunch of points you typically get this
in science class so here you see a bunch
of blue points and there’s an
x-coordinate and a y-coordinate it
should be intuitive to you that as we
add more blue points that red line
starts to fit them better and better
some of you in pre-algebra might
remember y equals MX plus B right
you got a intercept you have a slope
there and as we add more points or teach
the computer the line gets better and
better now you might look at this and
think how on earth is this machine
learning but think about this for a
moment that I can give the computer that
x-coordinate on the bottom and it’ll
give me its best guess of what the
y-coordinate should be right I give it
an X and it gives me a why isn’t that
amazing but actually it’s just a simple
formula y equals MX plus B this is
linear regression now this is the
simplest possible form of machine
learning but already you can start to
see the pattern so you give it more data
the model gets better and better and
some
interesting about how we’re going to be
approaching machine learning in the
future imagine that the future of
computer science is not just going to be
about learning Java or C++ or Python a
future direction for computer science
might be becoming a teacher how do we
teach the computer with the right points
the best points
not so much bias representative data so
a large component of computer science in
the future is going to be about being a
great teacher think about that for a
moment
now all of this seems sophisticated and
all of this is taking off there’s a huge
amount of hype here
why is artificial intelligence taking
off so much right now there’s three key
reasons the first is we have incredible
hardware now hardware from companies
down the road like Nvidia and Nvidia
makes gamer boards and display boards
for gamers let play for tonight maybe
you’ve heard of Fortnight and it’s not
just for tonight it’s any first-person
shooter boy you want to get those pixels
as fast as possible onto the screen and
it turns out the same hardware that we
try to make into better and better video
boards that same hardware is also useful
for teaching computers with machine
learning in fact people like me now buy
these video boards at such a level that
it’s actually hard for gamers to buy
these sports right because they’re
actually more useful to us and NVIDIA is
one of many companies what happens to be
one not too far from here
that I’m showing so we got great
hardware we also have great software
software from companies like Google or
universities like Berkeley and Stanford
it’s not just software it’s free
software anyone in this entire audience
can go download the software every logo
you see here for free you don’t have to
pay for the best software that the best
companies and universities actually use
how amazing is that data science is now
democratized that any of us can actually
get this software and just start to use
and just start to play so you got
hardware you got software and we got a
lot of data now
to teach these computers with you know
one of the most amazing tricks with data
is instead of going to go gather data
how can you convince people to give you
data for free right learn from others as
they’re at doing their job I actually
gather their data and then teach the
computer so here are just a couple of
representative examples Amazon we buy
things I’m sure many of you do too on
Amazon but and in the course of buying
products the computer learns for you
they ordered this and then they ordered
that a week later right and maybe
someone else might be interested in
ordering this other product a week later
right you see the recommendations all
the time on the top right is Watson
playing in jeopardy train with all the
literature and all the articles of the
world right playing a game show but
still important project on the bottom
left there is way mo one of many
companies within a 25 mile radius of
where we’re standing trying to make a
car that drives itself right you just
heard another talk about self-driving
cars and a lot of these cars now are
learning from the human drivers and then
trying to pick that up from what they’re
actually seeing in terms of behavior and
then Google of course learns from the
entire world’s literature on the web
every web page no matter how good or bad
it is Google knows how to learn from
that incredible amounts of data now but
you got great software and great
hardware as well what I’m gonna tell you
about is another amazing area where we
can use AI and machine learning because
we got the hardware we got the software
and we have incredible amounts of data
and that’s in biology and medicine I’m
showing here just what’s looks like a
regular cat scan of the brain and
whether it’s a cat scan or an MRI or a
cardiac study where we’re looking at how
the heart is beating boy we get a lot of
images in medicine a lot of data there
and it might be radiology like this or
pathology where we’re looking under the
microscope maybe it’s someone’s cancer
maybe they even tell if it’s a cancer so
we got a lot of image data and biology
medicine but it’s not just image data we
have all of these medical records now I
bet most of the younger folks in this
audience have never seen files like this
but some of the older folks in this
audience remember when they used to go
see a doctor they would take notes in
these kinds of file folders here
of course this is how doctors offices
used to be organized we move to
electronic medical records especially
because it’s really nice and pretty when
they look all organized it’s really a
mess when you have a patient that’s been
in the hospital for days or weeks or
months you don’t even know where to
store all the paper sometimes this
particular floor of the VA hospital
where this was taken was under so much
pressure from the weight of the paper
that the floor was starting to buckle of
course we’re gonna start to computerize
these records if you have any idea if
you have any doubt that your doctor is
using computers you might remember the
last time you saw your doctor you might
have seen the back of their head because
they were typing as they were talking to
you
that’s an electronic medical record
system another neat development I want
to share with you actually is that while
we keep all these records and we use
them in a safe responsible way more of
these records are actually gonna make it
out to you and many of you might know
many of you might use an iPhone I’m just
giving one example of how records are
gonna make it to you some of you might
remember this health app for the longest
time I never use this app right it
counted some steps and might look at
your sleep patterns if you were that
sophisticated with your iPhone but now
actually the second tab over you click
on health data and it actually figures
out where you are and all the hospitals
around you and makes a guess maybe one
of these hospitals or clinics is where
you get your care and literally you can
login with your credentials there and
all of that health did and I loop goes
right into your iPhone and the future I
predict another year or two we’re gonna
see all a bunch of health apps helping
you figure out what to do with all this
health data
I couldn’t even show any of this a year
ago this is all brand new right now on
your iPhone you probably don’t even know
this app does this it’s an amazing time
for health data another grand source of
health data is in DNA data what is DNA
it’s the same core elements the building
blocks of all of your cells you have six
billion letters six billion base pairs
three from your mother three from your
father
those two come together and make you who
you are right all the different genes
recipes that your cells need to make on
the left there is how we used to
sequence DNA in 2003 we finished a
genome project that was after 13 years
three billion dollars to get that
project done many countries involved on
the top right now is my hand holding a
USB stick sequencer this literally is
powered by USB power right instead of
factories like the left you’d actually
get a entire genome sequence just with a
USB stick it’s a no-brainer this is
gonna affect medicine so you got
electronic health records you got all
this image data you got all this DNA
data what an amazing time it is to be a
computer scientist in this field of
biology and medicine today and all this
data is a sitting there waiting for
someone to figure out what to do what to
do with our patients today now it’s
amazing to get all that data from one
patient it’s amazing to get all that
data from one Hospital what I’m gonna
explain to you is something even bigger
than that I’m really lucky that I get to
be the chief data scientist now for the
entire University of California Health
System now many of you know what the
University of California is right we
have a lot of graduates here from market
I go to Berkeley or go to UCLA those are
the two major universities I think we
send folks to but they go to Santa Cruz
they go to other campuses as well the
University of California has ten
campuses and three national labs it’s
incredible but of those ten campuses six
of them are medical that there’s a
medical school but we also have pharmacy
schools and veterinary schools and
nursing schools but got six major
medical campuses and that what did they
well campuses they are UCLA and UCSF are
in the top ten according to u.s. News a
war report and we train half of the
medical students and residents in the
state of California all those trainees
actually come out of the University of
California but we have nearly 100,000
doctors taking care of fifteen million
patients 5% of the u.s. population ends
up getting some care in the University
California and all of that data is there
for us to learn what’s working in
medicine what’s missing in medicine and
what might be excessive in medicine what
are things we don’t need to do anymore
and I’m not just gonna walk you through
a couple editor Actos two anecdotes
showing you how we’re doing this
actually at University of California San
Francisco
the first is explained to see it’s
called type-2 diabetes usually called
adult onset diabetes sadly more and more
kids are getting this and this happens a
lot because you either get older or you
get obese and there’s obesity epidemic
going on
in the United States and type 2 diabetes
is nasty it’s not a great condition to
have it eventually can rob you of your
eyesight and your kidney health and your
heart how do we treat Type 2 diabetes
well this is the graphic that comes out
of American Diabetes Association and the
way to think about this is patients
start at the top and they work their way
down at the top it says you should eat
better and lose weight maybe exercise of
course that’s a no-brainer but if and
when that fails you got a big metformin
there at the top and if and when that
fails you got these six boxes in the
middle that you can choose six
categories or drugs and when those fail
you can add the other five categories
right and then when everything fails you
go to metformin and insulin injected
insulin at the bottom right so you can
kind of see the path of what you’re
supposed to take with patients so I was
staring at this diagram I was noticing
those six boxes in the middle there I’m
sick u-bolt which one would I choose if
our diabetes doctor well the graphic
there says the choice depends on the
patient and the disease in other words
it gives you no clue which box you
should choose but more and more it’s
important to actually figure out which
box to choose because some of those
boxes in the middle cost 200 times more
than some of the other boxes in the
middle yet they’re all in the same kind
of nice square shape same pastelly
colors you’d never really have a guess
that some of those boxes cost 200 times
some more we used to call these diabetes
donuts but then we realized I would be
inappropriate for diabetes so now we
call these life savers right and
actually you can read this like a pie
chart and pie would also be
inappropriate for diabetes there’s
twelve thousand and seven patients here
with type-2 diabetes in the middle there
and this is the first drug we choose a
third of them you can see are in yellow
if you follow that at the top it says
metformin okay that’s great that was the
right answer there right and a third of
them at the bottom there are on insulin
wow there’s someone starting them on
insulin first so that’s interesting
and they got the gray is
final ureas and all the other colors
there are different combinations of
drugs patients are starting on two or
three drugs that’s pretty heavy amount
of drugs to use on a patient in the
beginning and the black bar means there
are so many little slices there I can’t
even show you them all so this is a
first move we made the patient goes
comes back in 90 days here’s the next
move we made so everyone in yellow that
has just the yellow by itself that means
that first dose of metformin we picked
was perfect
this patient has never had to change
that dose yellow to yellow means we are
still on the same drug metformin but we
had to change the dose and every other
color to any other color means we added
a drug subtracted a drug changed the
drug it didn’t work out just right
patient goes home comes back in 90 days
here’s the next move patient goes home
comes back in the next 90 days and
here’s the fourth move then this is
diabetes played out over the first four
moves but then you realize at the top it
says we have sixteen hundred and forty
different ways to play this game sixteen
hundred different ways our doctors are
taking care of type 2 diabetes probably
too many what’s the future where’s this
patient going to be in 90 days what’s
gonna happen to this patient in the next
year and what are we gonna do about it
and that to me is gonna be the new
definition of a data-driven Medical
Center right one that knows how to use
the data to take care of all 15 million
of its patients that’s why I’m so
thrilled to be at University of
California and so thrilled to tell you
this is a modern era of using data in
medicine right it’s not about
stethoscopes and bandages more and more
it’s gonna be about data learn about AI
learn about computer science but learn
about biology and medicine too
biology is extremely molecular today and
medicine is very digital patients
families doctors and nurses are all
counting on the next generation to come
up with all of the solutions we need we
need new drugs we need new Diagnostics
we need better ways to take care of
patients I really really hope some of
you will join us on this journey to
create a much better medical system
keeping patients of the world healthy and happy
magazine or newspaper today without
seeing articles about Big Data
artificial intelligence machine learning
and there’s a lot of fear factor going
on right there’s a lot of fear that our
elections are going to be impacted that
our livelihoods are being threatened
that our economy is under attack in fact
it really seems like some of these
articles are proposing that data is the
next gold or the next oil and in fact
interning says that if data is the next
oil then artificial intelligence is the
next engine it really seems like big
data is threatening our privacy and
artificial intelligence is ready to take
our jobs today I’m going to try to
reassure you about artificial
intelligence big data and machine
learning and maybe convince you this is
the most amazing time to be an innovator
in this space it’s not about fear it’s
about the opportunity right now and it’s
an amazing opportunity at that but first
let’s make sure everyone is level set it
on exactly what do these terms mean
because they’re often used
interchangeably the first term is
artificial intelligence and that’s maybe
the more grander of the terms artificial
intelligence just means mimicking
aspects of human intelligence it might
be driving a car translating a language
it might be recognizing faces in a photo
one kind of artificial intelligence is
machine learning where we’re gonna teach
that computer to do those skills using
data now it might be obvious of course
we’re gonna use data to teach a computer
what else would you use but you could
also use rules you could go gather a
roomful of experts and ask them what the
rules are to learn something but machine
learning is about using the data in fact
are two kinds of machine learning
they’re supervised and unsupervised
supervised means that we know the right
answer we know who’s in this picture and
the computer can get it right or wrong
and we can measure accuracy sensitivity
specificity we can come up with the
score how good is the computer
unsupervised means we really don’t know
what’s going on in this data computer
help us find something find some
patterns some interesting way to
organize the data there’s no right or
wrong there in unsupervised and then one
kind of machine learning is deep
learning and deep learning means to use
different neurons to kind of look like
the human brain try to mimic the human
neural architecture to do the machine
learning it crudely mimics that at least
it did at the beginning now now these
are the terms that we use in this field
and it’s amazing they still seem
sophisticated but part of why I’m on
stage is to demystify these concepts
because in the end these are software
tools that you borrow and put into your
software as you’re writing some code in
fact you can go to Amazon and buy
dummies books on these subjects right
now I just checked they’re about
nineteen dollars each
right even dummies can learn about this
I guess now if these terms still seem
sophisticated to you let me introduce
you to the most basic form of machine
learning almost all of us are taught in
eighth or ninth or 10th grade the
concept that we can fit a line to a
bunch of points you typically get this
in science class so here you see a bunch
of blue points and there’s an
x-coordinate and a y-coordinate it
should be intuitive to you that as we
add more blue points that red line
starts to fit them better and better
some of you in pre-algebra might
remember y equals MX plus B right
you got a intercept you have a slope
there and as we add more points or teach
the computer the line gets better and
better now you might look at this and
think how on earth is this machine
learning but think about this for a
moment that I can give the computer that
x-coordinate on the bottom and it’ll
give me its best guess of what the
y-coordinate should be right I give it
an X and it gives me a why isn’t that
amazing but actually it’s just a simple
formula y equals MX plus B this is
linear regression now this is the
simplest possible form of machine
learning but already you can start to
see the pattern so you give it more data
the model gets better and better and
some
interesting about how we’re going to be
approaching machine learning in the
future imagine that the future of
computer science is not just going to be
about learning Java or C++ or Python a
future direction for computer science
might be becoming a teacher how do we
teach the computer with the right points
the best points
not so much bias representative data so
a large component of computer science in
the future is going to be about being a
great teacher think about that for a
moment
now all of this seems sophisticated and
all of this is taking off there’s a huge
amount of hype here
why is artificial intelligence taking
off so much right now there’s three key
reasons the first is we have incredible
hardware now hardware from companies
down the road like Nvidia and Nvidia
makes gamer boards and display boards
for gamers let play for tonight maybe
you’ve heard of Fortnight and it’s not
just for tonight it’s any first-person
shooter boy you want to get those pixels
as fast as possible onto the screen and
it turns out the same hardware that we
try to make into better and better video
boards that same hardware is also useful
for teaching computers with machine
learning in fact people like me now buy
these video boards at such a level that
it’s actually hard for gamers to buy
these sports right because they’re
actually more useful to us and NVIDIA is
one of many companies what happens to be
one not too far from here
that I’m showing so we got great
hardware we also have great software
software from companies like Google or
universities like Berkeley and Stanford
it’s not just software it’s free
software anyone in this entire audience
can go download the software every logo
you see here for free you don’t have to
pay for the best software that the best
companies and universities actually use
how amazing is that data science is now
democratized that any of us can actually
get this software and just start to use
and just start to play so you got
hardware you got software and we got a
lot of data now
to teach these computers with you know
one of the most amazing tricks with data
is instead of going to go gather data
how can you convince people to give you
data for free right learn from others as
they’re at doing their job I actually
gather their data and then teach the
computer so here are just a couple of
representative examples Amazon we buy
things I’m sure many of you do too on
Amazon but and in the course of buying
products the computer learns for you
they ordered this and then they ordered
that a week later right and maybe
someone else might be interested in
ordering this other product a week later
right you see the recommendations all
the time on the top right is Watson
playing in jeopardy train with all the
literature and all the articles of the
world right playing a game show but
still important project on the bottom
left there is way mo one of many
companies within a 25 mile radius of
where we’re standing trying to make a
car that drives itself right you just
heard another talk about self-driving
cars and a lot of these cars now are
learning from the human drivers and then
trying to pick that up from what they’re
actually seeing in terms of behavior and
then Google of course learns from the
entire world’s literature on the web
every web page no matter how good or bad
it is Google knows how to learn from
that incredible amounts of data now but
you got great software and great
hardware as well what I’m gonna tell you
about is another amazing area where we
can use AI and machine learning because
we got the hardware we got the software
and we have incredible amounts of data
and that’s in biology and medicine I’m
showing here just what’s looks like a
regular cat scan of the brain and
whether it’s a cat scan or an MRI or a
cardiac study where we’re looking at how
the heart is beating boy we get a lot of
images in medicine a lot of data there
and it might be radiology like this or
pathology where we’re looking under the
microscope maybe it’s someone’s cancer
maybe they even tell if it’s a cancer so
we got a lot of image data and biology
medicine but it’s not just image data we
have all of these medical records now I
bet most of the younger folks in this
audience have never seen files like this
but some of the older folks in this
audience remember when they used to go
see a doctor they would take notes in
these kinds of file folders here
of course this is how doctors offices
used to be organized we move to
electronic medical records especially
because it’s really nice and pretty when
they look all organized it’s really a
mess when you have a patient that’s been
in the hospital for days or weeks or
months you don’t even know where to
store all the paper sometimes this
particular floor of the VA hospital
where this was taken was under so much
pressure from the weight of the paper
that the floor was starting to buckle of
course we’re gonna start to computerize
these records if you have any idea if
you have any doubt that your doctor is
using computers you might remember the
last time you saw your doctor you might
have seen the back of their head because
they were typing as they were talking to
you
that’s an electronic medical record
system another neat development I want
to share with you actually is that while
we keep all these records and we use
them in a safe responsible way more of
these records are actually gonna make it
out to you and many of you might know
many of you might use an iPhone I’m just
giving one example of how records are
gonna make it to you some of you might
remember this health app for the longest
time I never use this app right it
counted some steps and might look at
your sleep patterns if you were that
sophisticated with your iPhone but now
actually the second tab over you click
on health data and it actually figures
out where you are and all the hospitals
around you and makes a guess maybe one
of these hospitals or clinics is where
you get your care and literally you can
login with your credentials there and
all of that health did and I loop goes
right into your iPhone and the future I
predict another year or two we’re gonna
see all a bunch of health apps helping
you figure out what to do with all this
health data
I couldn’t even show any of this a year
ago this is all brand new right now on
your iPhone you probably don’t even know
this app does this it’s an amazing time
for health data another grand source of
health data is in DNA data what is DNA
it’s the same core elements the building
blocks of all of your cells you have six
billion letters six billion base pairs
three from your mother three from your
father
those two come together and make you who
you are right all the different genes
recipes that your cells need to make on
the left there is how we used to
sequence DNA in 2003 we finished a
genome project that was after 13 years
three billion dollars to get that
project done many countries involved on
the top right now is my hand holding a
USB stick sequencer this literally is
powered by USB power right instead of
factories like the left you’d actually
get a entire genome sequence just with a
USB stick it’s a no-brainer this is
gonna affect medicine so you got
electronic health records you got all
this image data you got all this DNA
data what an amazing time it is to be a
computer scientist in this field of
biology and medicine today and all this
data is a sitting there waiting for
someone to figure out what to do what to
do with our patients today now it’s
amazing to get all that data from one
patient it’s amazing to get all that
data from one Hospital what I’m gonna
explain to you is something even bigger
than that I’m really lucky that I get to
be the chief data scientist now for the
entire University of California Health
System now many of you know what the
University of California is right we
have a lot of graduates here from market
I go to Berkeley or go to UCLA those are
the two major universities I think we
send folks to but they go to Santa Cruz
they go to other campuses as well the
University of California has ten
campuses and three national labs it’s
incredible but of those ten campuses six
of them are medical that there’s a
medical school but we also have pharmacy
schools and veterinary schools and
nursing schools but got six major
medical campuses and that what did they
well campuses they are UCLA and UCSF are
in the top ten according to u.s. News a
war report and we train half of the
medical students and residents in the
state of California all those trainees
actually come out of the University of
California but we have nearly 100,000
doctors taking care of fifteen million
patients 5% of the u.s. population ends
up getting some care in the University
California and all of that data is there
for us to learn what’s working in
medicine what’s missing in medicine and
what might be excessive in medicine what
are things we don’t need to do anymore
and I’m not just gonna walk you through
a couple editor Actos two anecdotes
showing you how we’re doing this
actually at University of California San
Francisco
the first is explained to see it’s
called type-2 diabetes usually called
adult onset diabetes sadly more and more
kids are getting this and this happens a
lot because you either get older or you
get obese and there’s obesity epidemic
going on
in the United States and type 2 diabetes
is nasty it’s not a great condition to
have it eventually can rob you of your
eyesight and your kidney health and your
heart how do we treat Type 2 diabetes
well this is the graphic that comes out
of American Diabetes Association and the
way to think about this is patients
start at the top and they work their way
down at the top it says you should eat
better and lose weight maybe exercise of
course that’s a no-brainer but if and
when that fails you got a big metformin
there at the top and if and when that
fails you got these six boxes in the
middle that you can choose six
categories or drugs and when those fail
you can add the other five categories
right and then when everything fails you
go to metformin and insulin injected
insulin at the bottom right so you can
kind of see the path of what you’re
supposed to take with patients so I was
staring at this diagram I was noticing
those six boxes in the middle there I’m
sick u-bolt which one would I choose if
our diabetes doctor well the graphic
there says the choice depends on the
patient and the disease in other words
it gives you no clue which box you
should choose but more and more it’s
important to actually figure out which
box to choose because some of those
boxes in the middle cost 200 times more
than some of the other boxes in the
middle yet they’re all in the same kind
of nice square shape same pastelly
colors you’d never really have a guess
that some of those boxes cost 200 times
some more we used to call these diabetes
donuts but then we realized I would be
inappropriate for diabetes so now we
call these life savers right and
actually you can read this like a pie
chart and pie would also be
inappropriate for diabetes there’s
twelve thousand and seven patients here
with type-2 diabetes in the middle there
and this is the first drug we choose a
third of them you can see are in yellow
if you follow that at the top it says
metformin okay that’s great that was the
right answer there right and a third of
them at the bottom there are on insulin
wow there’s someone starting them on
insulin first so that’s interesting
and they got the gray is
final ureas and all the other colors
there are different combinations of
drugs patients are starting on two or
three drugs that’s pretty heavy amount
of drugs to use on a patient in the
beginning and the black bar means there
are so many little slices there I can’t
even show you them all so this is a
first move we made the patient goes
comes back in 90 days here’s the next
move we made so everyone in yellow that
has just the yellow by itself that means
that first dose of metformin we picked
was perfect
this patient has never had to change
that dose yellow to yellow means we are
still on the same drug metformin but we
had to change the dose and every other
color to any other color means we added
a drug subtracted a drug changed the
drug it didn’t work out just right
patient goes home comes back in 90 days
here’s the next move patient goes home
comes back in the next 90 days and
here’s the fourth move then this is
diabetes played out over the first four
moves but then you realize at the top it
says we have sixteen hundred and forty
different ways to play this game sixteen
hundred different ways our doctors are
taking care of type 2 diabetes probably
too many what’s the future where’s this
patient going to be in 90 days what’s
gonna happen to this patient in the next
year and what are we gonna do about it
and that to me is gonna be the new
definition of a data-driven Medical
Center right one that knows how to use
the data to take care of all 15 million
of its patients that’s why I’m so
thrilled to be at University of
California and so thrilled to tell you
this is a modern era of using data in
medicine right it’s not about
stethoscopes and bandages more and more
it’s gonna be about data learn about AI
learn about computer science but learn
about biology and medicine too
biology is extremely molecular today and
medicine is very digital patients
families doctors and nurses are all
counting on the next generation to come
up with all of the solutions we need we
need new drugs we need new Diagnostics
we need better ways to take care of
patients I really really hope some of
you will join us on this journey to
create a much better medical system
keeping patients of the world healthy and happy
Please follow and like us: