Skip to content Skip to menu
This website uses cookies to help us understand the way visitors use our website. We can't identify you with them and we don't share the data with anyone else. If you click Reject we will set a single cookie to remember your preference. Find out more in UCL's privacy notices.


Research theme


We use neuroimaging to ask:

  • How and where memories are formed and stored?
  • What are the mechanisms by which networks of neurons support the encoding, recall and imagination of events?
  • How do memories change over time and why do we forget?
  • Why do some people have a better memory than others?
  • How does memory change over the lifespan from babies, through childhood and old age?
  • How do disorders like epilepsy, dementia and limbic encephalitis affect memory?
  • How and why do brain structures involved in memory also support other functions like thinking about the future and spatial navigation?

Our goal is to provide:

A detailed understanding of how the brain supports memory so that we can help those who experience problems with recalling the past.



Losing the ability to remember our past experiences dislocates us from life’s timeline and precludes independent living.  Despite their obvious importance, it is unclear how these memories are enabled by the brain.

By conducting detailed neuroscientific studies of memories to establish precisely how they are built, how they are re-constructed during recollection and how neural representations of memories change over time, we will be in a better position to intervene as early as possible when memory ability starts to decline in the context of brain injury or disease.

Our research into the neural mechanisms that support memory aims to benefit patients with:

  • Epilepsy, to minimise damage to memory-relevant brain areas during surgery
  • Alzheimer’s disease and other forms of dementia, to detect decline as early as possible so that interventions can be made, and to maximise memory ability as the disease progresses
  • Post-Traumatic Stress Disorder, to understand flashbacks and suppressed memories so that their effects can be mitigated
  • Limbic encephalitis, and conditions affecting oxygen supply to the brain (e.g. stroke, respiratory arrest, heart attacks), to understand the nature of the memory loss and devise means to used preserved aspects of memory and cognition to offset some of the memory deficits

Our work also has implications beyond patients. For example, by helping healthy people to understand more about their memory, they can better maintain and possibly improve it throughout the lifespan. Similarly, our research may also help students to maximise their learning potential.

Recent findings

  • We discovered that licensed London taxi drivers have bigger hippocampi than control volunteers (e.g. Maguire et al., 2000; Woollett and Maguire, 2011 ). This work engaged not only scientists, but also the public and media world-wide.  It emphasised the capacity of the adult human brain for plasticity and life-long learning
  • We conducted the first systematic study showing that amnesic patients with focal bilateral hippocampal damage not only have difficulty recalling the past, they also cannot imagine the future or any type of fictitious scene or experience (e.g. Hassabis et al., 2007)
  • We spearheaded the use of decoding methodologies in fMRI studies of memory (e.g. Bonnici et al., 2012). This has allowed us to predict, from patterns of activity across hippocampal voxels, which particular memory a person is recalling, where they are in a virtual environment and how their autobiographical memory representations change over time.  This is a significant step in the search for the elusive engram, or memory trace, in the brain
  • We used this decoding approach to detect which parts of the hippocampus are still able to lay down memories in patients with temporal lobe epilepsy, which may be useful in helping to predict the effects of hippocampal surgery (e.g. Bonnici et al., 2013)
  • We have started to unravel the functions of specific parts of the hippocampus by showing, for example, the front – anterior – part seems to be particularly important for recalling past experiences and imagining the future (e.g. Zeidman and Maguire, 2016)
  • We have defined a genre of neuroimaging studies that use naturalistic paradigms, including virtual reality (e.g. Burgess et al., 2001; Spiers and Maguire, 2006).  These innovations allow us to study the brain in its natural context and address directly the sorts of real-world memory problems that patients encounter
  • Using MEG decoding we were able to show that learned sequences are encoded and performed using a ‘competitive queuing’ mechanism, in which items are pre-activated to a level corresponding to their order in the sequence, using a template for sequences in parahippocampal areas that generalises over different items used (Kornysheva et al., 2019)
  • Using fMRI and a new associative memory paradigm, we showed how multi-modal sequential experience is encoded into coherent events which are recalled holistically via hippocampal pattern completion of the event and neocortical reinstatement of all of its constituent parts (Horner et al., 2015)
  • We developed a neural-level computational model of how spatial context is encoded from egocentric sensory experience, stored as allocentric representations in medial temporal lobe, and reconstructed in egocentric imagery (Becker & Burgess, 2001; Byrne et al., 2007; Bicanski and Burgess, 2018). We verified several predictions, including the use of boundaries in forming spatial context (Bird et al., 2010) and the role of retrosplenial cortex in translating between ego- and allo-centric representations (Lambrey et al., 2011), and the discovery of the predicted cell types in rodents (boundary vector cells, Lever et al., 2009; object-vector cells Hoydal et al., 2019;  egocentric boundary vector cells Alexander et al., 2019)
  • We extended the above model to capture how the encoding and imagery of events is affected by negative emotional content in post-traumatic stress disorder (Brewin et al. 2010; Bisby and Burgess, 2017). We verified the predicted deficit in allocentric spatial memory in PTSD patients (Smith et al., 2015) and the predicted opposing effects of negative emotion on hippocampal versus amygdala processing of associative versus item information (Bisby et al., 2016).

Click here to find out more about MEMO, our big study into the origin of individual differences in memory.

Click here if you are an A-Level student and want to find resources relating to our London taxi driver work.

Click here to find detailed anatomical images relating to our tutorial on how to segment human hippocampal subfields from MR images.


Teams in this research area