Lecture 6: Attention Flashcards
Change blindness
attention can select only a subset of the sensory input for further processing.
we may be completely blind for unattended portions of the input
arousal vs. selective attention
Arousal is the global physiological and psychological state of an organism. From deep sleep to hyperalertness. involves ascending reticular activating systems at the brain stem.
selective attention is not global, but can act at any level of global state (arousal). selective attention is the allocation of resources among relevent inputs, thoughts, actions, while ignoring irrelevant or distracting ones. involves widespread cortical and subcortical networks.
Distinctions with selective attention:
overt vs. Covert attention
overt: attention is directed to the same location as the eyes.
Covert: attention is directed at another location than the eyes (example; boring lecture)
Voluntay vs. reflexive attention
voluntary: goal driven, top-down or endogenous attention is intentionally
reflexive: stimulus driven, bottom up, exogenous attention.
neuropsychology of attention
Neglect
attentional problems in the left visual field.
problems controlateral to (right hemispheric) lesion (after stroke) of parietal, frontal, temporal cortices. in particular regions centered around right temporoparietal junction.
normal vision and no blindness in left visual field.
but patients act as if left regions of space and left parts of objects do not exist. and they have limited awareness of their deficit
extinction
patients can detect left visual field stimuli normally when presented in isolation but not when presented simultaneously with stimuli in the right visual field.
can be partially overcome by directing patients attention to neglected locations
balint’s syndrome
inability to attend to more than one object at a time
severe problem after bilateral occipitoparietal leasions
simultanagnosia
difficulty in perceiving more than one object at the same time
optic ataxia
problems with visual guidance of reaching and grasping
ocular apraxia
problems making voluntary eye movements
The levels of attentional selection
Early selection (Broadbent)
Attention selects inputs before perceptual analysis is complete.
no higher-level semantics (meaning) processing of unattended inputs
The levels of attentional selection
Late selection
attention selects inputs only after perceptual analysis is complete.
partial or full higher-level, semantic processing of unattended inputs.
attenuation model (Treisman)
unattended inputs are not fully blocked from higher-level analysis but degraded.
cocktail party effect (Cherry)
one can focus on a single conversation while filtering out irrelevent chatter/music
however, one typically catches ones name despite being engaged in an intense conversation
dichotic listening tas (Cherry)
input to one ear need to be voiced and input to other ear is ignored.
hardly any memory for input from ignored ear.
not noticed in ignored ear: change in language, repeating the same word over and over
noticed in ignored ear: human voice or not, change in speakers gender
Cueing tasks (Posner)
Responses to a (visual) target stimulus under different conditions of attention
Three conditions: cues can either correctly (validly), incorrectly (invalidly), or not (neutrally) predict target
Cueing tasks (Posner)
Voluntary (endogenous) variant of the cueing task
Cue needs to be voluntarily processed (interpreted) by participant (they can choose to use it or not)
Cueing tasks (Posner)
Reflexive (exogenous) variant of the cueing task
Cue reflexively (automatically) attracts attention (participants cannot choose to use it or not)
Biased competition model
Competition in receptive fields (RFs)
-Multiple stimuli can fall into a neuron’s RF, and they “compete” to control the neuron’s firing
-RFs become larger when climbing up the visual cortical hierarchy
-There should be greater competition at later stages (more stimuli in RFs)
-Attention resolves competition in favor of the attended stimulus
-Effects of attention should be greater at later stages (more competition)
Greater competition at later stages
FMRI study comparing SEQ (sequential) vs. SIM (simultaneous) presentation of stimuli
Greater competition in SIM, reflected in reduced V4 but not V1 activity
Subcortical effects of voluntary visual spatial attention
Earliest effects in thalamus
- Focused spatial attention modulates activity in lateral geniculate nucleus (LGN) of the thalamus
- Reflects both initial attentional gating by LGN and later feedback from V1 to LGN
Reflexive visual spatial attention
Sensory-evoked ERPs measured with EEG
- Event-related potentials (ERPs) recorded during exogenous cueing task
- With short cue-target interval (ISI below 200 ms): larger P1 in cued trials (like endogenous variant)
- However, with long cue-target interval (ISI longer than 200 ms, inhibition of return), effect reverses: larger P1 in non-cued trials
- P1 reflects behavior both in exogenous and endogenous cueing task → similar effects of voluntary and reflexive attention on early visual processing
- But different control mechanisms!
Visual search
Pop-out vs. conjunction search
Pop-out search
- Target stimulus differs from distractors by a single feature only (e.g., color)
- Search is easy and search slopes are flat (independent of the number of distractors in display)
Conjunction search
- Target stimulus differs from distractors by more than a single feature (e.g., color and letter identity)
- Search is more difficult and search slopes are steep (dependent on the number of distractors in display)
- Why is conjunction search slow? According to feature integration theory (Treisman), because it requires spatial attention
Attention in conjunction search
Feature integration theory (Treisman)
- Individual features can be analyzed without attention and in parallel (in separate feature maps)
- Efficient pop-out search because differences in one feature are detected automatically
- Slow conjunction search because differences in one feature are not enough
- Instead, different features need to be “bound” together by the “glue” of spatial attention (in map of locations)
- Spatial attention is a serial bottleneck, therefore conjunction search is inefficient
Object attention
An object-based attention system
Behavioral evidence for object attention (Duncan)
- Box (small/large, left/right opening) with overlaid bar (dotted/dashed, left/right tilt)
- Task: judge one or two features
- Two judgments on the same object are as fast as one judgment
- Two judgments on two different objects are slower than one judgment
- Processing limitation in attending two objects (remember Bálint’s syndrome) → An object-based attention system in addition to a space-based system
Behavioral evidence for object attention: Cueing
- Standard voluntary arrow-cueing task but with wrench-like objects figures
- Spatial attention: faster responses to validly cued than to invalidly cued targets
- Object attention: faster responses to invalidly cued targets on the cued than on the uncued object
- Object attention “spreads” within the object, inaddition to the spatial “spotlight” of attention
FMRI evidence for object-spreading of attention
- Spatial attention: increased early visual cortex activity for cued vs. uncued targets
- Object attention: greater activity for uncued targets on the cued than on the uncued object
Object attention independent of spatial attention
FMRI: Ruling out spatial attention
- Superimposed transparent faces and houses, either moving or static
- Selective attention to either face, house, or motion (constant physical stimulation)
- Activity in face-sensitive fusiform face area (FFA) highest with attention to faces
- Activity in house-sensitive parahippocampal place area (PPA) highest with attention to houses
- Activity in motion-sensitive area MT/MST highest with attention to motion
- Demonstrates object-driven attentional effects independent of spatial attention
Consequences vs. sources of voluntary spatial attention
Target vs. cue processing
- So far, we have discussed the consequences of attention by looking at target processing
- To reveal the sources of attention, check what happens before the target appears, after the cue is shown
- Cues evoke anticipatory activity in specific visual cortex regions where the target is going to appear (“priming” of processing)
- Which areas control such preparatory attentional processes? (remember neglect!)
Effect of frontal control areas on sensory areas
FEF effects on V4 responses
- Microstimulation in FEF: map saccade vector of FEF neuron
- Stimulate FEF neuron below saccade threshold: no eye movement, but shift of attention to saccade goal (red arrow →)
- Record from V4 neuron with RFs (black circle) overlapping or not with saccade goal of FEF neuron
- Increased spiking with overlap of V4 RF and FEF saccade goal
Dorsal vs. ventral attentional control networks
Dorsal attention network: Voluntary attention
- Brain regions respond to cues when participants actively use the cues (rather than passively view them)
- Intraparietal sulcus (IPS), superior parietal lobule (SPL), frontal areas including frontal eye field (FEF)
- Similar control areas for feature and object attention
Dorsal vs. ventral attentional control networks
Ventral attention network: Reflexive attention
- Brain regions respond to salient, unexpected, novel stimuli that reorient attention
- Right lateralized temporoparietal junction (TPJ), ventral frontal cortex (VFC) incl. inferior and middle frontal gyrus (IFG, MFG)
- Damage results in neglect
