Visual agnosia

From Psy3242

Patients with visual agnosia have the inability to make sense of certain aspects of incoming visual stimuli. More specifically, visual agnosics often do not grasp the “bigger picture” of what they are looking at. For instance, many visual agnosics can describe the color, texture, or even the shape of an object, yet cannot say exactly what the object is. So while all of the correct visual stimuli is being received, the legion suffered by visual agnosics disallow them to recognize familiar objects and faces. Research on visual agnosia shows us that different parts of the brain compute different aspects of the same information retrieved from the retina.

      Visual agnosia comes about from a legion in the posterior occipital lobe and/or the temporal lobe of the brain, often caused by a stroke.  In Goodale et al (1992), researchers postulated that two streams of vision give rise to this unique disorder.  Basically, they believed that a dorsal stream was responsible for more primal aspects of vision (i.e. aiming movement), and a ventral stream was more responsible for the higher cognitive perceptions of the world.  They dubbed these pathways the action stream and the perceptual stream, respectively.

      In their study, patient DF suffered from irreversible brain damage due to carbon monoxide poisoning in 1988, and MRI scans showed damage to the lateral occipital cortex (a ventral stream structure).  As the researchers expected, DF could not describe objects’ size, shape, or orientation, yet demonstrated appropriate grasping motions when reaching for those same objects.  Apparently, DF had some sort of underlying visual conception of the object, but one that did not reach conscious interpretation.  

Visual agnosics were classically subcategorized as either one of two types:

Apperceptive agnosia:

     Individuals diagnosed with this form were able to identify parts of an object, just as the presence of a handle on a coffee cup, but were unable to identify the cup itself.  In Benson & Greenberg (1969), Mr. S described a safety-pin as silver and shiny but could not recognize the object, nor could he recognize letters, numbers, or faces.

Associative agnosia:

      Individuals diagnosed with this form were able to copy objects, pick out objects similar to other objects, and even categorize them, but could still not identify what the object was.  Researchers suggested that the problem was in linking perceptual information to semantic content.

      The basic observed difference between the two was in the ability to copy drawings.  The neurological basis of this comes from the extent of the damage to the ventral processing stream.  Basically, apperceptive agnosics display damage to an earlier stage of the ventral processing stream, which is why associative agnosics have a better, although still impaired, ability to recognize objects.

Modern View:

      More research on visual agnosia suggests that these classifications are too basic, as many patients have sensory deficits working in tandem with these perceptual deficits (i.e. blind spots called scotomas).  Three patients exemplify the complicated nature of visual agnosia:

JL- who could recognize objects in standard orientations but had trouble recognizing them when presented end-on.

AB- pt. could draw/match/ID unusual objects, but profoundly impaired at object/picture naming and describing functionality. Implies semantic memory deficits.

HJA- pt. could define a carrot when asked verbally, but could not ID a picture of one. Could name objects by touch that he could not ID verbally. Implies intact memory but problems visually activating it.

      Although many modern theories have surfaced to help explain the disorder, Ellis and Young’s ideas merit special attention as they attempted to integrate all other models.  They proposed a set of sequential stages in visual perception that lead to correct object recognition:

2D- ‘primal sketch’ includes information about boundaries, contours and brightness fluctuations, but not overall form. 2.5D- viewer-centered information about form and contour but not object constancy or perceptual classification. 3D- true object (rather than viewer) centered mental representation, independent of viewer’s position, specifying the real 3D shape of an object from any view, enabling true object recognition.