Feedback as the way forward: sensory predictions in the primate face processing hierarchy

• Applicant: Dr. Caspar M. Schwiedrzik
• Subject Area: Cognitive, Systems and Behavioural Neurobiology
• Term: since 2018
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 405847822

The world is full of structure. Thunder follows lightning and traffic lights switch from red to green. The brain uses such patterns to predict what will happen next. By anticipating instead of reacting, we can, for example, drive off just as the traffic light turns green and thus avoid loud honking. Prediction improves perception. In disorders such as autism, the ability to make predictions appears to be impaired. However, little is known about how predictions are implemented in the brain. In particular, it is unclear whether brain areas such as the hippocampus are involved in predictive processes in addition to cortical areas. This is particularly relevant in the case of predictions that are not purely sensory, e.g., predictions based on abstract, conceptual information. We want to test whether purely sensory and conceptual predictions are realized by feedback from the hippocampus to visual cortical areas. This hypothesis goes well beyond classical theories of predictive processing, which are commonly restricted to purely cortical computaitons. We will investigate this in macaques as well as in humans. We will use the visual face processing system as a model. This system is located in the visual cortex and has connections to the hippocampus. We will investigate whether the hippocampus learns the models of the environment necessary to predict visual information, how predictions are exchanged between hippocampus and cortex, how this influences information processing and whether the hippocampus is causally relevant for predictions. To this end, we will perform parallel electrophysiological recordings in both macaques (Aim 1) and epilepsy patients (Aim 2), which will provide information on the localization and dynamics of information flow between hippocampus and cortex during predictable and unpredictable stimulus sequences. We will investigate the causal relevance of hippocampal feedback for prediction by reversible inactivation using electrophysiological and pupillometric markers. These studies across species (humans, monkeys) and levels of abstraction (sensory, conceptual) will provide a clear test of the role of the hippocampus in prediction and reveal a new quality of information processing beyond current, purely cortical theories of predictive processing. In this way, we gain insight into the mechanisms of prediction, a core task of the brain, and into the information processing roots of deficits in conditions such as autism.

• DFG Programme: Independent Junior Research Groups