Thesis defense Silvy Collin (Donders series 323)

Our real-life experiences are continuous, and form a coherent whole. These rich and continuous real-life experiences need to be captured in a meaningful neural structure to allow us to remember them as a coherent whole. In my thesis, I aim to answer the question: Does the brain represent episodic memories in dynamic, hierarchical event networks? To answer this question, I first investigated how hierarchical event networks are formed, and then I explored the dynamics of event networks by investigating neural mechanisms underlying memory updating. The results showed that episodic memory representations progressively increase in scale along the hippocampal long axis which may enable the formation of memory hierarchies. Event elements from likely future events are already represented in the brain in anticipation to future events. These event networks are malleable after their initial formation; the hippocampal-cortical memory system can integrate new events. Furthermore, humans are able to voluntarily influence which items they integrate with each other using neurofeedback based on their brain activity patterns, which is another indication that memory networks are malleable. Together, I conclude that events in memory are represented as dynamic, malleable, hierarchical networks. These scientific insights add to the fundamental knowledge about episodic memory formation in the human brain.