Space and Memory
We study the neural mechanisms of spatial and episodic memory and model-based learning in healthy volunteers.
Many of our experiments are driven by, and feed into, our development of a comprehensive neural-level model of the different computations across brain regions that support spatial memory and imagery (Bicanski and Burgess, 2018). These address many of the key topics in cognitive neuroscience of memory:
Scanning at 7T
We have extended our previous finding on hippocampal ‘pattern completion’, (Horner et al., 2015), by using 7T fMRI to show that this is specifically associated with hippocampal region CA3 (Grande et al., 2019; in collaboration with Magdeburg DZNE). Current work characterises the grid firing patterns using 7T fMRI in more detail (in collaboration with Oxford WIN).
Emotion and memory
We extended the above model to show memory is affected by negative emotional content in post -traumatic stress disorder (PTSD; Bisby and Burgess, 2017). We identified the neural systems learning about spatial contextual fear in an ecologically valid setting (Suarez-Jimenez et al., 2017). Current work extends this by investigating the model’s prediction that deliberate and intrusive memories in PTSD reflect opposing patterns of (increased or decreased) activity in hippocampus versus amygdala.
Theta rhythmicity, Epilepsy and Schizophrenia
We used intra-cranial recording in patients with Epilepsy to show that increases in hippocampal theta power precede navigational trajectories and predict their length (Bush et al., 2018). With MEG, we found that medial temporal-medial prefrontal theta coupling is associated with spatial navigation (Kaplan et al., 2017). Current work is finding that this coupling is reduced in Schizophrenia (Adams et al., submitted) and that performing hippocampal-dependent tasks promotes the occurrence of inter-ictal ‘spikes’ in Epilepsy (Vivekananda et al., 2019).
Hippocampal processing, pathology and model-based learning
Our hippocampal-dependent spatial tasks (developed using fMRI) are being used for detection of impairments in Alzheimer’s (Howett et al., 2019) and Huntington’s (Harris et al., 2019) diseases. We also used one of these tasks to show that ‘model-based’ learning strategies correlate with hippocampal-dependent spatial processing in healthy volunteers, and are specifically impaired in temporal lobectomy patients (Vikbladh et al., 2019). We have developed MEG-decoding methods to identify the brain encoding of sequences (Korneysheva et al., 2019) which will be used to detect sequential ‘replay’ during model based planning.
- A large-scale neurocomputational model of spatial cognition integrating memory with vision Neural Networks DOI: 10.1016/j.neunet.2023.08.034
- Association Between False Memories and Delusions in Alzheimer Disease. JAMA Psychiatry DOI: 10.1001/jamapsychiatry.2023.1012
- View all publications by the Space and Memory team