Human (ERP and imaging) and monkey (cell recording) data together

1
Human (ERP and imaging) and monkey (cell
recording) data together

• 1. Modality specific extrastriate cortex is
  modulated by attention (V4, IT, MT).
• 2. V1 is modulated when task conditions are
  demanding in cell studies, but disagreement
  between ERP and fMRI for V1 may reflect both
  initial (ERP) effects and later (fMRI) effects.
  Have to see it before you attend to it?
• 3. Attention Mechanisms include both enhancement
  and inhibition. Increased neuronal gain

2
Modeling Selective Visual Attention

• Robert Desimone and John Duncan(1995)
• Biased Competition Model
• Hypothesis
• Multiple stimuli in the visual field activate
  populations of neurons that automatically engage
  in competitive interaction, which are assumed to
  be through intracortical connections.
• When attention is directed to a stimulus, this is
  thought to be accompanied by feedback signals
  generated within areas outside the classical
  visual system.
• These signals bias the competition. As a result,
  neurons responding to attended stimulus remain
  active while suppressing neurons responding to
  unattended stimulus.

3
Duncan and Desimone proposal for attention
selection

• Competition results in few or one stimulus
  actively represented at a time in distributed
  representations (lateral inhibition, winner take
  all effect). This prevents cross talk or
  interference problems.
• Pattern completion highlights commonalities.
  Attending to a color will bring up all stimuli
  that have that color.
• Winner determined both by bottom up effects
  (intensity and novelty) and top down activation
  from higher areas.

4
(No Transcript)
5
(No Transcript)
6
More about biased competition
One theory that brings together all of the
reviewed attention effects (top-down biases, gain
modulation, enhancement and suppression) is
Desimone and Duncans biased competitionmodel
of attention. The theory rests on three
assumptions. First, given the limits on our
ability to process several stimuli at once,
visual objects compete for representational
resources, and only one or a small number of
stimuli can be represented at one time. As the
neural representations of visual stimuli are
highly distributed, competitive processing occurs
in many of the brain areas sensitive to visual
input. Second, the competition is integrated
across several areas, such that the
neural populations that represent different
aspects of a single object interact in a
mutually facilitatory fashion. The gain in
response to the selected object is accompanied
by suppressed processing in the neural
populations representing features of
different objects. Therefore, as a winner
emerges in one system, the same object
becomes dominant across the distributed
network. Last, the competition can be biased
not only by bottom-up factors (for example,
stimulus intensity), but also by top-down
influences that are based on current task
demands. Top-down bias is reflected in neural
priming (enhanced processing) of populations
representing the relevant object
attributes, resulting in a competitive advantage
for the relevant stimulus. An important
challenge for this theory (and other theories of
attention) is to explain precisely how the
distributed neural populations responding to a
single object know that they are representing
the same object and so should enhance each other
while suppressing the neural representations of
other objects (the binding problem).
7
Modeling Selective Visual Attention
Figure from Neural Mechanisms of Selective Visual
Attention by Robert Desimone and John Duncan
8
Controlling attention The top downPrefrontal
cortex

• Prefrontal cortex is called the executive system
  of brain and has major role in working memory
  (maintaining representations while we need to
  keep thinking about them) consistent with a
  control of attention role.
• Dorsolateral prefrontal cells maintain activity
  during active attention to a location in absence
  of stimulus, and level of activity is related to
  level of attention so it could bias earlier
  areas. Ventromedial areas maintain activity
  during active attention to objects.
• Bloodflow studies provide best evidence, but need
  good designs to be convincing. Best control is
  exogenous (nonpredictive) cue like at bright box
  at some location versus endogenous pointing cue
  at fixation. Right dorsolateral prefrontal
  bloodflow is unique to endogenous (top down)
  condition.

9
(No Transcript)
10
More top down The posterior parietal cortex

• Neurons in posterior parietal cortex increase
  firing for attended stimuli and locations.
• Attention to spatial locations increases
  bloodflow in posterior parietal cortex.
• Visual neglect or extinction occurs with right
  posterior parietal damage.

11
Why do we need posterior parietal cortex?

• Need to mediate between high level
  representations of objects in space that guide
  our topdown allocation and retinotopically
  mapped modality specific representations of
  visual stimuli that are subject to competition
  effects
• We dont want attention to jump when we move our
  eyes, we also want to attend to other aspects
  besides visual. Unlike attention experiments
  discussed so far we dont keep our eyes on a
  constant fixation point. We move around so
  correspondence between retinal location and
  location in the world is constantly shifting.
• We need to translate between world and retinal
  coordinates.

12
Posterior parietal cortex neurons encode for
intention to move

• Egocentric space is represented. Eye movements,
  head movements, limb movements that will get you
  to what you are interested in.
• Salient (important, attended to stimuli) are
  represented.
• So interface between what we want and how we get
  to it.L6Action.swf

13
Parietal cortex translates between world and
retinotopic co-ordinates

• Parietal neurons modulate firing to receptive
  field stimuli depending on fixation or eye
  position. This isnt all you need to figure out
  where the stimulus is in head-centered space but
  if you have a population of these (distributed
  representation) then you can pinpoint one
  location

14
Parietal cortex also has nonretinotopic fields
(LIP)

• Cells encode the memory of the location of the
  field and shift with eye movement (before the
  eyes get there!).

15
A planned shift in the visual world is seen by
parietal neurons before it happens
16
Supramodel attentional control

• Need to co-index attention for visual, auditory
  and tactile qualities.Need integration of
  multiple sensory and motor representations. See
  this in VIP. Hemispatial neglect is all about
  loss of supramodal attentional control (p
  207-208.)

17
Desimone 2005 Parallel searches then serial
searches so not a single spotlight model and not
a simple binding by location model.
18
New parallel and serial model

• Throughout the period of searching, neurons gave
  enhanced responses and synchronized their
  activity in the gamma range whenever a preferred
  stimulus in their receptive field matched a
  feature of the target, as predicted by parallel
  models.
• Neurons also gave enhanced responses to candidate
  targets that were selected for saccades, or
  foveation, reflecting a serial component of
  visual search.
• Thus, serial and parallel mechanisms of response
  enhancement and neural synchrony work together to
  identify objects in a scene.

19

• Features (parallel)
• The feature-related enhancement we observed is
  likely the result of a combination of
  feature-selective responses in the visual cortex,
  including V4, and top-down feedback from
  structures involved in working memory and
  executive control, such as the prefrontal cortex
  and possibly the parietal cortex. Such feedback
  must be capable of targeting neurons with the
  appropriate feature preferences throughout the
  visual field map.
• Location (serial)
• The saccade-related enhancement, on the other
  hand, likely originates from feedback to V4
  neurons with RFs at particular locations,
  originating from structures with spatial
  attention and oculomotor functions such as the
  frontal eye field and the lateral intraparietal
  area. These areas are thought to represent a
  salience map in which stimuli are represented
  according to their behavioral relevance
  independent of their features

20
More top down The posterior parietal cortex

• Neurons in posterior parietal cortex increase
  firing for attended stimuli and locations.
• Attention to spatial locations increases
  bloodflow in posterior parietal cortex.
• Visual neglect or extinction occurs with right
  posterior parietal damage.

21
Hemispatial neglect
22
(No Transcript)
23
Left Visual Extinction
Extinguished
Right
Left
Right