Enhancedamygdalaresponsestoemotional versusneutralschematic facialexpressions Christopher I. Wright, 1,2,CA Brian Martis, 1,3 Lisa M. Shin, 1,4 HÔkan Fischer 1,5 and Scott L. Rauch 1 1 Psychiatric Neuroimaging Research Group and Nuclear Magnetic Resonance Center,Massachusetts General Hospital,Harvard Medical School, 13th St., Bldg149,CNY-9,Charlestown,MA 02129; 2 Brigham Behavioral Neurology Group,Departmentof Neurology, Brigham and Women’s Hospital, Boston,MA02115; 3 Departmentof Psychiatry,Universityof Illinois at Chicago,Chicago,IL 60612; 4 DepartmentofPsychology,TuftsUniversity,Medford, MA 02155,USA; 5 ARC-Aging Research Center,Karolinska Institute and Stockholm University, Stockholm, Sweden CA,1 Corresponding Author and Address Received31January2002;accepted18February2002 Human facial emotional expressions are complex. This may con- foundstudiesexaminingbrainresponsesto thesestimuliincontrol andclinicalpopulations.However,severallinesofevidencesuggest that a few elementary facial features convey the gist of emotional expressions.Using fMRI,weassessedbrainresponses tolinedraw- ings of emotionally valenced (i.e. angry and happy) and neutral facesin healthy human subjects. Signi¢cantly increased fMRI signal was found in the amygdala, hippocampus and prefrontal cortex in response to emotional vs neutral schematic faces. Although direct comparisons of schematic and human faces will be needed, these initialresults suggest that schematic facesmaybeuseful for study- ing brain responses to emotional stimuli because of their sim- plicity relative to human faces. NeuroReport 13:785–790  c 2002 Lippincott Williams & Wilkins. Keywords: Cerebral cortex; Faces; Functional magnetic resonance imaging; Human; Limbic system INTRODUCTION Functional neuroimaging and lesion studies have identified neural systems responsive to human emotional facial expressions. In this regard, the amygdala has been consistently implicated in the processing of human emo- tional facial expressions. It is activated in response to overtly [1–5] or covertly presented [6,7] emotional expressions, and selective damage to the amygdala leads to specific deficits in the recognition of threat-related emotional expressions [8,9]. In addition, a variety of cortical areas (i.e. prefrontal, insular and inferior temporal cortex) have been associated with processing of emotional facial stimuli [7,10,11]. The studies mentioned above, and recent work in clinical populations [12,13], have demonstrated the utility of using human emotional face stimuli in elucidating the neural substrates of emotional processing and its dysregulation. However, human faces are visually complex, and studies involving different sets of human emotional face stimuli raise the possibility of confounding effects, for example due to race [14,15]. Thus, a simplified version of emotionally expressive face stimuli might help to minimize the variance associated with such confounds. Line drawings or schematic face expression stimuli might be suitable in this regard, as a subset of simple features appear to convey the necessary information for recognition of facial emotion [16]. In support of this notion is the fact that certain facial expressions, and their meanings, are acknowledged across a wide range of cultures [16]. More- over, relatively simple geometric patterns or line drawings of emotional expressions provide enough information to reliably indicate the meaning of an expression [17–20]. Furthermore, behavioral studies demonstrate the salience of such schematic faces, showing enhancement of directed attention [21] and increased efficiency of visual search [22] for emotionally valenced (angry or happy) vs neutral schematic expressions. It is not known whether the human amygdala, or other brain areas, are responsive to simple line drawings of faces that convey emotion. Such information is important as an initial step towards understanding whether schematic emotional face stimuli might serve as reliable probes for studying emotional processing in heterogenous popula- tions. Therefore, we evaluated human brain responses to simple line drawings of emotional and neutral facial expressions using fMRI. MATERIALS ANDMETHODS Subjects: Sixteen healthy right-handed subjects (eight male, eight female), mean (7 s.d.) age of 30 7 11.6 years (range 20–59), were studied. Subjects with previous or current psychiatric, neurological or medical disease or current use of psychoactive medication or substance abuse were excluded. Written informed consent was obtained. This study was approved and conducted in accordance with 0959-4965  c Lippincott Williams & Wilkins Vol 13 No 6 7 May 2002 785 BRAIN IMAGING NEUROREPORT