WEBVTT

00:00:00.760 --> 00:00:03.240
Paying close attention to something:

00:00:03.280 --> 00:00:04.520
not that easy, is it?

00:00:05.520 --> 00:00:10.536
It's because our attention is pulled
in so many different directions at a time,

00:00:10.560 --> 00:00:14.640
and it's in fact pretty impressive
if you can stay focused.

00:00:16.360 --> 00:00:20.416
Many people think that attention
is all about what we are focusing on,

00:00:20.440 --> 00:00:25.240
but it's also about what information
our brain is trying to filter out.

00:00:26.320 --> 00:00:29.040
There are two ways
you direct your attention.

00:00:29.600 --> 00:00:31.160
First, there's overt attention.

00:00:31.640 --> 00:00:35.776
In overt attention,
you move your eyes towards something

00:00:35.800 --> 00:00:37.360
in order to pay attention to it.

00:00:38.360 --> 00:00:40.336
Then there's covert attention.

00:00:40.360 --> 00:00:44.376
In covert attention,
you pay attention to something,

00:00:44.400 --> 00:00:45.960
but without moving your eyes.

00:00:47.040 --> 00:00:48.680
Think of driving for a second.

00:00:50.960 --> 00:00:53.976
Your overt attention,
your direction of the eyes,

00:00:54.000 --> 00:00:55.656
are in front,

00:00:55.680 --> 00:00:57.456
but that's your covert attention

00:00:57.480 --> 00:01:00.560
which is constantly scanning
the surrounding area,

00:01:01.600 --> 00:01:03.480
where you don't actually look at them.

00:01:05.519 --> 00:01:07.456
I'm a computational neuroscientist,

00:01:07.480 --> 00:01:10.576
and I work on cognitive
brain machine interfaces,

00:01:10.600 --> 00:01:13.640
or bringing together
the brain and the computer.

00:01:14.720 --> 00:01:16.320
I love brain patterns.

00:01:16.720 --> 00:01:18.416
Brain patterns are important for us

00:01:18.440 --> 00:01:21.936
because based on them
we can build models for the computers,

00:01:21.960 --> 00:01:23.376
and based on these models

00:01:23.400 --> 00:01:27.616
computers can recognize
how well our brain functions,

00:01:27.640 --> 00:01:29.240
and if it doesn't function well,

00:01:30.080 --> 00:01:34.000
then these computers themselves
can be used as assistive devices

00:01:34.760 --> 00:01:35.960
for therapies.

00:01:36.480 --> 00:01:38.120
But that also means something,

00:01:39.360 --> 00:01:41.856
because choosing the wrong patterns

00:01:41.880 --> 00:01:43.776
will give us the wrong models

00:01:43.800 --> 00:01:45.456
and therefore the wrong therapies.

00:01:45.480 --> 00:01:46.680
Right?

00:01:47.640 --> 00:01:49.296
In case of attention,

00:01:49.320 --> 00:01:50.600
the fact that we can

00:01:51.800 --> 00:01:55.296
shift our attention not only by our eyes

00:01:55.320 --> 00:01:56.640
but also by thinking --

00:01:57.440 --> 00:02:01.520
that makes covert attention
an interesting model for computers.

00:02:02.280 --> 00:02:05.736
So I wanted to know
what are the brainwave patterns

00:02:05.760 --> 00:02:09.440
when you look overtly
or when you look covertly.

00:02:10.440 --> 00:02:12.200
I set up an experiment for that.

00:02:12.960 --> 00:02:15.696
In this experiment
there are two flickering squares,

00:02:15.720 --> 00:02:19.080
one of them flickering
at a slower rate than the other one.

00:02:20.600 --> 00:02:24.416
Depending on which of these flickers
you are paying attention to,

00:02:24.440 --> 00:02:28.400
certain parts of your brain
will start resonating in the same rate

00:02:29.200 --> 00:02:30.640
as that flickering rate.

00:02:32.000 --> 00:02:34.936
So by analyzing your brain signals,

00:02:34.960 --> 00:02:38.000
we can track where exactly
you are watching

00:02:38.760 --> 00:02:40.320
or you are paying attention to.

00:02:43.000 --> 00:02:47.216
So to see what happens in your brain
when you pay overt attention,

00:02:47.240 --> 00:02:50.496
I asked people to look directly
in one of the squares

00:02:50.520 --> 00:02:51.800
and pay attention to it.

00:02:52.760 --> 00:02:58.056
In this case, not surprisingly,
we saw that these flickering squares

00:02:58.080 --> 00:03:00.016
appeared in their brain signals

00:03:00.040 --> 00:03:02.400
which was coming
from the back of their head,

00:03:03.560 --> 00:03:06.960
which is responsible for the processing
of your visual information.

00:03:08.280 --> 00:03:10.616
But I was really interested

00:03:10.640 --> 00:03:13.800
to see what happens in your brain
when you pay covert attention.

00:03:14.480 --> 00:03:18.376
So this time I asked people
to look in the middle of the screen

00:03:18.400 --> 00:03:20.280
and without moving their eyes,

00:03:21.120 --> 00:03:23.840
to pay attention
to either of these squares.

00:03:25.120 --> 00:03:26.736
When we did that,

00:03:26.760 --> 00:03:30.696
we saw that both of these flickering rates
appeared in their brain signals,

00:03:30.720 --> 00:03:31.920
but interestingly,

00:03:32.640 --> 00:03:36.176
only one of them,
which was paid attention to,

00:03:36.200 --> 00:03:37.856
had stronger signals,

00:03:37.880 --> 00:03:40.136
so there was something in the brain

00:03:40.160 --> 00:03:42.696
which was handling this information

00:03:42.720 --> 00:03:48.920
so that thing in the brain was basically
the activation of the frontal area.

00:03:50.440 --> 00:03:53.416
The front part of your brain
is responsible

00:03:53.440 --> 00:03:56.320
for higher cognitive functions as a human.

00:03:57.160 --> 00:04:01.600
The frontal part,
it seems that it works as a filter

00:04:02.640 --> 00:04:07.016
trying to let information come in
only from the right flicker

00:04:07.040 --> 00:04:08.680
that you are paying attention to

00:04:09.400 --> 00:04:13.360
and trying to inhibit the information
coming from the ignored one.

00:04:15.400 --> 00:04:20.696
The filtering ability of the brain
is indeed a key for attention,

00:04:20.720 --> 00:04:23.496
which is missing in some people,

00:04:23.520 --> 00:04:26.000
for example in people with ADHD.

00:04:26.640 --> 00:04:31.656
So a person with ADHD
cannot inhibit these distractors,

00:04:31.680 --> 00:04:36.440
and that's why they can't focus
for a long time on a single task.

00:04:37.600 --> 00:04:39.136
But what if this person

00:04:39.160 --> 00:04:42.696
could play a specific computer game

00:04:42.720 --> 00:04:45.600
with his brain connected to the computer,

00:04:46.440 --> 00:04:48.560
and then train his own brain

00:04:49.360 --> 00:04:51.800
to inhibit these distractors?

00:04:53.680 --> 00:04:56.160
Well, ADHD is just one example.

00:04:57.200 --> 00:05:00.456
We can use this cognitive
brain machine interfaces

00:05:00.480 --> 00:05:02.680
for many other cognitive fields.

00:05:03.760 --> 00:05:05.536
It was just a few years ago

00:05:05.560 --> 00:05:11.280
that my grandfather had a stroke
and he lost complete ability to speak.

00:05:12.640 --> 00:05:15.976
He could understand everybody,
but there was no way to respond,

00:05:16.000 --> 00:05:18.480
even not writing
because he was illiterate.

00:05:20.000 --> 00:05:22.520
So he passed away in silence.

00:05:24.800 --> 00:05:27.136
I remember thinking at that time,

00:05:27.160 --> 00:05:31.056
what if we could have a computer

00:05:31.080 --> 00:05:32.440
which could speak for him?

00:05:33.840 --> 00:05:36.056
Now, after years that I am in this field,

00:05:36.080 --> 00:05:38.400
I can see that this might be possible.

00:05:40.240 --> 00:05:43.096
Imagine if we can find brainwave patterns

00:05:43.120 --> 00:05:46.560
when people think
about images or even letters,

00:05:47.720 --> 00:05:50.656
like the letter A generates
a different brainwave pattern

00:05:50.680 --> 00:05:52.400
than the letter B, and so on.

00:05:52.960 --> 00:05:56.640
Could a computer one day
communicate for people who can't speak?

00:05:57.640 --> 00:05:59.080
What if a computer

00:05:59.960 --> 00:06:04.520
can help us understand
the thoughts of a person in a coma?

00:06:05.840 --> 00:06:07.456
We are not there yet,

00:06:07.480 --> 00:06:10.216
but pay close attention.

00:06:10.240 --> 00:06:11.936
We will be there soon.

00:06:11.960 --> 00:06:13.456
Thank you.

00:06:13.480 --> 00:06:16.880
(Applause)