Connectome cost conservation model of skill learning

Información del proyecto

COSMOS

Identificador del acuerdo de subvención: 101054909

DOI

10.3030/101054909

Fecha de la firma de la CE 15 Junio 2022

Fecha de inicio 1 Octubre 2022

Fecha de finalización 30 Septiembre 2027

Financiado con arreglo a

• European Research Council (ERC)

Coste total

€ 2 484 375,00

Aportación de la UE

€ 2 484 375,00

Coordinado por

TEL AVIV UNIVERSITY
Israel

Periodic Reporting for period 1 - COSMOS (Connectome cost conservation model of skill learning)

Período documentado: 2022-10-01 hasta 2025-03-31

Resumen del contexto y de los objetivos generales del proyecto

The COSMOS project—Connectome Cost Conservation Model of Skill Learning—aims to revolutionize our understanding of neuroplasticity by focusing on the brain’s connectome, a comprehensive map of its structural and functional connections. Neuroplasticity, the brain’s ability to reorganize itself in response to learning and experience, is a cornerstone of cognitive neuroscience. However, current research often examines isolated neural or behavioral elements, missing a global perspective on how the brain rewires itself as a network. COSMOS addresses this gap by leveraging cutting-edge neuroimaging techniques, including diffusion and functional MRI, to explore how learning reshapes the connectome.
The project investigates neuroplasticity across diverse skill domains—programming, rock climbing, trumpet playing, and martial arts (Brazilian Jiu-Jitsu)—which demand distinct cognitive, motor, and sensory skills. By analyzing both naive learners and skilled professionals, COSMOS seeks to uncover unique network patterns that support skill acquisition. Furthermore, the project aims to develop predictive models of learning outcomes based on individual brain connectomes, offering transformative insights for applications in education, rehabilitation, and even neurological disease management.
Beyond scientific exploration, COSMOS embodies an interdisciplinary approach, integrating neurobiology, psychology, computer science, and advanced data analytics. Its results are expected to redefine how neuroplasticity is understood and harnessed, providing empirical tools for assessing and enhancing learning in various contexts.

Trabajo realizado desde el comienzo del proyecto hasta el final del período abarcado por el informe y los principales resultados hasta la fecha

COSMOS is structured around five interlinked work packages (WPs), each designed to address specific research objectives:
1. WP1: Acquisition, Analysis Pipeline, and Data Management. The foundation for the project was established through a robust MRI protocol incorporating advanced diffusion and resting-state fMRI sequences. A comprehensive data management system was implemented, featuring:
o Dedicated servers for raw, processed, and analyzed data.
o A project dashboard and mobile app enabling seamless collaboration and data tracking.
o Tools for GDPR-compliant data sharing, ensuring accessibility for future research.
These advancements have streamlined data collection, analysis, and dissemination, supporting the project's long-term objectives.

2. WP2: The Skilled Brain Network. This phase mapped the connectomes of skilled professionals across four domains, identifying distinct neural network patterns associated with each skill. The findings confirmed that skills such as programming, rock climbing, and trumpet playing strengthen unique brain regions. A classification model trained on these connectomes achieved 87% accuracy in predicting an individual’s skill domain, a significant leap forward in neuroimaging research. These results lay the groundwork for understanding skill-specific brain adaptations.
See attached figure showing the professional brain. (A) Connectome nodes that uniquely identify each profession projected on the brain surface. (B) Classification model confusion matrix reaching 87% accuracy in profession classification in the test set.

3. WP3: Skill Learning and Plasticity. COSMOS has enrolled 631 naive participants, resulting in over 1,200 MRI scans documenting pre- and post-training brain states. Initial analyses show that learning strengthens intra-hemispheric connectivity while weakening inter-hemispheric links. These findings align with the hypothesis that learning optimizes network efficiency. Although more data are being collected, preliminary power analyses suggest that group sizes of 50 are sufficient for statistically robust insights across most tasks, with the exception of trumpet playing.

4. WP4: A Probabilistic Model of Plasticity. Using Monte Carlo simulations, this work package models brain connectivity changes as stochastic processes influenced by skill learning. While still in its early stages, this modeling approach promises to reveal how the brain balances efficiency and resource allocation during neuroplasticity.

5. WP5: Learning Predictions. Behavioral metrics for assessing learning success have been developed for programming and rock climbing but remain challenging for trumpet playing and martial arts. The focus is on creating machine learning models that predict learning outcomes from baseline connectomes and observed connectivity changes.

Avances que van más allá del estado de la técnica e impacto potencial esperado (incluida la repercusión socioeconómica y las implicaciones sociales más amplias del proyecto hasta la fecha)

COSMOS has made groundbreaking contributions to the field of neuroplasticity, advancing both methodology and theory:
1. Naturalistic fMRI: By using task-specific stimuli during scans, this approach enhances the relevance of functional connectivity analyses. For example, trumpet players showed distinct brain synchronization while watching performances of wind instruments, highlighting task-specific network engagement.
2. Time Delay Imaging (TDI): This innovative method incorporates axon diameter estimates into connectivity analyses, providing a physiologically meaningful framework for understanding brain network efficiency. TDI-based connectomes demonstrate stronger correlations with functional connectivity than conventional methods, paving the way for more accurate models of brain function.
These advances have broad implications, from refining our understanding of neuroplasticity to developing practical tools for predicting individual learning trajectories. The integration of these methodologies with the project's vast dataset ensures that COSMOS remains at the forefront of neuroscience research.

Impact and Future Directions
The COSMOS project has profound implications for multiple fields:
• Education and Training: Insights into how the brain adapts to skill learning can inform personalized learning strategies, optimizing outcomes in both academic and professional settings.
• Rehabilitation: Predictive models based on connectomes could guide interventions for patients recovering from brain injuries or undergoing therapy for neurodegenerative diseases.
• Basic Science: COSMOS bridges the gap between neurobiology and psychology, offering a unified framework for studying brain-behavior relationships.
To maximize the impact of its findings, the project has prioritized open science and data sharing. A dedicated database and accompanying app will allow researchers worldwide to access anonymized data, fostering collaboration and accelerating innovation. Publications detailing the project’s methodologies and results are already underway, ensuring timely dissemination of its contributions.

The professional brain