The roles of “face” and “non-face” areas during individual face perception: Evidence by fMRI adaptation in a brain-damaged prosopagnosic patient
Introduction
Functional neuroimaging studies of the human visual cortex have identified several areas that play an important role in object perception and recognition (Grill-Spector and Malach, 2004). These regions respond more strongly when viewing pictures of object shapes than control stimuli preserving low-level visual information such as textures, visual noise, or scrambled objects. Among these areas, the lateral occipital cortex (LOC; e.g. Grill-Spector et al., 2001, Grill-Spector et al., 1999, Grill-Spector et al., 1998, Malach et al., 1995) shows greater fMRI response to objects (including faces) than these control stimuli, but this area does not show systematic larger responses to some object categories than others. It is located anterior to retinotopic visual areas, extending both ventrally (vLOC) on the lateral bank of the fusiform gyrus and dorsally (dLOC) in two anatomically segregated subregions. Anterior to the vLOC, a region of the lateral fusiform gyrus, the ‘fusiform face area’, or ‘FFA’ (e.g. Kanwisher et al., 1997, Sergent et al., 1992) also responds more strongly to objects than control stimuli, but shows a preferential response for faces above all other object categories. Larger responses to faces than objects are also consistently observed in the inferior occipital gyrus, the so-called ‘occipital face area’ (‘OFA’; Gauthier et al., 2000b) generally posterior to, and partially overlapping with the vLOC.
FMRI-adaptation (Grill-Spector et al., 2006a, Grill-Spector and Malach, 2001) or repetition-suppression (Henson and Rugg, 2003) studies show a larger response in the LOC to novel objects than to repeated objects (Avidan et al., 2002, Grill-Spector et al., 1999, Sayres and Grill-Spector, 2006) and a correlation of that response with recognition performance (Avidan et al., 2002, Bar et al., 2001, Grill-Spector et al., 2000, James et al., 2000). Similar effects have been observed in both the ‘FFA’ and ‘OFA’ for faces (e.g. Eger et al., 2004b, Gauthier et al., 2000b, Gilaie-Dotan and Malach, 2007, Grill-Spector et al., 2004, Loffler et al., 2005, Rotshtein et al., 2005, Schiltz and Rossion, 2006). These observations indicate that the LOC and both of the ‘face areas’ are directly involved in the perception of individual exemplars of objects and faces respectively.
An unresolved issue is whether a high-level visual area that does not respond preferentially to one category, i.e. the LOC for faces, nevertheless plays a role in discriminating individual members of that category. According to a strict localizationist view, the answer to this question is negative. This view suggests a dominant or exclusive role of areas responding preferentially to faces, such as the ‘FFA’, for face perception (Kanwisher et al., 1997, Kleinschmidt and Cohen, 2006, Spiridon and Kanwisher, 2002). In contrast, according to a more distributed view, a large part of the ventral stream, including the ‘FFA’, contributes to face perception (Haxby et al., 2001, Tovee, 1998). That is, visual areas that do not respond preferentially to faces, such as the LOC, would also play a role in face perception. Moreover, areas responding preferentially to faces, such as the ‘FFA’, would also contribute to object perception.
Here we aimed to shed light on this issue by recording fMRI-adaptation in a well described brain-damaged patient, PS, who is no longer able to perceive and recognize individual faces, i.e. acquired prosopagnosia (Bodamer, 1947, Quaglino et al., 2003, Sergent and Signoret, 1992). Contrary to most cases of prosopagnosia, the patient’s ability to recognize and discriminate non-face objects, even at the individual level, is remarkably preserved (Rossion et al., 2003, Schiltz et al., 2006). Her prosopagnosia follows a dominant right hemispheric lesion in the inferior occipital cortex, which damaged the territory of the right ‘OFA’, where no face preferential activation can be observed. Despite this lesion, the patient has a right ‘FFA’ as significant and large in size as in normal controls (Rossion et al., 2003, Sorger et al., 2007), a finding that has also been reported in another brain-damaged case of prosopagnosia with bilateral inferior occipital lesions encompassing the ‘OFA’ (Steeves et al., 2006). However, there is a lack of release from identity adaptation to individual faces in the patient PS’s right ‘FFA’, in agreement with her behavioral impairment at discriminating individual faces (Schiltz et al., 2006). Most recently, we have mapped the low- and high-level visual areas of the patient PS with respect to her lesions, and observed a right vLOC area of equal size and level of activation than normals, next to the dominant lesion causing prosopagnosia (Sorger et al., 2007).
In the present study, we took advantage of this pattern of structurally damaged (right ‘OFA’) and intact (right ‘FFA’ and LOC) tissue in PS’s brain to test two hypotheses.
First, we asked whether the patient’s ‘FFA’, which does not show release from adaptation to face identity (Schiltz et al., 2006), may nevertheless play a role in her preserved ability to discriminate individual exemplars of objects. To do that, we used an experiment in which all normal participants showed release from identity adaptation to both objects and faces in the ‘FFA’, and tested it in the same area in the patient’s brain. We found for the first time that release from adaptation was significant for objects in both normal viewers and the patient’s ‘FFA’, still in the absence of any effect for faces for the patient.
Second, a whole-brain analysis of the patient’s data in the same experiment revealed a significant release from adaptation to individual exemplars of faces in the right vLOC only, an area that does not respond stronger to faces than objects. More specifically, both objects and faces showed release from adaptation effects in all normal participants, and in the patient’s brain, in the right vLOC. These results suggest that following brain damage, the prosopagnosic patient is unable to recruit her dominant face perception system in the ‘FFA’ to discriminate individual faces, but rather relies on an area that does not respond preferentially to faces. This is in line with her performances during individual face discrimination tasks, which are seriously deficient, but yet largely above chance level.
These observations were tested further in two fMRI experiments using an event-related paradigm. In experiment 2, pictures of faces and objects were used by pairs and showed the same results as in the block design used in experiment 1. In the third experiment, an active discrimination of facial identity was performed.
Section snippets
Subjects
The prosopagnosic patient PS has already been described in detail in previous studies (Caldara et al., 2005, Rossion et al., 2003, Schiltz et al., 2006, Sorger et al., 2007). Briefly, PS was born in 1950 and sustained a closed head injury in 1992 that left her with extensive lesions of the left mid-ventral (mainly fusiform gyrus) and the right inferior occipital cortex. Minor damages to the left posterior cerebellum and the right middle temporal gyrus were also detected (see Sorger et al., 2007
Results
In the face localizer, there was a significantly larger response to faces compared to objects in the right middle fusiform gyrus (‘FFA’) for all control subjects (Talairach coordinates: 36 ± 4, − 48 ± 6, − 15 ± 4; mean cluster size: 717 voxels ± 491) and for PS (35, − 53, − 20; 479 voxels) (Table 1). In the object localizer, there was a significantly larger response to objects compared to scrambled objects with an activation corresponding to the ventral part of the right occipital complex (vLOC) for both
Discussion
Our three fMRI-adaptation experiments carried out with normal participants and the brain-damaged prosopagnosic patient PS show two novel observations.
- 1.
In the same high-level visual area, i.e. the ‘FFA’, the processes allowing for the discrimination of individual members of a visual category can be preserved for non-face objects but is impaired for faces. This suggests that faces and non-face objects responses in this area are functionally independent.
- 2.
High-level visual areas that do not respond
Conclusions
Our results show that the right fusiform gyrus (‘FFA’) in the patient’s brain can be impaired for discriminating individual items of its preferential category, faces, while being able to carry out interindividual object discrimination, in line with the behavior of the patient. Thus, the same brain area can be functional for one category and disrupted for another. Second, we observed a larger signal to different faces than same faces in the ventral part of the lateral occipital complex (vLOC),
Acknowledgments
We are grateful to PS for her especially great patience during the experiments and to all our control subjects. LD was supported by a research grant (FSR) from the Université catholique de Louvain; BR is supported by the Belgian National Foundation for Scientific Research (FNRS).
References (73)
- et al.
Analysis of the neuronal selectivity underlying low fMRI signals
Curr. Biol.
(2002) - et al.
Cortical mechanisms specific to explicit visual object recognition
Neuron
(2001) - et al.
Prosopagnosia and facial discrimination
J. Neurol.
(1972) - et al.
The functional magnetic resonance imaging hemodynamic response to faces remains stable until the ninth decade
NeuroImage
(2003) - et al.
Face recognition and postero-inferior hemispheric lesions
Neuropsychologia
(1997) - et al.
Investigation of the single case in neuropsychology: confidence limits on the abnormality of test scores and test score differences
Neuropsychologia
(2002) - et al.
Recognition of unfamiliar faces in prosopagnosia
Neuropsychologia
(1990) - et al.
Evidence for perceptual deficits in associative visual (prosop)agnosia: a single-case study
Neuropsychologia
(2004) - et al.
Prosopagnosia and alexia without object agnosia
Cortex
(1998) - et al.
Scale invariant adaptation in fusiform face-responsive regions
NeuroImage
(2004)