Retinotopic coding links perception and memory in the brain, study finds

A recent study published in Natural neuroscience Retinotopic coding suggests that the cerebral cortex determines how information from the retina is processed.

Research: A retinotopic code forms connections between perceptual and memory systems. Image credit: Lia Koltryina / Shutterstock.com

Introduction

How does a sensory signal interact with brain neurons to form a memory? Despite research findings showing that the precision with which these memory and emotion-linked representations occur can be explained by patterned activities, this has long been a point of interest.

Conventionally, it is assumed that there is no common coding between these two types of cortical regions that govern perception and memory. Retinotopy describes how visual input from the retina is mapped to specific neurons, particularly those related to vision.

This retinotopic organization has been proposed to be replaced by abstract amodal coding in memory regions. The default mode network (DMN) is an example of how higher-order memory structures in the cortical cortex occur by signal propagation through visual areas.

To date, it is not clear how these two types of information can interact meaningfully based on different neural codes. Recent studies have shown that the DMN and similar higher-level cortical regions exhibit retinotopy in the form of population receptive fields (pRFs) evoked by visual cues that generate inverse response amplitudes.

What did the study show?

The present study suggests that retinotopy exists at all levels of cortical neural processing layers, thereby allowing perception and memory areas to be functionally linked with structured associations. This is based on functional magnetic resonance imaging (fMRI) findings that map visual pRFs.

Also, higher concentrations of negative pRFs were found in memory regions compared to cognitive regions. Thus, pRFs in memory areas (PMA) may bridge the gap between perceptual areas (SPAs) responsible for spatial memory and temporal memory areas by using retinotopic coding as a common substrate for the communication of these areas.

Similar locations in the visual field are represented by -pRFs attached to SPAs and PMAs. This may be due to signals originating in the SPAs and transmitted to the PMAs.

The amplitude of -pRF activity in PMAs was inversely related to +pRFs in SPAs during participants' recall tasks of familiar visual scenes such as kitchens. This reflects the phenomenon of bottom-up perceptual and top-down internal memory tasks, and supports the researchers' hypothesis.

The results also indicate spatially specific inhibitory interactions of these two types of regions that arise when structured by retinotopic neural coding. Antagonistic dynamics were stronger between +/-pRFs attached to the same region of the spatio-visual signal during both perception and memory tasks.

However, since these two tasks involve situations that involve maximally antagonistic dynamics of the visual and memory systems, the scientists tested activity levels during a task in which the two systems interact. This resulted in the same spatially specific inhibitory association of PRF activity in these regions, thus suggesting how these neurons normally interact in the natural setting.

The dynamic interplay between these regions is built upon the scaffolding of retinotopic coding, which determines the direction and magnitude of interactions between the recollection and perception systems within the neural network.

What are the implications?

Our findings challenge conventional views of brain organization…A shared retinotopic code between extrinsic (perceptual) and intrinsic (memory) regions of the brain structures their reciprocal activity..”

These connected regions form a mutually inhibitory set of paired memory or memory-linked and perceptual regions that display opposite responses in specific regions. Areas associated with this function show such responses during perception, where information is transported to the brain for processing, and recall when retrieving stored information.

Even at high levels of brain processing, visual inputs from the environment are clearly mapped to specific regions by inverse retinotopic coding. This may improve understanding of how the DMN functions during externally driven tasks.

Further research is needed to determine the extent to which retinotopic coding guides these types of interactions between perceptual and mnemonic neural regions and whether they are limited to scenes or extend to all areas in the cortex.