User:T.calder/sandbox

The fusiform face area (FFA) is a brain area located in the fusiform gyrus and involved in the visual processing of human faces. There is also evidence that the FFA is involved in identifying objects from other categories, especially those for which an individual is highly knowledgeable.

Localization


The proposed fusiform face area is located in the medial area of the fusiform gyrus, in the ventral pathway of the temporal lobe. While the FFA is found in both hemispheres of the brain, it is larger in the right hemisphere. A homologous brain area also involved in face perception has been found in the superior temporal sulcus of macaque monkeys.

Much research on the FFA has been conducted using imaging studies, often involving functional Magnetic Resonance Imaging (fMRI). Prior to the development of fMRI, Positron Emission Tomography (PET) was commonly used. A common research paradigm involves monitoring FFA activity in response to visual stimuli. Additionally, a significant amount of research has been done on patients with severe deficits in their ability to recognize faces (a condition called called prosopagnosia), which is often associated with damage to the FFA.

Function
The functional role of the FFA has been contested, and there are currently two competing hypothesis regarding its function. The face-specificity hypothesis proposes that the FFA is a specialized organ for the visual recognition of human faces. This theory suggests that the FFA is a domain specific area, meaning that it is only involved in processing faces and not other categories of objects. The expertise hypothesis proposes that the FFA is involved in visual recognition of any categories for which the viewer has an 'expert' knowledge of. In the case of the expertise hypothesis, face recognition is believed to be just one of many domains for which the FFA is sensitive.

Face-Specificity Hypothesis
A large body of research supports the hypothesis that the FFA is involved in recognizing and identifying faces. In 1992, a PET study conducted by Justine Sergent found that certain brain areas, including the medial fusiform gyrus of the temporal lobe, showed increased activation in response to faces but not to other objects. In 1997, Nancy Kanwisher demonstrated that parts of the medial fusiform gyrus activated in response to intact faces but not to scrambled faces, houses, or human hands. She called this area the fusiform face area, and proposed that it is a domain specific cognitive module for the processing of human faces.

Since the late 1990s, several studies have attempted to determine exactly what categories the FFA is involved in visually processing. In 1999 an fMRI study found that the FFA is activated more when viewing human faces than when viewing whole human bodies, animals, or inanimate objects. Animals activated the FFA more than non-living objects, especially when animals were presented with visible faces. The FFA also does not activate in response to non-animate living objects, such as flowers. These two studies show that the FFA is not broadly sensitive to living things, but rather specifically responds to faces.

The FFA is sensitive to the common geometric configuration of human facial features, and not individual features themselves. A study by Hadjikhani, Kveraga, Naik, and Ahlfors (2009) found that non-face objects that look like human faces do activate the FFA. The FFA is activated in response to upright but not inverted Mooney faces, demonstrating that it is not activated by individual facial features but the global pattern of a face.

Case Studies
Several case studies have confirmed the importance of the FFA for overt face processing. Patient F.E., who suffers from severe prosopagnosia as a result of bilateral damage to both his fusiform gyri, was found to lack overt face recognition, although he was able to covertly recognitize faces. This is consistent with the finding that visual attention modulates activity in the FFA. The posterior cingulate and orbistrofrontal gyri have been proposed as an extended face system, allowing for some degree of face recognition even when the FFA is non-functioning.

A second, constasting case study involved a patient known as C.K. Patient C.K. suffers from severe agnosia, the inability to visually identify objects, but has an intact ability to recognize faces. However, patient C.K. has demonstrated poor performance on face recognition tasks that do not normally activate the FFA, such as viewing inverted faces. These tasks activate areas involved in recognizing non-face objects.

The contrasting patterns of performance between patient F.E. (who can recognize objects but not faces) and C.K. (who can recognize faces but struggles with other objects) indicates a double dissociation between face and object recognition. That is, that the the inability to perceive faces among prosopagnosia patients is due to specific damage to their face-recognition module, and not a general cognitive impairment.

Role of Attention
Activity in the FFA is stronger when viewed faces are the focus of attention. Wojciulik, Ewa, Nancy Kanwisher, and Driver (1998) found that FFA activity decreased when viewing faces were presented outside a subject's focus of attention, but there was no difference when viewing houses. They proposed that despite the sensitivity of the FFA for faces, recognizing faces still requires overt attention. A separate study found that activation of the FFA in response to faces occurred even when faces were presented too briefly for the participate to be consciously aware of them, suggesting that processing in the FFA is automatic. The results of these two studies are not necessarily contradictory, as it is possible that face processing is largely automatic, but only when the face is attended to.

Inversion Effect
One commonly found effect is that face processing differs whether faces are presented upright or inverted. Inverted faces are less likely to be recognized as faces. This is commonly called the inversion effect. When viewing inverted faces, visual processing is shifted from holistic processing (viewing the face as a combined whole) to part-based processing (viewing the face in terms of individual features), which impairs performance on face identification tasks.

Studies have found that there is decreased FFA activity in response to inverted faces, when compared to upright faces. This has been taken as evidence that the FFA is sensitive to the global pattern of human facial features, and not to individual features themselves. This is consistent with the finding that the FFA activates in response to non-face stimuli that have a similar configuration to human faces.

Development
Experience in early infancy plays a role in both the structure and function of the FFA. The sensitivity and size of the FFA has been found to increase with age, being more sensitive in adults than children. However, there is no change in size or sensitivity for the fusiform body area (a similar area involved in recognizing human body shapes), suggesting the the FFA's specialty develops over early childhood and has a more delayed developmental trajectory than the FBA.

The developmental trajectory of the FFA is comparable to the developmental trajectory of facial recognition abilities themselves. One study found that newborn infants show no preference for faces of differing races, while 3-month old infants prefer to look at the faces that match their own race. This study indicates that while some face sensitivity is present at birth, it is refined over time in response to environmental experience, a pattern of development that mimics that of the FFA.

Emotional Processing
In addition to identifying faces, the FFA is involved in processing the emotional content of faces. The FFA follows a similar pattern of activation to the amygdala (a part of the brain sensitive to fearful stimuli) in response to negative facial expressions. Activity in both the amygdala and the FFA follows an inverted U-shaped curve, initially increasing and then decreasing (or habituating) over time. In addition to emotional recognition, the FFA is involved in emotional learning involving faces. One study found increased activation in both the FFA and the amygdala when participants viewed faces which had been previously associated with negative emotions through classical conditioning. Despite being involved in processing facial expressions, emotional perception is preserved when the FFA is severely damaged, indicating that other areas are also invovled.

Expertise Hypothesis
A second hypothesis is that the FFA is not specialized for facial recognition, but rather for 'expert' recognition of visual objects. That is, the FFA is visual recognition module for that responds to object categories for which an individual has a large deal of knowledge, and is involved in making distinctions based on very fine details. This hypothesis does not suggest that the FFA is unimportant for face recognition, but rather that faces are only one of the expert domains that the FFA is involved in recognizing.

There is evidence that the FFA is involved in processing non-face object categories. A 2006 study using high-resolution fMRI found that there exist several clusters within the FFA which are selective for object categories other than faces. These categories were sculptures, cars, and animals. While face-selective areas were more common, these areas were not any more individually sensitive to faces than other object-related areas were to their specific category. It has been suggested that previous imaging studies showing that the FFA activited only in response to faces could have been hindered by technological limitations, including poor imaging resolution.

There is evidence that the FFA is sensitive to categories for which the viewer is knowledge. Increased activation of the FFA has been found in high-level chess players when they viewed common chess-board configurations, but not for novice players, supporting the expertise hypothesis. Gauthier et al (1999) found that when participants were trained to recognize novel objects called “greebles”, FFA activation in response to greebles increased. This effect was found only for upright greebles, mimicing the finding that the FFA only responds to faces presented upright and is involved in holistic processing. This is also consistent with the finding that face recognition is partially shaped by experience, which suggests that the FFA has some degree of plasticity which may extend to non-face categories.

However, an alternative explanation for the greebles effect was proposed by Brants, Wagemans, and Op de Beeck (2011), who found that FFA activation in response to greebles was not determined by training, but instead by how closely the greeble resembled a human face. Additionally, they found an inversion effect occured for greebles before any training had occurred, again casting doubt on the hypothesis that the FFA responds to highly-knowledgeable object categories.

Due to conflicting and incomplete evidence, the face-specificity versus expertise debate has yet to be resolved.

Involvement in Mental Disorders
The FFA has been implicated in several mental disorders which are characterized either by difficulty with overtly recognizing faces, or with social deficits partially caused by difficulty with face perception. The FFA has been specifically implicated in prosopagnosia and autism spectrum disorder.

Prosopagnosia
The disorder most prominently associated with the FFA is prosopagnosia, the specific inability to consciously recognize faces. While individuals with prosopagnosia are able to visually identify a human face as being a face, they are largely unable to differentiate among faces or associate faces with individuals. Prosopagnosia can be present from birth (developmental prosopagnosia) or can occur as a result of brain damage (acquired prosopagnosia). Not only have deficits in the FFA been implicated as a cause of prosopagnosia, but prosopagnosia patients have been used to test the facial-specificity hypothesis.

For many patients, acquired prosopagnosia has been associated with lesions to the right FFA, but not with lesions to more anterior temporal areas of the fusiform gyrus. A study by Busigny and Rossion (2010) found that acquired prosopagnosia patients are impaired in their ability to discriminate between faces, but not cars. These findings demonstrate that prosapagnosia is a specific deficit in facial recognition ability, and not a more general deficit in the ability to compare visually similar objects based on their fine details.

Studies involving prosopagnosia patients support the hypothesis that the FFA is needed for overt, but not covert, facial recognition. A study by Barton, Press, Keenan, and O'Connor (2002) involved briefly (500ms) presenting images of faces of famous actors and politicians to individuals with prosopagnosia. Participants were then asked to identify if the face belonged to an actor or a politician. It was found that the brief face presentation had a priming effect on prosopagnosia patients, even when patients denied being consciously aware of the identity of the face. Clearly, some capacity to recognize faces is present in the absence of a functioning FFA.

Furl et al (2011) measured activity in the fusiform gyrus in response to a task that required participants to differentiate between faces as well as cars. Performance for individuals with development prosopagnosia was impaired for face differentiation, but not car differentiation, and FFA activity showed a positive, linear correlation with facial recognition ability. Taken together, studies on prosopagnosia largely support the hypothesis that the FFA is required for face perception.

Autism Spectrum Disorder
Autism spectrum disorder (ASD) is a developmental disorder marked by severe social difficulties. Some of these deficits are related to face processing (e.g. impaired understanding of facial expressions and abnormal eye-gaze). Klin et al (1999) studied the ability of autistic children to correctly match faces when the same was was presented a second time from either a different angle or with a different facial expression. Children with autism showed significantly poorer performance on both tasks, suggesting these social deficits are related to difficulties processing faces.

Poor performance on face recognition tasks by autistic patients has been directly linked to decreased activity in the FFA. Pierce et al. (2001) found that autistic individuals showed little FFA activation in response to faces. Instead, activation was found for the frontal cortex and primary visual cortex, suggesting that autistic individuals process faces outside of the FFA. A similar pattern of decreased activation was found for the amygdala, which is consistent with the finding that both the amygdala and FFA respond to the emotional content of faces.

One study found that inducing autistic individuals to make eye-contact with fearful faces normalizes activity in the FFA. This suggests that impaired emotional recognition partially results from a failure to focus attention on the eyes. However, a study of adult autistic patients found that facial affect recognition training regimes designed to increase facial emotional recognition ability did not increase activity in the FFA.