L12 - Visual Processing: Object Representations and Neuropsychological Disorders Flashcards
1
Q
Object representations in LOC
What is the fMRI adaptation paradigm?
A
- An imaging paradigm pioneered by Malach, Grill-Spector and colleagues
- Gets around some of the limitations in spatial resolution of fMRI - already has good spatial resolution but still contains thousands of neurons in closest
- Allows us to discern if in a particular area there are neural populations with different response properties vs a homogeneous neuronal population
- A way of probing the nature of the object representations in a particular area (as there is a cluster of neurons in that space)
- Are the invariant with respect to size, location, orientation, etc?
- Gets around some of the limitations in spatial resolution of fMRI - already has good spatial resolution but still contains thousands of neurons in closest
- BOLD response decreases when a stimulus is repeatedly
- approximates neural adaptation – repeated firing “tires” or “sharpens” the neurons
- Perhaps neurons are becoming more efficient so response is not as intense when you repeat the stimuli
- Measures neural activity indirectly quite well
- If you adapt, then change some aspect of the stimulus (e.g. size, orientation, etc) and you see recovery from adaptation, this implies = new population of neurons is responding to this new attribute
- If there is no recovery from adaptation, this implies that the same (adapted) neural population responds irrespective of change in size, orientation, etc
- approximates neural adaptation – repeated firing “tires” or “sharpens” the neurons
e.g. showing stimulus
- Homogenous population (right hand-side) which are size-invariant - fMRI response will remain at that adapted level
- Neurons on the left - more finely tuned, the green neurons - fresh response dependent on size of stimulus
2
Q
Object representations in LOC
What did they find with activity invariant neurons with object representation in the LOC?
A
Invariant in …
- Size
- Position in VF
- Image format (grayscale pic vs line drawing, shape defined by motion, luminance, stereo cues)
- Visual or tactile input - not much recovery from adaptation
- But more sensitive to viewpoint and illumination - could be a very different shape reaching the brain
-
LOC = high-level object shape representation
- Not as bothered about lower level surfaces
3
Q
LO and LOtv
How is LO different to VOT in response profile?
A
(LO) Lateral division
- More sensitive to changes in location and size than VOT
- More sensitive to 2D shape features - e.g. a line drawing
- A sub-region called LOtv (Lateral Occipital Tactile Visual area) is activated equally by visual and haptic input (Amedi et al, 2002) , not much recovery from adaptation
- LO codes the geometry of the shape
LOtv
- Visual-tactile area LOtv activated by both seen and palpated objects, but not by characteristic sounds of objects (e.g. a telephone ring)
- No green activation in LOtv
BUT
- LOtv also activated by soundscapes or auditory “shapes” (Amedi et al., 2007)
- 3 dimensions
- Sound frequency
- Time (left/right)
- Brightness conveyed by loudness
- Soundscape = carries information about the shape you are looking at
4
Q
LO and LOtv
What was Amedi et al. study on how to use sound device to perceive soundscapes?
A
- 2 blind and 5 sighted subjects trained to use vOICe to recognise objects
- 5 subjects with no vOICe training were taught to simply associate objects with soundscapes
- After vOICe training, LOtv was activated by vOICe objects, compared to scrambled vOICe sounds or other object-specific sounds (e.g. telephone ring)
- LOtv was not activated by vOICe objects in the subjects with no vOICe training
- LOtv - Very sensitive to the geometry of the shape
- LOtv is a multimodal shape area
5
Q
VOT (Ventral division)
What is the profile of VOT?
A
- More invariant to changes in location and size than LO
- Less recovery from adaptation when change surface contours - generalises more between photographs and line drawings
- Sensitive to perceived 3D shape (despite different 2D contours)
- Not activated by haptic input
- Correlates with recognition performance
- Codes a more abstract identity representation and mediates awareness whereas Lotv more about geometry of the shapes
6
Q
Disorders of Object Processing
What is agnosia?
A
“not knowing” or “loss of knowledge”
7
Q
What is visual agnosia?
A
- A failure to make sense of visual information to know what it represents
- Person is not blind, elementary visual function is intact (e.g. normal contrast sensitivity)
- Can recognise things from other modalities
8
Q
What are the different classifications of visual organisations? (Lissauer (1890))
A
- Apperceptive
- Recognition deficits linked to problems in perceptual processing
- Associative
- Patient can derive a normal visual representations but cannot link them to information stored in memory
- A normal percept stripped of its meaning
- Patient can derive a normal visual representations but cannot link them to information stored in memory
9
Q
What might be examples of what will occur in apperceptive agnosias?
A
10
Q
What is visual form agnosia?
A
- very severe form, caused by widespread bilateral damage to the occipital lobes (usually from CO poisoning) or head injury
- cannot discriminate even simple shapes (e.g. square from a triangle), cannot copy drawings, cannot read, cannot recognise faces
- Lower level might be in tact
11
Q
What is an example case of visual form agnosia?
A
Case DF - Milner et al., 1991
DF’s lesion on a rendered surface of the brain, compared to the location of LOC in normal subjects. - Damage LOC - get this form of agnosia
12
Q
What are associative agnosias?
A
- Failures of recognition that cannot be attributed to faulty perception
- Patients can copy drawings, discriminate shapes, segment images
- Cannot identify objects
- Due to disconnection between intact perceptual input and memory? (peripheral deficit)
- Or loss of stored object representations? (central deficit)
13
Q
What different stimuli do associative agnosia target?
A
- Visual object agnosia – inability to recognise objects
- Prosopagnosia – inability to recognise faces
- Alexia – inability to recognise words - associated with lesions in left-hemisphere
- Topographical agnosia – inability to recognise familiar environments and landmarks
- Lesions causing associative agnosias tend to be in the occipital and temporal regions (vOT?) – generally bilateral
- Objects (and especially words) more in the left hemisphere
- Faces more in the right hemisphere (right Fusiform Gyrus)
14
Q
What are simple dissociations?
A
- Patient can do A but not B
- Could conclude that A and B tap into different (and independent) processes
- BUT this could be due to a difference in task difficulty!
15
Q
What are double dissociations?
A
- One patient can do A but not B
- A different patient can do B but not A
- Much more powerful evidence that A and B are independent
