Elsevier

NeuroImage

Volume 19, Issue 3, July 2003, Pages 877-883
NeuroImage

Regular article
The functionally defined right occipital and fusiform “face areas” discriminate novel from visually familiar faces

https://doi.org/10.1016/S1053-8119(03)00105-8Get rights and content

Abstract

Neuroimaging (PET and fMRI) studies have identified a set of brain areas responding more to faces than to other object categories in the visual extrastriate cortex of humans. This network includes the middle lateral fusiform gyrus (the fusiform face area, or FFA) as well as the inferior occipital gyrus (occipital face area, OFA). The exact functions of these areas in face processing remain unclear although it has been argued that their primary function is to distinguish faces from nonface object categories—“face detection”—or also to discriminate among faces, irrespective of their visual familiarity to the observer. Here, we combined the data from two previous positron emission tomography (PET) studies to show that the functionally defined face areas are involved in the automatic discrimination between unfamiliar faces and familiar faces. Consistent with previous studies, a face localizer contrast (faces–objects) revealed bilateral activation in the middle lateral fusiform gyrus (FFA, BA37) and in the right inferior occipital cortex (OFA, BA19). Within all the regions of the right hemisphere, larger levels of activation were found for unfamiliar as compared to familiar faces. These results suggest that the very same areas involved in categorizing faces at the basic or individual level, play a role in differentiating familiar faces from new faces, showing an overlap between visual and presemantic mnesic representations of faces in the right hemisphere.

Introduction

Understanding how the human brain perceives and recognizes faces has become one of the most exciting and debated areas of research in cognitive neuroscience (e.g., Kanwisher 2000, Haxby et al 2000, Tarr and Gauthier 2000. Pursuing this general goal, neuroimaging studies have localized a set of brain areas in the human visual extrastriate cortex that respond more during the presentation of faces than to other object categories. These areas are located mainly in the middle lateral fusiform gyrus (the fusiform face area (FFA), Kanwisher et al., 1997), and in the inferior occipital gyrus (infOg, occipital face area (OFA), e.g., Halgren et al., 1999), with a right hemisphere advantage. Being located beyond the retinotopic striate and extrastriate visual cortices, the exact anatomical borders of these regions cannot be defined with respect to other visual areas (Halgren et al., 1999). Accordingly, their localization is based on functional criteria, i.e., a significantly larger response for pictures of faces than other objects (Kanwisher et al., 1997). An important and unresolved issue is whether the face areas are involved in perceptual aspects of face processing only, helping to discriminate between faces and other object categories (“face detection,” Tong et al., 2000) or between different unfamiliar faces (Gauthier et al., 2000), or if they also contribute to presemantic face recognition, i.e., the discrimination between a face previously seen and a novel face. Despite the computational complexity of this task, humans are particularly efficient at it, even when the “familiar” face has been encountered just once, months or years ago (Bruce and Young, 1998). Because recognizing familiar faces and encoding new faces in memory appear to be the primary functions of the face processing system, it would be surprising that the areas forming the core of the face processing system (Haxby et al., 2000) do not play any role in these functions. In addition, the inability to recognize and encode faces in memory is exactly what defines prosopagnosia—a deficit usually observed following lesions of the occipitotemporal cortex—but these patients do not usually present difficulties at discriminating faces from other object categories (e.g., Bruyer et al., 1983).

In healthy humans, the processing of familiar and unfamiliar faces has been compared in a number of imaging experiments Haxby et al 2000, Henson et al 2000, Henson et al 2002, George et al 1999, Gorno-Tempini et al 1998, Gorno-Tempini and Price 2001, Katanoda et al 2000, Leveroni et al 2000, Rossion et al 2001, Sergent et al 1992, Wiser et al 2000, but none of these studies tested the hypothesis that the exact same functional regions that respond more to faces than to other object categories, are also involved in familiar vs novel face discrimination. In addition, in the previous literature, there are a number of task and stimuli confounds that prevent a straightforward interpretation of activation differences produced by familiar and unfamiliar faces in visual areas (see Rossion et al., 2001).

Here, we tested the hypothesis that the functionally defined face areas differentiate between visually familiar and novel faces, using a multistudy analysis of the data of two previously reported PET experiments Rossion et al 2000a, Rossion et al 2001 carried out on the same eight male subjects. The data from the first study (Rossion et al., 2000a) were used to functionally define the areas responding significantly more to faces than to objects in this group of subjects, and these regions were then tested for familiar/unfamiliar face differences in the same subjects scanned later.

Section snippets

Participants

Eight right-handed adult male subjects (age range: 22–25) took part in both PET studies, the two scanning sessions being performed at 4 to 6 weeks interval time.

Stimuli

Details of stimulation can be found in Rossion et al. (2000a) for the localizer task, and in Rossion et al. (2001) for the familiar/unfamiliar face comparisons. In the first study (localizer task), subjects underwent 12 scans, half of them with presentations of faces, and the other half with objects. In the second study, 30 novel

Discussion

The main conclusion of our multistudy analysis is that the functionally defined right fusiform face area and also the right occipital face area, at a lower level, are involved in the discrimination of familiar from unfamiliar individual faces. These results complement our previous study (Rossion et al., 2001), in which we identified a network of visual areas activated by faces compared to a baseline resting condition and modulated by visual familiarity of the faces. In that study, the

Acknowledgements

Bruno Rossion is supported by the National Funds for Scientific Research (Belgium). This work was supported by grant FRSM 3.4520.98 and grant ARC ARC 01/06-267 (Communauté Française de Belgique—Actions de Recherche Concertées). We thank two anonymous referees for helpful suggestions on a previous version of this manuscript.

References (43)

  • A.R. Damasio et al.

    Prosopagnosiaanatomic basis and behavioral mechanisms

    Neurology

    (1982)
  • H.D. Ellis et al.

    Repetition priming and face processing priming occur within the system that responds to the identity of a face

    Q. J. Exp. Psychol. A

    (1990)
  • M.J. Farah

    Visual AgnosiaDisorders of Object Recognition and What They Tell Us About Normal Vision

    (1990)
  • P.T. Fox et al.

    Enhanced detection of focal brain response using intersubject averaging and change-distribution analysis of subtracted PET images

    J. Cereb. Blood Flow Metab.

    (1988)
  • K.J. Friston et al.

    Statistical parametric maps in functional imaginga general linear approach

    Hum. Brain Mapp.

    (1995)
  • I. Gauthier et al.

    Activation of the middle fusiform area increases with expertise in recognizing novel objects

    Nat. Neurosci.

    (1999)
  • I. Gauthier et al.

    The fusiform “face area” is part of a network that processes faces at the individual level

    J. Cogn. Neurosci.

    (2000)
  • N. George et al.

    Contrast polarity and face recognition in the human fusiform gyrus

    Nat. Neurosci.

    (1999)
  • M.L. Gorno-Tempini et al.

    The neural systems sustaining face and proper name processing

    Brain

    (1998)
  • M.L. Gorno-Tempini et al.

    Identification of famous faces and buildings—a functional neuroimaging study of semantically unique items

    Brain

    (2001)
  • E. Halgren et al.

    Location of human face-selective cortex with respect to retinotopic areas

    Hum. Brain Mapp.

    (1999)
  • Cited by (158)

    View all citing articles on Scopus
    View full text