Computational Modeling of Grid-Like Representations in the Entorhinal Cortex

Core Concepts

The activity of grid cells in the entorhinal cortex can give rise to hexadirectional modulations of population-level signals, which may be observed indirectly using functional magnetic resonance imaging (fMRI) in humans. This study quantitatively evaluates three hypotheses on how the activity of individual grid cells could translate into these macroscopic grid-like representations.

Abstract

This study aimed to quantitatively evaluate three hypotheses on the emergence of macroscopic grid-like representations in the human entorhinal cortex using computational modeling:

Conjunctive Grid by Head-Direction Cell Hypothesis:

Grid cells with head-direction tuning aligned to the grid axes can produce hexadirectional modulation of population activity.
The magnitude of hexasymmetry depends on the precision of head-direction tuning and the alignment of preferred head directions to grid axes.

Repetition Suppression Hypothesis:

Firing-rate adaptation in grid cells can lead to stronger repetition suppression for movements aligned vs. misaligned with grid axes.
This results in hexadirectional modulation of population activity, but the effect is sensitive to the subject's navigation pattern.

Structure-Function Mapping Hypothesis:

Clustered grid phase offsets of anatomically adjacent grid cells can produce hexadirectional modulation.
The magnitude and apparent preferred grid orientation depend on the subject's starting location relative to the grid fields.
With more realistic, weakly clustered grid phases, the hexasymmetry is reduced and largely explained by the hexasymmetry of the navigation path itself.

The simulations suggest that the conjunctive grid by head-direction cell hypothesis can produce the strongest and most robust hexadirectional modulation. In contrast, the structure-function mapping and repetition suppression hypotheses are more sensitive to the subject's navigation pattern. These findings highlight the importance of quantifying the biological properties of grid cells in humans to further elucidate the emergence of macroscopic grid-like representations.

Stats

"Grid cells may allow the navigating organism to perform vector computations and may thus constitute an essential neural substrate for different types of spatial navigation including path integration, though their exact functional role still remains unclear."
"Studies in rodents have demonstrated that the firing fields of anatomically adjacent grid cells do not only have similar spacing and orientation, but also a similar grid phase offset to a reference location."
"Previous studies presented several qualitatively different hypotheses on how the activity of single grid cells translates into a macroscopically visible hexadirectional fMRI signal."

Quotes

"Grid cells are neurons that activate whenever an animal or human traverses the vertices of a triangular grid tiling the entire environment into equilateral triangles."
"To provide possible explanations for the emergence of macroscopic grid-like representations, previous studies presented several qualitatively different hypotheses on how the activity of single grid cells translates into a macroscopically visible hexadirectional fMRI signal."
"Our simulations suggest that hexadirectional modulation is best explained by the conjunctive grid by head-direction cell hypothesis, which can produce the strongest and most robust hexasymmetry."

Key Insights Distilled From

Quantitative modeling of the emergence of macroscopic grid-like representations

by Bin Khalid,I... at www.biorxiv.org 12-20-2022

https://www.biorxiv.org/content/10.1101/2022.12.20.521210v2

Deeper Inquiries

What other neural mechanisms or computational principles beyond the three hypotheses considered in this study could potentially contribute to the emergence of macroscopic grid-like representations

In addition to the three hypotheses explored in the study, several other neural mechanisms or computational principles could potentially contribute to the emergence of macroscopic grid-like representations. One such mechanism could involve interactions between grid cells and other types of spatially tuned neurons, such as place cells or border cells. These interactions could lead to the formation of higher-order spatial representations that exhibit grid-like properties. Additionally, the influence of oscillatory rhythms, such as theta oscillations, on the activity of grid cells and their network dynamics could play a role in shaping macroscopic grid-like representations. Furthermore, the impact of neuromodulatory systems, such as the cholinergic or dopaminergic systems, on the activity of grid cells and their coordination within the entorhinal cortex could also contribute to the emergence of grid-like patterns at a macroscopic level.

How might the findings from this computational modeling study be used to design future empirical studies that could provide more definitive evidence for or against the proposed hypotheses

The findings from this computational modeling study can be leveraged to design future empirical studies that could provide more definitive evidence for or against the proposed hypotheses. One approach could involve conducting neuroimaging studies, such as functional magnetic resonance imaging (fMRI) or electroencephalography (EEG), in conjunction with behavioral tasks that manipulate the subjects' navigation patterns. By comparing the neural activity patterns observed in these studies with the predictions generated by the computational models, researchers can test the validity of the different hypotheses. Additionally, conducting single-cell recordings in animal models or human patients, when feasible, could provide direct evidence of the neural mechanisms underlying macroscopic grid-like representations. By combining these empirical approaches with computational modeling, researchers can gain a more comprehensive understanding of the neural basis of grid-like representations in the brain.

Given the importance of grid cells for spatial navigation, how might a deeper understanding of the mechanisms underlying macroscopic grid-like representations inform our broader understanding of the functional role of grid cells in cognition and behavior

A deeper understanding of the mechanisms underlying macroscopic grid-like representations can significantly inform our broader understanding of the functional role of grid cells in cognition and behavior. By elucidating how grid cells interact with other neural populations and how their activity is modulated by various factors, such as movement direction, environmental cues, and neuromodulatory signals, researchers can uncover the precise computational functions of grid cells in spatial navigation and memory processes. This knowledge can have implications for various fields, including neuroscience, psychology, and artificial intelligence. For example, insights into the functional role of grid cells could inspire the development of more biologically inspired navigation algorithms for autonomous robots or inform the design of spatial memory tasks in cognitive psychology experiments. Ultimately, a deeper understanding of macroscopic grid-like representations can provide valuable insights into the neural mechanisms underlying complex cognitive functions and behaviors.