Department of Radiology and Medical Informatics, University of Geneva, Geneva 1211, Switzerland; Neuro-X Institute, École Polytechnique Fédérale de Lausanne (EPFL), Geneva 1202, Switzerland.
Center of Precision Rehabilitation for Spinal Pain, School of Sport, Exercise and Rehabilitation Sciences, College of Life and Environmental Sciences, University of Birmingham, Edgbaston B15 2TT, United Kingdom.
Neuroimage. 2023 Jul 15;275:120174. doi: 10.1016/j.neuroimage.2023.120174. Epub 2023 May 16.
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.
绘制驱动人类行为的神经模式是神经科学的一个关键挑战。即使是我们日常生活中最简单的动作，也源于中枢神经系统（CNS）中多个神经结构的动态和复杂相互作用。然而，大多数神经影像学研究都集中在研究大脑机制，而脊髓在塑造人类行为方面与大脑的协同作用在很大程度上被忽视了。尽管最近出现了可以同时针对大脑和脊髓的功能磁共振成像（fMRI）序列，为在 CNS 的多个层面研究这些机制开辟了新的途径，但迄今为止的研究仅限于推断性的单变量技术，无法充分揭示潜在神经状态的复杂性。为了解决这个问题，我们建议超越传统分析，而是使用基于创新的共激活模式（iCAP）的基于数据的多元方法来利用脑脊髓信号的动态内容。我们在一项同时采集大脑-脊髓 fMRI 数据的运动序列学习（MSL）中证明了这种方法的相关性，以强调大规模 CNS 可塑性如何支持早期技能获取的快速提高以及延长练习后的缓慢巩固。具体来说，我们发现了皮质、皮质下和脊髓功能网络，这些网络用于以高精度解码学习的不同阶段，从而描绘出有意义的学习进展的脑脊髓特征。我们的研究结果提供了令人信服的证据，表明神经信号的动态与基于数据的方法相结合，可以用于解开 CNS 的模块化组织。虽然我们概述了该框架探测运动学习神经相关性的潜力，但它的多功能性使其广泛适用于探索其他实验或病理条件下脑脊髓网络的功能。
