Neural mechanisms of memory processes

The overall goal of the cognitive neuroscience of memory is to provide a link between the cellular mechanisms of learning and memory as they are investigated in animals, and human neuroimaging studies on complex memory phenomena. This goal can only be reached by an integration of various methods ranging from in vitro and in vivo electrophysiology in animals to intracranial EEG recordings in patient populations and to noninvasive EEG and fMRI in healthy human participants. Moreover, approaching the mechanisms of memory processes beyond mere neural correlates requires one to test specific models, which can be either derived from animal data, computational neuroscience or cognitive psychology.

In my talk, I present evidence from several sources that a late step of long-term memory formation, memory consolidation, relies on a reactivation of stimulus-specific neural activity patterns. Rodent recordings have provided convincing evidence that sequences of action potentials in spatially-selective hippocampal pyramidal cells (“place cells”) re-occur spontaneously during subsequent quiet resting state and sleep. However, declarative memory for individual items is difficult, if not impossible, to test in rodents. In an intracranial EEG study in presurgical epilepsy patients, we showed that stimulus-specific activity patterns are replayed in humans as well. Furthermore, in a related simultaneous EEG/fMRI study in healthy participants, we found that reactivation is actually predictive of memory accuracy for individual items.

Taken together, these studies provide converging evidence that memory consolidation relies on a reactivation of previously acquired memory traces. More generally, they illustrate how an interdisciplinary combination of methods across species allows one to test specific models of memory processes.