HomePublications

Investigating holistic face processing within and outside of face-responsive brain regions

Research output: Contribution to journalArticle

Open Access permissions

Open

Documents

DOI

Authors

  • Celia Foster
  • Isabelle Bülthoff
  • Andreas Bartels
  • Mintao Zhao

Organisational units

Abstract

It has been shown that human faces are processed holistically (i.e. as indecomposable wholes, rather than by their component parts) and this holistic face processing is linked to brain activity in face-responsive brain regions. Although several brain regions outside of the face-responsive network are also sensitive to relational processing and perceptual grouping, whether these non-face-responsive regions contribute to holistic processing remains unclear. Here, we investigated holistic face processing in the composite face paradigm both within and outside of face-responsive brain regions. We recorded participants’ brain activity using fMRI while they performed a composite face task. Behavioural results indicate that participants tend to judge the same top face halves as different when they are aligned with different bottom face halves but not when they are misaligned, demonstrating a composite face effect. Neuroimaging results revealed significant differences in responses to aligned and misaligned faces in the lateral occipital complex (LOC), and trends in the anterior part of the fusiform face area (FFA2) and transverse occipital sulcus (TOS), suggesting that these regions are sensitive to holistic versus part-based face processing. Furthermore, the retrosplenial cortex (RSC) and the parahippocampal place area (PPA) showed a pattern of neural activity consistent with a holistic representation of face identity, which also correlated with the strength of the behavioural composite face effect. These results suggest that neural activity in brain regions both within and outside of the face-responsive network contributes to the composite-face effect.

Details

Original languageEnglish
Article number117565
JournalNeuroImage
Volume226
Early online date19 Nov 2020
DOIs
Publication statusE-pub ahead of print - 19 Nov 2020
Peer-reviewedYes

View graph of relations

ID: 185189586