Project Details
Projekt Print View

Reactivation in cortical and subcortical systems during consolidation of declarative memory - investigations in wakefulness and sleep

Subject Area General, Cognitive and Mathematical Psychology
Term from 2008 to 2015
Project identifier Deutsche Forschungsgemeinschaft (DFG) - Project number 65348950
 
Final Report Year 2017

Final Report Abstract

During the current reporting period (2011-2016), the work of my group had a double focus, one on neuroscientific theory, the other on methodological development. They were represented in the composition of the group, which consists currently of four PhD students, two psychologists/neuroscientists and two engineers. The overarching goal of the present project was to detect memory reactivation during sleep. To be able to pinpoint specific memories, we decided to use the electroencephalogram (EEG). Because there were no readily available methods to detect memory-related patterns of activity in EEG, one goal of the present application was to develop the tools required for detection of memory reactivation during sleep. The methods part of the project focussed on understanding the distribution of the outcomes of machine learning algorithms (classification accuracies), how these can be tested for significance, and on developing applicable algorithms to use MVPA for continuous EEG data. The progress of methods development was made possible by the close interaction of the psychologists, who provided experimental EEG data and insight into the psycho-physiological generation of these signals, and the engineers, who had a strong background in machine learning and its application to electrical signals. These experiments led to the publication of a number of high-impact publications (e.g. in Nature Communications). The second major line of research was centred on the related question where memories are actually stored in the brain. In a series of fMRI experiments, we tried to locate the areas of the brain that are involved when a new memory trace is formed. In three separate experiments, using five different declarative memory tasks, we found that the hippocampus is consistently involved only during early encoding of a memory, whereas the posterior parietal cortex, particularly the precuneus, becomes more active over repeated encounters with the new material. Moreover, precuneus activity remains stable over days and is also related to behavioural performance. It shows therefore all necessary signs of a memory representation. Although our first results on this topic were already submitted several years ago, only now the community starts to accept the finding that the parietal cortex plays an important role in semantic long-term memory. Therefore, our first publication with such a finding appeared in PNAS only last year. Apart from these central projects, a number of smaller side-projects and collaborations were pursued. For example a number of experiments on the use of transcranial magnetic stimulation (TMS) were performed. Another series of experiments tried to find behavioural signs for an effect of sleep on the semanticization of episodic memories, on the extraction of gist and on problem solving. All of these phenomena have been hypothesized in recent years and some initial evidence exists. However, we were unable to find reliable evidence that sleep actually transforms the contents of memory, as it would be required for semanticization or problem solving to occur. Publication of these null findings is planned in the near future.

Publications

 
 

Additional Information

Textvergrößerung und Kontrastanpassung