Transcript cog neuro - Carnegie Mellon University

Spatial representation and
parietal cortex
Marlene Behrmann
Department of Psychology,
CMU and CNBC
Contact: [email protected]
268-2790
•URL: http://www.cnbc.cmu.edu/~behrmann
Division of labour in human
visual cortical system
Mishkin and Ungerleider
The parietal lobes:
• “crossroads of the brain”
(Critchley, 1953)
• well situated
topographically
• multimodal
Inferior
Parietal lobule
Supramarginal Angular
gyrus
gyrus
• requisite cortical and subcortical connectivity
• Distribution of MCA, little collateral supply
Neglect
Federico Fellini
(1920-1993)
Neglect
• Asymmetry in incidence: RH (66%) in humans
not monkeys
• Inferior parietal lobule
– Areas 39 and 40;
– non-human primate analog: IPL (7) vs STS
– bimodal: short-lived vs persistent
• Affects different sensory modalities
• Not sensory deficit
*
*
Not just parietal
• Distributed network: (Mesulam; Heilman)
– dorsolateral prefrontal, medial frontal (cingulate,
thalamus, basal ganglia, white matter).
– Same network activated in eye movement studies
• Close relationship between attention and eye movements
(Corbetta et al.)
• Other terminology
– Extinction
– Allesthesia
– anosagnosia
Extent of effect
• Other sensory modalities
– Auditory
– Olfactory
– Tactile
• Mental imagery
– Piazza del Duomo (Bisiach and Luzzatti)
• Affects output: not surprisingly
Vision and eye movements
(Behrmann et al.)
• 45 x 36 degrees visual angle
• magnetic scleral coil in right eye
• indicate number of As in display
Lesions:
neglect patients
Lesions:
BD controls
with hemianopia
Location and duration of fixations
Eye movements reflect left-sided neglect
(a) Fixations
(b) Duration
50
Percentage
40
30
20
Controls
10
Neglect
Hemianopic
0
1
2
3
4
1
Quartile Bands
2
3
4
QuickT ime™ and a TI FF (Uncompressed) decompressor are needed to see this picture.
Qu ickTime™ and a TIFF (Unc ompressed) de compressor are needed to see this picture.
What can we learn about parietal
cortex?
• What determines what information is
neglected?
• What gives rise to neglect?
• What happens to the information that is
neglected?
What can we learn about parietal
cortex?
• What determines what information is
neglected?
• What gives rise to neglect?
• What happens to the information that is
neglected?
Possible frames of reference:
what defines ‘left’?
Visual neglect in allocentric
coordinates
Depiction of environmental neglect
Visual neglect in allocentric and
egocentric coordinates
Tactile and visual neglect
Moscovitch and Behrmann
Tactile neglect in allocentric co-ordinates
Mean neglect errors (maximum =10)
5
Thumb
4
Little finger
3
2
1
0
Palm down
Palm up
Side of hand
Range of possible frames of
reference
Neglect with respect to object
midline
• Target
• Copy
Behrmann and Tipper
right object
right space
left object
right space
Object-based neglect: inhibition for
right and facilitation for left targets
1500
Static
RT (in msecs)
1250
Rotating
1000
750
500
Left
Right
Side of target
Directional selectivity of neurons
Olson and Gettner (1995, 1998)
A
Fixation spot: 200 msec
B
Sample bar: 500 msec
C
Cue: 300 msec
D
Delay: 400-600 msec
E
Fixation spot of f
Target bar on
F
Response
L/R of bar
L/R eye movement
Neurons fire for left or right of bar
independent of direction of movement
Left with Respect to Bar
Rig ht wi th Respect to Bar
Cue
on
Eye Movement
50 0 mse c
Cue
on
Eye Movement
50 0 mse c
Paradigm: object and environ- neglect
square
circle
Simultaneous object- and
environ-based neglect
Circles
Squares
(object-based coordinates) (envir-/viewer-based coordinates)
RT (in msecs)
750
650
550
450
350
Left
Right
Left
Side of target
Right
Multiple reference frames in eye
movements too
BLOCK
DEJECT
FLAP
FARM
FEVER
BRAIN
HEIGHT
HOLDER
TREAD
FLAME
BRIGHT
SLEDGE
CAGE
WHIP
TEAR
Normal: no errors in reading
Neglect: errors and eye movements
Another example
BLOCK
DEJECT
FLAP
FARM
FEVER
BRAIN
HEIGHT
X
X
HOLDER
TREAD
SLEDGE
FLAME
right
BRIGHT
X
CAGE
X
TEAR
WHIP
Hemianopic: no reading errors
Frames of reference
• Egocentric (dependent on viewer)
• But also allocentric (independent on viewer)
– Not only in vision, also in tactile
• Multiple coordinate frame
– Also evident in eye movements
What can we learn about parietal
cortex?
• What determines what information is
neglected?
• What gives rise to neglect?
• What happens to the information that is
neglected?
What gives rise to neglect?
gradient
Gradient consistent with neuronal distribution:
68 bilateral, 29 contra, 3 ipsi
Suggests competition too:
bad on left, too good on right
Visual search paradigm
I. ÔPop-outÕTask.
Target defined by a distinctive feature:
specifically, an intersecting line.
II. ÔSerialÕTask.
Target defined by the lack
of a distinctive feature.
Normal subjects
Feature search
Conjunction search
a. Normal controls
4500
4580
left target
right target 3580
no target
2580
1500
1580
500
580
121
10
5
6
5
6
0
1
12
10
2500
0
ms
3500
Patients with LHD
ms
b. LHD N-
4500
4580
3500
3580
2500
2580
1500
1580
500
580
ms
3500
3580
2500
2580
1500
1580
500
580
10
5
4580
0
10
5
4500
0
c. LHD N+
Feature search
a. RHD N-
Conjunction search
4500
4580
3500
3580
left target
right target
ms
no target
2500
2580
1500
1580
500
580
Patients
with
RHD
b. RHD N+
6
6
11
1
11
6
1
ms
1
4580
4500
3500
3580
2500
2580
1500
1580
500
580
ms
4500
4580
3500
3580
2500
2580
1500
1580
500
580
1
6
11
11
6
1
c. RHD N++
12 1
display size
6
12
So:
• Spatial deficit apparent in visual search
• Generally scaled with severity of neglect
– But more in RHD than LHD
– Some competition: better on right side as
neglect severe in conjunction
What can we learn about parietal
cortex?
• What determines what information is
neglected?
• What gives rise to neglect?
• What happens to the information that is
neglected?
Fate of neglected information
• Do patients process unattended information
normally?
– failure to reach consciousness?
– degraded processing?
Volpe, Ledoux and Gazzaniga
(1979)
Same or different?
Name the objects
Response: different
Response: A star
Unconscious processing/ failure to explicitly report
information that is available
Priming paradigm
McGlinchey Berroth et al.
related
GOSE
unrelated
THIP
Normal subjects: faster lexical decision time if related than unrelated
Results
• Patients faster to say ‘yes’ when semantic related
pic on right (eye-nose) compared to eye-ship
• Patients faster to say ‘yes’ when semantic related
pic on left
• SAME AMOUNT OF PRIMING/FACILITATION
FROM BOTH SIDES! Even though report the
information.
• Patients process neglected info normally.
Results … Continued
• BUT: normals show double priming on
left than right
• Cannot conclude processing is normal!
• Priming not as demanding as explicit
report.
What can we learn about parietal
cortex?
• What determines what information is neglected?
– Multiple reference frames, demands of task
• What gives rise to neglect?
– Competition between residual activated neurons, spatial and
temporal factors relevant
• What happens to the information that is neglected?
– Seems to be activated to some extent, not fully
Alternative view of parietal cortex:
“how” not “where”
• Milner and Goodale
– Parietal cortex involved in on-line ballistic
movements
Where the action is!
Grasping
matching
What can we learn about parietal
cortex?
• What determines what information is neglected?
– Multiple reference frames, demands of task
• What gives rise to neglect?
– Competition between residual activated neurons, spatial and
temporal factors relevant
• What happens to the information that is neglected?
– Seems to be activated to some extent, not fully