Brain Networks in Learning

Project Information

BNL

Grant agreement ID: 895465

DOI

10.3030/895465

Project closed

EC signature date 28 April 2020

Start date 1 March 2021

End date 28 February 2023

Funded under

• EXCELLENT SCIENCE - Marie Skłodowska-Curie Actions

Total cost

€ 165 085,44

EU contribution

€ 165 085,44

Coordinated by

IDRYMA TECHNOLOGIAS KAI EREVNAS
Greece

Periodic Reporting for period 1 - BNL (Brain Networks in Learning)

Reporting period: 2021-03-01 to 2023-02-28

Summary of the context and overall objectives of the project

Learning is a complex process that involves different stages. It is thought to create a trace in the brain that serves as our memory. How we learn and memorize information is a hugely important question since this mechanism supports all knowledge from experience. Many studies have shown which parts of the brain are involved in different stages of learning leading up to the formation of a memory trace. However, we know little about how the brain changes during the transition from the early to the later stages of learning. In this study, we investigated how the activity and the connections between different parts of the brain change during the transition from early to late stages of verbal learning using functional magnetic resonance imaging (fMRI). We found which specific areas of the brain become more active at different stages of learning to memorize a list of words, and we observed widespread changes in the connections between distinct parts of the brain. These changes were also related to how accurately the participants were able to learn. Some of the most important brain areas involved in this process were the right dorsomedial prefrontal cortex, the left temporal cortex, the posterior hippocampus, and the globus pallidus. We also found that lasting changes in the connections between different brain areas could be detected when comparing periods of rest before and after learning, particularly in areas such as the temporo-parietal junction, which is involved in internal narration. Overall, our study provides new insights into how the brain changes during verbal learning and helps us better understand this process.

Work performed from the beginning of the project to the end of the period covered by the report and main results achieved so far

We designed and executed an experiment where the participants had to read (distractor words) or memorize (target words) a list of words while lying inside an fMRI scanner. All words were drawn from a comprehensive curated corpus of the Greek language to match several statistical features of the words included in the auditory verbal learning task (VLT). We selected words to match the words in VLT for their number of syllables, length, frequency of occurrence in the corpus, frequency of their letter pairs, and syllable frequency. The words selected included commonly used words with little distinguishing characteristics.
We recorded brain activation while the subjects were instructed to read distractor words or to memorize the target list of words. We compared the changes in activation of different brain areas estimated using generative statistical models between the two conditions to uncover the set of brain areas associated with learning. Early stages of learning were associated with activation changes in visual and temporal cortical areas of the left hemisphere and the right dorsomedial prefrontal cortex. As the transition to the late phase was taking place, the areas with affected activation included additionally part of the left somatomotor cortex and the right amygdala. The later stages of learning were associated with changes in the activation of the right hemisphere’s precuneus/posterior cingulate cortex.
The changes in activations were accompanied by widespread changes in connectivity. Brain areas playing a key role in connectivity changes throughout the process of learning included both hemispheres pHIP and GP, parts of the right hemisphere’s thalamus and amygdala, and the right hemisphere’s OFC and dmPFC.
The most predictive areas for learning through their changes in activation included frontal areas (left lPFC and PFC, and right frontal operculum/insula), right pHIP and temporal cortex, as well as left post-central and parietal cortical areas. The most predictive areas for learning through their changes in connectivity included frontal cortical areas (left precentral ventral, frontal operculum/insula, and PFC and right dmPFC), left temporal and post-central cortical areas, right somatomotor cortex, both hemispheres precuneus/posterior cingulate cortex as well as both hemispheres’ pHIP, the left GP, the left NAc (core) and the left medial amygdala.
The areas participating in the presumed memory trace following the learning task include early visual and temporal pole cortical areas of both hemispheres, as well as the left inferior parietal lobule and angular gyrus.
Our results are largely consistent with the existing literature although they extend previous results on verbal learning in healthy adults in temporal precision of changes in activation during the learning processes and the maximal inclusion of brain areas in our dynamic connectivity models.
Considered jointly with existing literature on the functional role of the affected brain areas, this evidence suggests that the memory trace of the verbal learning task consists of the engagement of a set of areas related to a process of internal visual narration.
We participated extensively in dissemination activities with the participation of the general public including high-school students, scientific writing in a personal blog and the local media, teaching and student supervision. The project’s software code and data will be publicly released before the publication of our results. The project’s results will be submitted as a pre-print and an open-access peer-reviewed publication.

Progress beyond the state of the art and expected potential impact (including the socio-economic impact and the wider societal implications of the project so far)

In this research project our goal was to describe the formation and spread of the distributed memory trace throughout the brain’s networks during verbal learning. Very few studies have described the transition throughout the brain-wide states leading to the formation of the memory trace in terms of changes in activation and connectivity of both cortical and subcortical brain areas based on changes in activation and connectivity. Our use of principled inference based on generative statistical models provides an easily extendable analytical approach that can be adapted to study any dynamic process underlying flexible behavioural adaptations. In summary, both our results on verbal learning and our methodological innovations have prepared the stage for future studies to expand our understanding on learning and other behavioural adaptations.

activations-90dpi-8bit.png