Shortcuts for fear in hierarchical visual systems

Abstract

It is crucial to our survival that we can rapidly predict, detect, and respond to potential threats in our environment. As highly visual creatures, we have developed sophisticated neural networks for processing visual information about the world around us. Both the visual system and the amygdala are arguably the most thoroughly studied components of the brain and yet there is considerable controversy over how these two systems interact to rapidly respond to signs of threat. The work presented in this thesis aimed to resolve this controversy by investigating the structural and effective neural connectivity underlying rapid responses to fearful faces, with a particular focus on a pathway from the superior colliculus to the amygdala via the pulvinar that bypasses the visual cortex. This thesis also aimed to extend our understanding of how neural shortcuts for fear may influence the overall, hierarchical structure of the human brain, particularly regarding its ability to predict and to produce our conscious visual experience.In Chapter 2, I used a combination of magnetoencephalography (MEG) and dynamic causal modelling (DCM) to show that a subcortical connection between the pulvinar and the amygdala rapidly conveys both coarse (i.e. low spatial frequencies) and fine (i.e. high spatial frequencies) visual information about neutral and fearful faces. These results offer new evidence for a rapid but functionally-indistinct pulvino-amygdala connection that may receive multiple sources of input from early visual areas.In Chapter 3, I investigated how the structural connectivity between subcortical visual areas and the amygdala relates to neural activity and behaviour evoked during face perception. I reconstructed white matter fibres from the superior colliculus, the pulvinar, and the amygdala, using diffusion-weighted images (DWI) from over 600 participants. For the pulvino-amygdala connection, greater fibre density correlated with stronger functional coupling (as given by dynamic causal modelling for fMRI) during face perception, as well as more accurate fearful face recognition. These novel relationships between anatomy and function provide unparalleled evidence for a subcortical pulvino-amygdala connection involved in fearful face processing.Finally, in Chapter 4, I investigated whether our conscious perception of faces differs depending on our prior expectations for threat. I discovered that prior expectations accelerate conscious access to neutral but not fearful faces, which were the most consciously-accessible stimuli overall (Experiment 1), especially when unexpected (Experiment 2). Model-based electroencephalography (EEG) analysis revealed that prior expectations hasten stimulus encoding in the visual cortex, while surprising fearful faces elicit accelerated evidence accumulation in the right inferior frontal gyrus. This newly discovered influence of aversive prediction errors on conscious perception has strong implications for how emotion and prediction networks interact in the brain.In conclusion, this thesis significantly furthers our understanding of how visual processing rapidly interacts with the amygdala to facilitate fearful face processing. Furthermore, these findings open up a new area of investigation into how the predictive brain operates under threat.