Visual Agnosias

Visual Agnosias

1. a deficit in object recognition

   • damage to extrastriate areas (e.g. V2, V3), but V1 must be intact.

2. Lissauer (1890)

   • Apperceptive
     • unable to perceive a coherent percept of object structure
       • cannot identify or copy shapes
         • Less bad at letter copying as V1 (undamaged) can process simple stimuli such as straight lines
       • have difficulty with fragmented or unusual views of objects (e.g. probs with Warringtons "unusual views" test and Gollins fragmented figures)
     • "more general"
     • Milner et al. (1991), Sparr et al. (1991): MRI = damage to lateral occipital areas
       • most commonly due to an anoxic episode (e.g. carbon monoxide poisoning)
         • damage to watershed areas (where arteries join) as they are most vulnerable brain tissues
       • disruption at the early stages of visual processing
     • Benson & Greenberg (1969) studied MR S: unable to recognise faces, identify line drawings/simple shapes - unablt to distinguish between the majority of the rectangles in the Efron Test (better with real objects as he could use colour and texture cues, but still guesses).
       • BUT: he could navigate corridors successfully in his wheelchair
         • Similarly, Campions (1987) patient RC could negotiate obstacles in the room
           • its assumed in these cases the dorsal stream is still intact. (Not the case for everyone - others might have more extensive lesions which encompass ventral + dorsal structures
       • Also tested extensively by Efron = argued Mr S disorder was one of visual contour
   • Associative
     • objects can be perceived but not recognised
     • damage to inferior temporal regions
     • can copy line drawings quite well but does not recognise
     • disruption at higher cognitive level
       • semantic system operates as the final stage of recognition. therefore, any unequivocally 'associational' agnosias must result from the disconnection or structures dependant on ventral stream outputs
     • unable to access semantic information associated with percepts
     • no problems with unusual views based on perceptual cues but fail on matching based on functional similarity (e.g. does not relate an open umbrella to a closed one)
     • usually from stroke
   • Humphreys & Riddoch (1987): this binary scheme of classification is not perfect for the pattern of object recognition disorders described in more recent literature.
     • Regardless, L's distinction was useful and remains a convenient, if crude, initial means of classification.

3. Patient D.F. visual form agnosia

   • NOT a low-level visual deficit (can detect light throughout visual field-Milner et al., 1991) and has well preserved colour discrimination and HIGH spatial frequency grating patterns
     • E.g. Milner & Heywood (1989): she was much better at differentiating between fine shades of colour than between comparably fine shades of grey
   • BUT when allowed to explore tactual information (e.g. 3D shapes) she performs well
     • similarly, good at writing single letters from dictation
     • Humphrey et al. (1994) for 'natural objects' she uses colour and fine texture to help identify (vs. outline shape)
     • would pick up a knife by the blade if that was the closest/easiest as doesnt recognise what it is
   • deficits largely restricted to form perception
     • impaired processing of low spatial frequency gratings
     • ability to discriminate between between simple shapes is grossly impaired
     • severely impaired on Efron test
   • preserved visuomotor ability
     • Carey et al. (1996): grasp everyday objects with remarkable accuracy, navigates obstacles and can follow a moving light with her eyes
     • Milner et al., (1991): she could not report the orientation of the slot but could 'post' her hand or a card through without difficulty
       • attempts to match the card to the slot without actually moving it towards it were near random
     • Goodale et al. (1991): DF showed excellent visual control of anticipatory hand posture when asked to pick up blocks of different sizes even when she could not discriminate between perceptually
       • again, when asked to make perceptual judgements about object width her responses were wildly innaccurate
     • thought that the ventral stream is severely compromised and so residual motorskills are dependent on dorsal stream
       • visual inputs (orientation,shape,size) arise from V1 or tectopulvinar route, probs both
     • although can grab cross, DF does not rotate the hand to easiest-pick-up position and does not spread finders to engage accurately
     • visuomotor skills aid her object recognition. Relies upon PROPRIOCEPTORY FEEDBACK
       • this visuomotor system is independent of object recognition systems
     • compared to optic ataxic patients
       • DF orients hand and anticipatory opens (O.A's dont)
       • DF cannot match objects (O.A.'s can)
       • DF cannot recognise objects (O.A. can)
       • DF (visual form agnosia) = ventral damage; Optic ataxia = dorsal
   • Milner & Goodale (2006): likely lateral prestriate cortical damage = necrosis to areas V2,3,4. MRI = V1 is largely intact
     • Although interblobs of V1 receive magno + parvo inputs (Nealey & Maunsell, 1994) its assumbed that neither mago nor parvo interblob channel in DF are able to send info to the ventral stream
       • thought that DFs dorsal stream receives info about orientation and size info via tectopulvinar and retinotectal pathways rather than via the striate cortex (e.g. why visuomotor abilities have been left intact)
       • size and orientation info comes via dorsal (ventral = disrupted)
         • how come she can still discriminate colour if ventral stream is disrupted?
           • not enough parvo cells in the dorsal stream to account for this
             • Livingstone & Hubel (1988): model accounts for intact connections from V2 to V4 & IT?
               • theory that parvo cells were 'more resilient' was also tested but didnt amount to much
   • MRI scan = existence of dense lesions in the occipital cortex bilaterally

4. Milner & Goodale (2006): assumptions

   • Ventral stream (perception): objec-centred coding, perceptual constancies, colour & object perception and "off-line" functions
   • Dorsal stream (action): viewer centred coding, transient, many movement-related units and "on-line" functions

5. Category-specific Agnosias

   • Warrington & Shallice (1984): worse at recognising living things vs. non-living
     • Martin & Caramazza (2003): specifically four-legged animals
   • Evidence for "category-based" explanations
     • reports specific kind of semantic object representation structure
     • Patient EW: poor at recognising one type of living thing (animals) but not another (fruit/veg) or non-living things.
       • this deficit occurs in both visual and functional tasks
   • Evidence for "property-based" organisation
     • visual (colour, texture, shape) versus finctional areas of the brain (uses, places found)
     • difficult to reconcile with category specific deficits.... patient EWs deficits are difficult to explain using this model
       • BUT: maybe because animals dont have a specific function? or perhaps its requires global rather than local features processing?

6. Neuroimagine evidence

   • evidence for semantic category dissociation
   • differential cortical representations of faces vs. objects (fusiform face area)
   • differential representation of human body vs. other stimuli
   • differential representation of faces, houses, tools , words, etc (e.g. FFA, parahippocampal place area, visual word form area)
   • Chao et al. (2002) different cortical representation of tools versus animals (cat. spec. pattern); network of regions involved
   • Newer model: sparse vs population codes
     • Haxby et al argue for strong population coding - category viewed can be identified by a pattern of neural responses
       • TMS argues for sparse coding