Elsevier

NeuroImage

Volume 32, Issue 4, 1 October 2006, Pages 1905-1917
NeuroImage

Multiple brain networks for visual self-recognition with different sensitivity for motion and body part

https://doi.org/10.1016/j.neuroimage.2006.05.026Get rights and content

Abstract

Multiple brain networks may support visual self-recognition. It has been hypothesized that the left ventral occipito-temporal cortex processes one's own face as a symbol, and the right parieto-frontal network processes self-image in association with motion–action contingency. Using functional magnetic resonance imaging, we first tested these hypotheses based on the prediction that these networks preferentially respond to a static self-face and to moving one's whole body, respectively. Brain activation specifically related to self-image during familiarity judgment was compared across four stimulus conditions comprising a two factorial design: factor Motion contrasted picture (Picture) and movie (Movie), and factor Body part a face (Face) and whole body (Body). Second, we attempted to segregate self-specific networks using a principal component analysis (PCA), assuming an independent pattern of inter-subject variability in activation over the four stimulus conditions in each network. The bilateral ventral occipito-temporal and the right parietal and frontal cortices exhibited self-specific activation. The left ventral occipito-temporal cortex exhibited greater self-specific activation for Face than for Body, in Picture, consistent with the prediction for this region. The activation profiles of the right parietal and frontal cortices did not show preference for Movie Body predicted by the assumed roles of these regions. The PCA extracted two cortical networks, one with its peaks in the right posterior, and another in frontal cortices; their possible roles in visuo-spatial and conceptual self-representations, respectively, were suggested by previous findings. The results thus supported and provided evidence of multiple brain networks for visual self-recognition.

Introduction

It has been established that the cognitive process specific to self-recognition entails ontogenetically and phylogenetically higher levels of development, as shown by intensive research of human infants (Anderson, 1984) and animals (Gallup, 1982). Evidence from recent cognitive neuroscience research suggests that the self-specific process during visual self-recognition is not unitary but involves multiple independent processes sustained by discrete brain mechanisms (see Gillihan and Farah, 2005 for review). Split brain patients could recognize their own face in either hemisphere; however, two hemispheres appear to have different sensitivities to one's own face and familiar faces (Turk et al., 2002, Keenan et al., 2003, Uddin et al., 2005b). Hemispheric dominancy in self-recognition has, however, been a matter of controversy; some researchers argue of right-hemisphere dominancy (Keenan et al., 2001, Decety and Sommerville, 2003), while others insist on the opposite (Turk et al., 2003, Brady et al., 2004). This controversy appears to be reconciled by assuming that distinct networks in the two hemispheres have different sensitivities to self-images depending on the experimental design. On the other hand, the lateral parietal and frontal cortices have been implicated in the processing of coincidence between the proprioception of one's own action and the visual perception of motion (motion–action contingency) (Fink et al., 1999, Farrer et al., 2003), which plays a key role during the infantile development of self-recognition in the mirror (Bigelow, 1981). On the contrary, activation of medial cortical structures, such as the cingulate cortex and medial prefrontal cortex during the processing of self-related (-relevant or -referential) stimuli (Vogeley and Fink, 2003, Northoff and Bermpohl, 2004) has often been reported in functional imaging studies. Furthermore, it is interesting to note cases of demented patients who combed or shaved in front of a mirror, which is regarded as a sign of visual self-recognition in infant and animal studies, but did not recognize their own mirror images (Phillips et al., 1996, Breen et al., 2001). This observation may show a conceptual limitation of assuming a single self-specific process in visual self-recognition.

Recently, a hypothesis of functionally and anatomically two discrete self-specific networks for visual self-recognition has been proposed based on the results of an fMRI experiment and the findings of previous neuropsychological and functional imaging studies (Sugiura et al., 2005). In this study, activation during recognition of one's own face, but not during recognition of familiar and learned faces, was observed in the left fusiform gryus and the right occipito-parietal and frontal-opercular regions. It has been speculated that the left fusiform gyrus processes one's own face as a symbol, based on the coincidence of the activation focus with the area involved in the processing of a visual word form (Cohen et al., 2000, Simons et al., 2003), neuropsychological data of a case showing alexia and impaired visual self-recognition (Gallois et al., 1988), and a fixed sociobehavioral meaning of one's own face in contrast to the episodic nature of familiar faces. The involvement of the right parieto-frontal regions in the processing of motion–action contingency, as suggested in previous functional imaging studies (Fink et al., 1999, Farrer et al., 2003), has led to the hypothesis that this right parieto-frontal network plays a critical role in the infantile development of self-recognition in the mirror and continues to play a role in visual self-representation over adulthood. Given that self-specific activation of the left fusiform gyrus is consistent with the results of two previous studies (Sugiura et al., 2000, Kircher et al., 2000), and that of the right parieto-frontal network is congruent with the results of another study (Uddin et al., 2005a), these two networks may have different sensitivities to self-images under different experimental conditions.

In this study, we performed an event-related fMRI to identify anatomically and functionally distinct brain networks for visual self-recognition, assuming that these networks may have different sensitivities to the features of self-image. First, we tested the following two predictions based on the hypothesis of two self-specific networks, which involve the left fusiform gyrus and the right parieto-frontal regions (Sugiura et al., 2005):

  • 1)

    If the network including the left fusiform gyrus processes a self-image as a visual symbol, as suggested by the coincidence of the location of the activation foci with that of the letter- or word-form-processing area (Sugiura et al., 2005), this network is more sensitive to a static picture of a face alone than to movies or a whole body. This is based on the fact that symbols, as well as letters and words, are usually static rather than moving, and when a face is presented as a symbol, it is usually presented as a single image such as in an identification card.

  • 2)

    If the network composed of the right parietal and frontal cortices is engaged in the processing of motion–action contingency during the infantile development of visual self-representation, the network is predicted to show a higher sensitivity to a movie than to a picture, and to a whole body than to a face alone. This is because this developmental process takes place while an infant observes his/her own whole-body action.

We accordingly prepared four stimulus conditions, that is, a movie of a whole body, a movie of a face, a picture of a whole body, and a picture of a face, for each subject, a familiar person, and an unfamiliar person. Self-specific activation was defined as differential activation for self relative to a familiar person; areas showing greater activation for familiar than unfamiliar persons were excluded to ensure that activation observed was not caused by greater familiarity with self than with a familiar person. Self-specific activation was compared across stimulus conditions, to examine whether the assumed networks (the left fusiform gyrus and the right parieto-frontal regions) exhibit the predicted profiles of sensitivity to the four stimulus conditions.

Second, aside from the two predictions for the first analysis, we attempted to identify functionally and anatomically independent multiple self-specific networks using an exploratory multivariate approach. We assumed that each network shows a different sensitivity to self-image under the four stimulus conditions, and the pattern of sensitivity varies among individuals. For example, the network processing one's own face as a symbol might prominently be activated under the Face conditions in some subjects, and the network processing self-image in association with motion–action contingency might be activated under the Body conditions in other subjects. Such inter-subject variability likely exists because people have different mirror viewing habits, daily physical activities, and degrees of awareness of their own faces and bodies. A principal component analysis (PCA) was used to summarize the patterns of inter-subject variability in self-specific activation under the four stimulus conditions in regions of interest (ROIs) where self-specific activation was detected in the first conventional subtraction analysis. Each principal component is expected to represent a distinct pattern of inter-subject variability of self-specific activation in an independent network. Each principal component or network is interpreted on the basis of existing knowledge of regions involved and correlations between behavioral data and principal component scores. This relatively new approach (Sugiura et al., in press) conforms to the analysis of functional connectivity (Friston et al., 1993, Friston and Büchel, 2004) and may be regarded as a ROI-based extension of the approach presented by Horwitz (1994).

Section snippets

Subjects

Forty-two healthy right-handed volunteers (27 males and 15 females, aged 18 to 24 years) participated in the study (six males and five females were excluded in the analysis, see fMRI measurement and image preprocessing). They applied in response to our advertisement in groups of two or three wherein each individual in a group knew each other. All subjects had normal vision and none had a history of neurological or psychiatric illness. Handedness was evaluated using the Edinburgh Handedness

Behavioral data

The percent correct and mean reaction time during familiarity judgment (Table 2) were analyzed using a two-way analysis of variance. The effect of stimulus condition only was significant for the percent correct (F(3,6) = 12.2, P < 0.01), and the effects of both stimulus condition and factor Person were significant for the mean reaction time (F(3,6) = 493.4, P < 0.001; and F(2,6) = 24.8, P < 0.01, respectively). Comparisons across the stimulus conditions were performed between Motion types for

Discussion

Self-specific activation was observed in a ventral occipito-temporal region adjacent to the fusiform gyrus (FG/ITG) bilaterally, the right parietal and frontal cortices, which is consistent with previous functional imaging findings on visual self-recognition (Sugiura et al., 2000, Sugiura et al., 2005, Kircher et al., 2000, Uddin et al., 2005a). Consistent with the first prediction, self-specific activation in the left ventral occipito-temporal cortex was greater during the Face condition than

Conclusion

Self-specific activation in the left ventral occipito-temporal cortex was higher during the observation of a face than during the observation of a whole body, when stimuli were presented as still pictures, rather than as movies. The results are consistent with the hypothesis that the left ventral occipito-temporal cortex processes one's own face as a symbol (Sugiura et al., 2005). The PCA results of the inter-subject variability in the ROIs showing self-specific activation revealed two

Acknowledgments

We would like to thank all our colleagues in NICHe, Tohoku University, for their support in the fMRI experiment, and Wataru Suzuki for his helpful comments on the manuscript. This research project was supported by the RISTEX/JST, the CREST/JST, a Grant-in-aid for Scientific Research on Priority areas (C)-Advanced Brain Science Project (MEXT), and the 21st Century COE Program (MEXT) entitled, “A Strategic Research and Education Center for an Integrated Approach to Language and Cognition” (Tohoku

References (60)

  • G. Northoff et al.

    Cortical midline structures and the self

    Trends Cogn. Sci.

    (2004)
  • K.N. Ochsner et al.

    The neural correlates of direct and reflected self-knowledge

    NeuroImage

    (2005)
  • R.C. Oldfield

    The assessment and analysis of handedness: the Edinburgh inventory

    Neuropsychologia

    (1971)
  • T.W. Schmitz et al.

    Metacognitive evaluation, self-relevance, and the right prefrontal cortex

    NeuroImage

    (2004)
  • O. Simon et al.

    Topographical layout of hand, eye, calculation, and language-related areas in the human parietal lobe

    Neuron

    (2002)
  • J.S. Simons et al.

    Neural mechanisms of visual object priming: evidence for perceptual and semantic distinctions in fusiform cortex

    NeuroImage

    (2003)
  • M. Sugiura et al.

    Passive and active recognition of one's own face

    NeuroImage

    (2000)
  • M. Sugiura et al.

    Cortical mechanisms of visual self-recognition

    NeuroImage

    (2005)
  • L.Q. Uddin et al.

    Self-face recognition activates a frontoparietal “mirror” network in the right hemisphere: an event-related fMRI study

    NeuroImage

    (2005)
  • L.Q. Uddin et al.

    Split-brain reveals separate but equal self-recognition in the two cerebral hemispheres

    Conscious. Cogn.

    (2005)
  • K. Vogeley et al.

    Neural correlates of the first-person-perspective

    Trends Cogn. Sci.

    (2003)
  • K. Vogeley et al.

    Mind reading: neural mechanisms of theory of mind and self-perspective

    NeuroImage

    (2001)
  • J.R. Anderson

    The development of self-recognition: a review

    Dev. Psychobiol.

    (1984)
  • A.E. Bigelow

    The correspondence between self- and image movement as a cue to self-recognition for young children

    J. Gen. Psychol.

    (1981)
  • D.J. Bolger et al.

    Cross-cultural effect on the brain revisited: universal structures plus writing system variation

    Hum. Brain Mapp.

    (2005)
  • N. Breen et al.

    Mirrored-self misidentification: two cases of focal onset dementia

    Neurocase

    (2001)
  • C. Büchel et al.

    The functional anatomy of attention to visual motion. A functional MRI study

    Brain

    (1998)
  • R.B. Cattell

    The scree test for the number of factors

    Multivariate Behav. Res.

    (1966)
  • L. Cohen et al.

    The visual word form area. Spatial and temporal characterization of an initial stage of reading in normal subjects and posterior split-brain patients

    Brain

    (2000)
  • P.E. Downing et al.

    A cortical area selective for visual processing of the human body

    Science

    (2001)
  • Cited by (0)

    View full text