Brain Areas Devoted to Vision

1
Brain Areas Devoted to Vision
2
The Retina An Approachable Part of the Brain

• - John Dowling

3
(No Transcript)
4
Spatial ResolutionThe next slide (from Anstis)
shows the fall-off in spatial resolution in the
periphery. By fixating at the center, you should
be able to see all the letters equally well.
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
Two types of Photoreceptors

• Cones
• Less sensitive to light
• Three types (usually), responding differently to
  different wavelengths
• More heavily represented in fovea
• Rods
• More sensitive to light
• Only one type no color differences
• More heavily represented in the periphery

10
(No Transcript)
11
(No Transcript)
12
False color image showing the arrangement of
cones in a macaque monkey at a location 1.4 deg
nasal from the central fovea.
13
Because there are only 3 cone types, the retina
can be tricked into thinking a particular
wavelength is present with the appropriate
combination of three other wavelengths.
14
Colorblindness

• Not really blindness at all
• Inability to distinguish certain colors
• usually red and green
• Often due to differences in which cone types are
  present.
• Usually in males.

15
(No Transcript)
16
(No Transcript)
17
(No Transcript)
18
Retinal ganglion cells respond best to
center-surround patterns.
19
Changes in ganglion cell membrane potential
accompanying center and surround responses.
20
(No Transcript)
21
(No Transcript)
22
There are a number of different types of retinal
ganglion cells.

• Magnocellular
• Larger
• Lower spatial resolution
• Faster and more transient responses
• Important for motion perception
• Projects more strongly to dorsal (WHERE) pathway
• Parvocellular
• Smaller
• Higher spatial resolution
• Slower and more sustained responses
• Important for color and form perception
• Projects more strongly to ventral (WHAT) pathway

23
Color Opponency in Parvocellular Retinal Ganglion
Cells
24
(No Transcript)
25
The axons of the retinal ganglion cells lead from
the retina to the Lateral Geniculate Nucleus
(LGN) of the thalamus.
26
(No Transcript)
27
(No Transcript)
28
LGN

• Each LGN represents only the contralateral
  hemifield.
• 6 layers, each receiving inputs from only one
  eye.
• Each layer organized as spatial map.
• Like retinal ganglion cells, LGN cells generally
  respond best to center-surround stimuli.

29
LGN

• The top 4 layers have parvocellular neurons
  making up the P pathway.
• Smaller cells, slower responses
• Carry info from cones about color
• 80 of LGN neurons in macaque
• The botton 2 layers have magnocellular neurons
  making up the M pathway.
• Larger cells, faster responses
• Important for detecting motion
• No info about color

30
90 of the axons in the optic nerve connect to
the LGN, and indirectly to V1.There is also an
older visual pathway that passes through the
Superior Colliculus rather than through the LGN.

• No color information
• No complex form information.
• Important for eye movements.
• Connects with cortical visual areas.

31
(No Transcript)
32
(No Transcript)
33
Axons from LGN neurons connect to Primary Visual
Cortex (area V1). Their axons terminate in layer
4C.The input to V1 is still monocular. V1 has
many monocular cells, but also binocular cells,
which receive inputs from both eyes.Beyond V1,
all cells receive inputs from both eyes.
34
Orientation Selectivity

• Many cells in V1 and higher areas are tuned to
  respond to edges with a particular orientation.

35
Tuning curve and corresponding polar plot from 2
macaque V1 cells (one orientation selective, the
other nonselective) in response to drifting bars
of varied orientation and direction.
OBorientation bias. DBdirection bias.
36
Different Types of Cells in V1

• Simple
• Edge of specific orientation at specific location
  in receptive field.
• Complex
• Edge of specific orientation anywhere within
  receptive field.
• Hypercomplex
• Edge of specific orientation anywhere within
  receptive field.
• End-stopping Edge must end within receptive
  field.

37
Simple Cell in V1
38
Complex Cell in V1
39
Ocular dominance and orientation-selective
columns in V1.
40
Blobs and Interblob Regions

• In layers 2 and 3 of area V1.
• Mainly part of the P pathway.
• Blobs
• Cells that respond to color differences.
• Interblob Regions
• More sensitive to orientation and form.

41
M pathway represented mainly in layer 4B of area
V1.

• Transient responses
• Not sensitive to color differences

42
(No Transcript)
43
(No Transcript)
44
The separate pathways seen in V1 are are also
found in V2.
Thick Stripe
Interstripe
Thin Stripe
45
Pathways in V2

• Thick Stripes
• Motion
• Form
• Thin Stripes
• Color
• Interstripes
• Form

46
(No Transcript)
47
(No Transcript)
48
Stimuli used to map cortical visual areas.

• These stimuli create waves of activity that
  travel across each visual area.
• The timing of this activity wave can be used to
  construct a retinotopic map of each area.

Stimuli used to create traveling waves of neural
activity in retinotopically organized cortex are
shown. The stimuli were composed of a
contrast-reversing checkerboard pattern
flickering at 8 Hz. (A) The expanding ring
stimulus is shown at five moments in time
spanning one stimulus cycle. (B) At each location
within the visual field, the stimulus follows a
square-wave alternation between the
contrast-reversing rings and the uniform gray
field. The expanding ring stimulus was delayed in
the periphery relative to the center hence, the
temporal phase of the square-wave alternation
varied as a function of distance from the center
of the visual field. Two stimulus cycles are
shown. (C) The rotating wedge stimulus is shown
at five moments in time spanning one stimulus
cycle. From Engel, Glover, Wandel (1997).
Cerebral Cortex, 7, 181-192.
49
Mapping One Visual Hemifield
upper vertical midline
eccentricity
horizontal midline
fovea
lower vertical midline
50
How a quadrant of retinal space is mapped onto
cortical space.
Grill-Spector Malach (2004). Annual Review of
Neuroscience, 27, 649-677.
51
Mapping Retinal Space onto Cortical Space in V1,
V2, etc.

• Left and right hemifields represented in
  contralateral hemispheres.
• Much more cortical area devoted to fovea than to
  periphery.
• Some maps have split along horizontal midline,
  separating hemifield into quadrants.
• Each quadrant mapped onto a rectangular stripe.

52
V1 and V2 in Human Occipital Cortex

• Showing posterior portion of inside surface of
  left hemisphere.
• Upper visual field represented in inferior
  region lower field in superior.

53
Flattened Maps of Visual Cortical Regions in
Human and Macaque
54
The Ventral Portion of Human Visual Cortex,
showing activation according to eccentricity.
Eccentricity distance from fovea.
55
These visual areas are just the beginning. There
are over 30 visual areas in humans.These
different visual areas respond differently to
different visual properties.

• MT responds to motion.

• V4/V8 responds to color.

56
(No Transcript)
57
(No Transcript)
58
Connections Between Areas in Macaque Visual Cortex
WHERE
WHAT
59
WHAT and WHERE Pathways
WHERE
WHAT
60
WHERE Pathway
WHAT Pathway

• Representing locations of things.
• Includes representation of speed and direction of
  motion.
• Important for manipulating objects.
• Damage can cause spatial neglect.

• Object recognition.
• Receives information about color, orientation,
  form.
• Damage can cause visual agnosia.

61
Attention
62
Convergence within the What Pathway

• Higher levels of the hierarchy have larger and
  larger receptive fields.
• At the top levels , each neuron can receive input
  from many different locations and many different
  objects.
• Potential for interference and crosstalk.

63
Convergence within the What Pathway

• The flow of information through the What Pathway
  has to be regulated to prevent interference.
• Information from some objects must be blocked so
  that information from other objects can be fully
  processed.
• Gating of information.

64
The Need for Selection

• Visual processing requires a mechanism for
  selecting some inputs and inhibiting others.
• This selection process is often referred to as
  attention.
• The same sort of selection process is needed
  outside visual processing.
• Within the auditory system and other perceptual
  systems.
• To select between different perceptual
  modalities.
• To select internal memories and thoughts for
  attention.

65
Two Different Types of Visual Attention

• Overt Attention
• Shifts of eye position
• Covert Attention
• Selecting certain locations or objects within a
  fixation, without moving the eyes.

66
Helmholtz1894
67
Two Different Factors Affect how Attention is
Allocated

• Voluntary Attention
• Governed by the subjects goals
• Endogenous
• Top-down
• Reflexive Attention
• Driven by salient aspects of the stimulus
• Exogenous
• Automatic
• Bottom-up

68
The neural responses to perceptual stimuli are
shaped by attentional selection.

• Effects of attention demonstrated experimentally.
• Same stimuli used in different conditions.
• Only the allocation of attention is different
  across conditions.
• Differences in neural response show effects of
  attention.

69
ERP Experiments
70
Measuring Endogenous Attention with ERPs

• Stimulus on left in both conditions.
• In one condition, subject expects stimulus on
  left and attends there.
• In other, attention is on right.
• Response in occipital lobe contralateral to
  stimulus is greater when that location is
  attended.

71
Measuring Attention with ERPs
72
Effects of Endogenous Attention on P1

• Shows the difference between cued and uncued
  responses.
• Note also differences between upper and lower
  visual field.

73
Exogenous Attentions Effect on P1
Inhibition of return at longer delays.
74
Exogenous Attentions Effect on P1
75
Later Attentional Effects

• P1 occurs 100 msec or left after stimulus
  appears.
• Another component, N2pc, occurs 200-300 msec
  after stimulus.
• pc posterior contralateral
• Stronger on attended side than on unattended
  side.
• Can be used to track movement of attention from
  one side to the other.

76
Measuring Shift of Attention from One Side to the
Other

• Visual search task.
• Target has gap on left.
• On 75 of trials, target is red.
• On 25 of trials, target is blue.
• Attend first to red, then blue.

77
Measuring Shift of Attention from One Side to the
Other

• On some trials, red and blue will be on same side
  of display.
• Attention should stay on that side.
• On other trials, red and blue will be on opposite
  sides.
• Attention should shift from one side to the other.

78
Measuring Shift of Attention from One Side to the
Other
79
ERP studies show the temporal dynamics of
attention. To get better information about the
spatial layout of the brain images involved, we
need studies using PET and fMRI.
80
Two general findings emerge from brain imaging
studies of visual attention.

• Brain areas devoted to specific visual features
  become more activated when those features are
  relevant to the task.
• Location
• Color
• Shape
• Motion
• Some brain areas serve to control attention.
• Parietal cortex
• Pulvinar
• Superior colliculus

81
Attention Modulates Visual Processing in
Extrastriate Areas

• When task requires attention to color, activity
  increases in areas devoted to color.
• Similar pattern for motion and shape.

82
Attention Modulates Visual Processing

• Attentional modulation is strongest in higher
  areas of the ventral (what) pathway.
• Harder to find attentional modulation in V1.
• Attentional modulation found recently in LGN.
• Connections from cortex back down to LGN
• No connections back to retina to allow
  attentional effects there.

83
Distractor Interference

• Spatial attention can help to control
  interference from distractor objects.
• Compare brain activity when target object appears
  alone, and when it appears with distractors.
• When distractors are present and target is
  attended, response is similar to when target
  appears alone.

84
Additional evidence on attentional control of
distractor interference comes from single-cell
recordings.
85
In one block of trials, the monkey attends to
stimuli with one color. In another block, monkey
attends to the other color.
86

• Two stimuli with different colors in the neurons
  receptive field.
• One (effective) triggers a strong response in
  this neuron.
• The other (ineffective) does not.

87

• When ineffective stimulus is attended, neural
  response is weak, similar to response when
  effective stimulus is not present.

88
Attentional effects are seen in higher level
visual areas as well.First, a word about two
higher level visual areas in humans.
89

• Fusiform Face Area (FFA)
• Parahippocampal Place Area (PPA)

90
Two Objects Occupying Same Region Produce Rivalry
91
Activity in FFA and PPA trades off according to
reported perception.
92
Object-Based AttentionActivity of higher visual
areas gated by attention to objects.
93
Attention modulates the activity of different
visual brain areas.Other brain areas become
active when attentional control is necessary.
94
Parietal Cortex

• Task requires shifting attention to different
  locations.
• Shifts generate activity in posterior parietal
  cortex.
• Similar activity during visual search, which also
  requires attentional shifts.

95
Posner Cuing Paradigm

• Cue indicates where to attend.
• Response faster when stimulus appears at cued
  location.

96
For patients with parietal lobe damage, the delay
for a stimulus at an unexpected location
contralateral to lesion is much greater.
Difficulty in shifting attention contralateral to
lesion comparable to neglect.
97
Parietal damage can also cause spatial neglect
and extinction.
98
There is some evidence of neglect in imagery as
well.
99
Many Methods Contributing to Attention Studies

• ERP
• PET
• fMRI
• Dipole modeling to link ERP and fMRI results
• Single cell recording
• Chemical activation and inhibition
• Patient studies

100
Brain Areas Involved in Visual Attention