Prosopagnosia: neuro-functional basis of face recognition impairment

Get a pdf of this research program (updated July 2015)

Following brain damage (trauma, stroke ...), some people may present with great difficulties in recognizing familiar faces and encode new faces in memory. This visual recognition impairment does not appear to be due to sensory visual defects or intellectual disorders, and recognition of other people through other modalities, the voice for instance, is preserved. This condition was termed prosopagnosia by Bodamer (1947) and first described by Wigan in 1844, and then Quaglino & Borelli in 1867 (paper translated by Ellis & Florence, 1990).

Cases of acquired prosopagnosia with preserved visual functions are extremely rare, and most cases have also large deficits in object recognition. In fact, the question of whether prosopagnosia can be 'pure' (without object recognition impairments) is still a matter of debate for some authors, even though there are a few cases described in the literature who seem to have normal object recognition (as reviewed by a recent paper in our laboratory, Busigny et al., 2010).

Studying cases of prosopagnosia can be particularly interesting, for two reasons:

1. They can help us understanding better the location and the critical role(s) of the brain areas involved in normal face recognition (see e.g., Rossion et al., 2003; Schiltz et al., 2006; Dricot et al., 2008).

2. The kind of visual cues that they still can or cannot extract and remember on faces may help us understanding better how normal people recognize faces (i.e., what kind of information/processes are particularly important) (see e.g., Caldara et al., 2005; Van Belle et al., 2010).

Because the patients are extremely rare and all have different associated impairments, we advocate the detailed investigations of single-cases of prosopagnosia rather than group studies of prosopagnosic patients.

For instance, we have been studying a fascinating case of prosopagnosia, PS, who sustained brain damage in 1992 at the age of 41. She has recovered all neuropsychological functions but has a massive prosopagnosia.

For a review on prosopagnosia with an emphasis on PS' case, see Mayer & Rossion, 2007, or in French: Rossion (2008): La reconnaissance des visages (Cerveau & Psycho, N°25). 

We have been testing PS over 15 years (first tests early 2000), and together with many colleagues, we have published nearly 20 papers on her case, as summarized below. We have also tested other cases of prosopagnosia, or visual agnosia (NS, Delvenne et al., 2004). In a nutshell, this work has allowed us to draw a number of following conclusions (described in the research program and in the review paper of Rossion, 2014).

* Brain-damaged patients with visual recognition impairments limited ot the category of faces do exist ("pure prosopagnosia"): PS (Rossion et al., 2003; Busigny et al., 2010) and other cases such as GG and LR (Busigny et al., 2010; Busigny et al., 2014). It cannot be accounted for a problem at using fine-grained details to discriminate  visual shapes.

* Such patients tend to rely less on the region of the eyes of faces and more on the mouth (Caldara et al., 2005), and their patterns of eye gaze fixations agree with that (Orban de Xivry et al., 2008).

* They are impaired at perceiving individual faces holistically: absence (or reduction) of inversion effect (Busigny & Rossion, 2010), composite and whole-part effects (Ramon et al., 2010; see also Ramon & Rossion, 2010). Our work with gaze-contingency demonstrates more strikingly this lack of holistic face perception (Van Belle et al., 2010; see this research program here: https://face-categorization-lab.webnode.com/research/eye-movements-gaze-contingency-and-face-perception/)

* They can still perceive a face as a face using a holistic representation: they see Mooney faces, faces in Arcimboldo paintings for instance (Rossion et al., 2011).

In summary, at the functional level, the problem of such patients is not that they can't perceive holistically or that the can't use fine-grained details to individualize items, as previously thought: it is the combination of the two to achieve a holistic representation at a sufficient level of detail to individualize the face that is problematic (Rossion, 2014).

At the neural level now:

* Such patients can have lesions at various localizations of the (right) ventral occipito-temporal cortex, indicating that many regions are critical for face perception. The system is widely distributed, but fragile: a small lesion, or a transient interruption through electrical stimulation (Jonas et al., 2012; Jonas et al., 2015), can cause prosopagnosia.

* The right hemisphere is critical, not the left hemisphere, except in some left handed patients (Rossion, 2014)

* FMRI studies of PS have shown that her main lesion in the right hemisphere destroyed part of the inferior occipital cortex, where preferential processing for faces is usually observed in neuroimaging studies of normal subjects ('occipital face area', 'OFA').  However, she has a preserved right fusiform gyrus and preferential 'FFA' activation for faces (Rossion et al., 2003;  Rossion et al., 2011 ), suggesting that in the normal brain the 'FFA' is not dependent on inputs from the posteriorly located 'OFA' as would be the case in a stricht hierarchical  model for face processing.

* Nevertheless,  PS 's 'FFA' does not discriminate between individual faces in line with the behavior of the patient (Schiltz et al., 2006). These findings have been replicated with the visual agnosic patient DF (Steeves et al., 2006; Steeves et al., 2009).

Related Papers/Articles

Jonas, J., Rossion, B., Brissart, H., Frismand, S., Jacques, C., Hossu, G., Colnat-Coulbois, S., Vespignani, Vignal, Maillard, L. (in press). Beyond the core face-processing network: intracerebral stimulation of a face-slective area in the right anterior fusiform gyrus elicits transient prosopagnosia. Cortex, in press. [PDF] [ VIDEOS ]

Van Belle, G., Lefevre, P., & Rossion, B. (2015). Face inversion and acquired prosopagnosia reduce the size of the perceptual field of view. Cognition, 136, 403-408. [PDF]

Jonas, J., Rossion, B., Krieg, J., Koessler, L., Colnat-Coulbois, S., Maillard, L., Frismand, S., Colnat-Coulbois, S., Vignal, J.-P., Vespignani, H., Jacques, C., Brissart, H., Maillard, L. (2014). Intracerebral electrical stimulation of a face-selective area in the right inferior occipital cortex impairs individual face discrimination. NeuroImage, 99, 487-497. [PDF] [videos]

REVIEW Rossion, B. (2014). Understanding face perception by means of prosopagnosia and neuroimaging. Frontiers in Bioscience (Elite Ed.); 6-308-317. [PDF]

Busigny, T., Van Belle, G., Jemel, B., Hosein, A., Joubert, S., Rossion, B. (2014). Face-specific impairment in holistic perception following focal lesion of the right anterior temporal lobe. Neuropsychologia, 56, 312-333. [PDF]

Jonas, J., Descoins, M., Koessler, L., Colnat-Coulbois, S., Sauvee, M., Guye, M., Vignal, J-P., Vespignani, H., Rossion, B., Maillard, L. (2012). Focal electrical intracerebral stimulation of a face-sensitive area causes transient prosopagnosia. Neuroscience, 222, 281-288. [PDF] [video 1] [video 2]

Quadflieg, S., Todorov, A., Laguesse, R., Rossion, B. (2012). Normal face-based judgments of social characteristics despite impaired holistic face processing. Visual Cognition, 20, 865-882. [PDF]

Prieto, E.A., Caharel, S., Henson, R., Rossion, B. (2011). Early (N170/M170) face-sensitivity despite right lateral occipital brain damage in acquired prosopagnosia. Front. Hum. Neurosci. 5:138. doi: 10.3389/fnhum.2011.00138 [PDF]

• The first report of a face-sensitive N170 effect over the right hemisphere of the prosopagnosic patient PS, despite extensive damage in this hemisphere and no evidence of right OFA activation in numerous previous studies. The N170 also shows the amplitude increase and delay to inverted faces. A M170 is also disclosed in a MEG study. Interestingly, the component is absent in the left hemisphere, where she has another lesion in the middle fusiform gyrus (no left FFA). This observation suggests that the OFA is not necessary to observe early sensitivity to faces in the right occipito-temporal cortex.

• In this paper, we show that the pure case of prosopagnosia GG, who has unilateral right hemispheric damage (lingual, parahippocampal and medial part of the fusiform gyrus) also presents with a relatively larger impairment in recognizing faces when preventing him from seeing the central feature of the face (contingent mask) than when restricting his perception to one feature at a time (contingent-window). This is the same pattern of performance as patient PS (Van Belle et al., 2010, paper below), despite almost no overlap between their brain damage.

Rossion, B., Dricot, L., Goebel, R., Busigny, T. (2011). Holistic face categorization in higher-level cortical visual areas of the normal and prosopagnosic brain: towards a non-hierarchical view of face perception. Frontiers in Human Neuroscience, 4:225. doi: 10.3389/fnhum.2010.00225. [PDF]

• Patient PS, pure prosopagnosia. This paper reports (1) a full behavioral study in which we show that she categorizes Mooney and Arcimboldo as faces just like normal observers; (2) an fMRI study in normal participants in which it is shown that contrary to full face photographs, such face stimuli activate the FFA without leading to face-sensitive responses in the inferior occipital cortex (no OFA). (3) This is completed by the same neuroimaging study of PS, showing that she recruits the middle fusiform gyrus (FFA) to perceive these stimuli as faces, just like normal observers. Overall, these observations support the view that face-related activation may emerge in the middle fusiform gyrus independently of any face-sensitive activation in the posetriorly located inferior occipital cortex, especially if the whole configuration of the stimulus is critical for the stimulus to be seen as a face.

Jiang, F., Blanz, V., Rossion, B. (2011). Holistic processing of shape cues in face identification: evidence from face inversion, composite faces and acquired prosopagnosia. Visual Cognition, 19, 1003-1034.[PDF]

Busigny, T. & Rossion, B. (2011). Holistic processing impairment can be restricted to faces in acquired prosopagnosia: Evidence from the global/local Navon effect. Journal of Neuropsychology, 5, 1-14. [PDF]

• Patient PS, pure prosopagnosia. Demonstration of a normal Navon effect with hierarchical letters, while this case of prosopagnosia has been shown to be impaired at holistic processing of individual faces. This study shows that impairment in holistic face processing is not just an impairment at general holistic processing.

Busigny, T., Joubert, S., Felician, O., Ceccaldi, M., Rossion, B. (2010). Holistic perception of the individual face is specific and necessary: evidence from an extensive case study of acquired prosopagnosia. Neuropsychologia, 48, 4057-4092.  [PDF]

• First report of the case of prosopagnosia GG (right unilateral damage, stroke), showing that his impairment is restricted to faces and concerns the ability to individualize faces based on a holistic representation. Extensive investigation (24 behavioral experiments).

Van Belle, G., de Graef, P., Verfaillie, K., Busigny, T., Rossion, B. (2010). Whole not hole: expert face recognition requires holistic perception. Neuropsychologia, 48, 2609-2620. [PDF]

• The first study to use gaze-contingency in a face perception task, performed here in normal observers and prosopagnosic patient PS. The results show a striking dissociation between a condition in which only one fixated feature is revealed at a time (window: relatively less impairment for the prosopagnosic patient than normal observers) and a condition is which the fixated feature is masked (mask: much larger impairment for the prosopagnosic patient).

Busigny, T., Graf, M., Mayer, E., Rossion, B. (2010). Acquired prosopagnosia as a face-specific disorder: Ruling out the general visual similarity account . Neuropsychologia, 48, 2051-2067. [PDF]

• Patient PS, pure prosopagnosia. Demonstration that her ability to discriminate highly similar visual items is as good as normal observers, with simple shapes and common objects. The still popular view that prosopagnosia can be accounted for by the fact that faces form a particularly visually homogenous category does not hold.

Busigny, T. & Rossion, B. (2010). Acquired prosopagnosia abolishes the face inversion effect. Cortex, 46, 965-981. [PDF]

• Patient PS, tested with 5 tasks showing a lack of inversion effect: performance is not better or faster for upright than inverted face in the Benton test, delayed and simultaneous matching tasks, and familiarity decisions. No advantage for inverted faces is found for the patient though. A review of the literature shows that cases of inversion superiority in prosopagnosia are very rare and can probably be accounted for by low-level visual defects.

Ramon, M., & Rossion, B. (2010). Impaired processing of relative distances between features and of the eye region in acquired prosopagnosia—two sides of the same holistic coin? Cortex, 46, 374-389. [PDF]

• Patient PS, tested in an experiment in which various aspects of individual faces were diagnostic for individualization (mouth, eyes, relative distances between features, …). The patient did not show the same profile of performance as normal observers in normal circumstances, being relatively more impaired at processing the eyes and relive distances. However, when the advantage provided by holistic processing was reduced by indicating the nature and location of the diagnostic cues, the patient’s profile of performance looked like normal observers. This study suggest that holistic processing is the cause of the increased difficulties at extracting certain cues (eyes, relative distances) on faces.

Ramon, M., Busigny, T., Rossion, B. (2010). Impaired holistic processing of unfamiliar individual faces in acquired prosopagnosia. Neuropsychologia, 48, 933-944. [PDF]

• Patient PS, tested in 5 experiments showing a lack of whole-part and composite face effects when individualizing faces. The study also shows that there is – unfortunately - a large degree of interindividual variability in the magnitude (and presence) of such effects in a single experiment so that it is difficult to make strong conclusions from the absence of the effect in prosopagnosia based on a single experiment. Yet, the last experiment of the paper relies on a more sensitive composite paradigm and shows a clear absence of holistic processing for the patient PS.

Rossion, B., Kaiser, M.D., Bub, D., Tanaka, J.W. (2009). Is the loss of diagnosticity of the eye region a common aspect of acquired prosopagnosia?Journal of Neuropsychology, 3, 69-78. [PDF]

Steeves, J., Dricot, L., Goltz, H., Sorger, B., Peters, J., Milner, D., Goodale, M.-A., Goebel, R., Rossion, B. (2009). Abnormal face identity coding in the middle fusiform gyrus of two brain-damaged prosopagnosic patients. Neuropsychologia, 47, 2584-2592. [PDF]

• Cases PS (prosopagnosia) and DF (aperceptive visual agnosia) : absence of release from adaptation in the FFA of the patients for different faces. Presence of release from adaptation for nonface objects for patient PS but not DF. Shed light on the neural circuitry subtending face and object recognition.

Orban de Xivry, J.-J., Ramon, M., Lefèvre, P., Rossion, B. (2008). Reduced fixation on the upper area of personally familiar faces following acquired prosopagnosia. Journal of Neuropsychology, 2, 245-268. [PDF]

• First eye movement study of PS (prosopagnosia), who was involved in a familiar face recognition task (children of the kindergarten where she works). The results show that she focuses most of the time on the mouth (60%). More strikingly, she fixates exactly on each facial feature (mouth, left eye, right eye) while a normal observer who is familiar with the faces fixates in between features, in the centre of the face below the eyes during face identification (“center of mass” of the individual face).

Dricot, L., Sorger, B., Schiltz, C., Goebel, R., Rossion, B. (2008). The roles of “face” and “non-face” areas during individual face perception: evidence by fMRI adaptation in a brain-damaged prosopagnosic patient. NeuroImage, 40, 318-332. [PDF]

• Previously, we showed that PS’ right FFA does not show any release from adaptation when different individual faces are presented (Schiltz et al., 2006). This observation is replicated here, but the question is: why does the patient perform better than chance level at individual face discrimination? What are the neural substrates of this performance? We find release to individual face adaptation in the ventral part of the right lateral occipital cortex (vLOC) of PS, spared by the lesion. Hence, she appears to rely on a general visual area to individualize faces rather than on her residual face processing system.

Rossion, B. (2008). Constraining the cortical face network by neuroimaging studies of acquired prosopagnosia. NeuroImage, 40, 423-426. [PDF]

• A brief commentary and summary of our research showing how combining single case studies of prosopagnosia and neuroimaging can greatly inform about the neural basis of face processing

Sorger, B., Goebel, R., Schiltz, C., Rossion, B. (2007). Understanding the functional neuroanatomy of prosopagnosia . NeuroImage, 35, 836 - 852. [PDF] 

• A detailed anatomical and functional study of the brain of a rare case of pure prosopagnosia (PS). Localization and extent of lesions, retinotopy, LOC, MT localizers.

Mayer, E. & Rossion, B. (2007). Prosopagnosia. In O. Godefroy and J. Bogousslavsky. The Behavioral and Cognitive Neurology of Stroke : Cambridge University Press. pp. 315-334. 

Schiltz C, Sorger B, Caldara R, Ahmed F, Mayer E, Goebel R, Rossion B. (2006). Impaired face discrimination in acquired prosopagnosia is associated with abnormal response to individual faces in the right middle fusiform gyrus. Cerebral Cortex, 16, 574-586. [PDF] [slideshow summary]

• First evidence that the (right) FFA does not show release from adaptation to individual faces in 17 acquired prosopagnosia (PS). Block design and event-related experiments, multiple sessions. Suggest that the damaged right OFA is necessary to help the FFA individualizing faces.

Caldara, R., Schyns, P., Mayer, E., Smith, M., Gosselin, F., Rossion, B. (2005). Does prosopagnosia take the eyes out from faces? Evidence for a defect in the use of diagnostic facial information in a brain-damaged patient. Journal of Cognitive Neuroscience, 17, 1652-1666. [PDF]

• Patient PS. First evidence for a defect at extracting diagnostic information from the eyes of faces in acquired prosopagnosia, by means of response classification (‘Bubbles’). Like the observation of a preserved right FFA in the patient’s brain (Rossion et al., 2003), this was also an unexpected result: we were just interested in defining what kind of information was no longer diagnostic of identity for the prosopagnosic, without a priori.

Delvenne, J.-F., Seron, X., Coyette, F., Rossion, B.(2004). Evidence for perceptual deficits in associative visual (prosop)agnosia: a single-case study. Neuropsychologia, 42, 597-612. [PDF]

• Patient NS. A remarkable case of visual agnosia, with lesions sparing the occipital cortex. A series of behavioral experiment shows that despite a classification as a case of associative agnosia, the patient presents with impairments of perceptual nature. Impairment at integrating features into a coherent whole (integrative visual agnosia), no inversion effect.

Rossion, B., Caldara, R., Seghier, M., Schuller, A.-M., Lazeyras, F., Mayer, E. (2003). A network of occipito-temporal face-sensitive areas besides the right middle fusiform gyrus is necessary for normal face processing. Brain Journal of Neurology, 126, 2381-2395. [PDF] [slideshow summary]

• First behavioral and neuroimaging study of the case of PS, showing that she presents with a severe and highly specific impairment at face recognition, including individual face matching/discrimination. In fMRI, she presents with a normal (magnitude, size, location) preferential activation for faces in the right middle fusiform gyrus (FFA), despite a lesion of the right inferior occipital cortex and no OFA. This unexpected finding has led us to question the hierarchical view of the neural circuitry of face perception (OFA to FFA). It has been replicated since then in many studies listed above, by other groups of researchers also, and in another case of acquired prosopagnosia.