TEDSalon London Spring 2011
Neil Burgess: How your brain tells you where you are
Filmed:
Readability: 4.2
1,458,267 views
How do you remember where you parked your car? How do you know if you're moving in the right direction? Neuroscientist Neil Burgess studies the neural mechanisms that map the space around us, and how they link to memory and imagination.
Neil Burgess - Neuroscientist
At University College in London, Neil Burgess researches how patterns of electrical activity in brain cells guide us through space. Full bio
At University College in London, Neil Burgess researches how patterns of electrical activity in brain cells guide us through space. Full bio
Double-click the English transcript below to play the video.
00:15
When we park in a big parking lot,
0
0
2000
00:17
how do we remember where we parked our car?
1
2000
2000
00:19
Here's the problem facing Homer.
2
4000
3000
00:22
And we're going to try to understand
3
7000
2000
00:24
what's happening in his brain.
4
9000
2000
00:26
So we'll start with the hippocampus, shown in yellow,
5
11000
2000
00:28
which is the organ of memory.
6
13000
2000
00:30
If you have damage there, like in Alzheimer's,
7
15000
2000
00:32
you can't remember things including where you parked your car.
8
17000
2000
00:34
It's named after Latin for "seahorse,"
9
19000
2000
00:36
which it resembles.
10
21000
2000
00:38
And like the rest of the brain, it's made of neurons.
11
23000
2000
00:40
So the human brain
12
25000
2000
00:42
has about a hundred billion neurons in it.
13
27000
2000
00:44
And the neurons communicate with each other
14
29000
3000
00:47
by sending little pulses or spikes of electricity
15
32000
2000
00:49
via connections to each other.
16
34000
2000
00:51
The hippocampus is formed of two sheets of cells,
17
36000
3000
00:54
which are very densely interconnected.
18
39000
2000
00:56
And scientists have begun to understand
19
41000
2000
00:58
how spatial memory works
20
43000
2000
01:00
by recording from individual neurons
21
45000
2000
01:02
in rats or mice
22
47000
2000
01:04
while they forage or explore an environment
23
49000
2000
01:06
looking for food.
24
51000
2000
01:08
So we're going to imagine we're recording from a single neuron
25
53000
3000
01:11
in the hippocampus of this rat here.
26
56000
3000
01:14
And when it fires a little spike of electricity,
27
59000
2000
01:16
there's going to be a red dot and a click.
28
61000
3000
01:19
So what we see
29
64000
2000
01:21
is that this neuron knows
30
66000
2000
01:23
whenever the rat has gone into one particular place in its environment.
31
68000
3000
01:26
And it signals to the rest of the brain
32
71000
2000
01:28
by sending a little electrical spike.
33
73000
3000
01:31
So we could show the firing rate of that neuron
34
76000
3000
01:34
as a function of the animal's location.
35
79000
2000
01:36
And if we record from lots of different neurons,
36
81000
2000
01:38
we'll see that different neurons fire
37
83000
2000
01:40
when the animal goes in different parts of its environment,
38
85000
2000
01:42
like in this square box shown here.
39
87000
2000
01:44
So together they form a map
40
89000
2000
01:46
for the rest of the brain,
41
91000
2000
01:48
telling the brain continually,
42
93000
2000
01:50
"Where am I now within my environment?"
43
95000
2000
01:52
Place cells are also being recorded in humans.
44
97000
3000
01:55
So epilepsy patients sometimes need
45
100000
2000
01:57
the electrical activity in their brain monitoring.
46
102000
3000
02:00
And some of these patients played a video game
47
105000
2000
02:02
where they drive around a small town.
48
107000
2000
02:04
And place cells in their hippocampi would fire, become active,
49
109000
3000
02:07
start sending electrical impulses
50
112000
3000
02:10
whenever they drove through a particular location in that town.
51
115000
3000
02:13
So how does a place cell know
52
118000
2000
02:15
where the rat or person is within its environment?
53
120000
3000
02:18
Well these two cells here
54
123000
2000
02:20
show us that the boundaries of the environment
55
125000
2000
02:22
are particularly important.
56
127000
2000
02:24
So the one on the top
57
129000
2000
02:26
likes to fire sort of midway between the walls
58
131000
2000
02:28
of the box that their rat's in.
59
133000
2000
02:30
And when you expand the box, the firing location expands.
60
135000
3000
02:33
The one below likes to fire
61
138000
2000
02:35
whenever there's a wall close by to the south.
62
140000
3000
02:38
And if you put another wall inside the box,
63
143000
2000
02:40
then the cell fires in both place
64
145000
2000
02:42
wherever there's a wall to the south
65
147000
2000
02:44
as the animal explores around in its box.
66
149000
3000
02:48
So this predicts
67
153000
2000
02:50
that sensing the distances and directions of boundaries around you --
68
155000
2000
02:52
extended buildings and so on --
69
157000
2000
02:54
is particularly important for the hippocampus.
70
159000
3000
02:57
And indeed, on the inputs to the hippocampus,
71
162000
2000
02:59
cells are found which project into the hippocampus,
72
164000
2000
03:01
which do respond exactly
73
166000
2000
03:03
to detecting boundaries or edges
74
168000
3000
03:06
at particular distances and directions
75
171000
2000
03:08
from the rat or mouse
76
173000
2000
03:10
as it's exploring around.
77
175000
2000
03:12
So the cell on the left, you can see,
78
177000
2000
03:14
it fires whenever the animal gets near
79
179000
2000
03:16
to a wall or a boundary to the east,
80
181000
3000
03:19
whether it's the edge or the wall of a square box
81
184000
3000
03:22
or the circular wall of the circular box
82
187000
2000
03:24
or even the drop at the edge of a table, which the animals are running around.
83
189000
3000
03:27
And the cell on the right there
84
192000
2000
03:29
fires whenever there's a boundary to the south,
85
194000
2000
03:31
whether it's the drop at the edge of the table or a wall
86
196000
2000
03:33
or even the gap between two tables that are pulled apart.
87
198000
3000
03:36
So that's one way in which we think
88
201000
2000
03:38
place cells determine where the animal is as it's exploring around.
89
203000
3000
03:41
We can also test where we think objects are,
90
206000
3000
03:44
like this goal flag, in simple environments --
91
209000
3000
03:47
or indeed, where your car would be.
92
212000
2000
03:49
So we can have people explore an environment
93
214000
3000
03:52
and see the location they have to remember.
94
217000
3000
03:55
And then, if we put them back in the environment,
95
220000
2000
03:57
generally they're quite good at putting a marker down
96
222000
2000
03:59
where they thought that flag or their car was.
97
224000
3000
04:02
But on some trials,
98
227000
2000
04:04
we could change the shape and size of the environment
99
229000
2000
04:06
like we did with the place cell.
100
231000
2000
04:08
In that case, we can see
101
233000
2000
04:10
how where they think the flag had been changes
102
235000
3000
04:13
as a function of how you change the shape and size of the environment.
103
238000
3000
04:16
And what you see, for example,
104
241000
2000
04:18
if the flag was where that cross was in a small square environment,
105
243000
3000
04:21
and then if you ask people where it was,
106
246000
2000
04:23
but you've made the environment bigger,
107
248000
2000
04:25
where they think the flag had been
108
250000
2000
04:27
stretches out in exactly the same way
109
252000
2000
04:29
that the place cell firing stretched out.
110
254000
2000
04:31
It's as if you remember where the flag was
111
256000
2000
04:33
by storing the pattern of firing across all of your place cells
112
258000
3000
04:36
at that location,
113
261000
2000
04:38
and then you can get back to that location
114
263000
2000
04:40
by moving around
115
265000
2000
04:42
so that you best match the current pattern of firing of your place cells
116
267000
2000
04:44
with that stored pattern.
117
269000
2000
04:46
That guides you back to the location that you want to remember.
118
271000
3000
04:49
But we also know where we are through movement.
119
274000
3000
04:52
So if we take some outbound path --
120
277000
2000
04:54
perhaps we park and we wander off --
121
279000
2000
04:56
we know because our own movements,
122
281000
2000
04:58
which we can integrate over this path
123
283000
2000
05:00
roughly what the heading direction is to go back.
124
285000
2000
05:02
And place cells also get this kind of path integration input
125
287000
4000
05:06
from a kind of cell called a grid cell.
126
291000
3000
05:09
Now grid cells are found, again,
127
294000
2000
05:11
on the inputs to the hippocampus,
128
296000
2000
05:13
and they're a bit like place cells.
129
298000
2000
05:15
But now as the rat explores around,
130
300000
2000
05:17
each individual cell fires
131
302000
2000
05:19
in a whole array of different locations
132
304000
3000
05:22
which are laid out across the environment
133
307000
2000
05:24
in an amazingly regular triangular grid.
134
309000
3000
05:29
And if you record from several grid cells --
135
314000
3000
05:32
shown here in different colors --
136
317000
2000
05:34
each one has a grid-like firing pattern across the environment,
137
319000
3000
05:37
and each cell's grid-like firing pattern is shifted slightly
138
322000
3000
05:40
relative to the other cells.
139
325000
2000
05:42
So the red one fires on this grid
140
327000
2000
05:44
and the green one on this one and the blue on on this one.
141
329000
3000
05:47
So together, it's as if the rat
142
332000
3000
05:50
can put a virtual grid of firing locations
143
335000
2000
05:52
across its environment --
144
337000
2000
05:54
a bit like the latitude and longitude lines that you'd find on a map,
145
339000
3000
05:57
but using triangles.
146
342000
2000
05:59
And as it moves around,
147
344000
2000
06:01
the electrical activity can pass
148
346000
2000
06:03
from one of these cells to the next cell
149
348000
2000
06:05
to keep track of where it is,
150
350000
2000
06:07
so that it can use its own movements
151
352000
2000
06:09
to know where it is in its environment.
152
354000
2000
06:11
Do people have grid cells?
153
356000
2000
06:13
Well because all of the grid-like firing patterns
154
358000
2000
06:15
have the same axes of symmetry,
155
360000
2000
06:17
the same orientations of grid, shown in orange here,
156
362000
3000
06:20
it means that the net activity
157
365000
2000
06:22
of all of the grid cells in a particular part of the brain
158
367000
3000
06:25
should change
159
370000
2000
06:27
according to whether we're running along these six directions
160
372000
2000
06:29
or running along one of the six directions in between.
161
374000
3000
06:32
So we can put people in an MRI scanner
162
377000
2000
06:34
and have them do a little video game
163
379000
2000
06:36
like the one I showed you
164
381000
2000
06:38
and look for this signal.
165
383000
2000
06:40
And indeed, you do see it in the human entorhinal cortex,
166
385000
3000
06:43
which is the same part of the brain that you see grid cells in rats.
167
388000
3000
06:46
So back to Homer.
168
391000
2000
06:48
He's probably remembering where his car was
169
393000
2000
06:50
in terms of the distances and directions
170
395000
2000
06:52
to extended buildings and boundaries
171
397000
2000
06:54
around the location where he parked.
172
399000
2000
06:56
And that would be represented
173
401000
2000
06:58
by the firing of boundary-detecting cells.
174
403000
2000
07:00
He's also remembering the path he took out of the car park,
175
405000
3000
07:03
which would be represented in the firing of grid cells.
176
408000
3000
07:06
Now both of these kinds of cells
177
411000
2000
07:08
can make the place cells fire.
178
413000
2000
07:10
And he can return to the location where he parked
179
415000
2000
07:12
by moving so as to find where it is
180
417000
3000
07:15
that best matches the firing pattern
181
420000
2000
07:17
of the place cells in his brain currently
182
422000
2000
07:19
with the stored pattern where he parked his car.
183
424000
3000
07:22
And that guides him back to that location
184
427000
2000
07:24
irrespective of visual cues
185
429000
2000
07:26
like whether his car's actually there.
186
431000
2000
07:28
Maybe it's been towed.
187
433000
2000
07:30
But he knows where it was, so he knows to go and get it.
188
435000
3000
07:33
So beyond spatial memory,
189
438000
2000
07:35
if we look for this grid-like firing pattern
190
440000
2000
07:37
throughout the whole brain,
191
442000
2000
07:39
we see it in a whole series of locations
192
444000
3000
07:42
which are always active
193
447000
2000
07:44
when we do all kinds of autobiographical memory tasks,
194
449000
2000
07:46
like remembering the last time you went to a wedding, for example.
195
451000
3000
07:49
So it may be that the neural mechanisms
196
454000
2000
07:51
for representing the space around us
197
456000
3000
07:54
are also used for generating visual imagery
198
459000
4000
07:58
so that we can recreate the spatial scene, at least,
199
463000
3000
08:01
of the events that have happened to us when we want to imagine them.
200
466000
3000
08:04
So if this was happening,
201
469000
2000
08:06
your memories could start by place cells activating each other
202
471000
3000
08:09
via these dense interconnections
203
474000
2000
08:11
and then reactivating boundary cells
204
476000
2000
08:13
to create the spatial structure
205
478000
2000
08:15
of the scene around your viewpoint.
206
480000
2000
08:17
And grid cells could move this viewpoint through that space.
207
482000
2000
08:19
Another kind of cell, head direction cells,
208
484000
2000
08:21
which I didn't mention yet,
209
486000
2000
08:23
they fire like a compass according to which way you're facing.
210
488000
3000
08:26
They could define the viewing direction
211
491000
2000
08:28
from which you want to generate an image for your visual imagery,
212
493000
3000
08:31
so you can imagine what happened when you were at this wedding, for example.
213
496000
3000
08:34
So this is just one example
214
499000
2000
08:36
of a new era really
215
501000
2000
08:38
in cognitive neuroscience
216
503000
2000
08:40
where we're beginning to understand
217
505000
2000
08:42
psychological processes
218
507000
2000
08:44
like how you remember or imagine or even think
219
509000
3000
08:47
in terms of the actions
220
512000
2000
08:49
of the billions of individual neurons that make up our brains.
221
514000
3000
08:52
Thank you very much.
222
517000
2000
08:54
(Applause)
223
519000
3000
ABOUT THE SPEAKER
Neil Burgess - NeuroscientistAt University College in London, Neil Burgess researches how patterns of electrical activity in brain cells guide us through space.
Why you should listen
Neil Burgessis is deputy director of the Institute of Cognitive Neuroscience at University College London, where he investigates of the role of the hippocampus in spatial navigation and episodic memory. His research is directed at answering questions such as: How are locations represented, stored and used in the brain? What processes and which parts of the brain are involved in remembering the spatial and temporal context of everyday events, and in finding one's way about?
To explore this space, he and his team use a range of methods for gathering data, including pioneering uses of virtual reality, as well as computational modelling and electrophysiological analysis of the function of hippocampal neurons in the rat, functional imaging of human navigation, and neuropsychological experiments on spatial and episodic memory.
A parallel interest: Investigating our human short-term memory for serial order, or how we know our 123s.
Neil Burgess | Speaker | TED.com