Current Research Directions

Introduction

The resulting profound memory impairment following patient H.M.'s bilateral medial temporal lobe (MTL) removal marked the beginning of a new age in memory research. Specifically, it was this loss that made it clear to researchers that the MTL was necessary for long-term memory formation. However, despite the fact that the MTL and, specifically, the hippocampus, is arguably the most studied structure in the brain, the specific contributions of individual structures within the MTL and how they interact still remain to be well described.

Item and Relational Encoding

Based both on behavioral data suggesting that human recognition memory may be supported by at least two distinct processes and neuroscience data from both single unit physiology and lesion experiments suggesting that subregions in the MTL may contribute differentially to memory, our recent work has focused on elucidating the specific contributions of different MTL subregions (for review, see Davachi, 2006). Specifically, we have combined the hippocampal relational memory theory first proposed by Cohen and Eichenbaum (1993) with emerging knowledge that MTL cortical regions (in particular, the perirhinal and parahippocampal cortices) that input to the hippocampus may support the representation and/or maintenance of domain-specific information. Recent papers that have contributed to this working model are Davachi et al (2003), Staresina and Davachi (2006, 2008), Awipi and Davachi (2008) and  Dougal et al (2007).

See our working model below (taken from Davachi, 2006):

High resolution brain imaging

Myriad computational models of hippocampal function have proposed ways in which the hippocampus might be able to support both the encoding of new information and the retrieval of memory. We are using high resolution imaging to look for pattern separation and completion signals in the human hippocampus.

Figure above: High Resolution Map showing parahippocampal cortical activation during a memory retrieval task

Memory consolidation

A fundamental question regarding memory is how does memory change with time? and what are the operations that influence the longevity or durability of a memory. We have developed a behavioral paradigm that decreases the forgetting rate of associative information (Litman and Davachi, 2008) and are using this paradigm to ask what kinds of brain changes support enhanced memory durability.

Event Memory

What is an episode? Does the way in which we parse events over the course of our days influence the organization and long-term memory? Motivated by research examining the effects of event perception on the short-term accessibility of information, we are asking how event segmentation influences long-term memory.

Social and Emotional Memory

The majority of our previous work has focused on memory for neutral, non-social information (for e.g. words, objects, scenes).  An important new focus in the lab is on using the knowledge we've gained about memory formation and retrieval to ask a more difficult question - how and when are these processes employed in the real-world?