Research reportSuccessful episodic memory retrieval of newly learned faces activates a left fronto-parietal network
Introduction
Face recognition is a basic human ability with a high degree of social relevance. Even newborns are able to specifically distinguish faces from other objects [32] and age-related decrements in face recognition have been reported [1]. The fusiform gyrus, part of the medial temporal cortex, is specialized for face perception [17] and lesions to occipito-temporal brain areas can lead to an impairment in face recognition (prosopagnosia). Fronto-temporal brain damage can cause syndromes such as the capgras delusion where previously familiar faces can be recognized but lack a feeling of familiarity [8], [29]. Behavioural studies indicate that the cognitive processes for recognising familiar faces are distinct from recognising newly learned faces. An example of this is the ability of subjects to recognise familiar persons even from very low quality images, in contrast to their relatively poor ability to recognize newly learned unfamiliar faces [6], [15]. The mechanisms underlying the recognition of unfamiliar faces are of particular interest for forensic reasons (e.g., eye witnessing) and might be useful in developing automatic, computer-based methods of person identification.
Functional imaging studies that have investigated the neural correlates of familiar face recognition have found medial temporal activation, including the hippocampus [23]. The cerebral regions differentially involved in the recognition of newly learned faces, however, remain elusive. Brain imaging can facilitate the formulation of a theoretical model of unfamiliar face recognition [15]. It is only recently that event-related functional magnetic resonance imaging (fMRI) allowed the distinction between cerebral activation during the retrieval of newly learned items and that of distractor faces in a recognition memory task. We used this technique to investigate brain activation during correct recognition of newly learned faces compared to correct identification of distractor faces. We hypothesized that fusiform and medial temporal areas are involved in face processing per se [17]. Event-related potential (ERP) studies of face recognition demonstrated a parietal and a fronto-central activation when previously learned faces were compared with unknown faces [10], [11]. Similarly, left frontal and parietal areas are most consistently exhibiting stronger activity for old vs. new items during verbal recognition memory tasks in functional imaging studies [27]. Based on these findings, we hypothesized that left parietal and medial frontal cortical regions are involved in the recognition of newly learned faces.
Section snippets
Subjects
Six male and six female right-handed subjects [26], aged 20–32 (mean 25, S.D. 3.3) years were studied. Subjects were excluded if they had any medical, neurological or psychiatric illness, past or present, or if they were taking medication. Their informed consent was obtained according to the declaration of Helsinki. The local Ethical Committee approved the study.
Encoding procedure
Thirty photographs of male or female faces were presented, randomly intermixed, each for 4 s as the subjects lay in the scanner. The
Behavioral data
Subjects correctly recognized 75.8% (S.D. 12.6%) of the old faces and correctly rejected 85.5% (S.D. 84%) of the new faces (p=0.059; paired t test). The reaction time for rejecting a new face (mean 1.70 s; S.D. 0.35 s) was significantly slower (p=0.002; paired t test) than correctly recognizing an old face (mean 1.47 s; S.D. 0.28 s).
Functional imaging data
For the face recognition task vs. baseline (main effect), positive signal changes were detected bilaterally in visual response regions including the primary visual
Discussion
In this study, the correct recognition of newly learned faces was correlated with positive signal change in the left parietal and left medial frontal/anterior cingulate cortex. As reaction time was longer for the rejection of unlearned faces, it is unlikely that this result is due to a higher effort for recognizing newly learned faces. When face processing was compared to a low-level baseline, positive signal change was present in the primary visual, fusiform and hippocampal regions, thus
References (32)
Autonomic recognition of names and faces in prosopagnosia: a neuropsychological application of the Guilty Knowledge Test
Neuropsychologia
(1984)Facial recognition 1990
Cortex
(1990)- et al.
Human and automatic face recognition: a comparison across image formats
Vis. Res.
(2001) - et al.
Capgras delusion: a window on face recognition
Trends cogn. sci.
(2001) - et al.
Dissociating memory processes involved in direct and indirect tests with ERPs to unfamiliar faces
Cogn. Brain Res.
(2001) - et al.
Recognition of unfamiliar faces
Trends Cogn. Sci.
(2000) - et al.
The distributed human neural system for face perception
Trends Cogn. Sci.
(2000) - et al.
Neural substrates for the recognition of newly learned faces: a functional MRI study
Neuropsychologia
(2000) - et al.
Recognizing one's own face
Cognition
(2001) - et al.
Improvement of the acquisition of a large amount of MR images on a conventional whole body system
Magn. Reson. Imaging
(1999)