The role of emotion processing areas in childrenʼs face perception network: A functional magnetic resonance imaging pilot study in 7- to 9-year-old children

Introduction Paediatric functional magnetic resonance imaging (fMRI) studies on face perception mainly focused on the location of regions in the so-called core network of face processing located in the occipito-temporal cortex. Little research has been done on the core networkʼs counterpart, the extended face processing network. The present fMRI study seeks to close this research gap by examining activation patterns within the extended face processing network during emotional face processing in children. Regarding the localisation of the core network regions, recent literature points to an increasing specification of typical face-selective activation patterns (Aylward et al., 2005, Gathers et al., 2004, Paul et al., 2003) in the core network. Conversely, previous studies detected more wide-spread activation patterns in the extended part of the face system in children. Faces are rarely neutral, but convey emotions – in everyday life, in fact, usually more than just one. The processing of the perceived emotions takes place in the extended part of the face network. This ability also develops with age, but so far it is unclear what the developmental differences look like at the neural level.

Methods Using functional Magnetic Resonance Imaging, we investigated brain activity in 7- to 9-year-old children (n = 8) during face processing (using faces with neutral, sad and fearful expression) and compared their neural response with an adult group (n = 10). In order to investigate the processing of the different facial expressions, the neural response of the groups was examined depending on the presented emotion category. Regions of interest were the amygdala, the insula and the inferior frontal gyrus, which were associated with emotion processing based on imaging studies with adults.

The experimental design is based on a face localizer paradigm which presented grey-scale photographs of faces (fearful, sad and neutral expression; Karolinska Directed Emotional faces (KDEF) dataset, http://www.emotionlab.se/resources/kdef) and houses. To capture the activity of smaller, subcortical regions, such as the amygdala, not the whole brain but only a brain section was measured. The area covered parts of the frontal and temporal lobe, as well as limbic structures and ranged up to the occipital lobe.

Results Our results were partly consistent with those of previous studies, since largely distributed neural activation patterns within the childrenʼs extended face perception network were active. Three different areas were identified to exhibit a significant heightened BOLD-response in 7- to 9-year-old children compared to adults: the left ventro-lateral prefrontal cortex and the bilateral medial temporal lobe which in the left hemisphere extended to the anterior insula. As reported by other studies, more wide-spread patterns compared to the adult group were found. The activation pattern particularly included the pre-defined ROIs amygdala, insula and inferior frontal gyrus. Notably, in contrast to adults, children showed a higher BOLD response within these ROIs and predominantly showed activation peaks in left frontal lobe.

Conclusion The observed activation differences indicate an important functional development in frontal and prefrontal regions and the limbic system in the course of emotional face perception across age. The data distribution of the children shows more variance in all three presented emotions than that of the adults, which illustrates the process of specialization of the face network that has not yet been completed in the children sample. The strong involvement of emotion-processing areas could be an indication that children are more emotionally stimulated by the presented faces than the adults. Moreover, an increased activity in the medial frontal cortex in children compared to adults, especially when processing neutral faces, points to more intensive evaluation processes and an ongoing acquisition of the ability to categorize emotions.