WEBVTT

00:00:00.080 --> 00:00:01.670
The following
content is provided

00:00:01.670 --> 00:00:03.820
under a Creative
Commons license.

00:00:03.820 --> 00:00:06.550
Your support will help MIT
OpenCourseWare continue

00:00:06.550 --> 00:00:10.150
to offer high quality
educational resources for free.

00:00:10.150 --> 00:00:12.710
To make a donation or to
view additional materials

00:00:12.710 --> 00:00:16.620
from hundreds of MIT courses,
visit MIT OpenCourseWare

00:00:16.620 --> 00:00:17.275
at ocw.mit.edu.

00:00:26.180 --> 00:00:28.250
PROFESSOR: So
today, we are going

00:00:28.250 --> 00:00:32.330
to talk about the midget
and parasol channels.

00:00:32.330 --> 00:00:41.080
I will begin by reminding you
of what these two channels are

00:00:41.080 --> 00:00:43.630
that I very briefly
talked about before.

00:00:43.630 --> 00:00:50.010
Namely, if you look at the
Golgi stain of cross sections

00:00:50.010 --> 00:00:53.740
of the retina, when you look
at just the ganglion cells,

00:00:53.740 --> 00:01:00.310
it was discovered
initially by the inventors,

00:01:00.310 --> 00:01:05.060
if you will, of the
Golgi stain, namely,

00:01:05.060 --> 00:01:07.350
Golgi himself and Cajal.

00:01:07.350 --> 00:01:09.220
Cajal did this, initially.

00:01:09.220 --> 00:01:12.440
And then, subsequently, a
fellow whose name is down here,

00:01:12.440 --> 00:01:16.280
Polyak, has used the same
technique to study the retina.

00:01:16.280 --> 00:01:20.700
And what he discovered
was-- not Polyak initially,

00:01:20.700 --> 00:01:23.490
but Cajal initially
discovered, and then was

00:01:23.490 --> 00:01:26.480
verified by Polyak using
these anatomical techniques,

00:01:26.480 --> 00:01:28.770
namely, the Golgi
stain-- that there's

00:01:28.770 --> 00:01:32.700
one class of cells that have
very small dendritic arbors,

00:01:32.700 --> 00:01:35.820
as you can see here,
and another set that

00:01:35.820 --> 00:01:38.310
has much, much larger arbors.

00:01:38.310 --> 00:01:41.010
Now, once this was
discovered, subsequently when

00:01:41.010 --> 00:01:44.350
it became possible to
record from individual cells

00:01:44.350 --> 00:01:46.170
and to study their
receptive fields

00:01:46.170 --> 00:01:50.990
and their basic organization,
and also to look

00:01:50.990 --> 00:01:52.920
at them from a
different point of view,

00:01:52.920 --> 00:01:55.980
it was, first of
all, shown that these

00:01:55.980 --> 00:01:58.250
are two very distinct
classes of cells

00:01:58.250 --> 00:02:01.910
in terms of the dendritic
arbors they create.

00:02:01.910 --> 00:02:03.950
And as I mentioned
to you before,

00:02:03.950 --> 00:02:06.450
these cells are
called parasol cells

00:02:06.450 --> 00:02:09.340
because the dendritic arbors
look like an umbrella.

00:02:11.060 --> 00:02:14.220
And at comparable
eccentricities,

00:02:14.220 --> 00:02:17.500
the midget cells are three
times smaller in diameter

00:02:17.500 --> 00:02:20.410
than are the so-called
parasol cells.

00:02:20.410 --> 00:02:23.140
And then I showed you
some data the last time

00:02:23.140 --> 00:02:26.020
indicating that these
two types of cells

00:02:26.020 --> 00:02:30.090
conduct at different velocities
to the central nervous system,

00:02:30.090 --> 00:02:34.260
with these bigger cells
having bigger axons conducting

00:02:34.260 --> 00:02:37.260
significantly faster
than the midget cells,

00:02:37.260 --> 00:02:41.070
and that you get a
distribution when you record

00:02:41.070 --> 00:02:47.790
from all the axons in the
retinal geniculate pathway

00:02:47.790 --> 00:02:52.300
that they form separate
populations because, initially,

00:02:52.300 --> 00:02:54.670
the argument was that
maybe this is just

00:02:54.670 --> 00:02:56.880
a continuous
population of cells.

00:02:56.880 --> 00:02:58.410
But it turns out
these are indeed

00:02:58.410 --> 00:03:02.350
two very distinct classes
on the anatomical level

00:03:02.350 --> 00:03:05.510
as well as the conduction
velocity level.

00:03:05.510 --> 00:03:09.500
Then when it became possible
to record from single neurons

00:03:09.500 --> 00:03:11.215
and to study their
receptive fields,

00:03:11.215 --> 00:03:13.350
it was shown-- I
showed you this picture

00:03:13.350 --> 00:03:15.780
before as well--
that you have, first

00:03:15.780 --> 00:03:18.680
of all, center-surround
organization as discovered

00:03:18.680 --> 00:03:22.470
by Kuffler; and secondly, that
the midget and parasol cells

00:03:22.470 --> 00:03:26.390
are very different in size as
far as the receptive fields

00:03:26.390 --> 00:03:29.930
are concerned as well
as the dendritic arbors;

00:03:29.930 --> 00:03:32.130
and that the midget
cells in central retina

00:03:32.130 --> 00:03:36.120
receive an input from
just a single cone,

00:03:36.120 --> 00:03:39.060
whereas the parasol cells
got a multiple input.

00:03:39.060 --> 00:03:41.990
And another distinction
that was, at that time, made

00:03:41.990 --> 00:03:44.230
is that the neuronal
responses are

00:03:44.230 --> 00:03:47.330
such that midget cells respond
in a much more sustained

00:03:47.330 --> 00:03:49.780
fashion when you
activate the center

00:03:49.780 --> 00:03:52.750
mechanism than do
the parasol cells.

00:03:52.750 --> 00:03:55.070
So these respond much
more transiently.

00:03:55.070 --> 00:03:59.590
So this, then, were
initial cues as to what

00:03:59.590 --> 00:04:02.670
might be involved as to why
we have these two systems.

00:04:02.670 --> 00:04:04.740
And there were all
kinds of hypotheses.

00:04:04.740 --> 00:04:10.290
And then, eventually,
in the '60s and '70s

00:04:10.290 --> 00:04:14.550
and especially in the '80s and
'90s, all kinds of experiments

00:04:14.550 --> 00:04:18.029
were carried out to
try to establish just

00:04:18.029 --> 00:04:21.430
why these two separate
systems have evolved.

00:04:22.770 --> 00:04:28.880
So now, when the anatomical
work was further progressing,

00:04:28.880 --> 00:04:31.890
it was found, as I have
mentioned also the last time

00:04:31.890 --> 00:04:36.020
to you, that each cone in
the retina-- certainly true

00:04:36.020 --> 00:04:39.930
for the red and the
green cones, which

00:04:39.930 --> 00:04:42.400
would be equivalent
to the midget system--

00:04:42.400 --> 00:04:46.530
gives rise to an ON and
an OFF bipolar cell.

00:04:46.530 --> 00:04:49.110
Those are the ones we
discussed the last time.

00:04:49.110 --> 00:04:51.160
What I did not
discuss the last time,

00:04:51.160 --> 00:04:52.960
and I'm not going to
discuss today either

00:04:52.960 --> 00:04:56.900
but will the next time we
talk about color perception,

00:04:56.900 --> 00:04:59.560
is that the blue
cones-- how many of you

00:04:59.560 --> 00:05:02.973
remember what is the frequency
of blue cones in the retina?

00:05:02.973 --> 00:05:03.780
AUDIENCE: 1 in 8.

00:05:03.780 --> 00:05:04.870
PROFESSOR: 1 in 8.

00:05:04.870 --> 00:05:05.440
Very good.

00:05:06.530 --> 00:05:07.620
Excellent.

00:05:07.620 --> 00:05:11.480
So that being the case, that
already makes them different

00:05:11.480 --> 00:05:13.960
because the numerosity's
so much lower.

00:05:13.960 --> 00:05:16.620
And then people have done
all kinds of recordings,

00:05:16.620 --> 00:05:19.430
and to this day,
it's not quite clear

00:05:19.430 --> 00:05:21.380
what the connection
or pattern is.

00:05:21.380 --> 00:05:23.410
But as I said, next time,
we will talk about it

00:05:23.410 --> 00:05:25.570
in some detail when
we talk about color.

00:05:25.570 --> 00:05:28.280
But here, don't get frightened,
we will discuss this

00:05:28.280 --> 00:05:29.110
the next time.

00:05:29.110 --> 00:05:32.300
But at any rate, we
have the blue cones.

00:05:32.300 --> 00:05:35.440
And when we talk
about color opponency

00:05:35.440 --> 00:05:37.940
that I will explain
the next time,

00:05:37.940 --> 00:05:42.230
we think of the opponency
to blue being yellow,

00:05:42.230 --> 00:05:45.090
and the opponency
to red being green.

00:05:45.090 --> 00:05:47.690
And so the assumption
here was made

00:05:47.690 --> 00:05:49.540
that the connections
in this case,

00:05:49.540 --> 00:05:53.475
somehow, must take place
to create color opponency.

00:05:55.580 --> 00:05:58.950
And since we have only three
kinds of cones in the retina,

00:05:58.950 --> 00:06:01.430
in the primate-- red,
green, and blue--

00:06:01.430 --> 00:06:03.330
the opponency here
must, somehow,

00:06:03.330 --> 00:06:05.340
involve both the
red and green cones.

00:06:05.340 --> 00:06:08.960
So that's the complication that
we will discuss the next time.

00:06:08.960 --> 00:06:10.605
And so the assumption
was made that we

00:06:10.605 --> 00:06:14.160
have some so-called yellow/blue
cells and blue/yellow cells.

00:06:14.160 --> 00:06:16.660
And so I just want
to keep you puzzled

00:06:16.660 --> 00:06:20.470
as to what this
means, especially

00:06:20.470 --> 00:06:23.500
what it means with respect
to how we can see colors.

00:06:23.500 --> 00:06:25.570
And that's what we will
talk about the next time.

00:06:26.580 --> 00:06:29.610
So now, here we have,
by contrast-- and we

00:06:29.610 --> 00:06:33.010
go back here-- here, this
is, then, the midget system.

00:06:34.450 --> 00:06:36.190
And I showed you
the two extremes,

00:06:36.190 --> 00:06:38.850
which are very clear,
and in between, which

00:06:38.850 --> 00:06:43.090
would be somehow
creating the blue system.

00:06:43.090 --> 00:06:46.780
So that's the midget system
with very small receptive fields

00:06:46.780 --> 00:06:48.860
and very small cells.

00:06:48.860 --> 00:06:52.560
Now, by contrast, when you look
at the parasol system, what

00:06:52.560 --> 00:06:57.190
you see here is, as indicated
by the picture I showed you

00:06:57.190 --> 00:07:00.480
just a minute ago, is that the
receptive fields are bigger.

00:07:00.480 --> 00:07:02.390
The ganglion cell's
a lot bigger.

00:07:02.390 --> 00:07:06.180
And the dendritic arbors
are much more extensive.

00:07:06.180 --> 00:07:09.360
Now, that means, by the
way-- I shouldn't even say

00:07:09.360 --> 00:07:11.650
"by the way"-- that
means, very importantly,

00:07:11.650 --> 00:07:18.720
that the ON and OFF bipolar
cells that connect with the ON

00:07:18.720 --> 00:07:26.560
and OFF cells of the parasol
system are also much bigger

00:07:26.560 --> 00:07:30.010
and have much more extensive
arbors, dendritic arbors,

00:07:30.010 --> 00:07:32.905
because they sample many
cells instead of just one.

00:07:34.880 --> 00:07:38.420
But overall, what this
means is that when

00:07:38.420 --> 00:07:41.930
you look at the bipolar
cells in the retina,

00:07:41.930 --> 00:07:45.400
there are about three
times as many bipolar cells

00:07:45.400 --> 00:07:46.650
than there are photoreceptors.

00:07:47.666 --> 00:07:49.290
Now, if you think
about this-- somebody

00:07:49.290 --> 00:07:55.770
remembered the last
time-- that in the retina,

00:07:55.770 --> 00:08:00.060
we have about 50
million cones, and we

00:08:00.060 --> 00:08:04.155
have about 120 to 150 rods.

00:08:06.350 --> 00:08:08.990
And so, then, if you think
about the fact, at least,

00:08:08.990 --> 00:08:12.881
for the cone system, we have
three times as many bipolar

00:08:12.881 --> 00:08:13.380
cells.

00:08:13.380 --> 00:08:15.965
This only applies fully
to the cone system.

00:08:19.110 --> 00:08:24.840
That still means that we
have 150 million bipolar

00:08:24.840 --> 00:08:25.770
cells in the retina.

00:08:25.770 --> 00:08:27.269
So it's absolutely
incredible, then,

00:08:27.269 --> 00:08:29.740
this tiny little thing that
you have in your head, which

00:08:29.740 --> 00:08:32.040
is less than an
inch in diameter,

00:08:32.040 --> 00:08:35.235
you have these millions and
millions and millions of cells.

00:08:35.235 --> 00:08:35.735
Amazing.

00:08:37.340 --> 00:08:41.010
So this is then the arrangement
for the parasol system.

00:08:41.010 --> 00:08:43.970
And now, we can
progress and make

00:08:43.970 --> 00:08:51.046
another point about how
clever the wiring became

00:08:51.046 --> 00:08:52.170
in the course of evolution.

00:08:53.290 --> 00:08:55.755
The cleverness here
is that, as you know,

00:08:55.755 --> 00:09:01.090
we only have a single layer
of photoreceptors, which,

00:09:01.090 --> 00:09:05.400
outside the fovea, has
a mix of rods and cones.

00:09:05.400 --> 00:09:08.050
But then if you look at the
receptive fields of ganglion

00:09:08.050 --> 00:09:12.390
cells, what you find is that
they have overlapping receptive

00:09:12.390 --> 00:09:15.500
fields for many
of the attributes.

00:09:15.500 --> 00:09:18.180
So you have overlapping
for ON and OFF, of course.

00:09:18.180 --> 00:09:20.460
And that's obvious
from the wiring.

00:09:20.460 --> 00:09:23.450
But you also have
overlap for the midget

00:09:23.450 --> 00:09:24.635
and the parasol cells.

00:09:25.860 --> 00:09:29.510
And lastly, there's an
overlap-- even if you just

00:09:29.510 --> 00:09:32.010
look at the rod
input-- which results

00:09:32.010 --> 00:09:36.060
in realizing that the
receptive fields become larger

00:09:36.060 --> 00:09:44.270
during dark adaptation by virtue
of the connections of the rod

00:09:44.270 --> 00:09:48.350
photoreceptors to the rod
bipolars than to the A,

00:09:48.350 --> 00:09:52.680
to amacrine cells, which then
connect with the same ganglion

00:09:52.680 --> 00:09:54.770
cells as do the cones.

00:09:55.950 --> 00:09:59.560
So this is, then, an
incredibly clever arrangement

00:09:59.560 --> 00:10:03.245
which enables you to see
things extremely well.

00:10:06.260 --> 00:10:10.390
And it's an incredible
feat of wiring

00:10:10.390 --> 00:10:13.500
that accomplish this
incredible arrangement.

00:10:13.500 --> 00:10:17.780
So now, another important
thing to consider, in addition

00:10:17.780 --> 00:10:20.730
to what I've said so
far about the midget

00:10:20.730 --> 00:10:23.505
and the parasol
cells, is that they

00:10:23.505 --> 00:10:25.460
have distribution over space.

00:10:25.460 --> 00:10:28.575
If you look at it from, say,
from the center of the eye,

00:10:28.575 --> 00:10:31.480
meaning the fovea,
going to the periphery,

00:10:31.480 --> 00:10:36.770
the ratio of midget and parasol
cells changes dramatically.

00:10:36.770 --> 00:10:39.300
In the center
here, in the fovea,

00:10:39.300 --> 00:10:41.445
you have a huge difference
in number of cells.

00:10:42.870 --> 00:10:47.490
But as you get to the periphery,
these two types of cells

00:10:47.490 --> 00:10:49.970
eventually become
equally numerous.

00:10:49.970 --> 00:10:52.610
Now, to show this
in relationship

00:10:52.610 --> 00:10:55.270
to the lateral
geniculate nucleus--

00:10:55.270 --> 00:10:57.925
again, to repeat, in the fovea,
you have an 8 to 1 ratio;

00:10:57.925 --> 00:11:00.620
in the periphery, they
have 1 to 1 ratio.

00:11:00.620 --> 00:11:05.450
And this is directly reflected
in the lateral geniculate

00:11:05.450 --> 00:11:06.350
nucleus.

00:11:06.350 --> 00:11:09.500
Everything I've shown you
so far about the geniculate

00:11:09.500 --> 00:11:11.970
was a six-layered structure.

00:11:11.970 --> 00:11:16.150
But those six layers
exist only to about

00:11:16.150 --> 00:11:18.850
an 18-degree eccentricity
from the fovea.

00:11:20.900 --> 00:11:24.640
And after that, the
lateral geniculate nucleus

00:11:24.640 --> 00:11:26.790
becomes a four-layered
structure.

00:11:26.790 --> 00:11:30.120
And as done here in
a schematic fashion,

00:11:30.120 --> 00:11:34.450
the four layers for
the left and right eyes

00:11:34.450 --> 00:11:37.610
are pretty much
equal in numerosity.

00:11:37.610 --> 00:11:40.680
And that creates,
then, the 1 to 1 ratio.

00:11:40.680 --> 00:11:43.760
So that makes one think,
why do we have this shift?

00:11:43.760 --> 00:11:47.510
There must be something very
important in central vision

00:11:47.510 --> 00:11:49.920
for which the midget
system is good for.

00:11:49.920 --> 00:11:55.880
And in peripheral retina,
the parasol system

00:11:55.880 --> 00:11:57.580
becomes more important.

00:11:57.580 --> 00:12:00.290
So we are going to
pay attention to that

00:12:00.290 --> 00:12:08.585
as we examine what the functions
are of these two systems.

00:12:09.640 --> 00:12:11.470
So how do we go
about finding out

00:12:11.470 --> 00:12:13.140
the functions of
these two systems?

00:12:13.140 --> 00:12:14.780
Well, there are several methods.

00:12:14.780 --> 00:12:16.730
The first one I'm
going to tell you about

00:12:16.730 --> 00:12:23.680
is to record from individual
neurons-- say, in the cortex--

00:12:23.680 --> 00:12:26.980
and determine what
kinds of inputs

00:12:26.980 --> 00:12:31.260
they get from the midget
and the parasol cells.

00:12:31.260 --> 00:12:36.820
So to do that, you first want
to look at the exact projections

00:12:36.820 --> 00:12:39.630
of the retinal
ganglion cells again.

00:12:39.630 --> 00:12:42.460
Just to remind you, here's the
lateral geniculate nucleus,

00:12:42.460 --> 00:12:45.440
again, in the region where
there are six layers.

00:12:45.440 --> 00:12:47.870
I already showed you
this picture before.

00:12:47.870 --> 00:12:51.070
And I told you that the
parvocellular layers project

00:12:51.070 --> 00:12:55.000
to 4C alpha; the magnocellular
layers to 4C beta;

00:12:55.000 --> 00:12:58.510
and those other cells, which
we'll talk about eventually,

00:12:58.510 --> 00:13:02.600
its lateral layers project to
the upper parts of the cortex.

00:13:02.600 --> 00:13:05.260
So now, we have a
good idea that even

00:13:05.260 --> 00:13:10.400
if the inputs to the
visual cortex-- meaning,

00:13:10.400 --> 00:13:16.080
in this case, of course,
the V1-- the two systems are

00:13:16.080 --> 00:13:20.139
separate in the input
layers, in 4C alpha and beta.

00:13:20.139 --> 00:13:21.680
And then the question
comes up, well,

00:13:21.680 --> 00:13:25.560
what happens when you look at
cells above and below the input

00:13:25.560 --> 00:13:26.580
layers?

00:13:26.580 --> 00:13:27.740
Do they converge?

00:13:27.740 --> 00:13:28.710
Or what's going on?

00:13:28.710 --> 00:13:31.400
So that's the kind of
questions we're going to ask.

00:13:31.400 --> 00:13:35.250
And we are going to
look at this more

00:13:35.250 --> 00:13:38.620
carefully here to find
out just how do you

00:13:38.620 --> 00:13:40.480
do an experiment like that.

00:13:40.480 --> 00:13:50.330
Well, one thing you can do here,
you can use a reversible agent

00:13:50.330 --> 00:13:53.080
that you can inject either
in the parvocellular

00:13:53.080 --> 00:13:55.800
or magnocellular layers
of the geniculate.

00:13:55.800 --> 00:13:59.070
And if you do that, if in
the parvocellular layers,

00:13:59.070 --> 00:14:00.435
this would be a blocking agent.

00:14:01.500 --> 00:14:03.500
Xylocaine is something
that's used frequently.

00:14:03.500 --> 00:14:05.230
There are several other agents.

00:14:05.230 --> 00:14:08.420
And so if you inject
that substance in here,

00:14:08.420 --> 00:14:12.840
you render the cells in the
geniculate unresponsive.

00:14:14.940 --> 00:14:17.640
It's to serve the same idea
as what we talked about

00:14:17.640 --> 00:14:21.740
with APB in the last time,
except this is not quite as

00:14:21.740 --> 00:14:25.040
neat because it's
not a pure chemical

00:14:25.040 --> 00:14:27.540
treatment like you did with APB.

00:14:27.540 --> 00:14:30.900
Secondly, you can do it in the
magnocellular layers as well

00:14:30.900 --> 00:14:32.960
and block that region.

00:14:32.960 --> 00:14:34.690
Now, if you do this
kind of experiment,

00:14:34.690 --> 00:14:37.090
let's first talk about
what happens in area of V1.

00:14:38.820 --> 00:14:41.810
Now, this is a very difficult
and complicated experiment.

00:14:43.130 --> 00:14:51.210
Sometimes, you can spend days
recording from a single animal

00:14:51.210 --> 00:14:53.900
because, first of all,
what you have to do

00:14:53.900 --> 00:14:57.490
is you have to put in electrodes
into the lateral geniculate

00:14:57.490 --> 00:15:02.787
nucleus, which is way down,
preferably into both regions,

00:15:02.787 --> 00:15:05.120
but there are many experiments
just to one or the other.

00:15:06.160 --> 00:15:07.905
And then put an
electrode into V1.

00:15:08.950 --> 00:15:11.650
Now, the next important
task, of course,

00:15:11.650 --> 00:15:16.830
is that you've got to record
from V1 in the region to which

00:15:16.830 --> 00:15:18.810
these cells or
these cells project.

00:15:20.670 --> 00:15:24.220
So to find that overlap
in the receptive fields

00:15:24.220 --> 00:15:26.660
that you need to do when
you record here, here,

00:15:26.660 --> 00:15:29.390
and here is that you have
to take many, many electrode

00:15:29.390 --> 00:15:33.032
penetrations until, finally,
you have an overlap.

00:15:33.032 --> 00:15:34.490
And once you have
that overlap, you

00:15:34.490 --> 00:15:39.580
can be sure that when you
inactivate this region

00:15:39.580 --> 00:15:43.050
by injecting the
Xylocaine, you can

00:15:43.050 --> 00:15:46.390
assess what the
responses are in the V1

00:15:46.390 --> 00:15:50.200
before, during, and
after the injection.

00:15:50.200 --> 00:15:53.580
And most importantly, before
and during, naturally.

00:15:53.580 --> 00:15:54.670
So let's look at that.

00:15:56.900 --> 00:16:00.450
This has been done
studying many cells

00:16:00.450 --> 00:16:02.769
to get an overview
of what's going on.

00:16:02.769 --> 00:16:04.310
And I'm going to
show you, initially,

00:16:04.310 --> 00:16:08.640
just one example of
a cell to give you

00:16:08.640 --> 00:16:10.380
a feel for what that's like.

00:16:10.380 --> 00:16:12.900
So here we have
the cell respond.

00:16:12.900 --> 00:16:14.600
What you do is you
take a bar of light

00:16:14.600 --> 00:16:17.080
and move it across the
receptive field-- brrp,

00:16:17.080 --> 00:16:19.560
brrp-- for each edge.

00:16:19.560 --> 00:16:21.780
So those are the two
cumulative responses.

00:16:21.780 --> 00:16:22.280
[INAUDIBLE]

00:16:24.490 --> 00:16:25.520
That's the normal.

00:16:25.520 --> 00:16:29.460
Now, you inject into the
parvocellular portions

00:16:29.460 --> 00:16:30.346
of the geniculate.

00:16:30.346 --> 00:16:32.220
And lo and behold, the
cell keeps responding.

00:16:33.290 --> 00:16:35.540
Then you inject into
the magno portions.

00:16:35.540 --> 00:16:38.320
And then lo and behold, the
cell still keeps responding.

00:16:38.320 --> 00:16:41.552
But when you inject
into both-- bango!

00:16:41.552 --> 00:16:42.385
There's no response.

00:16:43.710 --> 00:16:47.440
So that means that
this particular cell

00:16:47.440 --> 00:16:53.914
gets a convergent input from
the midget and the parasol cells

00:16:53.914 --> 00:16:56.080
as they pass through the
lateral geniculate nucleus.

00:16:57.570 --> 00:17:01.030
Now, if one then does
a systematic study

00:17:01.030 --> 00:17:05.220
and then records from many
cells, what has been found here

00:17:05.220 --> 00:17:08.369
is that some cells do
get an exclusive input

00:17:08.369 --> 00:17:12.010
from the midget cells.

00:17:12.010 --> 00:17:15.069
Some get an exclusive input
from the parasol cells.

00:17:15.069 --> 00:17:18.369
And some are just like the one
I've shown you here-- namely,

00:17:18.369 --> 00:17:20.829
they get a convergent input.

00:17:20.829 --> 00:17:26.859
So that tells us, then, what
the very, very basic nature is

00:17:26.859 --> 00:17:30.930
of the input to the visual
cortex, you, to some degree,

00:17:30.930 --> 00:17:32.760
keep separate the two systems.

00:17:32.760 --> 00:17:34.370
And to some degree,
you also have

00:17:34.370 --> 00:17:35.910
a condition where
they are united.

00:17:37.380 --> 00:17:44.646
So that's what
happens in area V1.

00:17:45.920 --> 00:17:48.290
So now, the next thing
you are going to ask

00:17:48.290 --> 00:17:53.090
is, well, what happens
in other cortical areas?

00:17:53.090 --> 00:17:57.150
So here, then, I'm showing
you the same method picture.

00:17:58.350 --> 00:18:02.250
But now, you have
recording in V4 and MT,

00:18:02.250 --> 00:18:04.510
doing the similar
kinds of injections

00:18:04.510 --> 00:18:08.360
that I just described in the
lateral geniculate nucleus.

00:18:08.360 --> 00:18:13.790
So when one does this,
some interesting results

00:18:13.790 --> 00:18:15.190
had been obtained.

00:18:15.190 --> 00:18:19.280
And as always, before
this kind of experiment

00:18:19.280 --> 00:18:22.680
that actually tested
this question,

00:18:22.680 --> 00:18:24.530
there were, of
course, hypotheses.

00:18:24.530 --> 00:18:29.400
Some people
hypothesized that area

00:18:29.400 --> 00:18:32.680
of V4, which had
at one time been

00:18:32.680 --> 00:18:36.170
proclaimed to be a
color area-- and we'll

00:18:36.170 --> 00:18:41.350
come back to that shortly--
it was claimed, therefore,

00:18:41.350 --> 00:18:45.930
that this area gets input
only from the midget system.

00:18:45.930 --> 00:18:49.560
And this area-- since it
has motion-selective cells,

00:18:49.560 --> 00:18:56.835
if you remember-- only gets
input from the parasol cells.

00:18:58.150 --> 00:19:02.780
So, therefore, we can
now have the acid test,

00:19:02.780 --> 00:19:05.770
thanks to some
remarkable work that

00:19:05.770 --> 00:19:13.040
had been done, much
of it by John Maunsell

00:19:13.040 --> 00:19:15.640
over at Harvard,
who had actually

00:19:15.640 --> 00:19:18.360
worked here at MIT before then.

00:19:18.360 --> 00:19:20.910
And he did an
experiment like that.

00:19:20.910 --> 00:19:25.310
And here's an example of a
single cell recording from V4.

00:19:27.230 --> 00:19:30.770
This is a magno block, and
this is the parvo block.

00:19:30.770 --> 00:19:34.050
This is before the block,
and this is after the block.

00:19:34.050 --> 00:19:36.270
So what you can see
here, this cell, again,

00:19:36.270 --> 00:19:38.570
response to a bar
moving across-- brrp,

00:19:38.570 --> 00:19:39.880
brrp-- like that.

00:19:39.880 --> 00:19:42.490
Here are the two
responses shown here as

00:19:42.490 --> 00:19:45.120
to how they come across
the receptive field.

00:19:45.120 --> 00:19:52.830
And you can see a vigorous
response after you do this

00:19:52.830 --> 00:20:02.100
and before you inject
Xylocaine into the geniculate

00:20:02.100 --> 00:20:04.920
to block, in this
case, the magno system;

00:20:04.920 --> 00:20:07.090
and in this case,
the parvo system.

00:20:07.090 --> 00:20:09.470
This is backwards from the
way I usually presented.

00:20:09.470 --> 00:20:11.190
I usually like to
put parvo first.

00:20:11.190 --> 00:20:13.220
Preference-- it's the bias.

00:20:13.220 --> 00:20:14.850
At any rate, you
can see what happens

00:20:14.850 --> 00:20:18.950
is dramatic for this particular
cell after you injected,

00:20:18.950 --> 00:20:22.610
you blocked the magno system.

00:20:22.610 --> 00:20:26.190
The cell-- we said has
spontaneous activity--

00:20:26.190 --> 00:20:29.300
but it no longer
responds to the edges.

00:20:29.300 --> 00:20:32.510
By contrast, when
you do a parvo block,

00:20:32.510 --> 00:20:34.780
you only get a small effect.

00:20:34.780 --> 00:20:39.450
There is a reduction for
this particular cell,

00:20:39.450 --> 00:20:44.290
but it's still responding.

00:20:46.120 --> 00:20:49.390
So that is the example of
a recording in area of V4.

00:20:50.670 --> 00:20:54.180
Now, let us go and ask the
question, what about MT?

00:20:54.180 --> 00:20:55.340
Now, here's an example.

00:20:55.340 --> 00:20:59.580
Same arrangement-- magno first,
parvo second, before the block,

00:20:59.580 --> 00:21:00.730
after the block.

00:21:00.730 --> 00:21:03.670
In this case, in
MT, you get a cell

00:21:03.670 --> 00:21:07.720
that totally stopped
responding after magno block

00:21:07.720 --> 00:21:09.540
to this moving stimulus.

00:21:09.540 --> 00:21:13.130
And here, what you see
is after parvo block,

00:21:13.130 --> 00:21:14.300
the response continues.

00:21:15.650 --> 00:21:18.760
So now, this is just two cells.

00:21:18.760 --> 00:21:22.640
Now, to be sure that these
two cells are generally

00:21:22.640 --> 00:21:26.900
representative of these
observations, what you need

00:21:26.900 --> 00:21:31.830
to do is to collect this kind
of information from many cells,

00:21:31.830 --> 00:21:34.660
and then come up with a
qualitative statement.

00:21:34.660 --> 00:21:37.660
So what you can do is
you can, for each cell,

00:21:37.660 --> 00:21:39.850
when you do this
experiment, you can't

00:21:39.850 --> 00:21:42.640
determine how much the cell
fired here to the stimulus

00:21:42.640 --> 00:21:44.540
and how much it fired here.

00:21:44.540 --> 00:21:47.030
And then you can get a
ratio of that response.

00:21:47.030 --> 00:21:51.840
You can turn it into, like,
a percentage or maybe just

00:21:51.840 --> 00:21:54.650
to score them as 0 to 1 number.

00:21:55.810 --> 00:21:58.240
And then if you do that
for a whole bunch of cells,

00:21:58.240 --> 00:22:00.640
what you find is shown here.

00:22:00.640 --> 00:22:03.190
This is a bunch of cells in V4.

00:22:03.190 --> 00:22:05.460
This is a bunch of cells in MT.

00:22:05.460 --> 00:22:09.960
And what you can see here
is that in V4, first of all,

00:22:09.960 --> 00:22:16.530
you have a medium degree
of blockage, nothing major,

00:22:16.530 --> 00:22:17.680
but some blockage.

00:22:17.680 --> 00:22:20.320
But most importantly,
you get blockage

00:22:20.320 --> 00:22:23.490
both for parvocellular and
magnocellular inactivation.

00:22:24.690 --> 00:22:28.470
By contrast, in
MT, what you find

00:22:28.470 --> 00:22:32.480
is that when you block
magnocellular geniculate,

00:22:32.480 --> 00:22:36.600
most of the cells are
dramatically affected.

00:22:36.600 --> 00:22:44.830
Only a few cells are blocked
as a result of a parvocellular

00:22:44.830 --> 00:22:46.010
block.

00:22:46.010 --> 00:22:52.670
So the outcome of this is
that the hypothesis pertaining

00:22:52.670 --> 00:22:58.130
to MT, that it gets mostly an
input from the parasol system,

00:22:58.130 --> 00:22:59.410
is correct.

00:22:59.410 --> 00:23:03.160
But the idea that V4
gets an input only

00:23:03.160 --> 00:23:07.030
from the midget system
is obviously incorrect,

00:23:07.030 --> 00:23:10.410
showing that it gets
an input from both.

00:23:10.410 --> 00:23:16.230
So to then summarize
the wiring diagram here,

00:23:16.230 --> 00:23:18.830
what you have here is
the eye, of course.

00:23:18.830 --> 00:23:21.100
And then you have the
midget and parasol cells

00:23:21.100 --> 00:23:24.200
that project, respectively,
to the parvocellular

00:23:24.200 --> 00:23:27.850
and magnocellular layers of
the lateral geniculate nucleus,

00:23:27.850 --> 00:23:29.490
then they go off to the cortex.

00:23:29.490 --> 00:23:31.630
And if you remember,
in the cortex,

00:23:31.630 --> 00:23:34.920
they terminated
4C alpha and beta.

00:23:34.920 --> 00:23:38.000
And then, also, as I told
you, up here, you already

00:23:38.000 --> 00:23:41.640
have many cells above and
below the input layers

00:23:41.640 --> 00:23:46.070
that get a convergent input from
the midget and parasol cells.

00:23:46.070 --> 00:23:50.940
So what happens then, in terms
of the projections to higher

00:23:50.940 --> 00:23:55.310
cortical areas-- to talk
about V2, for example--

00:23:55.310 --> 00:23:57.480
you get some cells
which are purely

00:23:57.480 --> 00:24:00.500
driven by the midget
system, some cells purely

00:24:00.500 --> 00:24:05.850
driven by the parasol, and many
cells that are driven by both.

00:24:05.850 --> 00:24:08.570
So V2 becomes very
complicated because they

00:24:08.570 --> 00:24:10.320
have different subdivisions.

00:24:10.320 --> 00:24:12.830
Maybe in V2-- we talked
about it a little bit--

00:24:12.830 --> 00:24:17.080
that may be receiving
different input from area of V1

00:24:17.080 --> 00:24:18.910
in terms of whether
they are driven

00:24:18.910 --> 00:24:21.260
by the midget or
the parasol cells.

00:24:21.260 --> 00:24:26.560
Now, most notably, when
from V2 and from V1,

00:24:26.560 --> 00:24:29.510
you look at the
projections to area MT,

00:24:29.510 --> 00:24:32.180
the middle temporal
area, what you find

00:24:32.180 --> 00:24:34.990
is that this is
heavily dominated

00:24:34.990 --> 00:24:38.740
by the input from
the parasol cells.

00:24:38.740 --> 00:24:41.080
Then if you go beyond
that, this continues

00:24:41.080 --> 00:24:43.950
to the parietal lobe from MT.

00:24:43.950 --> 00:24:47.105
But then, when you
go from V2 to V4,

00:24:47.105 --> 00:24:49.520
then temporal, then the
frontal lobe, what you

00:24:49.520 --> 00:24:51.630
find that there's
a mix of inputs

00:24:51.630 --> 00:24:52.755
from both of these systems.

00:24:54.040 --> 00:24:57.330
So it highlights the fact
that, indeed, these things

00:24:57.330 --> 00:24:59.210
are quite complicated.

00:24:59.210 --> 00:25:04.410
And so we need to now
turn to a different method

00:25:04.410 --> 00:25:09.320
to try to ferret out beyond just
establishing the connections

00:25:09.320 --> 00:25:11.935
as to what on earth these
two systems are for.

00:25:15.790 --> 00:25:19.880
And needless to say, hypotheses
were rampant about this.

00:25:21.440 --> 00:25:23.050
And what I'm going
to do now, I'm

00:25:23.050 --> 00:25:26.090
going to tell you about
how one can go about

00:25:26.090 --> 00:25:29.990
and how many investigators have
gone about trying to determine

00:25:29.990 --> 00:25:33.809
what the functions are of
these two systems in processing

00:25:33.809 --> 00:25:34.600
visual information.

00:25:35.744 --> 00:25:36.660
So how do you do that?

00:25:38.150 --> 00:25:41.690
Well, the way you
do that is you can

00:25:41.690 --> 00:25:44.530
use what is called
lesion studies.

00:25:44.530 --> 00:25:46.640
What you can do is
you can selectively

00:25:46.640 --> 00:25:50.130
block either the
parvocellular or magnocellular

00:25:50.130 --> 00:25:53.470
systems at the level of
lateral geniculate nucleus

00:25:53.470 --> 00:25:57.975
because, by lucky happenstance,
the parvocellular layer is

00:25:57.975 --> 00:26:00.350
getting input from midget,
and the magnocellular layer is

00:26:00.350 --> 00:26:03.040
from the from the parasol cells.

00:26:03.040 --> 00:26:04.460
So you can make lesions, then.

00:26:04.460 --> 00:26:05.820
How do you make lesions?

00:26:05.820 --> 00:26:07.760
There are variety of
ways of making lesions.

00:26:07.760 --> 00:26:09.730
That's a huge field.

00:26:09.730 --> 00:26:13.300
It applies not only to vision,
to many, many other areas,

00:26:13.300 --> 00:26:20.250
to try to make carefully
selective lesions

00:26:20.250 --> 00:26:22.010
in various parts of the brain.

00:26:22.010 --> 00:26:24.390
Now, it's not easy to do
that in the geniculate.

00:26:24.390 --> 00:26:28.190
So what you have to
do is, again, you

00:26:28.190 --> 00:26:32.800
go in to the lateral geniculate
nucleus with a microelectrode.

00:26:34.500 --> 00:26:36.060
Once you do that--
you already know

00:26:36.060 --> 00:26:37.890
the layout of the
geniculate-- you

00:26:37.890 --> 00:26:40.790
can find out where the
receptive fields are located,

00:26:40.790 --> 00:26:42.560
and you can
determine whether you

00:26:42.560 --> 00:26:44.820
are recording from the
parvocellular or magnocellular

00:26:44.820 --> 00:26:45.319
layers.

00:26:45.319 --> 00:26:47.610
You're going to adjust
the depth of the electrode

00:26:47.610 --> 00:26:50.170
to be either in the
parvocellular or magnocellular

00:26:50.170 --> 00:26:51.280
layers.

00:26:51.280 --> 00:26:54.560
And once you have established
the receptive field location

00:26:54.560 --> 00:26:58.280
and were certain about in
what parts of the geniculate

00:26:58.280 --> 00:27:03.150
you are recording, you can
then proceed to make a lesion.

00:27:03.150 --> 00:27:05.510
Now, there are a number
of ways of making lesions.

00:27:05.510 --> 00:27:07.320
One of them is
called heat lesions.

00:27:07.320 --> 00:27:09.380
You should take a
metal microelectrode.

00:27:09.380 --> 00:27:12.370
You take it down
there, and then you

00:27:12.370 --> 00:27:17.140
pass some current to make
the tip of the electrode hot,

00:27:17.140 --> 00:27:20.630
you can affect, as you know,
maybe a millimeter area

00:27:20.630 --> 00:27:23.310
or something like that in the
lateral geniculate nucleus.

00:27:23.310 --> 00:27:25.560
I'm talking about
small, small areas.

00:27:25.560 --> 00:27:33.020
Another alternative that you
can use is to inject a chemical.

00:27:33.020 --> 00:27:36.760
One of those commonly used
is called Ibotenic acid,

00:27:36.760 --> 00:27:39.980
which is a very nice attribute,
that when it causes a lesion,

00:27:39.980 --> 00:27:42.495
the borders are nicely
clearly defined.

00:27:43.630 --> 00:27:47.760
So no matter how you
do this, then once it

00:27:47.760 --> 00:27:51.910
is done in a monkey, then you
can study a monkey for months

00:27:51.910 --> 00:27:54.480
on end to see what
the vision is,

00:27:54.480 --> 00:27:55.730
of how the vision is affected.

00:27:57.750 --> 00:28:03.730
And then after you've done
that, you then process the brain

00:28:03.730 --> 00:28:05.530
and look at the lateral
geniculate nucleus

00:28:05.530 --> 00:28:10.430
to see what was the size
and location of the lesion.

00:28:12.040 --> 00:28:15.110
Now then, in some cases, it's
"Oh, my god, you did both.

00:28:15.110 --> 00:28:16.280
Nyah, nyah, nyah."

00:28:16.280 --> 00:28:17.250
You screwed up.

00:28:17.250 --> 00:28:21.050
So months and months of
work goes down the toilet.

00:28:21.050 --> 00:28:23.680
But in some cases,
you get a good effect.

00:28:23.680 --> 00:28:25.420
And that, eventually,
if you do it

00:28:25.420 --> 00:28:27.860
several times on
several monkeys can

00:28:27.860 --> 00:28:30.920
result in a solid publication.

00:28:30.920 --> 00:28:34.870
So that is the basic process
for making the lesions.

00:28:35.880 --> 00:28:41.400
Now, let me then
move on and broach

00:28:41.400 --> 00:28:43.354
the next important
topic-- namely,

00:28:43.354 --> 00:28:44.520
what are you going to study?

00:28:45.680 --> 00:28:51.930
So what you have to study since
you want to have an open mind

00:28:51.930 --> 00:28:56.320
and you don't want to say, "oh,
yeah, everything in the brain

00:28:56.320 --> 00:28:59.040
does color" or
something like that,

00:28:59.040 --> 00:29:00.745
and so instead of
just studying color,

00:29:00.745 --> 00:29:02.620
you have to study
many other aspects.

00:29:06.200 --> 00:29:12.940
So I'm going to tell you
about several kinds of tests

00:29:12.940 --> 00:29:13.860
that have the newest.

00:29:13.860 --> 00:29:15.970
And I will explain
each of those to you.

00:29:15.970 --> 00:29:17.330
But first, I will list them.

00:29:18.750 --> 00:29:24.220
The behavioral task that
is used-- very important--

00:29:24.220 --> 00:29:29.620
is that you've got to be able
to confine the crucial stimulus

00:29:29.620 --> 00:29:34.100
either to the area
that you had blocked

00:29:34.100 --> 00:29:36.180
or the area that is intact.

00:29:37.480 --> 00:29:40.000
So one way to do that,
which is the easiest way,

00:29:40.000 --> 00:29:41.505
is to use a detection task.

00:29:41.505 --> 00:29:43.840
We already talked
about that briefly.

00:29:43.840 --> 00:29:45.610
The monkey first fixates.

00:29:45.610 --> 00:29:48.880
That confines his
looking to that location.

00:29:48.880 --> 00:29:51.570
And then you can present
a stimulus like this.

00:29:51.570 --> 00:29:53.770
And if he makes a
saccade to it, he

00:29:53.770 --> 00:29:56.020
gets a drop of apple
juice for a reward.

00:29:56.020 --> 00:29:57.800
Now, that is a procedure.

00:29:57.800 --> 00:29:59.300
And then, of course,
our next trial,

00:29:59.300 --> 00:30:00.750
it appears someplace else.

00:30:00.750 --> 00:30:03.370
On each trial, it's in
different locations.

00:30:03.370 --> 00:30:06.640
And on some trials, you
can present the stimulus

00:30:06.640 --> 00:30:08.675
in the area that
had been blocked.

00:30:08.675 --> 00:30:13.380
You get it from by
magnocellular or parvocellular

00:30:13.380 --> 00:30:18.200
lesions or infusions
or in an intact area.

00:30:18.200 --> 00:30:21.950
And that, then, enables
you to compare performance

00:30:21.950 --> 00:30:25.100
in regions where the monkey's
performance is normal

00:30:25.100 --> 00:30:29.150
because he's intact and in those
regions where he lacks either

00:30:29.150 --> 00:30:31.960
an input from the magnocellular
or from the parvocellular

00:30:31.960 --> 00:30:34.840
layers of the lateral
geniculate nucleus,

00:30:34.840 --> 00:30:38.250
meaning from the midget
and parasol cells.

00:30:38.250 --> 00:30:42.030
So now, another
task which is used

00:30:42.030 --> 00:30:46.190
to be able to carry out a
more thorough examination

00:30:46.190 --> 00:30:48.190
of visual capacities
is his discrimination.

00:30:49.200 --> 00:30:51.860
In which case,
after fixation, you

00:30:51.860 --> 00:30:54.190
have a whole bunch
of stimuli coming on.

00:30:54.190 --> 00:30:56.830
This is very effective
for studying color.

00:30:56.830 --> 00:30:59.690
And in this case, this
discrimination task often

00:30:59.690 --> 00:31:03.130
is called the so-called
oddities task.

00:31:03.130 --> 00:31:05.120
That's an easy way
to remember it.

00:31:05.120 --> 00:31:06.910
"Oddity" because this is odd.

00:31:06.910 --> 00:31:10.710
All the others, the distractors,
if you will, are the same.

00:31:10.710 --> 00:31:12.370
Only one stimulus is different.

00:31:13.299 --> 00:31:15.590
And, of course, the monkey
has to make a direct saccade

00:31:15.590 --> 00:31:19.570
to this stimulus to get a drop
of apple juice for reward.

00:31:19.570 --> 00:31:23.210
So that would be a so-called
discrimination task.

00:31:23.210 --> 00:31:26.600
So these are the two basic
tasks that can be used.

00:31:26.600 --> 00:31:30.780
And now, we can proceed and ask
the question of what exactly,

00:31:30.780 --> 00:31:33.910
what kinds of visual
capacities should be studied?

00:31:33.910 --> 00:31:37.065
So let me just say one
more thing, by the way.

00:31:37.065 --> 00:31:41.080
The way this is
done in a laboratory

00:31:41.080 --> 00:31:44.730
is that you have a performance
monitor, so-called,

00:31:44.730 --> 00:31:53.170
where the computer puts a square
around each of the dots that

00:31:53.170 --> 00:31:54.794
will have appeared--
in this case,

00:31:54.794 --> 00:31:56.460
this should be
discrimination-- or where

00:31:56.460 --> 00:31:59.150
they will appear singly
in the detection task.

00:31:59.150 --> 00:32:01.620
And if the monkey
makes a saccade that

00:32:01.620 --> 00:32:04.980
lends the saccadic eye movement,
which you are recording,

00:32:04.980 --> 00:32:08.980
into the square area
there, then automatically

00:32:08.980 --> 00:32:11.120
a drop of apple
juice is discharged.

00:32:11.120 --> 00:32:13.810
But if he were to make a
similar saccade to here,

00:32:13.810 --> 00:32:15.610
he will not get
rewarded because this

00:32:15.610 --> 00:32:17.412
is the only correct position.

00:32:17.412 --> 00:32:19.620
And on each trial, that's
going to be someplace else.

00:32:19.620 --> 00:32:21.279
Everybody understands
the method?

00:32:21.279 --> 00:32:22.445
It's fairly straightforward.

00:32:23.540 --> 00:32:28.770
So now, we can move on
and look at an example

00:32:28.770 --> 00:32:32.430
of the kinds of lesions
that this can create.

00:32:32.430 --> 00:32:34.620
Here is a lateral
geniculate nucleus again,

00:32:34.620 --> 00:32:36.150
the six-layered portion.

00:32:36.150 --> 00:32:39.910
And you can see this area
here, that this is a lesion.

00:32:39.910 --> 00:32:42.560
There are no cells
functional here,

00:32:42.560 --> 00:32:45.452
which affects only the
parvocellular layers.

00:32:45.452 --> 00:32:47.160
You see here the two
magnocellular layers

00:32:47.160 --> 00:32:48.120
are normal.

00:32:48.120 --> 00:32:52.070
So this is the lesion
area which affects

00:32:52.070 --> 00:32:57.430
a few degrees of visual
angle-- in this case,

00:32:57.430 --> 00:32:59.320
in the lower part
of the visual field.

00:32:59.320 --> 00:33:01.420
Now, here we have
a magno lesion.

00:33:01.420 --> 00:33:02.800
If you just look
at this quickly,

00:33:02.800 --> 00:33:04.630
you say, what lesion
are you talking about?

00:33:04.630 --> 00:33:06.820
But if you look closely,
can you see this here?

00:33:06.820 --> 00:33:08.630
There are no cells here.

00:33:08.630 --> 00:33:14.210
So this is the region where both
layers, magnocellular layers

00:33:14.210 --> 00:33:16.710
that get the input
from the parasol cells,

00:33:16.710 --> 00:33:17.710
had been blocked.

00:33:17.710 --> 00:33:19.730
So that is the block there.

00:33:19.730 --> 00:33:24.630
So this is, then, an example
of a successful lesion.

00:33:24.630 --> 00:33:26.790
And once you do these
kinds of experiments--

00:33:26.790 --> 00:33:29.320
some of these experiments
can take a couple of years,

00:33:29.320 --> 00:33:39.250
maybe longer-- you end up
having some monkeys who

00:33:39.250 --> 00:33:44.420
have neat lesions like
this, then it's a control.

00:33:44.420 --> 00:33:45.960
Let me just add
that, which I'm not

00:33:45.960 --> 00:33:48.440
going to show any pictures of.

00:33:48.440 --> 00:33:51.440
But you can also make,
on purpose, a lesion

00:33:51.440 --> 00:33:55.180
where you block both of
them-- both the magno

00:33:55.180 --> 00:33:56.440
and the parvocellular layers.

00:33:56.440 --> 00:33:59.220
You just block out a portion
of the lateral geniculate

00:33:59.220 --> 00:34:00.410
completely.

00:34:00.410 --> 00:34:02.390
Now, that's an
important control.

00:34:02.390 --> 00:34:06.050
For all kinds of experiments,
controls are essential.

00:34:06.050 --> 00:34:09.870
Now, let me tell you-- this
is available in the published

00:34:09.870 --> 00:34:12.179
material that you have
been asked to read--

00:34:12.179 --> 00:34:17.580
that when you block
both magno and parvo,

00:34:17.580 --> 00:34:20.170
maybe a quarter of the
lateral geniculate nucleus,

00:34:20.170 --> 00:34:26.750
whenever you present stimuli in
the area that has a topographic

00:34:26.750 --> 00:34:29.540
correspondence to
the lesion's region,

00:34:29.540 --> 00:34:31.300
the monkey cannot see a thing.

00:34:31.300 --> 00:34:34.449
He cannot perform any
of those tasks at all.

00:34:34.449 --> 00:34:38.560
So clearly, whatever
effect one gets here

00:34:38.560 --> 00:34:40.949
with these selective
lesions can then

00:34:40.949 --> 00:34:43.360
be ascribed-- if you see
deficit after this one

00:34:43.360 --> 00:34:46.810
or deficit after that one--
can be ascribed to what

00:34:46.810 --> 00:34:50.070
the midget and the
parasol cell systems do.

00:34:51.310 --> 00:34:53.500
So that, then, is the procedure.

00:34:53.500 --> 00:34:58.720
And now, we can move on and
list the perceptual functions

00:34:58.720 --> 00:35:00.380
that one wants to test.

00:35:00.380 --> 00:35:03.540
Now, think about it for
a minute, just quickly.

00:35:03.540 --> 00:35:05.040
What kinds of
functions would I want

00:35:05.040 --> 00:35:06.880
to test if I'm running
this experiment?

00:35:08.820 --> 00:35:11.750
How can we break down
the multitude of things

00:35:11.750 --> 00:35:16.010
that we have to process
into some basic functions?

00:35:16.010 --> 00:35:18.495
Well, first of all,
very important function,

00:35:18.495 --> 00:35:20.120
even though it may
have now [INAUDIBLE]

00:35:20.120 --> 00:35:22.390
is called contrast sensitivity.

00:35:22.390 --> 00:35:24.950
Because everything
you look at, the light

00:35:24.950 --> 00:35:28.590
reflects just about from
everything that you look at,

00:35:28.590 --> 00:35:34.490
and so the contrast
of the stimuli that's

00:35:34.490 --> 00:35:37.840
on a white sheet of paper
or a gray sheet of paper

00:35:37.840 --> 00:35:40.440
or when you look at
photographs is highly varied.

00:35:41.859 --> 00:35:43.400
And the question
is, how well can you

00:35:43.400 --> 00:35:45.720
see the different
levels of contrast?

00:35:45.720 --> 00:35:48.510
Another one, of course,
that's obvious is color.

00:35:48.510 --> 00:35:51.660
How well can you process
color information?

00:35:51.660 --> 00:35:53.495
And then another one is pattern.

00:35:54.980 --> 00:35:56.640
We often talk about
basic patterns.

00:35:57.920 --> 00:36:00.210
And the one I'm going
to tell you about mostly

00:36:00.210 --> 00:36:03.630
would be checkerboards
or something like that.

00:36:03.630 --> 00:36:06.680
Then we talk about texture.

00:36:06.680 --> 00:36:09.340
Most of the things that
we encounter in the world

00:36:09.340 --> 00:36:10.770
are textured.

00:36:10.770 --> 00:36:14.260
And so it's important for us
to be able to see texture.

00:36:14.260 --> 00:36:15.885
Then, of course,
shape, that's obvious.

00:36:17.040 --> 00:36:19.070
Then stereopsis.

00:36:19.070 --> 00:36:23.210
We will talk about stereopsis
in much more detail in a bit.

00:36:23.210 --> 00:36:27.100
What stereopsis involves, as
I've mentioned to you already,

00:36:27.100 --> 00:36:32.200
are the difference of the
input in the two eyes.

00:36:32.200 --> 00:36:35.590
And that difference is called
disparity, which the brain then

00:36:35.590 --> 00:36:39.130
interprets as depth.

00:36:39.130 --> 00:36:42.010
Another one, very important
because you often just see

00:36:42.010 --> 00:36:44.900
things appearing quickly
in very brief times,

00:36:44.900 --> 00:36:45.800
is to study flicker.

00:36:46.870 --> 00:36:49.200
And yet another one
is to study motion

00:36:49.200 --> 00:36:52.000
because that's so
central to our existence.

00:36:52.000 --> 00:36:54.060
And then to study brightness.

00:36:54.060 --> 00:36:57.560
And one thing I
haven't mentioned yet

00:36:57.560 --> 00:36:58.970
is scotopic vision.

00:36:58.970 --> 00:37:02.150
How well can you see on
the photopic and scotopic

00:37:02.150 --> 00:37:02.650
conditions?

00:37:04.140 --> 00:37:06.470
So those are, then,
the procedures.

00:37:06.470 --> 00:37:09.585
And so let us now look at the
first one of these-- contrast

00:37:09.585 --> 00:37:10.085
sensitivity.

00:37:11.430 --> 00:37:13.340
This has been
extensively studied,

00:37:13.340 --> 00:37:17.100
hundreds and hundreds of
papers, many of them in humans.

00:37:17.100 --> 00:37:20.420
And most commonly in these
papers, what they did

00:37:20.420 --> 00:37:23.400
was they used
sinusoidal gratings.

00:37:23.400 --> 00:37:27.170
And the way you do that,
then, in similar experiments

00:37:27.170 --> 00:37:32.570
as a detection experiment, you
present sinusoidal gratings

00:37:32.570 --> 00:37:36.370
with spatial frequency and whose
contrast you systematically

00:37:36.370 --> 00:37:36.870
vary.

00:37:37.900 --> 00:37:39.850
And when you do
that, you get what

00:37:39.850 --> 00:37:43.470
is called a contrast
sensitivity function.

00:37:43.470 --> 00:37:48.550
And so the most common one
that you read in many papers

00:37:48.550 --> 00:37:50.350
is plotted this way.

00:37:52.040 --> 00:37:55.234
This is spatial
frequency low, high.

00:37:56.340 --> 00:37:57.335
And this is contrast.

00:38:00.857 --> 00:38:03.620
Low, high.

00:38:05.810 --> 00:38:08.440
And then if you do that, then
you systematically studied

00:38:08.440 --> 00:38:10.540
this in humans,
interestingly enough,

00:38:10.540 --> 00:38:14.680
you get a function
like that-- meaning

00:38:14.680 --> 00:38:18.410
that in between levels
of spatial frequency,

00:38:18.410 --> 00:38:20.110
you see the best.

00:38:20.110 --> 00:38:23.260
And extreme levels, you
don't see quite as well.

00:38:23.260 --> 00:38:28.910
So one could use sinusoidal
gratings in animals.

00:38:28.910 --> 00:38:32.980
But sinusoidal gratings
aren't essential.

00:38:32.980 --> 00:38:39.910
What you can do instead is you
can present a checkerboard.

00:38:39.910 --> 00:38:42.842
And so here's an example,
same procedure as before.

00:38:42.842 --> 00:38:44.800
The monkey sees this and
makes a saccade to it.

00:38:44.800 --> 00:38:46.110
It gets you water.

00:38:46.110 --> 00:38:48.180
And then you vary
the spatial frequency

00:38:48.180 --> 00:38:49.410
and the contrast of it.

00:38:49.410 --> 00:38:50.520
So here is one.

00:38:51.720 --> 00:38:55.080
You can hardly see that
because the contrast is low.

00:38:55.080 --> 00:38:58.390
So now, to see this
overall, here's an example.

00:38:58.390 --> 00:39:00.380
Here, we vary the contrast.

00:39:00.380 --> 00:39:02.680
And here, we vary the
spatial frequency.

00:39:02.680 --> 00:39:05.770
And if you look at that, you
can see that in this region,

00:39:05.770 --> 00:39:08.520
depending on how far back
you are, this region,

00:39:08.520 --> 00:39:09.680
you can see the best.

00:39:09.680 --> 00:39:10.840
Here, it will be less so.

00:39:10.840 --> 00:39:13.210
And here, of course, it
drops off dramatically,

00:39:13.210 --> 00:39:16.390
just like that curve
I have drawn there.

00:39:16.390 --> 00:39:21.310
So this, then, enables one to
generate a so-called "contrast

00:39:21.310 --> 00:39:25.690
sensitivity"
function in a monkey

00:39:25.690 --> 00:39:28.690
in those regions of the
visual field that are intact

00:39:28.690 --> 00:39:30.780
and in those regions
of visual field

00:39:30.780 --> 00:39:33.340
in which you have either
magnocellular or parvocellular

00:39:33.340 --> 00:39:36.920
lesion that selectively
blocks the parasol

00:39:36.920 --> 00:39:38.850
and the midget systems.

00:39:38.850 --> 00:39:41.590
So that's the experiment,
then, for just studying

00:39:41.590 --> 00:39:42.990
contrast sensitivity.

00:39:42.990 --> 00:39:46.110
And if you do that, this is
the kind of result you get.

00:39:46.110 --> 00:39:47.900
This is the monkey's
normal performance.

00:39:47.900 --> 00:39:53.890
In this case, of four
spatial frequency levels.

00:39:53.890 --> 00:39:56.090
And you go up and
down with contrast,

00:39:56.090 --> 00:39:58.480
just like that
curve I drew there.

00:39:58.480 --> 00:40:01.140
So this is your contrast
sensitivity function.

00:40:01.140 --> 00:40:04.080
And it shows that
under normal conditions

00:40:04.080 --> 00:40:09.290
and after magnocellular lesion
that blocks the parasol system,

00:40:09.290 --> 00:40:17.970
there is no effect, meaning that
the parasol system doesn't seem

00:40:17.970 --> 00:40:21.030
to be too important for
contrast sensitivity.

00:40:21.030 --> 00:40:23.790
By contrast, there's
a huge effect,

00:40:23.790 --> 00:40:27.760
especially at high
spatial frequencies,

00:40:27.760 --> 00:40:33.880
after a parvocellular lesion
that blocks the midget system.

00:40:33.880 --> 00:40:35.910
So that, in essence
then, is what

00:40:35.910 --> 00:40:38.140
happens with
contrast sensitivity.

00:40:38.140 --> 00:40:42.870
Now, let us move on and say,
well, what about color vision?

00:40:42.870 --> 00:40:44.370
So how do you do
the color vision?

00:40:44.370 --> 00:40:46.020
I already told you that before.

00:40:46.020 --> 00:40:50.170
What you do is you
present, in this case,

00:40:50.170 --> 00:40:52.960
eight stimuli, one of which
is different from the others,

00:40:52.960 --> 00:40:55.180
and just have red
and green ones.

00:40:55.180 --> 00:40:56.390
This is the odd one.

00:40:56.390 --> 00:40:57.995
The monkey makes
a saccade to it.

00:40:59.610 --> 00:41:02.409
He gets a drop of apple
juice for a reward.

00:41:02.409 --> 00:41:04.200
Now, if you want to be
systematic about it,

00:41:04.200 --> 00:41:07.790
you can vary the degree
of color contrast.

00:41:07.790 --> 00:41:11.360
But the effect is so
dramatic that it was not

00:41:11.360 --> 00:41:12.560
necessary to do that.

00:41:12.560 --> 00:41:14.400
So let me show you
what the effect was.

00:41:14.400 --> 00:41:17.010
Here, we have a monkey's
normal performance

00:41:17.010 --> 00:41:21.380
when the test element
is blue, red, and green.

00:41:22.520 --> 00:41:26.000
This happens after a
parvocellular lesion,

00:41:26.000 --> 00:41:28.290
meaning when you block
the midget system,

00:41:28.290 --> 00:41:31.310
the monkey cannot
see colors at all.

00:41:31.310 --> 00:41:33.120
Just a total loss.

00:41:33.120 --> 00:41:35.300
Whereas, after
magnocellular lesion,

00:41:35.300 --> 00:41:38.260
his performance is
indistinguishable from normal.

00:41:38.260 --> 00:41:40.600
So this, then,
establishes the fact

00:41:40.600 --> 00:41:46.455
that color vision is controlled
by the midget system.

00:41:47.720 --> 00:41:49.920
Now, the fact that this
is the case perhaps

00:41:49.920 --> 00:41:52.220
is not that surprising
because I told you

00:41:52.220 --> 00:41:59.150
that when you look at the
cells in the midget system

00:41:59.150 --> 00:42:02.060
in central vision,
that most of them

00:42:02.060 --> 00:42:04.290
get an input from
just-- at least,

00:42:04.290 --> 00:42:07.980
the red and green ones-- get
an input from a single cone.

00:42:07.980 --> 00:42:10.600
So just looking at the
receptive field organization

00:42:10.600 --> 00:42:14.830
tells you that that
system is likely to be

00:42:14.830 --> 00:42:17.090
very important for
color processing.

00:42:17.090 --> 00:42:20.334
So that, then, is
confirmed by this kind

00:42:20.334 --> 00:42:21.250
behavioral experiment.

00:42:22.510 --> 00:42:25.690
So now, let us look at
brightness perception.

00:42:25.690 --> 00:42:29.711
Now, how is that different
from contrast sensitivity?

00:42:29.711 --> 00:42:31.710
It's different because,
in this case, what to do

00:42:31.710 --> 00:42:36.530
then is you use a
discrimination task like this.

00:42:36.530 --> 00:42:38.600
And I think most of you
can tell that this one is

00:42:38.600 --> 00:42:40.080
brighter than the others.

00:42:40.080 --> 00:42:42.680
Purposefully, I made
that a small difference,

00:42:42.680 --> 00:42:45.320
so you can appreciate the
fact that on each trial,

00:42:45.320 --> 00:42:48.130
we can vary the difference
between the distractors

00:42:48.130 --> 00:42:49.210
and the target.

00:42:49.210 --> 00:42:51.680
And you can generate
a curve seeing

00:42:51.680 --> 00:42:53.200
how much brightness
difference do

00:42:53.200 --> 00:42:58.600
you need to be able to perceive
a brightness difference.

00:42:58.600 --> 00:43:01.120
So if you do that
kind of experiment,

00:43:01.120 --> 00:43:02.370
one is in for a big surprise.

00:43:03.380 --> 00:43:08.140
This shows here what happens
after a parvocellular lesion.

00:43:08.140 --> 00:43:11.090
And this shows here what happens
after a magnocellular lesion.

00:43:11.090 --> 00:43:13.750
So when you block
the midget system,

00:43:13.750 --> 00:43:19.840
the performance is unaffected
as is the case with a magno

00:43:19.840 --> 00:43:24.200
lesion, meaning that
if you, obviously,

00:43:24.200 --> 00:43:29.000
that both the midget system
and the parasol system process

00:43:29.000 --> 00:43:36.770
information about brightness,
at least, at these low spatial

00:43:36.770 --> 00:43:41.480
frequencies that we have
used that I just showed you.

00:43:41.480 --> 00:43:45.210
So now, the other surprise
was that, then, the question

00:43:45.210 --> 00:43:46.790
became, well, what
if you do this not

00:43:46.790 --> 00:43:52.810
under photopic conditions and
do under scotopic conditions?

00:43:52.810 --> 00:43:55.860
And again, there's
no effect, meaning

00:43:55.860 --> 00:43:59.140
that the unique
inputs from the rods

00:43:59.140 --> 00:44:01.970
and the unique
inputs from the cones

00:44:01.970 --> 00:44:04.420
must go into both
of the systems.

00:44:04.420 --> 00:44:06.430
Now, the reason
that was surprising

00:44:06.430 --> 00:44:08.600
is because a couple of
papers have been published,

00:44:08.600 --> 00:44:21.740
maybe about 15 to 20 years ago,
that claimed that the rods feed

00:44:21.740 --> 00:44:24.540
selectively into
the parasol system,

00:44:24.540 --> 00:44:26.370
not into the midget system.

00:44:26.370 --> 00:44:28.705
So this totally disprove that.

00:44:28.705 --> 00:44:32.930
And then subsequently, careful
anatomical experiments also

00:44:32.930 --> 00:44:37.730
established that both the
small and the large cells,

00:44:37.730 --> 00:44:41.550
magno ganglion cells,
receive convergent input

00:44:41.550 --> 00:44:46.700
from the rods and the
cones as I had diagrammed

00:44:46.700 --> 00:44:49.100
to you in the previous session.

00:44:50.570 --> 00:44:54.835
So then, let's go on and look at
pattern and texture perception.

00:44:57.750 --> 00:45:01.000
In this case, let me show you
the kind of experiment that's

00:45:01.000 --> 00:45:01.500
done.

00:45:01.500 --> 00:45:05.750
This is when you
look at patterns.

00:45:05.750 --> 00:45:08.750
One way to do it, go back
to those same checkerboards,

00:45:08.750 --> 00:45:11.150
but make them high
contrast and have one

00:45:11.150 --> 00:45:15.680
at a different spatial
frequency than the others.

00:45:15.680 --> 00:45:18.140
And then you can
systematically vary the degree

00:45:18.140 --> 00:45:21.290
of spatial frequency
difference between the targets

00:45:21.290 --> 00:45:22.880
and the distractors.

00:45:22.880 --> 00:45:25.060
So that would be
the method that is

00:45:25.060 --> 00:45:29.640
used to study this kind
of pattern perception.

00:45:29.640 --> 00:45:32.290
The other one is to
look at textures.

00:45:32.290 --> 00:45:34.790
And in this case, those
of you in the back

00:45:34.790 --> 00:45:36.080
probably can't see this.

00:45:36.080 --> 00:45:38.987
But those of you up front, can
you see this little area here,

00:45:38.987 --> 00:45:40.570
where the diagonal
lines are reversed?

00:45:41.600 --> 00:45:44.490
So what happens is, first, you
just present this whole thing.

00:45:44.490 --> 00:45:51.650
And then you present the reverse
patterns and the surround.

00:45:51.650 --> 00:45:53.530
And when you do
that, the monkey has

00:45:53.530 --> 00:45:55.554
to make a saccade
to that location

00:45:55.554 --> 00:45:56.970
and will get a
drop of apple juice

00:45:56.970 --> 00:45:59.050
again for that performance.

00:45:59.050 --> 00:46:01.430
And then on each trial,
this appears somewhere else

00:46:01.430 --> 00:46:06.470
in one of the four to eight
locations in the display.

00:46:06.470 --> 00:46:08.550
So that's the procedure.

00:46:08.550 --> 00:46:12.090
And then if you do
this, you, again,

00:46:12.090 --> 00:46:15.180
luckily get a very dramatic
effect, apparently.

00:46:15.180 --> 00:46:16.010
Here it is.

00:46:16.010 --> 00:46:19.220
We have here normal performance.

00:46:19.220 --> 00:46:22.700
Here's a parvocellular
lesion, magnocellular lesion.

00:46:22.700 --> 00:46:25.480
This is texture,
and this is pattern.

00:46:25.480 --> 00:46:28.050
Pattern tasks was, overall,
much more difficult

00:46:28.050 --> 00:46:30.030
than the texture
one in this case.

00:46:30.030 --> 00:46:33.030
Well, what you see here
that is really dramatic

00:46:33.030 --> 00:46:37.250
is that when you block
the midget system,

00:46:37.250 --> 00:46:41.350
there's a tremendous loss in
your ability to see patterns

00:46:41.350 --> 00:46:43.565
and in your ability
to see textures.

00:46:44.730 --> 00:46:49.850
So fine vision for
detail seems to be

00:46:49.850 --> 00:46:53.965
central for the processing
of the midget system.

00:46:56.260 --> 00:47:00.670
So that, then, is
the effect that one

00:47:00.670 --> 00:47:04.590
gets with texture
and with pattern.

00:47:05.720 --> 00:47:09.030
So now, let's move on and
talk about stereoscopic depth

00:47:09.030 --> 00:47:10.200
perception.

00:47:10.200 --> 00:47:11.900
It's a topic we
are going to look

00:47:11.900 --> 00:47:15.840
at in more detail
in a later session.

00:47:15.840 --> 00:47:17.960
Let me now introduce
this, first of all,

00:47:17.960 --> 00:47:24.672
by telling you that stereoscopic
depth perception resulted

00:47:24.672 --> 00:47:26.130
predominately by
virtue of the fact

00:47:26.130 --> 00:47:27.980
that the eye is
moving to the front

00:47:27.980 --> 00:47:30.685
so that there was a
major binocular overlap.

00:47:31.830 --> 00:47:35.720
And if you talk about a monkey,
for example, or even many,

00:47:35.720 --> 00:47:39.230
many animals that do have
stereoscopic vision, here's

00:47:39.230 --> 00:47:44.880
an example of looking at a tree
when just about the only cue

00:47:44.880 --> 00:47:50.025
you have here would be based on
stereopsis because all of them

00:47:50.025 --> 00:47:51.290
are equally dark.

00:47:51.290 --> 00:47:53.380
Which branch is in
front of which one?

00:47:53.380 --> 00:47:54.690
And it's very hard to tell.

00:47:54.690 --> 00:47:57.730
If your monkey will jump from
this branch to this branch,

00:47:57.730 --> 00:48:00.700
if he can't tell where they
are relative to each other,

00:48:00.700 --> 00:48:03.510
the monkey is going to
fall down and drop dread.

00:48:03.510 --> 00:48:06.880
So it is very
important for monkeys

00:48:06.880 --> 00:48:09.680
to have a highly functional
stereoscopic system, which

00:48:09.680 --> 00:48:10.770
they do.

00:48:10.770 --> 00:48:13.440
And so, how do we study this?

00:48:13.440 --> 00:48:16.490
Well, the way we
study this is to use

00:48:16.490 --> 00:48:19.550
what is called a
random-dot stereogram.

00:48:20.790 --> 00:48:24.350
Random-dot stereograms
were created

00:48:24.350 --> 00:48:31.230
once computers became a reality
many, many years back by Bela

00:48:31.230 --> 00:48:31.730
Julesz.

00:48:32.880 --> 00:48:36.150
And he came up with
the idea that if you

00:48:36.150 --> 00:48:38.640
use this kinds of
random-dot stereograms,

00:48:38.640 --> 00:48:40.790
there's no other depth cue.

00:48:40.790 --> 00:48:42.270
So, in other words,
you can study

00:48:42.270 --> 00:48:48.680
just the stereoscopic depth
perception aspect of it.

00:48:48.680 --> 00:48:49.770
Now, why is this?

00:48:49.770 --> 00:48:53.800
Because what you do here is
you look at this two displays,

00:48:53.800 --> 00:48:56.550
you present this to the left
eye and this to the right eye.

00:48:56.550 --> 00:48:59.872
The way this is
done is that you use

00:48:59.872 --> 00:49:01.080
what is called a stereoscope.

00:49:02.804 --> 00:49:04.720
I bet you most of you
have seen a stereoscope.

00:49:06.110 --> 00:49:07.380
Everybody see a stereoscope?

00:49:08.220 --> 00:49:08.720
No?

00:49:11.360 --> 00:49:14.940
So a stereoscope
used to be something

00:49:14.940 --> 00:49:18.790
that was extremely popular
starting in the 19th century.

00:49:20.300 --> 00:49:23.240
They created these handheld
devices with the two lenses

00:49:23.240 --> 00:49:24.410
that you look through.

00:49:24.410 --> 00:49:36.090
And then they created a camera
that had two lenses in it

00:49:36.090 --> 00:49:38.740
at the same distance,
roughly, as your two eyes.

00:49:38.740 --> 00:49:40.550
And so it took a
picture, thereby,

00:49:40.550 --> 00:49:47.145
creating two pieces,
one by each lens.

00:49:48.660 --> 00:49:51.210
And then, what you
did, since the two have

00:49:51.210 --> 00:49:54.090
a slightly different perspective
of what you're looking at,

00:49:54.090 --> 00:49:59.760
then the photographs are
put into the stereoscope

00:49:59.760 --> 00:50:00.350
to look at.

00:50:00.350 --> 00:50:02.220
These two images are fused then.

00:50:02.220 --> 00:50:03.860
Looks like a single image.

00:50:03.860 --> 00:50:06.920
And due to their
very disparities,

00:50:06.920 --> 00:50:08.140
you see real depth.

00:50:08.140 --> 00:50:09.015
It's really dramatic.

00:50:10.170 --> 00:50:13.610
And when we talk about
stereoscopic depth perception,

00:50:13.610 --> 00:50:16.220
I'm going to actually bring
some stereoscopes here

00:50:16.220 --> 00:50:19.840
and some examples so you can
see exactly what that is like.

00:50:22.160 --> 00:50:24.460
But at any rate, in this
case, what you can do

00:50:24.460 --> 00:50:27.430
is you can take a little
area here just like before.

00:50:27.430 --> 00:50:30.220
And you can take the
dots in this area.

00:50:30.220 --> 00:50:32.940
You move them a little bit
this way, a few pixels.

00:50:32.940 --> 00:50:35.230
And you move this a
bit this way, that way.

00:50:35.230 --> 00:50:37.380
And then if it's a
square area, you're

00:50:37.380 --> 00:50:39.460
going to see a
square sticking out.

00:50:39.460 --> 00:50:42.000
And the monkey will then
make a saccade to that.

00:50:42.000 --> 00:50:44.180
But with one eye, he
cannot do anything.

00:50:44.180 --> 00:50:49.370
He can't see a thing because
these random-dot stereograms

00:50:49.370 --> 00:50:52.195
only give you information
about disparity.

00:50:52.195 --> 00:50:53.820
So if you do that,
you're going to have

00:50:53.820 --> 00:50:55.720
the monkey make a
saccade to these.

00:50:55.720 --> 00:50:58.100
You can vary the
degree of disparity.

00:50:59.140 --> 00:51:02.440
And when you do that,
what you find is,

00:51:02.440 --> 00:51:08.000
quite dramatically, is that
after a parvo lesion, when

00:51:08.000 --> 00:51:11.250
you block the midget system,
there's a tremendous loss.

00:51:11.250 --> 00:51:14.685
The monkey
essentially, especially

00:51:14.685 --> 00:51:18.110
at smaller disparities,
has totally

00:51:18.110 --> 00:51:23.250
lost his ability to use
stereoscopic depth information.

00:51:23.250 --> 00:51:29.300
So there is no deficit, however,
after a magnocellular lesion.

00:51:29.300 --> 00:51:34.720
So therefore, you can say safely
that, especially at smaller

00:51:34.720 --> 00:51:41.270
disparities, what you have
is the midget system that

00:51:41.270 --> 00:51:45.320
performs this remarkable task
of seeing things in depth

00:51:45.320 --> 00:51:50.310
by virtue of disparity
in the two eyes produced

00:51:50.310 --> 00:51:51.770
by stereoscopic vision.

00:51:53.620 --> 00:51:55.990
So that's really a
very dramatic effect.

00:51:55.990 --> 00:51:59.815
Then the next thing we can
look at is motion perception.

00:52:01.190 --> 00:52:02.990
Again, this can be
done in various ways,

00:52:02.990 --> 00:52:04.740
but I will just show
you an example of it.

00:52:04.740 --> 00:52:06.750
So the monkey first fixates.

00:52:06.750 --> 00:52:08.890
Again, use random-dot
stereograms here.

00:52:09.910 --> 00:52:13.390
And then you set in motion
a small square area.

00:52:13.390 --> 00:52:13.890
Ready?

00:52:15.650 --> 00:52:16.810
I hope.

00:52:16.810 --> 00:52:17.470
There we go.

00:52:18.570 --> 00:52:21.600
Or you can do the high
spatial frequency like that.

00:52:22.660 --> 00:52:24.660
And then, of course, on
each trial, that appears

00:52:24.660 --> 00:52:25.740
someplace else.

00:52:25.740 --> 00:52:28.150
And you can vary
the velocity or you

00:52:28.150 --> 00:52:31.420
can vary the contrast
of the display

00:52:31.420 --> 00:52:35.360
and see how the monkey
performs in the intact portions

00:52:35.360 --> 00:52:37.830
of visual field
and how it performs

00:52:37.830 --> 00:52:40.410
in those portions of the
visual field that are blocked

00:52:40.410 --> 00:52:44.540
either for the midget system
or for the parasol system.

00:52:44.540 --> 00:52:45.850
Clear?

00:52:45.850 --> 00:52:47.830
So let's think about
it for a minute.

00:52:47.830 --> 00:52:49.413
What do you think
you're going to get?

00:52:51.770 --> 00:52:53.330
Here we go.

00:52:53.330 --> 00:52:55.610
Here we are-- motion detection.

00:52:55.610 --> 00:52:56.800
This is a parvo lesion.

00:52:56.800 --> 00:52:58.140
This is a magno lesion.

00:52:58.140 --> 00:53:01.720
And so what you get is
a dramatic deficit, not

00:53:01.720 --> 00:53:04.520
an all-amount deficit,
but a dramatic deficit

00:53:04.520 --> 00:53:07.440
in seeing motion
after the magno lesion

00:53:07.440 --> 00:53:09.150
but not after parvo lesion.

00:53:09.150 --> 00:53:14.510
So this says that,
indeed, the parasol system

00:53:14.510 --> 00:53:17.080
plays a very important role
in motion perception, not

00:53:17.080 --> 00:53:18.220
an exclusive role.

00:53:18.220 --> 00:53:19.910
There is still some
performance there

00:53:19.910 --> 00:53:26.550
at very, very low contrast here.

00:53:26.550 --> 00:53:29.120
And there's a big effect
at very high contrast,

00:53:29.120 --> 00:53:30.550
and monkey does better.

00:53:30.550 --> 00:53:35.140
So we can conclude
that the parasol system

00:53:35.140 --> 00:53:39.235
is very important for
motion perception.

00:53:40.420 --> 00:53:44.990
So now, the next thing we
can look at is flicker.

00:53:44.990 --> 00:53:48.160
And in this case, when
you study flicker,

00:53:48.160 --> 00:53:53.310
instead of using a
monitor, people used LEDs.

00:53:53.310 --> 00:53:54.850
Now, why do you think that is?

00:53:54.850 --> 00:53:56.755
Why don't they
just use monitors?

00:53:58.870 --> 00:54:00.460
Well, I mean, the
reason for that

00:54:00.460 --> 00:54:02.420
is very straightforward,
actually.

00:54:02.420 --> 00:54:04.900
When you use a
regular monitor, what

00:54:04.900 --> 00:54:06.360
is the frequency of a monitor?

00:54:06.360 --> 00:54:11.790
It's the same as the frequency
of the alternating current

00:54:11.790 --> 00:54:14.260
that you have, that
shine the lights up here.

00:54:14.260 --> 00:54:17.920
But what is that in
the United States?

00:54:17.920 --> 00:54:18.750
60 hertz.

00:54:20.260 --> 00:54:23.450
And so that's what happens
on a regular monitor.

00:54:23.450 --> 00:54:30.640
And every 1 over 60 hertz
is you shift the image.

00:54:32.160 --> 00:54:35.620
In Europe, the frequency
is actually 50 hertz.

00:54:35.620 --> 00:54:40.830
So at any rate, because
of that, there's

00:54:40.830 --> 00:54:42.930
a rather limited
range over which

00:54:42.930 --> 00:54:48.700
you can vary the ON and OFF
activity of a flickering spot.

00:54:48.700 --> 00:54:51.080
So if you use any of
this, however, you

00:54:51.080 --> 00:54:57.380
can use it in a huge,
huge range of small steps.

00:54:57.380 --> 00:54:59.890
So, therefore, this
is what the display

00:54:59.890 --> 00:55:02.960
looks like that people
use to study this.

00:55:02.960 --> 00:55:04.760
First, the monkey fixates here.

00:55:04.760 --> 00:55:08.636
And then one of these LEDs
will start flickering.

00:55:08.636 --> 00:55:10.260
And I'll show you
the flickering, which

00:55:10.260 --> 00:55:12.450
is not perfect here
on this monitor,

00:55:12.450 --> 00:55:14.150
but it will give you a sense.

00:55:14.150 --> 00:55:15.370
Everybody see that flicker?

00:55:16.520 --> 00:55:20.710
Now, the mean flicker
value of that location

00:55:20.710 --> 00:55:24.630
is the same as the yellow
lights that you see there.

00:55:24.630 --> 00:55:26.600
So if it flickers
at a high rate,

00:55:26.600 --> 00:55:29.000
you can't tell that it's
flickering because it's

00:55:29.000 --> 00:55:31.710
beyond the ability of
the eye to resolve it,

00:55:31.710 --> 00:55:35.460
and then you can't make a
saccade to that location.

00:55:35.460 --> 00:55:39.900
So if you look at that, what you
find here, after parvo lesion

00:55:39.900 --> 00:55:44.640
and after magno lesion,
you get a gigantic deficit

00:55:44.640 --> 00:55:45.925
after magno lesion.

00:55:46.930 --> 00:55:49.910
And that fits with what I
told you in the beginning--

00:55:49.910 --> 00:55:56.480
namely, that the
midget cells respond

00:55:56.480 --> 00:55:59.400
in a fairly sustained fashion
when the stimulus comes on.

00:55:59.400 --> 00:56:03.310
Whereas, the parasol
cells respond transiently,

00:56:03.310 --> 00:56:09.160
which makes them much more
readily available for motion

00:56:09.160 --> 00:56:10.110
and for flicker.

00:56:15.130 --> 00:56:20.480
So those, then, are
the major arrangements

00:56:20.480 --> 00:56:23.290
that we see with these
various experiments.

00:56:23.290 --> 00:56:27.145
And so we can summarize what
happens after parvocellular

00:56:27.145 --> 00:56:31.610
and magnocellular lesion
and say what happens here.

00:56:32.620 --> 00:56:35.220
So this one here is when
you block the midget system.

00:56:35.220 --> 00:56:38.850
This is where you block
the parasol system.

00:56:38.850 --> 00:56:41.480
And here are all the various
test that had been used.

00:56:41.480 --> 00:56:48.610
And you can see that
for the parasol system,

00:56:48.610 --> 00:56:52.470
the major deficit arises in
motion perception and flicker

00:56:52.470 --> 00:56:53.410
perception.

00:56:53.410 --> 00:57:00.020
So those two are very
important in the processing

00:57:00.020 --> 00:57:01.830
of the parasol system.

00:57:01.830 --> 00:57:05.790
Whereas, for the
midget system, you

00:57:05.790 --> 00:57:08.980
get lots of deficits in color
vision, texture perception,

00:57:08.980 --> 00:57:13.170
and fine pattern perception,
fine shape perception,

00:57:13.170 --> 00:57:15.760
in contrast sensitivity,
and in stereopsis.

00:57:17.360 --> 00:57:23.000
So this, then, gives you a sense
of what these two systems are

00:57:23.000 --> 00:57:26.730
important for in processing
visual information.

00:57:28.650 --> 00:57:35.330
And so, what you can do next
is to summarize what I just

00:57:35.330 --> 00:57:37.630
told you-- namely, that
the midget system is

00:57:37.630 --> 00:57:41.455
important for color, texture,
fine form, and fine stereo.

00:57:42.820 --> 00:57:45.830
The parasol system is
important for fast flicker

00:57:45.830 --> 00:57:48.470
and fast, low contrast motion.

00:57:48.470 --> 00:57:52.270
Both systems are capable of
doing brightness; coarse form;

00:57:52.270 --> 00:57:55.870
coarse stereo; slow
flicker; slow, high contrast

00:57:55.870 --> 00:57:58.400
motion; and scotopic vision.

00:57:58.400 --> 00:58:02.140
So it is not a
simple arrangement.

00:58:02.140 --> 00:58:03.330
It's a complex arrangement.

00:58:03.330 --> 00:58:07.740
There is overlap in what
both systems can do.

00:58:07.740 --> 00:58:09.720
So now, the big
question comes up.

00:58:11.540 --> 00:58:14.970
Why was it so important to
create both of these systems?

00:58:14.970 --> 00:58:17.810
Well, the number of
schemes that have evolved,

00:58:17.810 --> 00:58:19.810
certainly, the
motion is obvious.

00:58:19.810 --> 00:58:22.340
But one thing one
can come up with is

00:58:22.340 --> 00:58:26.440
a scheme of this sort that
has been proposed-- namely,

00:58:26.440 --> 00:58:31.240
that what happened as a result
of these two systems emerging,

00:58:31.240 --> 00:58:35.340
your ability to process
information has been extended.

00:58:35.340 --> 00:58:42.490
And for the midget system,
you expanded the ability

00:58:42.490 --> 00:58:46.000
to see up to high
spatial frequencies--

00:58:46.000 --> 00:58:48.320
very, very important attribute.

00:58:48.320 --> 00:58:52.000
And I should add, also
for the midget system,

00:58:52.000 --> 00:58:54.480
it became possible
for you to see color.

00:58:57.490 --> 00:59:01.050
Whereas, when you come
to the parasol system,

00:59:01.050 --> 00:59:03.520
it expanded the
range of your ability

00:59:03.520 --> 00:59:06.620
to process information
in the temporal domain--

00:59:06.620 --> 00:59:10.085
your ability to see very
fast motion, to see flicker.

00:59:11.310 --> 00:59:17.000
Now, let me add here that's
a very important attribute,

00:59:17.000 --> 00:59:24.330
by the way, for animals as
well as humans because one

00:59:24.330 --> 00:59:28.590
of the common things
that has emerged

00:59:28.590 --> 00:59:31.350
in the course of evolution
is what is called camouflage.

00:59:32.339 --> 00:59:34.255
I think I may have
mentioned that once before.

00:59:35.590 --> 00:59:39.520
That is that when
you or an animal,

00:59:39.520 --> 00:59:43.220
and the coloring of you,
which is done in thousands

00:59:43.220 --> 00:59:47.780
of animals, is made similar
to the background, then

00:59:47.780 --> 00:59:53.450
it's very difficult for you
to see that particular animal.

00:59:53.450 --> 00:59:55.380
So it has an
excellent camouflage.

00:59:55.380 --> 01:00:01.650
However, as soon as this
animal begins to move,

01:00:01.650 --> 01:00:05.170
the ability to camouflage
itself disappears.

01:00:05.170 --> 01:00:06.720
And animals have to move.

01:00:06.720 --> 01:00:11.830
And that's because
the parasol system

01:00:11.830 --> 01:00:15.390
is extremely
sensitive to motion.

01:00:15.390 --> 01:00:18.060
And so whenever an
animal begins to move,

01:00:18.060 --> 01:00:20.160
even though it has
excellent camouflage,

01:00:20.160 --> 01:00:23.050
it is predominantly
the parasol system

01:00:23.050 --> 01:00:28.460
that will destroy the
camouflage effect.

01:00:30.540 --> 01:00:32.660
I'm not saying
this is necessarily

01:00:32.660 --> 01:00:36.010
the only hypothesis, but
this is one hypothesis

01:00:36.010 --> 01:00:41.190
that had been advanced to try
to account for how sensible it

01:00:41.190 --> 01:00:42.815
is to have these two systems.

01:00:43.910 --> 01:00:45.780
Now, of course,
you could ask, how

01:00:45.780 --> 01:00:48.860
come this couldn't be just
done in one system-- all of it?

01:00:48.860 --> 01:00:51.950
Well, there are several
reasons for that--

01:00:51.950 --> 01:00:56.830
is that to combine a
situation where you have

01:00:56.830 --> 01:01:00.220
a sustained response and a
transient response into one

01:01:00.220 --> 01:01:01.170
is next to impossible.

01:01:02.710 --> 01:01:07.890
And to have cells that
are highly sensitive,

01:01:07.890 --> 01:01:13.620
or the parasol cells, you need
to have a convergent input

01:01:13.620 --> 01:01:15.780
from several photoreceptors.

01:01:15.780 --> 01:01:17.620
But as soon as
you have that, you

01:01:17.620 --> 01:01:21.080
lose your ability
to see fine detail.

01:01:21.080 --> 01:01:24.840
So these requirements
to do this and this

01:01:24.840 --> 01:01:28.230
are antagonistic to each other.

01:01:28.230 --> 01:01:31.300
And it's not possible to
do that in a single cell

01:01:31.300 --> 01:01:32.830
at the retinal level.

01:01:32.830 --> 01:01:37.010
So it was decided, inasmuch
as they should decide,

01:01:37.010 --> 01:01:39.050
that it will be
best just to create

01:01:39.050 --> 01:01:43.630
two separate systems however
complicated that might be.

01:01:44.730 --> 01:01:49.800
So that, then, brings me
to the overall summary

01:01:49.800 --> 01:01:52.540
of what I had covered today.

01:01:52.540 --> 01:01:54.840
First of all, I
told you-- that's

01:01:54.840 --> 01:01:57.610
the obvious part-- are two
major channels that originate

01:01:57.610 --> 01:02:00.540
in the retina are the
midget and the parasol.

01:02:00.540 --> 01:02:03.900
I should add here that there
are many other channels

01:02:03.900 --> 01:02:06.477
in the retina,
but the cells that

01:02:06.477 --> 01:02:08.310
do that-- and we'll
talk about some of those

01:02:08.310 --> 01:02:11.595
later-- are much less numerous.

01:02:13.550 --> 01:02:16.610
The overwhelming majority
of cells in the retina

01:02:16.610 --> 01:02:18.930
are either midget and parasol.

01:02:18.930 --> 01:02:22.090
Then in central retina,
the receptive field center

01:02:22.090 --> 01:02:24.400
of the midget cells
and parvocellular cells

01:02:24.400 --> 01:02:26.880
is comprised of
just a single cone.

01:02:30.500 --> 01:02:32.830
The midget cells and the
parvocellular cells--

01:02:32.830 --> 01:02:36.840
meaning that the midget projects
to the parvocellular layers.

01:02:36.840 --> 01:02:40.370
And I told you before that
the midget cells of the retina

01:02:40.370 --> 01:02:44.430
have receptive fields which are
quite similar to those, almost

01:02:44.430 --> 01:02:47.270
identical to those that you
see in the parvocellular

01:02:47.270 --> 01:02:49.850
layers of the lateral
geniculate nucleus.

01:02:49.850 --> 01:02:55.410
Then the parasol cells have
much larger receptive fields

01:02:55.410 --> 01:02:57.000
than do the midget cells.

01:02:57.000 --> 01:03:02.050
The cone input is mixed both
in the center and the surround.

01:03:02.050 --> 01:03:07.150
And that is the reason why this
system, the parasol system,

01:03:07.150 --> 01:03:09.230
cannot tell you
anything about color,

01:03:09.230 --> 01:03:12.510
but it is very sensitive and can
tell you about any change that

01:03:12.510 --> 01:03:17.140
occurs out there over time,
irrespective of the kind

01:03:17.140 --> 01:03:19.000
of color it has.

01:03:19.000 --> 01:03:22.940
Now, the midget and parasol cell
ratio from center to periphery

01:03:22.940 --> 01:03:25.900
changes from 8 to 1 to 1 to 1.

01:03:25.900 --> 01:03:29.100
Now, I should
reiterate that in that

01:03:29.100 --> 01:03:30.580
it is very important
for us to be

01:03:30.580 --> 01:03:32.390
sensitive to motion
in the periphery.

01:03:34.460 --> 01:03:41.810
That's if an animal is
threatened by some predator,

01:03:41.810 --> 01:03:48.030
it is highly desirable
to be able to very

01:03:48.030 --> 01:03:50.060
sensitive to any motion.

01:03:50.060 --> 01:03:52.730
And because of that,
you have a higher number

01:03:52.730 --> 01:03:56.900
of parasol cells in the
periphery than in the center.

01:03:56.900 --> 01:04:02.235
And so the ratio changes to
become equal in the periphery.

01:04:04.100 --> 01:04:06.290
Now, the midget
and parasol systems

01:04:06.290 --> 01:04:10.480
converge on some cells in V1.

01:04:10.480 --> 01:04:11.970
And the example,
the prime example,

01:04:11.970 --> 01:04:14.660
I've shown you when
they were separately

01:04:14.660 --> 01:04:18.420
blocked at the level the
geniculate, that the cell

01:04:18.420 --> 01:04:21.550
I showed you was one that did
receive a convergent input.

01:04:21.550 --> 01:04:25.110
But many other cells
receive single input.

01:04:25.110 --> 01:04:29.340
V4 receives inputs from both the
midget and the parasol cells.

01:04:29.340 --> 01:04:32.920
So it's not an area that only
deals with color, obviously.

01:04:32.920 --> 01:04:35.750
And what it will deal with,
we'll describe in more detail

01:04:35.750 --> 01:04:36.580
later.

01:04:36.580 --> 01:04:40.480
The major input to MT is
from the parasol cells.

01:04:40.480 --> 01:04:43.830
And that, often, MT
and MST have often

01:04:43.830 --> 01:04:46.965
been called the motion areas.

01:04:48.270 --> 01:04:51.276
But they also play
important role, by the way,

01:04:51.276 --> 01:04:57.520
in depth perception,
as we shall see.

01:04:59.690 --> 01:05:03.490
And then the midget
system extends the range

01:05:03.490 --> 01:05:06.880
of vision in the wavelength and
high spatial frequency domains.

01:05:08.130 --> 01:05:10.810
The parasol system
extends the range

01:05:10.810 --> 01:05:13.290
of vision in the high
frequency domain.

01:05:15.430 --> 01:05:18.970
And in the scheme
that I showed you

01:05:18.970 --> 01:05:22.245
is what leads to this
particular conclusion.

01:05:23.300 --> 01:05:27.660
So that, then, is the essence
of what I wanted to cover today.

01:05:28.850 --> 01:05:31.940
And I am now certainly
open to any questions

01:05:31.940 --> 01:05:35.770
that you might have about these
two fascinating systems that

01:05:35.770 --> 01:05:39.760
have evolved in the
retina over the millions

01:05:39.760 --> 01:05:41.376
and millions of
years of evolution.

01:05:48.666 --> 01:05:51.600
Well, I was so clear that
there are questions, huh?

01:05:54.990 --> 01:05:56.580
Well, I hope that
you sort of gotten

01:05:56.580 --> 01:06:00.820
a sense of what it is like to
do these kinds of experiments

01:06:00.820 --> 01:06:05.220
and how luckily, at least, in
some cases, these experiments

01:06:05.220 --> 01:06:09.620
can lead to nice discoveries
as to the workings

01:06:09.620 --> 01:06:11.005
of the visual system.

01:06:11.005 --> 01:06:14.070
Now, last time when we
talked about the APB, we did

01:06:14.070 --> 01:06:16.500
have a "magic bullet."

01:06:16.500 --> 01:06:21.490
This approach here is
not as neat, really,

01:06:21.490 --> 01:06:24.790
because you have to make
lesions rather than specifically

01:06:24.790 --> 01:06:26.723
affecting certain
neurotransmitters.

01:06:28.040 --> 01:06:31.130
And that could not be done
because the neurotransmitters

01:06:31.130 --> 01:06:34.390
for the midget and the
parasol cells are similar.

01:06:34.390 --> 01:06:36.250
So you can't use that
kind of procedure

01:06:36.250 --> 01:06:40.180
that we were able to use
miraculously for the ON and OFF

01:06:40.180 --> 01:06:40.680
channels.

01:06:42.510 --> 01:06:45.680
So that, in essence, is what
we are going to cover today.

01:06:47.150 --> 01:06:48.600
Did everybody sign?

01:06:48.600 --> 01:06:51.320
Any one of you not sign
the attendance sheet?

01:06:51.320 --> 01:06:53.960
If not, please come
up here and sign it.

01:06:53.960 --> 01:06:56.960
And then let me just say
again that next time,

01:06:56.960 --> 01:07:00.120
we are going to talk about
color-- another fascinating

01:07:00.120 --> 01:07:01.080
topic.

01:07:01.080 --> 01:07:06.180
And we are also going to
talk about visual adaptation.

01:07:08.820 --> 01:07:11.540
And if you can, if you
get a chance to do so,

01:07:11.540 --> 01:07:14.950
please try to read the
preparatory material which

01:07:14.950 --> 01:07:17.700
should make it easier
for you to not only

01:07:17.700 --> 01:07:19.450
comprehend but,
especially, to be

01:07:19.450 --> 01:07:22.270
able to memorize these facts.

01:07:22.270 --> 01:07:25.030
And the memorization is
a very important part

01:07:25.030 --> 01:07:27.080
of learning things
in any course.

01:07:27.080 --> 01:07:30.140
And it will, of
course, be essential

01:07:30.140 --> 01:07:32.395
when you get to the stage
of having to take the exams.

01:07:33.620 --> 01:07:36.150
Well, thank you so much.