Decoding cerebro-spinal signatures of human behavior: Application to motor sequence learning

TY - JOUR

T1 - Decoding cerebro-spinal signatures of human behavior: Application to motor sequence learning

AU - Kinany, N

AU - Khatibi, A

AU - Lungu, O

AU - Finsterbusch, J

AU - Büchel, C

AU - Marchand-Pauvert, V

AU - Van De Ville, D

AU - Vahdat, S

AU - Doyon, J

N1 - Copyright © 2023 The Author(s). Published by Elsevier Inc. All rights reserved.

PY - 2023/7/15

Y1 - 2023/7/15

N2 - Mapping the neural patterns that drive human behavior is a key challenge in neuroscience. Even the simplest of our everyday actions stem from the dynamic and complex interplay of multiple neural structures across the central nervous system (CNS). Yet, most neuroimaging research has focused on investigating cerebral mechanisms, while the way the spinal cord accompanies the brain in shaping human behavior has been largely overlooked. Although the recent advent of functional magnetic resonance imaging (fMRI) sequences that can simultaneously target the brain and spinal cord has opened up new avenues for studying these mechanisms at multiple levels of the CNS, research to date has been limited to inferential univariate techniques that cannot fully unveil the intricacies of the underlying neural states. To address this, we propose to go beyond traditional analyses and instead use a data-driven multivariate approach leveraging the dynamic content of cerebro-spinal signals using innovation-driven coactivation patterns (iCAPs). We demonstrate the relevance of this approach in a simultaneous brain-spinal cord fMRI dataset acquired during motor sequence learning (MSL), to highlight how large-scale CNS plasticity underpins rapid improvements in early skill acquisition and slower consolidation after extended practice. Specifically, we uncovered cortical, subcortical and spinal functional networks, which were used to decode the different stages of learning with a high accuracy and, thus, delineate meaningful cerebro-spinal signatures of learning progression. Our results provide compelling evidence that the dynamics of neural signals, paired with a data-driven approach, can be used to disentangle the modular organization of the CNS. While we outline the potential of this framework to probe the neural correlates of motor learning, its versatility makes it broadly applicable to explore the functioning of cerebro-spinal networks in other experimental or pathological conditions.

AB - Mapping the neural patterns that drive human behavior is a key challenge in neuroscience. Even the simplest of our everyday actions stem from the dynamic and complex interplay of multiple neural structures across the central nervous system (CNS). Yet, most neuroimaging research has focused on investigating cerebral mechanisms, while the way the spinal cord accompanies the brain in shaping human behavior has been largely overlooked. Although the recent advent of functional magnetic resonance imaging (fMRI) sequences that can simultaneously target the brain and spinal cord has opened up new avenues for studying these mechanisms at multiple levels of the CNS, research to date has been limited to inferential univariate techniques that cannot fully unveil the intricacies of the underlying neural states. To address this, we propose to go beyond traditional analyses and instead use a data-driven multivariate approach leveraging the dynamic content of cerebro-spinal signals using innovation-driven coactivation patterns (iCAPs). We demonstrate the relevance of this approach in a simultaneous brain-spinal cord fMRI dataset acquired during motor sequence learning (MSL), to highlight how large-scale CNS plasticity underpins rapid improvements in early skill acquisition and slower consolidation after extended practice. Specifically, we uncovered cortical, subcortical and spinal functional networks, which were used to decode the different stages of learning with a high accuracy and, thus, delineate meaningful cerebro-spinal signatures of learning progression. Our results provide compelling evidence that the dynamics of neural signals, paired with a data-driven approach, can be used to disentangle the modular organization of the CNS. While we outline the potential of this framework to probe the neural correlates of motor learning, its versatility makes it broadly applicable to explore the functioning of cerebro-spinal networks in other experimental or pathological conditions.

KW - Humans

KW - Brain/diagnostic imaging

KW - Spinal Cord/diagnostic imaging

KW - Learning/physiology

KW - Magnetic Resonance Imaging/methods

KW - Neuroimaging

U2 - 10.1016/j.neuroimage.2023.120174

DO - 10.1016/j.neuroimage.2023.120174

M3 - SCORING: Journal article

C2 - 37201642

VL - 275

SP - 120174

JO - NEUROIMAGE

JF - NEUROIMAGE

SN - 1053-8119

ER -