Individual differences in prefrontal coding of visual features

Qi Lin Presenter
RIKEN
Center for Brain Science
Wako, Saitama 
Japan
 
Wednesday, Jun 26: 11:30 AM - 12:45 PM
1897 
Oral Sessions 
COEX 
Room: Grand Ballroom 104-105 
Sitting at the apex of the cortical hierarchy (Young, 1992), the lateral prefrontal cortex (LPFC) is most commonly associated with complex and high-level cognition. Does LPFC also play a role in visual perception?
Evidence from nonhuman primates has characterized the representation of visual information in LPFC, even when there is no cognitive task involved (Riley et al., 2017; Tsao et al., 2008; Haile et al., 2019). In addition, feedback signals from LPFC seem necessary for recognizing objects under challenging conditions (Kar & DiCarlo, 2021). Yet to date, with few exceptions (e.g., Huth et al., 2012), how LPFC may support perception in humans remains understudied.
To fill in this gap, we built encoding models to relate visual features extracted from a deep neural network to predict brain activities in LPFC. We then contrasted the tuning profiles of LPFC for visual stimuli to those of the visual cortex. Strikingly, we found that the degrees of individual variability were higher in LPFC. That is, the stimuli that drove the maximal overall LPFC varied highly across individuals.