Amygdala activation at 3T in response to human and avatar facial expressions of emotions

Abstract

Facial expressions of emotions are important in nonverbal communication. Although numerous neural structures have been identified to be involved in emotional face processing, the amygdala is thought to be a core moderator. While previous studies have relied on facial images of humans, the present study is concerned with the effect of computer-generated (avatar) emotional faces on amygdala activation. Moreover, elicited activation patterns in response to viewing avatar faces are compared with the neuronal responses to human facial expressions of emotions. Twelve healthy subjects (five females) performed facial emotion recognition tasks with optimized 3T event-related fMRI. Robust amygdala activation was apparent in response to both human and avatar emotional faces, but the response was significantly stronger to human faces in face-sensitive structures, i.e. fusiform gyri. We suggest that avatars could be a useful tool in neuroimaging studies of facial expression processing because they elicit amygdala activation similarly to human faces, yet have the advantage of being highly manipulable and fully controllable. However, the finding of differences between human and avatar faces in face-sensitive regions indicates the presence of mechanisms by which human brains can differentiate between them. This mechanism merits further investigation.

Introduction

Recognizing the emotional expression in a face is a paramount social-cognitive competency of humans, which relies on a widespread neuronal network that involves cortical and subcortical-limbic structures (Adolphs, 2002). Studies have demonstrated that the amygdala is an important moderator of emotional processing, especially for the perception and recognition of emotional faces (see for review Calder et al., 2001). Several neuroimaging studies have highlighted amygdala activation in response to viewing fearful facial expressions (Adolphs et al., 1995, Irwin et al., 1996, Morris et al., 1996, Morris et al., 1998, Broks et al., 1998, Sprengelmeyer et al., 1999, Phillips et al., 2001) and even subliminal presentation of fearful expressions have been found to cause strong activation of the amygdala (Whalen et al., 1998, Morris et al., 1999). Showing fearful eyes alone also resulted in increased amygdala activation (Morris et al., 2002). This supports the view of an implicit and coarse appraisal role for the amygdala in the assessment of the valence of stimuli (threatening or pleasant) (LeDoux, 2000, Critchley et al., 2000) in particular when novel stimuli are presented (Schwartz et al., 2003). However, the amygdala's role in processing facial emotions other than fear is not as well documented, although recent data on amygdala activation in response to emotional and neutral facial expressions further indicate involvement of the amygdala in processing across emotions (Fitzgerald et al., 2006, Yang et al., 2002) based on its evaluation function.

Most studies on the perception of facial emotional expression have used two-dimensional (2D) photographs of real people. However, faces carry a lot more information than the emotion of the subject (e.g. facial shape and texture cues), which might perturb brain activity (Fink and Penton-Voak, 2002, Kranz and Ishai, 2006, Johnston, 2006 on facial appearance). In addition, the intensity of the facial expression of “real” people cannot be systematically controlled and varied. Moreover, photographing, lighting conditions and nonverbal signs, such as head tilts or head turns are possible confounds. Sato et al. (2002) used facial morphing techniques for the study of facial expression recognition in a bilateral amygdala-damaged patient and created facial expressions that blended happiness and fear, happiness and anger, or happiness and sadness for a subsequent categorization task. Facial morphing is a recognized method for the study of facial appearance and perception but it could distort reality and introduce artefacts, such as texture-smoothing from digital blending.

A possible means to circumvent these problems would be the use of three-dimensional (3D) virtual human face models (avatars). Avatars can be animated and deformed according to the experimenter's needs, yet still represent relevant action units (AUs; Ekman and Friesen, 1969, Ekman and Friesen, 1978) in the face. An avatar, derived from the Sanskrit where it was used to denote the bodily manifestation of an immortal being, refers in computing to the icon representing a user in internet forums or games. The term has also been adopted for computer-generated representations of humans (without being linked to a particular person), as applied here. Avatars allow the generation of stimuli with well defined facial appearance, for example faces systematically varying in age, sex and ethnicity. Photorealistic 3D models can be created by rendering under controlled lighting conditions, camera angle and focus length.

Useful applications of virtual characters have been investigated for several domains of human interaction (e.g. Morris et al., 2005) and communication (e.g. De Leo et al., 2003, Grammer and Oberzaucher, 2006). Bailenson et al. (2003) demonstrated that humans develop an experience of being with another person when confronted with avatars. Bailenson et al. (in press) showed that people were prepared to share most personal information with avatars that were low in realism and high in behavioral similarity. According to Ku et al. (2005), avatars have the potential to influence their subject's emotions dependent on gender of the avatar and intensity of expression. While some brain imaging studies have used morphed faces (e.g. Gur et al., 1994, Sato et al., 2002) or virtual environments (e.g. Morris et al., 2005) there is, to our knowledge, little known about the effect of avatar faces on brain response.

In the present study, we investigated neuronal activation, in particular amygdala activation, in response to expressions of emotions of real and virtual faces. Our aim was to determine whether facial expressions of avatars are able to cause amygdala activation similar to facial images of real people. We hypothesized that avatar facial expressions would elicit a similar amygdala response to images of human faces. We also hypothesized stronger activation of brain areas known to play a role in face processing to human stimuli, in particular elevated response of the fusiform gyrus and temporal regions.