Lecture 10

Higher level vision

Topics

  • Functional dissociation and single and double-dissociations
  • Apperceptive and Associative Agnosias
  • Prosopagnosia
  • Topographic and category-specific agnosias
  • How functional MRI and PET have been used to identify category-specific processing
  • Synesthesia

Preparation for class

  • Reading: PN6 Chapter 26: 596-599
  • Lecture 03.02 on methods in human neuroscience

Topic Slide

Oliver Sacks (1933-2015) was a British born and educated neurologist who spent most of his career in the United States where he wrote a series of popular books and New Yorker articles on unusual neurological syndromes. One such book, "The Man Who Mistook his Wife for a Hat" describes visual agnosias and prosopagnosia, the subject of today's lecture.

Elizabeth Warrington (b. 1931) is a British neuropsychologist who had a celebrated career at the National Hospital in London studying patients with category-specific visual agnosias and memory deficits.

Dissociations

Single and double-dissociations in cognitive neuropsychology

A single dissociation demonstrates that damage to brain area 'X' impairs function 'A' but not function 'B'. Although this might seem like good evidence that brain area 'X' is the brain loci for function 'A' and not function 'B', this form of evidence is weak. It is possible that functions 'A' and 'B' differ in difficulty. Function 'A' may be more difficult and thus easily disrupted by brain damage, while function 'B' is much easier and less easily disrupted.

A much stronger interpretation would be obtained from a double dissociation whereby damage to brain area 'X' impairs function 'A' but not function 'B', and damage to a different brain area 'Y' impairs function 'B' but not function 'A'. In this case, you have ruled out task difficulty or other general task-related differences as confounding factors.

In the case of brain lesions, double dissociations would require data from different subjects with different damage brain area 'X' and 'Y' and different deficits in function 'A' and 'B'. This might be demonstrable in an individual subject using transcranial brain stimulation (TMS) and/or functional neuroimaging (such as fMRI or PET). However, imaging data is correlative, and so strong inferences about causation can not usually be made without acquiring additional data from TMS or lesions.

Cognitive neuropsychology

The field of cognitive neuropsychology seeks to discover elemental psychological functions of human perception and cognition through patterns of functional dissociations in patients with brain damage. If one is interested only in psychological phenomena, then it doesn't matter what parts of the brain are damaged – only that functions can be dissociated.

The double dissociation is a powerful tool for carving the mind at its joints. However, this approach has its critics. For those interested in learning more about the strengths and weaknesses of this approach, the journal Cortex devoted a special issue to this topic.

Dissociation of activation in Dorsal and Ventral pathways using neuroimaging.

The logic of the double dissociation can be applied in neuroimaging studies. I provided an example from a 1994 PET study performed by Jim Haxby and colleagues. The full text of their article can be found here.

In the Haxby study, the identical stimulus array preferentially preferentially activated the dorsal or ventral pathway depending upon the task. When asked to judge the relative spatial location of items (faces) on the screen, the dorsal pathway was preferentially activated. When asked to judge the form of the the items, the ventral pathway was preferentially activated. Thus, two tasks that used the same stimuli and in the same subjects showed dissociated patterns of activation.

The double dissociation logic has been used in many neuroimaging studies to great effect. However, one should be aware that neuroimaging studies are correlative in nature. It is not known on the basis of neuroimaging activation studies alone whether the brain area active is necessary for the performance of the task in which it activates. This is a deficiency in all correlative methods, and thus evidence from converging modalities are needed to make strong causal inferences.

Apperceptive Agnosia

Agnosia comes from the Greek: 'a' meaning 'without' and 'gnosis' meaning 'knowledge'. It is used in brain research to describe deficits in perception and knowledge associated with brain damage in patients.

Many have used Lissauer’'s original distinction between apperceptive and associative agnosias. This taxonomy embodies an implicit hierarchy – that there is an initial perceptual analysis of objects before the objects are associated with other information (such as the names and uses of the objects).

This taxonomy ignores the role feedback might play between levels of analysis. Does perception exist in the complete absence of association? (example of word superiority in alexia where words help).

We will follow Martha Farah’s taxonomy from her book Visual Agnosias, 2nd edition, MIT press.

I offer this caution in evaluating the literature on agnosia: there are very few 'pure' cases and patients differ in the degree and quality of their deficits. There are individual cases that are grouped together under the same label where you may be more impressed by the differences than by the similarities.

As we will see, the study of agnosia moves very quickly from visual perception into attention, memory, and language. We will re-encounter some of the deficits discussed below when we discuss these other topics.

Deficits in apperceptive agnosia

Apperceptive agnosia is a visual form agnosia that reflects an underlying perceptual disturbance.

  • Perception is disrupted by low-level visual noise. I showed an example where a patient read the number '7415' instead of the word 'THIS', when the latter was interrupted by gaps.
  • Perception is disrupted when looking at ordinary objects presented in unusual views (stapler example) or in shadows (hats example).
  • Individuals with apperceptive agnosia have difficulty copying simple shapes and matching simple shapes.

Some apperceptive agnosics can copy simple drawings, but demonstrate problems in perceptual integration.

Simultanagnosia

Farah includes simultanagnosia as a type of apperceptive agnosia, and distinguishes between two types: dorsal simultanagnosia and ventral simultanagnosia.

Dorsal simultanagnosia

Patients with dorsal simultanagnosia can only attend to one percept at a time. When tested with individual objects, they can name or point to them. However, when two objects are presented simultaneously, they can only report one object while the other appears to fade.

When asked to draw an object, patients with dorsal simultanagnosia often make 'exploded' drawings, where parts of the objects are individual drawn and the parts are not integrated into a whole.

  • I provided an example of an 'exploded' drawing of a bicycle.

Patients with dorsal simultanagnosia often cannot localize objects in space. Perhaps because space is defined by relative location of objects. If you can only see one object, then it can't be localized.

Dorsal simultanagnosia is frequently discussed in the context of attention, and we will re-encounter it later in this semester when we discuss attention.

Ventral simultanagnosia

Ventral simultanagnosia is similar to the dorsal form, but the emphasis is upon reading and letter processing. Individuals with ventral simultanagnosia often read one letter at a time because they cannot 'see' the other letters.

We will discuss ventral simultanagnosia again in my lectures on language and a specific deficit in reading called alexia.

Associative agnosia

Associative agnosia reflects a perceptual categorization deficit. Individuals with associative agnosia have good perception and can make accurate drawings of objects, but cannot recognize the objects that they just drew. They appear to have a disconnection between perception and meaning. In most cases, other cues (auditory or sensory) can be used to identify the object.

A very interesting form of associative agnosia is category-specific agnosia. Here, the individual has lost knowledge of particular domains of knowledge. Some patients lose information about 'living' things, but not 'non-living' things; while other patients lose information about 'non-living' things but not 'living' things. This represents a double-dissociation about semantic knowledge. In extreme form, such as in some dementias, the individual appears to lose the entire representation of the object ("I don't know what a car is").

Prosopagnosia

A more specific types of agnosia is prosopagnosia, where an individual displays deficits in face perception and/or face recognition (who is it).

I showed some cortical stimulation data from my lab suggesting that the perception of a face can be interrupted with stimulation to the ventral visual pathway. This provides some weak evidence for an apperceptive prosopagnosia. However, in most cases, prosopagnosia refers to the inability to recognize who the person is. The individual sees that a face is a face, but cannot recognize familiar faces (such as their spouse, or even themselves in a mirror).

Deficits and abilities in prosopagnosia

I discussed several aspects of face processing in prosopagnosia:

  • The importance of configuration in faces, illustrated by the face inversion and the Thatcher illusion.
  • The prosopagnosic farmer who could, nonetheless, differentiate among individual sheep. This illustrates that prosopagnosia is not a general disability of individuation.
  • That prosopagnosics sometimes have 'partial knowledge' of an individual from face information, or 'guilty knowledge' as demonstrated by skin conductance changes.

Double-dissociation

Prosopagnosia in an individual patient is a single dissociation. Perhaps face processing is a very demanding skill, and that brain damage affects it more than other non-face processing.

Patient CK provides the second half of a double dissociation between face and object processing. CK can recognize faces but cannot recognize objects.

Other dissociations are found in the realm of memory. Some individuals with dense amnesias nevertheless can recognize faces. Other individuals cannot remember any new faces after brain damage, but show good recognition of faces learned before brain damage.

Face delusions

There are two interesting delusions related to face processing that give further insight into the functional components of face processing.

  • In the Capgras delusion, an individual believes that individuals that he or she can recognize have been replaced by imposters. I read a section from Ramachandran's book to illustrate this delusion. It has been speculated that Capgras delusion reflects a dissociation between familiarity and recognition. In other words, the recognized face does not activate the emotional knowledge associated with the individual. The face, thus, does not evoke the 'feeling' of the recognized person.
  • The Fregoli delusion is in some ways the flip side of the Capgras delusion, in that individuals report a strong sense of familiarity in the faces of strangers. Individuals suffering from this delusion report that they keep seeing the same individual. This can be associated with paranoia. The Fregoli delusion was named after a popular ‘quick-change artist'.

Physiological studies of face processing.

There has been a wealth of information on the localization of face processing in the brain.

  • Face cells in monkeys were first discovered by Charles Gross in the 1960s.
  • Direct recording of ERPs from the cortical surface have revealed focal potentials associated with face processing in the fusiform gyrus. Stimulation of those same fusiform sites result in transient prosopagnosia.
  • PET and functional MRI studies have revealed focal areas of the fusiform gyrus (and other brain regions) that are activated by face processing.
    • FMRI has shown that a network of brain areas are activated by faces that include sites outside of the fusiform gyrus. It may be that different aspects of face processing are represented in those areas. For example, the processing of face movements that underlie expression analysis may be localized in the posterior superior temporal sulcus, part of the dorsal visual pathway.

Topographic agnosia

Topographic agnosia refers to an individual's inability to recognize familiar places. Individuals suffering from topographic agnosia rely on street names and house numbers to find their way home.

Activation studies

Functional MRI (and some PET) studies have identified regions of the ventral brain associated with processing specific categories of stimuli.

  • Faces – 'fusiform face area'.
  • Objects – 'lateral object cortex', and 'mid-fusiform' object area.
  • Places – 'parahippocampal place area'.
  • Word form – 'word form area'.

Synesthesia

We have thus far focused upon 'dissociating' regions of the brain associated with different percepts. In the case of synesthesia, percepts that are normally separate (such as letters and colors) are combined. There are many types of synesthesia, however, the most common is Grapheme-Color, and Music-color.

In grapheme-color, different numbers and letters are 'seen' as having distinguishable colors. For example, the letter 'A' may be 'red' while the letter 'B' is 'green'.

I presented several tasks used to demonstrate synesthesia.

  • Target number 'pop-out'.
  • Part-whole phenomena.

I speculated on what might cause synesthesia. In the case of letters and colors, we know that ventral brain regions that are involved in letter and color processing are adjacent. This suggests a 'cross-wiring', with a sensory experience that is reminiscent of the phantom limb. However, some forms of synesthesia represent processes not known to be localized in adjacent brain regions (e.g., music-color).

Videos

Prerecorded lectures for 2020

Live lectures from prior semesters

The embedded video below is a recording of Lecture 12 given in Fall 2019.

Print Friendly, PDF & Email