What is visual perception? - Chapter 6

How does perception work in general?

All projection areas (primary sensory areas) are modality-specific. The primary auditory cortex is located in the Sylvian fissure (top part of the temporal lobe). The primary visual cortex has settled around the calcarine fissure (in the center of the occipital lobe). More complex processing of information takes place in the secondary areas, but only in the tertiary areas a link with other sensory information takes place. The main topic of this chapter is visual perception, partly because most of this research has been done in humans. A distinction can be made between lower-order visual information (such as the perception of color) and higher-order visual information (such as recognizing objects).

What is the physiological basis of object perception?

The visual perception starts with the capture of light information through the retina in the eyes. There are two types of light-sensitive cells in the retina: cones and rods. The cones are important for color perception while the rods (more sensitive to light information) are mainly used for perception when it's dark. There are three types of cones: short wavelengths (blue cones), medium wavelengths (green cones) and long wavelengths (red cones). The information is passed on to two types of ganglion cells: magnocellular cells (M cells) and parvocellular cells (P cells). P cells have a small receptive field and mainly pass on information relating to color perception. M cells pass on motion-related information. The bundled axons of these ganglion cells form the optical nerve (nervus opticus) The nerve pathways cross in the optical chiasma, so information from the left part of the retina of both eyes is combined with information from the right part of the retina of both eyes. The optic nerve transmits information to the NGL (nucleus geniculatus lateralis) of the thalamus where the information from the magnocellular and parvocellular system is processed in various layers. After processing in the NGL, the visual information is projected via optical radiation to the primary visual cortex / striate cortex / V1. 

Where are the 'what' and 'where' routes?

All visual information is mainly processed bottom-up, or sequentially: visual input is always converted in successive processing stages into a higher-order representation. Visual areas are indicated with codes V1 to V5. It is often referred to as an occipito-temporal (ventral) what-route that receives information from the parvocellular system of the NGL and an occipito-parietal (dorsal) where-route that receives input from the magnocellular system of the NGL. The ventral route is concerned with the recognition of objects and the processing of color, shape and texture (via V1 and V2 and V4 to specific parts of the cortex) while the where-route is involved in the visuospatial processing, the location of objects in space, and the guiding of visually controlled movements to these objects (via V1, V2, V3 and V5 to other specific parts of the brain).

Where are the specializations located in the visual cortex?

V3 is specialized in the perception of form; V4 is of great importance in color perception and V5 (also called MT, middle temporal) is mainly involved in the processing of movement. These areas are therefore not only involved in the what-route or where-route, but also in the processing of lower-order visual information.

What does the functional model for visual perception look like?

In order to see the world around us a complete scene, our brains must quickly and efficiently convert the two-dimensional, retinal image into a three-dimensional, internal representation and link this to our semantic knowledge. Almost all models of visual perception assume that there are several successive processing stages:

• First stage: a post-sensory analysis (primary sketch). Basic information is grouped based on similarities in contrast, color, texture, shape, orientation, and direction of movement: this allows a distinction to be made between a foreground and background. This corresponds to the Gestalt principles.

• Second stage: the perspective-dependent representations are converted into perspective-independent representations. This results in object constancy. The stage is called 'perceptual categorisation' by Warrington (1982).

• Third stage: the internal representation of the perceived object (percept) is linked to semantic knowledge. Top-down processes play a role. 

According to the model of Marr (1982), the first and last stage are correct, but the second stage is a 2½D sketch: during this intermediate stage a description is made of the surface of an object and its orientation with regard to ourselves. This enables us to interact with the object perceived. This intermediate link would make the conversion to a three-dimensional percept possible.

What impairments are possible in visual perception?

Problems in visual perception can be roughly divided into three categories:

1. Visual-field defects.

2. Lower-order visual disorders / elementary visual impairments / impairments in primary processing (this visual impairment often includes the term 'anopsia').

3. Higher order disorders are cognitive visual impairments that occur as a result of damage to areas outside the primary visual cortex (these disorders are often also called agnosias). According to Lissauer (1890) there are two types of agnosia, the apperceptive agnosia (the percept cannot be formed) and the associative agnosia (the formed percept cannot be associated with semantic knowledge).

Another way to categorize visual disturbances is based on the 'what' route versus the 'where' route. Three reasons for preferring such a distinction over the above-mentioned classification:

• Brain damage rarely corresponds to the anatomical and physiological boundaries between the different visual regions;

• After a stroke there is often a loss of tissue in one of the two hemispheres, while the other hemisphere remains intact;

• An impairment of primary processing also to some extent affects higher-order information processing in many cases. 

Visual-field defect

The following defects belong to the lower order visual disturbances. In the case of a visual-field defect, the damage to the optic nerve results in a visual acuity impairment in one eye. Damage before the optic chiasm, leads to a visual impairment for one eye. Damage after optic chiasm (such as damage to the NGL or V1) leads to homonymous vision loss. The extent of the field of vision loss depends on the extent of the damage. A hemianopsia is blindness for half the visual field. Quadrantanopsia means that there is a field of vision loss for a quarter of the visual field. Eventually, a scotoma is blindness for a small part of the field of view.

A disorder in visual acuity, contrast sensitivity and light-dark adaptation

Visual acuity is a measure of the smallest details that a person can distinguish. The formula that is used to calculate visual acuity (V) is: V = d / D, where d is the distance between the person being tested and the letter chart, and D is the distance at which a person with normal visual acuity can recognize a letter or symbol. V = 0.3 indicates that someone visually impaired. V ≤ 0.1 corresponds to severe visual impairment, and any value below 0.05 indicates blindness. In addition to visual acuity, the contrast sensitivity also determines the vision. More than three-quarters of patients with posterior brain damage suffer from decreased contrast sensitivity (blurred vision), while only 10 to 15% suffer from deteriorated visual acuity. Eventually, the total vision also includes the speed at which the patient can adapt to different light intensities (the light-dark adaptation).

Impairments in color perception

An impairment in color perception can be the consequence of damage to the eye, the optic nerve or the brain. If it is a matter of the latter (and then specifically damage to V4 in both hemispheres) we speak of a (cerebral) achromatopsia. Patients with achromatopsia only see pale or gray tones. The wavelengths are still observed by the retina but they are no longer processed. Often this disorder is accompanied by visual field defects and problems with face recognition and identification (prosopagnosia).

Impairments in movement perception

A movement detection impairment is also referred to as akinetopsia or movement blindness. It is a very rare disorder. Patient (LM) sustained this impairment following a stroke in the posterior part of both parietal lobes. 

Higher-order visual disorders

There are two types of higher-order visual disorders. An example of an apperceptive agnosia is the visual form agnosia. In this disorder all visual functions (such as sharpness and color perception) are intact, but the patient cannot recognize, match, copy or discriminate simple visual stimuli. Another apperceptive agnosia is the ventral stimulation agnosia: the patient cannot merge the independent detached parts into one percept. For this reason this disorder is also called an integrative agnosia (Riddoch & Humphreys, 1987). In an associative agnosia, three criteria must be met: (1) visual recognition problems (difficulty in naming or categorizing objects), (2) normal recognition in the other non-visual sensory modalities such as touch, and (3) intact lower-order visual perception. Someone with an apperceptive agnosia cannot copy the object while someone with an associative agnosia is able to do this, but has no idea what exactly he / she is drawing. Examples of associative agnosies are:

• Color agnosia: not being able to categorize, name and / or recognize colors. Often the object-color knowledge is impaired: the patient knows that bananas are colored yellow (because this is a common combination), but not that raspberries are red.

• Brightness agnosia: the inability to recognize brightness (for example, someone cannot see if the lights are on or off). Brightness information and color information are processed separately.

• Object agnosia: patients cannot name objects and are even unable to organize objects from the same category on the basis of semantic knowledge (on the basis of a similarity in form). This is the most serious visual associative agnosia. There are big differences between patients with optic aphasia (object anomia). Optic aphasia is limited to the inability to name objects. There is a connection between the percept and the semantic knowledge: the patient is capable of matching identical illustrations of objects and of categorizing objects on the basis of similarity of form, but they cannot do this on the basis of sematic category. The cause of this problem is usually sought in a disconnection between the visual system and the semantic system, commonly as a result of damage to the corpus callosum. Another 'disconnection syndrome' is pure alexia: the language comprehension, language production and writing skills are intact. However, patients have great difficulty reading. The disconnection is located between the visual cortex and the language areas in the left hemisphere, often due to a blocked or damaged corpus callosum so that the written information does not reach the left hemisphere.

What is prosopagnosia?

As with all other forms of agnosia, prosopagnosia is not an absolute defect: it takes a lot of effort and time to recognize faces. Even familiar people are often only recognized by the voice, clothing or the way they walk (but not by their face). In the most serious cases, the patient does not even recognize himself in the mirror. A prosopagnosia is usually the result of a bilateral lesion of the occipital-temporal region, but can also be the result of an unilateral lesion in the right hemisphere. The "fusiform face area" (FFA) is a specific area that is associated with the processing of the unchanging aspects of faces. There is almost never selective damage to this area, so prosopagnosia is often accompanied by problems in object perception or color perception. Patients suffering from apperceptive prosopagnosia cannot recognize a face as being a face; patients suffering from associative prosopagnosia can recognize a face as being a face, but they cannot distinguish a familiar face from an unknown face.

The face recognition model of Bruce and Young (1996)

The face recognition model of Bruce and Young (1986) states that recognizing a familiar face is independent of recognizing an emotional facial expression (double dissociation). According to Bruce and Young's model, we first make an independent internal representation: something that a patient with an apperceptive agnosia is not able to do. When the internal representation has been created, the recognition takes place according to the following three stages: 

A structural description - a ‘face recognition unit’ (FRU) - is made of the familiar face, so that the face can be distinguished from all other faces. There is an FRU for every known face that is activated as soon as this face is detected;

As soon as a FRU is activated, the corresponding "person identity node" (PIN) is also activated. This PIN contains information about the person. A PIN does not necessarily have to be activated by a FRU: the voice of the person or other personal information can also activate the PIN. According to this model, a patient with an associative prosopagnosia is not linked to the PIN.

After activating the PIN, the name of the person can be retrieved.

There is a lot of supporting (neuroimaging) evidence for this model.

What other visual impairments are known?

Blindsight

Sometimes patients with a (partial) loss of the visual field can still perceive the visual stimuli in the ‘blind’ field of vision at an unconscious and more rudimentary level of visual processing. This has been demonstrated with forced-choice paradigms in which the brightness, a basic form or a direction of movement must be indicated. This is possible because the subcortical brain regions involved in visual processing (such as the pulvinar) process and partially transmit such information to the brain regions involved in higher-order visual processing.

Visual hallucinations and illusions

A visual illusion is a distortion of the actual external stimulus; if there is no external stimulus present, this is referred to as a visual hallucination. The Charles Bonnet Syndrome (CBS) is caused by an impaired transmission of visual information to the visual cortex, which causes deafferentation symptoms: the visual cortex takes on a life of its own, as a result of which the patient sees things that are not there. The patient is aware that the images are not real and the images are almost never threatening. Anton's syndrome (anosognosia): Some patients that are cortically blind due to bilateral infarctions in the primary visual areas deny their blindness and confabulate a very detailed visual world. 30-50% of Parkinson's patients suffer from complex visual hallucinations in the last stage due to a disrupted dopaminergic system. They often also have lower-order visual disorders (such as impaired color perception). Patients who have had a stroke and have been damaged in the visual cortex suffer from palinopsia during the period shortly after they have sustained the lesion: they see images that they have just seen. For many patients, palinopsia is temporary.