Cookies   I display ads to cover the expenses. See the privacy policy for more information. You can keep or reject the ads.

Video thumbnail
Today almost anybody has an
electronic device that they
interact with multiple times
a day. So you could be typing
or you could be touching a
screen, and we believe that
there's hidden information there
that can start to shed light upon
the psychomotor effects of
diseases like Parkinson's disease
and other conditions.
We found a way to detect motor-
impairment by how the people
interact with their devices, not
with an app, but the way you type
on a keyboard on a daily basis.
There's subtleties in the way
that we type: the way that our
fingers interact with keyboards.
When your finger moves down
towards the key and senses that
your finger impacts the key and
that the key is depressed, your
brain understands this and then
sends back a signal to release
the finger. When psychomotor
performance is impeded that
time can fluctuate. And those
are the types of things that
we're looking into; those subtle
effects and how typing happens.
So we thought, "What is the
easiest way to test it. What is
the low-hanging fruit?" Well,
fatigue. Fatigue is influenced
by motor-impairment so we decided
to run a test. We asked people
to type during the day and then
we woke them up during the night.
They were just asked to type a
text, regardless of what the text
was, and we measured the timing
of keys. Some people were typing
really fast, some people were
typing really slow but this is
actually not what we are looking
at, but rather the actual
interaction with a single key.
We found that there is a very
strong difference between people
when they are fully awake and
when they are under the effects
of sleep inertia or fatigue. And
these effects can be quite
striking once we put them through
pattern-recognition and machine-
learning algorithms that we're
developing. And that allows us
to detect quite subtle changes
in a very obvious and reliable
way. Research indicates that the
current diagnosis of Parkinson's
disease actually happens about
five to ten years after the
disease onset. And the belief is
that if we can start to do early
detection of Parkinson's then we
can do things like improve
treatment development. And the
benefits of our technique are
that we can detect this
throughout a persons life without
interacting with or interfering
with their daily activities.