WHY DOES THE BRAIN HAVE SO MANY VISUAL AREAS

1
WHY DOES THE BRAIN HAVE SO MANY VISUAL AREAS?

• Neuroscience 500
• March 2005

Jody Culham Department of Psychology, University
of Western Ontario London, Ontario,
Canada culham_at_uwo.ca
2
Reading

• Kaas, J. H. (1989). Why does the brain have so
  many visual areas? Journal of Cognitive
  Neuroscience, 1, 121-135.
• Allman, J. (2000). Primate brains (Ch. 6, pp.
  122-157). Evolving Brains. New York Scientific
  American Library.
• Gazzaniga, M., Ivry, R. B. Mangun, G. R. (with
  L. Krubitzer) Evolutionary perspectives (Ch. 14,
  pp. 577-596 only). Cognitive Neuroscience The
  Biology of the Mind (2nd ed.). New York Norton.
• pdfs of papers and full color lecture slides are
  available on my web site
• http//defiant.ssc.uwo.ca/Jody_web/courses.htm

3
Topics

• How is the visual system organized into areas?
• What defines a visual area?
• How can we determine visual areas in the human?
• How do brains evolve?
• So why are there so many visual areas and how did
  they get that way?
• What questions remain?

4
Who knows, who cares, why bother?

• Vision is arguably one of the best understood
  neuroscience systems. Understanding vision often
  elucidates general principles of neural function.
• Cognitive neuroscience needs to use as much data
  from as many paradigms as possible to understand
  something as complicated as the brain. This
  poses concerns in comparing between species.
  Understanding homologies vs. new areas is
  valuable.
• Thinking about evolution and how the brain got
  that way is highly valuable to the
  neuroscientist.
• Neuroimaging studies often publish long lists of
  activated regions that are hard to compare
  between studies. My opinion is that imaging has
  been most successful where specific areas have
  been clearly identified and thoroughly studied.

5
Part 1 How is the visual system organized into
areas?
6
Primary Visual Pathway

• Retina
• Thalamus
• Lateral Geniculate Nucleus (LGN)
• divided into magno and parvo layers
• Primary visual cortex (V1)
• Extrastriate visual areas
• Each visual hemifield projects to the opposite
  hemisphere

7
Beyond V1

• primary visual cortex (V1)
• extrastriate cortex
• Over 30 visual areas
• Complex organization
• Visual areas make up 40 of monkey brain

Macaque Brain
Flattened Cortex
Source Van Essen et al., 2001
8
Why are there so many visual areas?

• MAGNO
• quick and dirty

• PARVO
• slow and detailed

Source Felleman Van Essen, 1991
Source Mapping the MInd cover image
9
Two visual pathways

• The two visual processing streams for different
  visual percepts
• What (ventral stream)- object recognition
• main input from slow and detailed parvo system
  
• Where or How (dorsal stream) - spatial
  perception, motor planning
• main input from quick and dirty magno system

Source Mishkin Ungerleider, 1982
10
Wiring the Streams
Dorsal
Ventral
11
The What Pathway

• Other Visual Areas
• contain more complex receptive fields
• Temporal Lobe
• contains many specialized areas for recognizing
  various things

body motion
faces
places
bodies
objects
12
Complex Receptive Fields
Face neurons in the monkey brain
13
The Where or How Pathway
grasping and reaching
attention
head movements

• Parietal Lobe
• contains many specialized areas for using vision
  to guide actions in space

motion perception
eye movements
14
Part 2 What Defines a Visual Area?
15
What is a Visual Area?

• Function
• an area has a unique pattern of responses to
  different stimuli
• Architecture
• different brain areas show differences between
  cortical properties (e.g., thickness of different
  layers, sensitivity to various dyes)
• Connectivity
• Different areas have different patterns of
  connections with other areas
• Topography
• many sensory areas show topography (retinotopy,
  somatotopy, tonotopy)
• boundaries between topographic maps can indicate
  boundaries between areas (e.g., separate maps of
  visual space in visual areas V1 and V2

16
Each visual area has a map
calcarine sulcus
horizontal meridian (HM)
VM
VM
HM
HM
VM
VM
left occipital lobe
left occipital lobe
right occipital lobe
right occipital lobe
vertical meridian (VM)

• Each visual area contains a map of visual space
• polar coordinates
• eccentricity runs posterior-anterior
• phase runs superior-inferior
• The map for visual area V1 lies along the
  calcarine sulcus

Source Jody Culham
17
Map boundaries divide visual areas
Consider just one hemifield/hemisphere
calcarine sulcus
V1 upper
horizontal meridian (HM)
V1 lower
VM
HM
VM
left occipital lobe
left occipital lobe
left occipital lobe
vertical meridian (VM)

• Adjacent maps are divided by specific boundaries
  (in vision, the horizontal and vertical meridia)
• Adjacent maps are mirror imaged

Source Jody Culham
18
MT A Case Study

• Middle temporal area of the macaque monkey
• Meets most criteria for an area
• Has an apparent human equivalent

19
MT Function

• Single unit recording
• Single neurons in MT are tuned to the direction
  of motion
• Neurons are arranged in direction hypercolumns
  within MT cortex

20
MT Function

• Lesions
• lesions to MT lead to deficits in perceiving
  motion
• Microstimulation
• stimulation of a neuron affects the perception of
  motion
• e.g., if you find a neuron with a preference for
  upward motion, and then use the electrode to
  stimulate it, the monkey becomes more likely to
  report upward motion

21
MT Architecture

• MT is stained with cytochrome oxidase (which
  indicates high metabolic activity)

22
MT Connectivity

• MT receives direct input from V1
• largely from the fast magno pathway cells
• MT projects to specific higher-level areas
• MT is an intermediate level visual area

23
MT Topography

• MT has a topographic representation of visual
  space

24
Part 3 How can we determine visual areas in the
human?
25
Tools for mapping human areas

• Neuropsychological Lesions
• Temporary Disruption
• transcranial magnetic stimulation (TMS)
• Electrical and magnetic signals
• electroencephalography (EEG)
• magnetoencephalography (MEG)
• Brain Imaging
• positron emission tomography (PET)
• functional magnetic resonance imaging (fMRI)

26
How well are human visual areas mapped?
27
An analogous mapping situation?
28
How can we map human (visual) areas?

• Look for homologues (or analogues) of known
  primate areas
• Example Human MT
• Look for areas that may participate in uniquely
  human abilities
• Example Language, calculation, social
  interaction areas

29
Back to our case study MT
MT
intermediate
V1
A patient with bilateral lesions to MT can no
longer perceive motion (Zihl et al., 1983)
A temporary disruption to human MT interferes
with motion perception (Beckers Zeki, 1995)
30
fMRI of Human MT
Video MT_on_rotatingbrain.mpg
Video V1MTmovie.mpg
fMRI in humans reveals MT
Moving vs. stationary dots activates V1 and
MT Flickering vs. stationary checkerboards
activates V1
31
Is Human MT an Area?

• Function
• properties of human MT are highly similar to
  macaque MT
• Architecture
• stains for cytochrome oxidase (like monkey MT)
• Connectivity
• not investigated
• Topography
• recent fMRI evidence finds a retinotopic map
  within human MT
• Evolutionary Relationships
• the existence of primate MT makes it likely that
  humans have a homologue
• the position of MT is considerably lower in
  humans than monkeys

predicted location
32
Part 4 How do Brains Evolve?
33
A Reminder About Evolution

• Evolution is a tree not a line

34
Evolution of Areas

• With evolution
• brains become larger (volume-wise)
• brains become more convoluted (more surface area)
• proportionately less cortex is devoted to primary
  somatosensory and motor areas
• there is more room for association cortex

35
Evolutionary Relationships

• Homology
• a structure, behavior or gene that has been
  retained from a common ancestor
• e.g., monkey hand and human hand

• Homoplasy (or Analogy)
• structures that look the same but do not
  necessarily have the same common ancestor
• e.g., bat wing and fly wing

36
Evolutionary Comparisons

• Knowledge of areas and topography can be useful
  in anticipating human areas
• Early visual areas show a good deal of similarity
  between human and monkey (based on retinotopic
  boundaries)

Macaque
Human
(fMRI)
(single neurons)
Tootell et al., 1996
37
MT Evolution

• MT exists in a variety of primate species
• In each of the three primate species, MT has the
  same types of cytoarchitecture and connections
• An MT-like area exists in tree shrews but its
  not yet clear whether it is homologous to primate
  MT

38
Primate vs. Cat Motion Areas

• A motion area (PMLS, posteromedial lateral
  suprasylvian area) exists in the cat
• Cat PMLS is unlikely to be a homologue of primate
  MT
• similar motion-responsive properties
• but does not exist in intervening species
• example of homoplasy
• illustrates independent evolution of similar areas

39
Part 5 So why are there so many visual areas and
how did they get there?
40
More brain, more visual areas
41
Why not one really big visual area?
V1
42
Why not a really big visual area?

• As areas become larger, longer interconnections
  are required
• Limits on cortical thickness and connections may
  constrain max area size

43
Parallel processing is more efficient

• Teach neural network to identify what and
  where
• One neural network with 18 nodes (neurons)
  devoted to both tasks
• versus
• One neural networks with two streams of 9 nodes
  each (total 18)
• After 300 training trials, the two stream model
  outperformed the single-system model

Rueckl, Cave Kosslyn, 1989
44
Different Tasks Require Different Information

• different regions may need to use different
  coding systems

dorsal stream viewer-centred
ventral stream object-centred
45
Wiring Constraints
David Van Essen proposes that as the brain
develops, areas that are richly interconnected
will be pulled together to form a gyrus (and
those that are weakly interconnected form sulci).
46
Sulcal Formation V1-V2
The V1/V2 border provides one example of two
richly interconnected areas that form a
gyrus. This arrangement also explains why maps
in V1 and V2 are mirror images of each other!
calcarine sulcus
Source Van Essen, 1997
47
Optimized Connections

• Multidimensional Scaling
• strength of connections can be used to infer
  spatial layout
• expected layout of visual areas matches anatomy
  amazingly well

Parietal
Occipital
Temporal
Malcolm Young
48
Its a small world after all

• Small world theory is being applied to
• social networks
• sexual networks
• terrorist networks
• Six degrees of Kevin Bacon
• Internet hyperlinks
• transportation networks
• THE BRAIN!

49
Big brain Small world
50
Optimized Connections

• Multidimensional Scaling
• strength of connections can be used to infer
  spatial layout
• expected layout of visual areas matches anatomy
  amazingly well

Parietal
Occipital
Temporal
Malcolm Young
51
General Organization Patterns
Malach et al., 2002, TICS
52
(No Transcript)
53
General Organization Patterns
Hasson et al., 2003, Neuron
54
Gene mutations
Pinky the Brain experiment
Chenn Walsh, 2002, Science

• simple gene mutations can have drastic
  consequences (e.g., changing one protein can
  change encephalization of a mouse)

55
Gene mutations

• Gene mutations can lead to a duplication of areas
• These areas can be mirror symmetric with existing
  areas
• Note This is one of many theories

56
Part 6 What questions remain?
57
Are humans just morphed monkeys?
Figure B shows the macaque monkey visual areas
from A morphed onto human cortex based on the
placement of sulcal landmarks (Van Essen et al.,
2001). fMRI activation suggests the areas have
moved somewhat. Can we assume humans are just
morphed monkeys? In some areas the human
cortical surface area is slightly larger than in
the macaque (e.g., visual cortex 2X) in others
it is considerably larger (e.g., parietal cortex
20X) Are individual areas larger? Are there
more areas?
58
Or are there expansion sources?
Frontal Eye Fields
MT
predicted location
predicted location

• most human functions seem to be located in
• inferior parietal cortex
• superior temporal cortex
• prefrontal cortex

Calculation
Language
59
Did new functions hijack old areas?

• ventral premotor cortex in monkeys
• mirror neurons
• ventral premotor cortex in humans
• mirror neurons
• Brocas area language
• Rizzolatti suggests speech evolved out of hand
  gestures and imitation

Rizzolattis Mirror Neurons
60
How much plasticity is there?
Experience seems to play a huge role
Are species differences due to genes or inputs?
61
How does Humpty Dumpty get put back together
again?

• The Binding Problem
• If an image (e.g., Humpty Dumpty) has each of its
  attributes (form, color, motion) processed in a
  different cortical area, how come we perceive the
  unified image?

• One possible solution
• Feedback loops may lead to synchronous firing
  between neuronal populations in different areas
  that encode the same attribute (e.g., spatial
  location)