we expect that artificial intelligence is all about robots and how these robots
is all about robots and how these robots
is all about robots and how these robots have complex Minds and how they should
have complex Minds and how they should
have complex Minds and how they should be able to interact with humans and
be able to interact with humans and
be able to interact with humans and interact with each other the movies
interact with each other the movies
interact with each other the movies wall-e and star wars are perfect
wall-e and star wars are perfect
wall-e and star wars are perfect examples of this however in reality
examples of this however in reality
examples of this however in reality artificial intelligence has a completely
artificial intelligence has a completely
artificial intelligence has a completely different meaning
different meaning
different meaning AI is all about classification being
AI is all about classification being
AI is all about classification being able to do tasks that a human could do
able to do tasks that a human could do
able to do tasks that a human could do visually but do it in a completely
visually but do it in a completely
visually but do it in a completely automated manner and a good example of
automated manner and a good example of
automated manner and a good example of this is in the security setting being
this is in the security setting being
this is in the security setting being able to do object classification
able to do object classification
able to do object classification detecting people detecting cars
detecting people detecting cars
detecting people detecting cars detecting fire hydrants dogs in just one
detecting fire hydrants dogs in just one
detecting fire hydrants dogs in just one image is one task that artificial
image is one task that artificial
image is one task that artificial intelligence can achieve however another
intelligence can achieve however another
intelligence can achieve however another example is that of one ring and this is
example is that of one ring and this is
example is that of one ring and this is in the medical sense one ring is an
in the medical sense one ring is an
in the medical sense one ring is an intelligent monitoring device for
intelligent monitoring device for
intelligent monitoring device for Parkinson’s disease and that’s why today
Parkinson’s disease and that’s why today
Parkinson’s disease and that’s why today I’ll be talking about how my first
I’ll be talking about how my first
I’ll be talking about how my first experience was with Parkinson’s disease
experience was with Parkinson’s disease
experience was with Parkinson’s disease inspired me to want to build something
inspired me to want to build something
inspired me to want to build something that can influence the lives of
that can influence the lives of
that can influence the lives of Parkinson’s patients and how I leveraged
Parkinson’s patients and how I leveraged
Parkinson’s patients and how I leveraged artificial intelligence the technologies
artificial intelligence the technologies
artificial intelligence the technologies behind it to build something that I
behind it to build something that I
behind it to build something that I could deploy to patients that they could
could deploy to patients that they could
could deploy to patients that they could benefit benefit benefit from it so first
benefit benefit benefit from it so first
benefit benefit benefit from it so first off what is Parkinson’s
off what is Parkinson’s
off what is Parkinson’s well Parkinson’s is a neurodegenerative
well Parkinson’s is a neurodegenerative
well Parkinson’s is a neurodegenerative disorder meaning that over time
disorder meaning that over time
disorder meaning that over time Parkinson’s patients will lose control
Parkinson’s patients will lose control
Parkinson’s patients will lose control of their movement meaning that they’ll
of their movement meaning that they’ll
of their movement meaning that they’ll have tremors
have tremors
have tremors they’ll have shaking they’ll have a lot
they’ll have shaking they’ll have a lot
they’ll have shaking they’ll have a lot of problems that make it significantly
of problems that make it significantly
of problems that make it significantly harder for them to lead independent
harder for them to lead independent
harder for them to lead independent lives and that’s why the biggest problem
lives and that’s why the biggest problem
lives and that’s why the biggest problem in current Parkinson’s disease
in current Parkinson’s disease
in current Parkinson’s disease Diagnostics is that they’re purely
Diagnostics is that they’re purely
Diagnostics is that they’re purely qualitative in the way that a physician
qualitative in the way that a physician
qualitative in the way that a physician has to sit down with the Parkinson’s
has to sit down with the Parkinson’s
has to sit down with the Parkinson’s patient in order to visually recognize
patient in order to visually recognize
patient in order to visually recognize the severity of motor symptoms and use
the severity of motor symptoms and use
the severity of motor symptoms and use that information to prescribe
that information to prescribe
that information to prescribe medications and this process has two
medications and this process has two
medications and this process has two major problems one it’s subject to
major problems one it’s subject to
major problems one it’s subject to extensive human error and secondly it’s
extensive human error and secondly it’s
extensive human error and secondly it’s very infrequent since a Parkinson’s
very infrequent since a Parkinson’s
very infrequent since a Parkinson’s patient live at most one or two
patient live at most one or two
patient live at most one or two monitoring appointments an entire month
monitoring appointments an entire month
monitoring appointments an entire month this prevents the physician
this prevents the physician
this prevents the physician from being able to get an accurate
from being able to get an accurate
from being able to get an accurate representation of how their patients
representation of how their patients
representation of how their patients motor symptoms are behaving when they’re
motor symptoms are behaving when they’re
motor symptoms are behaving when they’re at home and is therefore prevents them
at home and is therefore prevents them
at home and is therefore prevents them from being able to accurately prescribe
from being able to accurately prescribe
from being able to accurately prescribe medications to alleviate symptoms and
medications to alleviate symptoms and
medications to alleviate symptoms and that’s why I wanted to build something
that’s why I wanted to build something
that’s why I wanted to build something that could solve the solute install this
that could solve the solute install this
that could solve the solute install this problem and that’s why I built an
problem and that’s why I built an
problem and that’s why I built an intelligent wearable ring that can
intelligent wearable ring that can
intelligent wearable ring that can capture movement data for an entire day
capture movement data for an entire day
capture movement data for an entire day in a very passive manner and then
in a very passive manner and then
in a very passive manner and then analyze this data using machine learning
analyze this data using machine learning
analyze this data using machine learning so then I can diagnose the severity of
so then I can diagnose the severity of
so then I can diagnose the severity of motor symptoms at each and every hour of
motor symptoms at each and every hour of
motor symptoms at each and every hour of the day I’ll put a daily patient report
the day I’ll put a daily patient report
the day I’ll put a daily patient report a coherent report for the physician so
a coherent report for the physician so
a coherent report for the physician so then they can better prescribe
then they can better prescribe
then they can better prescribe medications and then overall use
medications and then overall use
medications and then overall use artificial intelligence to help the
artificial intelligence to help the
artificial intelligence to help the physician alleviate the symptoms of
physician alleviate the symptoms of
physician alleviate the symptoms of their patient so now I want to go back
their patient so now I want to go back
their patient so now I want to go back six years to when I was 10 years old
six years to when I was 10 years old
six years to when I was 10 years old where this entire journey actually began
where this entire journey actually began
where this entire journey actually began I was ten years old and I was just
I was ten years old and I was just
I was ten years old and I was just surfing YouTube like any ten-year-old
surfing YouTube like any ten-year-old
surfing YouTube like any ten-year-old would do and I came across a video of
would do and I came across a video of
would do and I came across a video of Muhammad Ali this was a video of him
Muhammad Ali this was a video of him
Muhammad Ali this was a video of him lighting the Olympic torch in Atlanta in
lighting the Olympic torch in Atlanta in
lighting the Olympic torch in Atlanta in 1996 during the Olympics and what I saw
1996 during the Olympics and what I saw
1996 during the Olympics and what I saw in this video was Muhammad Ali lighting
in this video was Muhammad Ali lighting
in this video was Muhammad Ali lighting the torch but his hands were shaking
the torch but his hands were shaking
the torch but his hands were shaking violently and I had no idea why that was
violently and I had no idea why that was
violently and I had no idea why that was happening so I called my dad over and I
happening so I called my dad over and I
happening so I called my dad over and I asked him why was his hand shaking and
asked him why was his hand shaking and
asked him why was his hand shaking and he told me that it’s because of
he told me that it’s because of
he told me that it’s because of Parkinson’s now I know I knew who
Parkinson’s now I know I knew who
Parkinson’s now I know I knew who Muhammad Ali was that he was a great
Muhammad Ali was that he was a great
Muhammad Ali was that he was a great boxer but I had absolutely no idea what
boxer but I had absolutely no idea what
boxer but I had absolutely no idea what Parkinson’s was so I decided to use the
Parkinson’s was so I decided to use the
Parkinson’s was so I decided to use the Internet I researched what Parkinson’s
Internet I researched what Parkinson’s
Internet I researched what Parkinson’s is and I found out about all the issues
is and I found out about all the issues
is and I found out about all the issues that it causes and that caused him and
that it causes and that caused him and
that it causes and that caused him and that made me think and be curious about
that made me think and be curious about
that made me think and be curious about how someone like Muhammad Ali can go
how someone like Muhammad Ali can go
how someone like Muhammad Ali can go about the regular day and have these
about the regular day and have these
about the regular day and have these shakings constantly happening how could
shakings constantly happening how could
shakings constantly happening how could they lead a normal lifestyle and that’s
they lead a normal lifestyle and that’s
they lead a normal lifestyle and that’s why Parkinson’s was always ingrained in
why Parkinson’s was always ingrained in
why Parkinson’s was always ingrained in my head I wanted to develop some
my head I wanted to develop some
my head I wanted to develop some solution for it and that’s why as I grew
solution for it and that’s why as I grew
solution for it and that’s why as I grew older I decided to volunteer at an
older I decided to volunteer at an
older I decided to volunteer at an Institute for Parkinson’s just to do
Institute for Parkinson’s just to do
Institute for Parkinson’s just to do simple filing work help out at the
simple filing work help out at the
simple filing work help out at the Institute talk to some patients but over
Institute talk to some patients but over
Institute talk to some patients but over there I was able to have firsthand
there I was able to have firsthand
there I was able to have firsthand experiences and talk with Parkinson’s
experiences and talk with Parkinson’s
experiences and talk with Parkinson’s patients I was able to learn their
patients I was able to learn their
patients I was able to learn their stories how there were adventurers
stories how there were adventurers
stories how there were adventurers before they got the disease how they
before they got the disease how they
before they got the disease how they wanted to climb mountains
wanted to climb mountains
wanted to climb mountains how they wanted to make companies be
how they wanted to make companies be
how they wanted to make companies be entrepreneurs but as soon as they were
entrepreneurs but as soon as they were
entrepreneurs but as soon as they were diagnosed with the disease that all
diagnosed with the disease that all
diagnosed with the disease that all prevented from being able to happen they
prevented from being able to happen they
prevented from being able to happen they had to they were unable to do simple
had to they were unable to do simple
had to they were unable to do simple tasks like eating food brushing their
tasks like eating food brushing their
tasks like eating food brushing their teeth being able to just drive their car
teeth being able to just drive their car
teeth being able to just drive their car those things became significantly harder
those things became significantly harder
those things became significantly harder and that’s why this was very touching to
and that’s why this was very touching to
and that’s why this was very touching to me I really wanted to build something
me I really wanted to build something
me I really wanted to build something that good it’s to some extent make their
that good it’s to some extent make their
that good it’s to some extent make their lives easier and be able to cope with
lives easier and be able to cope with
lives easier and be able to cope with their lives in a much simpler manner but
their lives in a much simpler manner but
their lives in a much simpler manner but I did not know how to do this at the age
I did not know how to do this at the age
I did not know how to do this at the age of 15 but naturally since I’ve
of 15 but naturally since I’ve
of 15 but naturally since I’ve participated in so many science fairs
participated in so many science fairs
participated in so many science fairs since first grade I started doing
since first grade I started doing
since first grade I started doing science fairs the answer came naturally
science fairs the answer came naturally
science fairs the answer came naturally the one the biggest thing that I’ve
the one the biggest thing that I’ve
the one the biggest thing that I’ve learned from participating in science
learned from participating in science
learned from participating in science fairs is that if you have a problem you
fairs is that if you have a problem you
fairs is that if you have a problem you should use whatever resources you have
should use whatever resources you have
should use whatever resources you have available to yourself and rapidly
available to yourself and rapidly
available to yourself and rapidly prototype keep building something even
prototype keep building something even
prototype keep building something even if it does not work at the beginning you
if it does not work at the beginning you
if it does not work at the beginning you just keep making new prototypes new
just keep making new prototypes new
just keep making new prototypes new versions and then from that at the end
versions and then from that at the end
versions and then from that at the end you’ll finally have something you can
you’ll finally have something you can
you’ll finally have something you can bring out to society that can actually
bring out to society that can actually
bring out to society that can actually help Society and that’s what so that
help Society and that’s what so that
help Society and that’s what so that resource that I figured out was
resource that I figured out was
resource that I figured out was artificial intelligence now at that time
artificial intelligence now at that time
artificial intelligence now at that time of age 15 I had no idea what artificial
of age 15 I had no idea what artificial
of age 15 I had no idea what artificial intelligence was but I watch tutorials I
intelligence was but I watch tutorials I
intelligence was but I watch tutorials I read blogs and I found out that that
read blogs and I found out that that
read blogs and I found out that that science was something purely magical
science was something purely magical
science was something purely magical simply being able to do automated
simply being able to do automated
simply being able to do automated diagnostics automatically making
diagnostics automatically making
diagnostics automatically making decisions seemed so cool to me that I
decisions seemed so cool to me that I
decisions seemed so cool to me that I made it my passion to read more and
made it my passion to read more and
made it my passion to read more and learn more about it so that I could
learn more about it so that I could
learn more about it so that I could become an expert in that field that’s
become an expert in that field that’s
become an expert in that field that’s why my goal became to use artificial
why my goal became to use artificial
why my goal became to use artificial intelligence to improve the lives of
intelligence to improve the lives of
intelligence to improve the lives of Parkinson’s patients and in order to do
Parkinson’s patients and in order to do
Parkinson’s patients and in order to do this I had three main stages that I had
this I had three main stages that I had
this I had three main stages that I had to achieve the first one wants to be
to achieve the first one wants to be
to achieve the first one wants to be able to train a machine learning model
able to train a machine learning model
able to train a machine learning model that can predict the severity of motor
that can predict the severity of motor
that can predict the severity of motor symptoms in an automated manner and then
symptoms in an automated manner and then
symptoms in an automated manner and then actually build some sort of a device
actually build some sort of a device
actually build some sort of a device that can capture movement data and then
that can capture movement data and then
that can capture movement data and then lastly go through a stage of prototype
lastly go through a stage of prototype
lastly go through a stage of prototype iteration constantly building new
iteration constantly building new
iteration constantly building new products and then in the end being able
products and then in the end being able
products and then in the end being able to build something that I can actually
to build something that I can actually
to build something that I can actually put on Parkinson’s patients give it to
put on Parkinson’s patients give it to
put on Parkinson’s patients give it to them so that then they’ll be able to
them so that then they’ll be able to
them so that then they’ll be able to benefit from that technology so from Xin
benefit from that technology so from Xin
benefit from that technology so from Xin learning I had 3 3 sub stages that I had
learning I had 3 3 sub stages that I had
learning I had 3 3 sub stages that I had to go
to go
to go the first one was to be able to find
the first one was to be able to find
the first one was to be able to find training data now in the training data
training data now in the training data
training data now in the training data what I wanted to do was to be able to
what I wanted to do was to be able to
what I wanted to do was to be able to see the various severe teas that are
see the various severe teas that are
see the various severe teas that are present in Parkinson’s disease be able
present in Parkinson’s disease be able
present in Parkinson’s disease be able to see that if I have movement data for
to see that if I have movement data for
to see that if I have movement data for Parkinson’s patient how can I build
Parkinson’s patient how can I build
Parkinson’s patient how can I build something that would be able to
something that would be able to
something that would be able to automatically distinguish between those
automatically distinguish between those
automatically distinguish between those severe ”tis so I found a dataset which
severe ”tis so I found a dataset which
severe ”tis so I found a dataset which had hours and hours of Parkinson’s
had hours and hours of Parkinson’s
had hours and hours of Parkinson’s disease data of varying disease severity
disease data of varying disease severity
disease data of varying disease severity and from that I extracted features from
and from that I extracted features from
and from that I extracted features from them meaning that I extracted the most
them meaning that I extracted the most
them meaning that I extracted the most important information just from raw
important information just from raw
important information just from raw acceleration data from someone’s hand so
acceleration data from someone’s hand so
acceleration data from someone’s hand so then from that meaningful representation
then from that meaningful representation
then from that meaningful representation I could teach a machine learning model
I could teach a machine learning model
I could teach a machine learning model to automatically see the differences in
to automatically see the differences in
to automatically see the differences in the severity and then automatically
the severity and then automatically
the severity and then automatically predict what that severity should be
predict what that severity should be
predict what that severity should be when it sees new Parkinson’s data so
when it sees new Parkinson’s data so
when it sees new Parkinson’s data so then with the algorithms made I also had
then with the algorithms made I also had
then with the algorithms made I also had to build a system that could capture
to build a system that could capture
to build a system that could capture movement data for an extended period of
movement data for an extended period of
movement data for an extended period of time
time
time and this was achieved through 3d
and this was achieved through 3d
and this was achieved through 3d printing another big fascination of mine
printing another big fascination of mine
printing another big fascination of mine just being able to visualize something
just being able to visualize something
just being able to visualize something put it into Google Sketchup and then
put it into Google Sketchup and then
put it into Google Sketchup and then send it off the 3d printer and the next
send it off the 3d printer and the next
send it off the 3d printer and the next day have it in your hand was just so
day have it in your hand was just so
day have it in your hand was just so magical to me that I just wanted to
magical to me that I just wanted to
magical to me that I just wanted to start as soon as I got the idea so
start as soon as I got the idea so
start as soon as I got the idea so that’s why I decided to use 3d printing
that’s why I decided to use 3d printing
that’s why I decided to use 3d printing to build something that a patient could
to build something that a patient could
to build something that a patient could wear and now this image shows all of my
wear and now this image shows all of my
wear and now this image shows all of my journey throughout my prototypes at the
journey throughout my prototypes at the
journey throughout my prototypes at the beginning I wanted to my goal was to
beginning I wanted to my goal was to
beginning I wanted to my goal was to simply build some algorithms but then I
simply build some algorithms but then I
simply build some algorithms but then I wanted to actually put it on to
wanted to actually put it on to
wanted to actually put it on to Parkinson’s patients I knew that if this
Parkinson’s patients I knew that if this
Parkinson’s patients I knew that if this was to be a reality I had to put it on
was to be a reality I had to put it on
was to be a reality I had to put it on patients get their data and see that was
patients get their data and see that was
patients get their data and see that was it was validated in order to do that I
it was validated in order to do that I
it was validated in order to do that I went through the first prototype just
went through the first prototype just
went through the first prototype just simply building an armband that could
simply building an armband that could
simply building an armband that could just capture data and then tell me if it
just capture data and then tell me if it
just capture data and then tell me if it was working or not so then from that I
was working or not so then from that I
was working or not so then from that I could get feedback from the patients and
could get feedback from the patients and
could get feedback from the patients and then move on to later later prototypes
then move on to later later prototypes
then move on to later later prototypes in which I really wanted to build
in which I really wanted to build
in which I really wanted to build something that was like jewelry so
something that was like jewelry so
something that was like jewelry so someone would want to wear so that’s why
someone would want to wear so that’s why
someone would want to wear so that’s why I decided to move on into a ring form
I decided to move on into a ring form
I decided to move on into a ring form factor so a patient would be like hey
factor so a patient would be like hey
factor so a patient would be like hey this is a wearable that I want to wear
this is a wearable that I want to wear
this is a wearable that I want to wear I’ll wear it every day and I’ll keep
I’ll wear it every day and I’ll keep
I’ll wear it every day and I’ll keep giving that data to the algorithms and
giving that data to the algorithms and
giving that data to the algorithms and then to my physician that they can help
then to my physician that they can help
then to my physician that they can help me get better prescriptions of
me get better prescriptions of
me get better prescriptions of medication
medication
medication so after that final prototype was made I
so after that final prototype was made I
so after that final prototype was made I knew that this is something I can bring
knew that this is something I can bring
knew that this is something I can bring out to Parkinson’s patients and in order
out to Parkinson’s patients and in order
out to Parkinson’s patients and in order to do that I had to go through
to do that I had to go through
to do that I had to go through Kickstarter this was a way that allowed
Kickstarter this was a way that allowed
Kickstarter this was a way that allowed me to connect with the community of
me to connect with the community of
me to connect with the community of Parkinson’s I was able to success
Parkinson’s I was able to success
Parkinson’s I was able to success succeed in my goal of first building
succeed in my goal of first building
succeed in my goal of first building something for Parkinson’s patients but
something for Parkinson’s patients but
something for Parkinson’s patients but then actually deploying it to them so
then actually deploying it to them so
then actually deploying it to them so that they could benefit from that
that they could benefit from that
that they could benefit from that technology so I built dozens and dozens
technology so I built dozens and dozens
technology so I built dozens and dozens of rings and gave it to Parkinson’s
of rings and gave it to Parkinson’s
of rings and gave it to Parkinson’s patients all around the world that now
patients all around the world that now
patients all around the world that now now they can actively use that with
now they can actively use that with
now they can actively use that with their physicians get better
their physicians get better
their physicians get better prescriptions prescriptions of
prescriptions prescriptions of
prescriptions prescriptions of medication and then be able to actually
medication and then be able to actually
medication and then be able to actually alleviate their symptoms and now I want
alleviate their symptoms and now I want
alleviate their symptoms and now I want to share just a one comment that came
to share just a one comment that came
to share just a one comment that came out from Kickstarter someone who really
out from Kickstarter someone who really
out from Kickstarter someone who really was able to use this product and I’m
was able to use this product and I’m
was able to use this product and I’m really happy that I was able to use
really happy that I was able to use
really happy that I was able to use artificial intelligence and use that
artificial intelligence and use that
artificial intelligence and use that technology to give benefit to these
technology to give benefit to these
technology to give benefit to these patients and the first one was that the
patients and the first one was that the
patients and the first one was that the person’s having deep brain stimulation
person’s having deep brain stimulation
person’s having deep brain stimulation surgery in which the doctors put
surgery in which the doctors put
surgery in which the doctors put electrodes in the patient’s brain so
electrodes in the patient’s brain so
electrodes in the patient’s brain so then their motor symptoms are alleviated
then their motor symptoms are alleviated
then their motor symptoms are alleviated and he plans on using this system to be
and he plans on using this system to be
and he plans on using this system to be able to monitor his tremors before he
able to monitor his tremors before he
able to monitor his tremors before he gets DBS surgery and then how it affects
gets DBS surgery and then how it affects
gets DBS surgery and then how it affects him after so it gives him that whole new
him after so it gives him that whole new
him after so it gives him that whole new way to see how that surgery benefitted
way to see how that surgery benefitted
way to see how that surgery benefitted him so now what’s this overall impact on
him so now what’s this overall impact on
him so now what’s this overall impact on Parkinson’s patients well it brings the
Parkinson’s patients well it brings the
Parkinson’s patients well it brings the doctors to the fingertips of a
doctors to the fingertips of a
doctors to the fingertips of a Parkinson’s patient through artificial
Parkinson’s patient through artificial
Parkinson’s patient through artificial intelligence I was able to bring all the
intelligence I was able to bring all the
intelligence I was able to bring all the analysis capabilities that a doctor can
analysis capabilities that a doctor can
analysis capabilities that a doctor can do to the fingertips of a patient it’s
do to the fingertips of a patient it’s
do to the fingertips of a patient it’s constantly following them constantly
constantly following them constantly
constantly following them constantly taking data constantly analyzing that
taking data constantly analyzing that
taking data constantly analyzing that data so you could give it to that
data so you could give it to that
data so you could give it to that physician so then they can better
physician so then they can better
physician so then they can better prescribe medications and help that
prescribe medications and help that
prescribe medications and help that patient so now through this experience
patient so now through this experience
patient so now through this experience with building something using artificial
with building something using artificial
with building something using artificial intelligence and working with
intelligence and working with
intelligence and working with Parkinson’s patients I was able to see
Parkinson’s patients I was able to see
Parkinson’s patients I was able to see that artificial intelligence can be used
that artificial intelligence can be used
that artificial intelligence can be used to solve many new issues let’s say if
to solve many new issues let’s say if
to solve many new issues let’s say if there’s other medical diagnostics that
there’s other medical diagnostics that
there’s other medical diagnostics that are right now inefficient expensive
are right now inefficient expensive
are right now inefficient expensive through artificial intelligence we can
through artificial intelligence we can
through artificial intelligence we can go back to the drawing board and figure
go back to the drawing board and figure
go back to the drawing board and figure out new and better ways to monitor those
out new and better ways to monitor those
out new and better ways to monitor those diseases we can see we can make new dev
diseases we can see we can make new dev
diseases we can see we can make new dev Isis we can make new algorithms to
Isis we can make new algorithms to
Isis we can make new algorithms to monitor those diseases in a way more
monitor those diseases in a way more
monitor those diseases in a way more efficient manner and that’s why I plan
efficient manner and that’s why I plan
efficient manner and that’s why I plan on continuing to work with patients
on continuing to work with patients
on continuing to work with patients continue to continue working on my
continue to continue working on my
continue to continue working on my artificial intelligence that I can
artificial intelligence that I can
artificial intelligence that I can improve Parkinson’s patients lives on a
improve Parkinson’s patients lives on a
improve Parkinson’s patients lives on a global scale thank you
”
Be First to Comment