Summary: Researchers have created a design that explains how spot cells in the hippocampus both store spatial and acute memories. Cortical cells and network cells in the entorhinal brain work together to form a scaffold that connects memories to particular neural patterns.
This model examines phenomena like progressive memory loss and the recall technique of storage palaces, as well as natural memory systems. Unlike older versions, it avoids a “memory cliff”, where adding new thoughts erases older ones, and preferably allows for consecutive and flexible understand. It combines decades of synaptic circuit research with insights into the coexistence of spatial and sequential memories.
Important Facts:
- Memory Mechanism: Place and network cells form a gallows linking geographical and acute thoughts.
- Memory Accuracy: The concept explains steady memory degradation and sequential understand, avoiding “memory rocks”.
- Memory Palaces: It shows how relating thoughts to geographical settings improves understand.
Origin: MIT
Researchers discovered brain cells, which store thoughts from a particular location, about 50 years ago. These cells also have a significant role in storing acute memories, or acute memories.
Although the mechanism by which place cells express geographical storage has been well-characterized, how they encode acute memories has remained a mystery.
How those location cells may be recruited to type acute memories, even when there is no geographic component, is explained by a new model developed by MIT researchers. In this design, location cell serve as a framework for linking thoughts as a linked line, along with network cells found in the entorhinal brain.
” This design is a first-draft design of the entorhinal-hippocampal acute memory circuit. It’s a solid base from which to learn about the nature of sequential memory. That’s the point I’m really excited about”, says Ila Fiete, a teacher of mind and mental science at MIT, a part of MIT’s McGovern Institute for Brain Research, and the top artist of the new research.
The model accurately replicates a number of characteristics of biological memory systems, including the large storage capacity, gradual degeneration of older memories, and the capacity of people who compete in memory competitions to store a lot of data in “memory palaces.”
MIT Research Scientist Sarthak Chandra and Sugandha Sharma PhD ‘ 24 are the lead authors of the study, which appears today in , Nature. Rishidev Chaudhuri, an assistant professor at the University of California at Davis, is also an author of the paper.
An index of memories
Place cells in the hippocampus collaborate closely with grid cells, a unique type of neuron that is arranged geometrically in a regular pattern of repeating triangles to encode spatial memory. A lattice of triangles representing a physical space is created by a population of grid cells.
In addition to helping us recall places where we’ve been, these hippocampal-entorhinal circuits also help us navigate new locations. These circuits are also crucial for creating episodic memories, which may have a spatial component but primarily consist of things like how you celebrated your last birthday or what you had for lunch yesterday, according to human patients.
According to Fiete,” the same hippocampal and entorhinal circuits are used for general episodic memory as well, not just for spatial memory.” What is the relationship between spatial and episodic memory that causes them to reside in the same circuit, you may ask?
This overlap in function has been attributed to two hypotheses, according to the proposal. One is that the circuit has a specialized space memory system because survival depends on it to remember where food was stored or where predators were seen. According to this theory, this circuit encodes spatial memory as a byproduct.
Because location is one of the components of many episodic memories, an alternative theory suggests that the circuit is not only designed to store episodic memories but also encodes spatial memories.
Fiete and her colleagues in this study made a third suggestion: that the peculiar tiling structure of grid cells and their interactions with the hippocampus are equally crucial for both episodic and spatial types of memory.
They used computational models that her lab has been developing over the past ten years to mimic how grid cells encode spatial information to create their new model.
It seemed like the right time to try to understand the interactions between the grid cells and the larger circuit, which includes the hippocampus, according to Fiete.
The researchers developed the idea that grid cells that interact with hippocampal cells can serve as a storage medium for episodic or spatial memory. A “well” is defined by each grid activation pattern, and these wells are regularly spaced out.
Each well serves as a pointer to a specific memory, which is stored in the synapses between the hippocampus and the sensory cortex, but they don’t store the content of a particular memory.
When the memory is later triggered by fragmentary pieces, grid and hippocampal cell interactions force the circuit state into the closest well, and the state at the bottom of the well connects to the appropriate area of the sensory cortex to fill in the details of the memory. The sensory cortex can store a lot of memory because it is much larger than the hippocampus.
” Conceptually, we can think about the hippocampus as a pointer network. It’s like an index that can be pattern-completed from a partial input, and that index then points toward sensory cortex, where those inputs were experienced in the first place”, Fiete says.
” The scaffold doesn’t contain the content, it only contains this index of abstract scaffold states”.
Additionally, sequence events can be linked to one another: Each well in the grid cell-hippocampal network effectively stores the information needed to activate the next well, allowing memories to be recalled in the appropriate sequence.
Modeling memory cliffs and palaces
The researchers ‘ new model, which is based on Hopfield networks, a type of neural network that can store and recall patterns, accurately replicates a number of memory-related phenomena.
Hopfield networks provide insight into how memories can be created by strengthening neuronal connections, but they don’t completely model how biological memory operates.
Every memory is meticulously recalled in Hopfield models until the required capacity is reached. At that point, no new memories can form, and worse, attempting to add more memories erases all prior ones.
This “memory cliff” doesn’t accurately represent what occurs in the biological brain, which gradually forgets the details of previously forgotten memories while adding fresh ones on top.
The new MIT model incorporates findings from decades of rodent recordings of grid and hippocampal cells as they forage and explore various environments. Additionally, it provides an explanation of the underlying mechanisms of a memory-building technique known as a memory palace.
One of the tasks in memory competitions is to learn the shuffled sequence of cards from one or more card decks. They typically do this by tying each card to a specific location in a memory palace, such as a childhood home or other familiar setting. When they need to recall the cards, they mentally wander through the house, imagining each card in its place as they go along.
Counterintuitively, adding the memory burden of associating cards with locations makes recall stronger and more reliable.
The MIT team’s computational model was able to carry out these tasks very well, suggesting that memory palaces take advantage of the memory circuit’s own strategy of associating inputs with a scaffold in the hippocampus, but one level down: Long-acquired memories that were previously acquired can now be used as a scaffold for new memories. This enables the preservation and recall of many more items in a series than would otherwise be possible.
The researchers now intend to expand on their model to find out how episodic memories can be transformed into cortical” semantic” memories, or the memory of facts that are dissociated from the specific context in which they were acquired ( for example, Paris is the capital of France ), how episodes are defined, and how brain-like memory models can be incorporated into contemporary machine learning.
Funding:
The research was funded by the U. S. Office of Naval Research, the National Science Foundation under the Robust Intelligence program, the ARO-MURI award, the Simons Foundation, and the K. Lisa Yang ICoN Center.
About this news about memory research
Author: Sarah McDonnell
Source: MIT
Contact: Sarah McDonnell – MIT
Image: The image is credited to Neuroscience News
Original Research: Closed access.
Ila Fiete and al., “spatial scaffolds in the hippocampus have episodic and associative memory..” Nature
Abstract
spatial scaffolds in the hippocampus have episodic and associative memory.
The brain’s hippocampal circuits enable the storage of sequential episodic memories and the creation of spatial maps for navigation.
Although there have been improvements made to the hippocampus’s ability to model spatial representations, we do not have reliable models of its contribution to episodic memory.
Here we present a neocortical–entorhinal–hippocampal network model that implements a high-capacity general associative memory, spatial memory and episodic memory.
The circuit, which we call vector hippocampal scaffolded heteroassociative memory ( Vector-HaSH), avoids the memory cliff of the previous memory model and instead exhibits a graceful trade-off between number of stored items and recall detail by factoring content storage from the dynamics of generating error-correcting stable states.
Even non-spatial episodic memory can be created using a pre-structured internal scaffold built on grid cell states because it abstracts the chaining problem into learning low-dimensional transitions.
Vector-HaSH reproduces several hippocampal experiments on spatial mapping and context-based representations, and provides a circuit model of the’ memory palaces’ used by memory athletes.
This work provides a comprehensive understanding of the hippocampus’s spatial mapping and associative and episodic memory functions.
