The lumbosacral spinal cord contains neural circuits crucial for locomotion, organized into rostrocaudal levels with distinct somatosensory and motor neuron pools that project to and from the muscles of the lower limbs. However, the specific spinal levels that innervate each muscle and the locations of neuron pools vary significantly between individuals, presenting challenges for targeted therapies and neurosurgical interventions aimed at restoring locomotion. Non-invasive approaches to functionally map the segmental distribution of muscle innervation-orprojectome-are therefore essential. Here, we developed a pipeline dedicated to record blood oxygenation level dependent (BOLD) signals in the lumbosacral spinal cord using functional magnetic resonance imaging (fMRI). We assessed spinal activity across different conditions targeting the extensor/flexor muscles of the right leg (ankle, knee, and hip) in 12 healthy participants. To enhance clinical relevance, we included not only active movements but also two conditions that did not rely on participants' performance: passive stretches and muscle-specific tendon vibration, which activates proprioceptive afferents of the vibrated muscle. BOLD activity patterns were primarily located on the side ipsilateral to the movement, stretch, or vibration, both at the group and participant levels, indicating the BOLD activity being associated with the projectome. The fMRI-derived rostrocaudal BOLD activity patterns exhibited mixed alignment with expected innervation maps from invasive studies, varying by muscle and condition. While some muscles and conditions matched well across studies, others did not. Significant variability among individual participants underscores the need for personalized mapping of projections for targeted therapies and neurosurgical interventions. To support the interpretation of BOLD activity patterns, we developed a decision tree-based framework that combines reconstruction of neural structures from high-resolution anatomical MRI datasets and muscle-specific fMRI activity to produce personalized projectomes. Our findings provide a valuable proof-of-concept for the potential of fMRI to map the lumbosacral spinal cord's functional organization, while shedding light on challenges associated with this endeavor.

In recent years, there has been a growing interest in investigating spontaneous brain activity, with a newfound focus on understanding the resting state activity of the spinal cord due to its importance in sensorimotor interactions. Previous studies have predominantly concentrated on the upper regions of the spinal cord. In particular one study explored functional dynamics in the cervical region utilizing the SpiCiCAP framework to reveal consistent neural activity patterns across individuals, aligning with established neuroanatomical features. The current study seeks to explore, for the first time, the spatial and temporal features of functional dynamics in the lower portion of the spinal cord (specifically, the lumbosacral region) using the same framework. Through the analysis, two distinct sets of components with varying levels of detail are identified, demonstrating generalizability across subjects. Overall, the BOLD signals cluster into regions corresponding to predefined spinal levels. However, as granularity increases, co-activation patterns within each level begin to segregate. Additionally, the study demonstrates that resting state activity within each spinal level exhibits a higher degree of coupling compared to other levels, shedding light on the intricacies of spinal cord functionality during periods of rest.

In the past decade, exploration of spontaneous blood-oxygen-level-dependent (BOLD) signal fluctuations has expanded beyond the brain to include the spinal cord. While most studies have predominantly focused on the cervical region, the lumbosacral segments play a crucial role in motor control and sensory processing of the lower limbs. Addressing this gap, the aims of the current study were twofold: first, confirming the presence and nature of organized spontaneous BOLD signals in the human lumbosacral spinal cord; second, systematically assessing the impact of various denoising strategies on signal quality and functional connectivity (FC) patterns. Given the susceptibility of spinal cord functional magnetic resonance imaging (fMRI) to noise, this step is pivotal to ensure the robustness of intrinsic FC. Our findings uncovered bilateral FC between the ventral and dorsal horns. Importantly, these patterns were consistently observed across denoising methods and demonstrating fair to excellent split-half temporal stability. Importantly, the evaluation of diverse denoising strategies highlighted the efficacy of physiological noise modeling (PNM)-based pipelines in cleaning the signal while preserving the strength of connectivity estimates. Together, our results provide evidence of robust FC patterns in the lumbosacral spinal cord, thereby paving the way for future studies probing caudal spinal activity.

The cerebellum is thought to help detect and correct errors between intended and executed commands 1 , 2 and is critical for social behaviours, cognition and emotion 3 – 6 . Computations for motor control must be performed quickly to correct errors in real time and should be sensitive to small differences between patterns for fine error correction while being resilient to noise 7 . Influential theories of cerebellar information processing have largely assumed random network connectivity, which increases the encoding capacity of the network’s first layer 8 – 13 . However, maximizing encoding capacity reduces the resilience to noise 7 . To understand how neuronal circuits address this fundamental trade-off, we mapped the feedforward connectivity in the mouse cerebellar cortex using automated large-scale transmission electron microscopy and convolutional neural network-based image segmentation. We found that both the input and output layers of the circuit exhibit redundant and selective connectivity motifs, which contrast with prevailing models. Numerical simulations suggest that these redundant, non-random connectivity motifs increase the resilience to noise at a negligible cost to the overall encoding capacity. This work reveals how neuronal network structure can support a trade-off between encoding capacity and redundancy, unveiling principles of biological network architecture with implications for the design of artificial neural networks.

Understanding the organizational principles of human brain activity at the systems level remains a major challenge in network neuroscience. Here, we introduce a fully data-driven approach based on graph learning to extract meaningful repeating network patterns from regionally-averaged timecourses. We use the Graph Laplacian Mixture Model (GLMM), a generative model that treats functional data as a collection of signals expressed on multiple underlying graphs. By exploiting covariance between activity of brain regions, these graphs can be learned without resorting to structural information. To validate the proposed technique, we first apply it to task fMRI with a known experimental paradigm. The probability of each graph to occur at each time-point is found to be consistent with the task timing, while the spatial patterns associated to each epoch of the task are in line with previously established activation patterns using classical regression analysis. We further on apply the technique to resting state data, which leads to extracted graphs that correspond to well-known brain functional activation patterns. The GLMM allows to learn graphs entirely from the functional activity that, in practice, turn out to reveal high degrees of similarity to the structural connectome. The Default Mode Network (DMN) is always captured by the algorithm in the different tasks and resting state data. Therefore, we compare the states corresponding to this network within themselves and with structure. Overall, this method allows us to infer relevant functional brain networks without the need of structural connectome information. Moreover, we overcome the limitations of windowing the time sequences by feeding the GLMM with the whole functional signal and neglecting the focus on sub-portions of the signals.

Summary The cerebellum is thought to detect and correct errors between intended and executed commands 1–3 and is critical for social behaviors, cognition and emotion 4–6 . Computations for motor control must be performed quickly to correct errors in real time and should be sensitive to small differences between patterns for fine error correction while being resilient to noise 7 . Influential theories of cerebellar information processing have largely assumed random network connectivity, which increases the encoding capacity of the network’s first layer 8–13 . However, maximizing encoding capacity reduces resiliency to noise 7 . To understand how neuronal circuits address this fundamental tradeoff, we mapped the feedforward connectivity in the mouse cerebellar cortex using automated large-scale transmission electron microscopy (EM) and convolutional neural network-based image segmentation. We found that both the input and output layers of the circuit exhibit redundant and selective connectivity motifs, which contrast with prevailing models. Numerical simulations suggest these redundant, non-random connectivity motifs increase discriminability of similar input patterns at a minimal cost to the network’s overall encoding capacity. This work reveals how neuronal network structure can balance encoding capacity and redundancy, unveiling principles of biological network architecture with implications for artificial neural network design.

Motivation: Spinal cord cross-sectional area (CSA) is an important biomarker for neurodegenerative and traumatic diseases. However, CSA measurements vary across MRI contrasts and imaging protocols, limiting its use in multi-center studies. Goal(s): The goal is to evaluate CSA variability using a novel contrast-agnostic segmentation method. Approach: We compared this method to the Spinal Cord Toolbox's DeepSeg, analyzing CSA across different sites, and MRI vendors. Additionally, we compared the segmentations in diverse datasets and pathologies. Results: The contrast-agnostic segmentation showed lower CSA variability, and superior performance in most cases, except for intramedullary cord compression, where the Spinal Cord Toolbox's DeepSeg was more accurate. Impact: The contrast-agnostic method yields reliable spinal cord CSA measurements, independent of MRI contrasts and vendors. This, combined with a soft segmentation output, can potentially detect subtle spinal cord atrophy in prospective multi-center cohorts.

In recent years, while the exploration of spontaneous brain activity has shifted from static methods (i.e., examining average connectivity patterns over the entire run) towards dynamic approaches (i.e., accounting for the non-stationarity of resting-state fluctuations), the non-stationary nature of intrinsic activity in the spinal cord has seldom been studied. Here, we propose to probe time-varying functional connectivity patterns in the spinal cord using a sliding-window correlation approach, as commonly employed in the brain. Our results suggest the potential of this approach to unravel ventral-ventral and dorsal-dorsal correlation patterns, while also emphasizing their time-varying nature.