Recent advances in imaging technology have enabled functional magnetic resonance imaging (fMRI) to extend beyond brain research, allowing non-invasive exploration of human spinal cord function. This thesis harnesses these technological developments to investigate the functional organization of the spinal cord, characterize individual connectivity patterns, and elucidate the mechanisms underlying brain-spine interactions.

This work makes three principal contributions. To begin with, it delivers the first systematic evidence of functional connectivity in the lumbosacral spinal cord, in both healthy individuals and patients with spinal cord injury, representing an incremental advance in spinal cord fMRI beyond the traditionally studied cervical region. Next, it introduces the novel concept of a "spine-print", showing that spinal functional connectivity patterns are individual-specific and stable across time, thus offering a potential biomarker for longitudinal monitoring. Finally, it applies computational modeling to brain-spine interactions, using Dynamic Causal Modeling (DCM) for generative, mechanistic inference and Graph Laplacian Mixture Models (GLMM) within Graph Signal Processing (GSP) for data-driven network discovery.

By integrating descriptive, individual, and mechanistic methodologies, this work advances the field of spinal cord fMRI while providing novel methodological and conceptual frameworks for studying central nervous system function in health and disease. The thesis represents a methodological breakthrough and advances the study of brain-spine interactions toward a comprehensive, systems-level perspective of the entire central nervous system.