The temporal dynamics of neural object representations

This doctoral project will investigate the temporal dynamics of object representation. Whereas many functional magnetic resonance imaging (fMRI) studies have already used multivariate analyses to look at information representation in neural signals, multivariate pattern analysis in the time domain has been lagging. Therefore, in our studies we will use electroencephalography (EEG) to research the temporal properties of neural activity, combined with multivariate machine learning techniques in order to dissociate between different object properties. The objects properties include, but are not limited to animacy, object shape and body topography. Based on previous fMRI studies, theories have arisen that the occipito-temporal cortex has category selective patches and spatially clusters along an animacy continuum. Therefore, we have decided that one of the things we will investigate are the animacy property and face/body division. Stimuli have been carefully chosen based on an intuitive taxonomy where some animals/objects rank higher on the animacy scale than others, resulting in a stimulus set that can be placed on an animacy continuum. For each animal/object we have also included images focused on either their faces or bodies as to possibly disentangle neural signals based on the face/body division. We will additionally be using representational similarity analysis (RSA) to compare the temporal neural signals to model representational dissimilarity matrices (RDM). Model RDMs will be constructed based on the animate-inanimate division, the face-body division and the taxonomic hierarchy.