Distributed harmonic patterns of structure-function dependence orchestrate human consciousness

Introduction

Luppi and colleagues frame the study around a central neuroscientific and clinical challenge: how conscious experience emerges from the interaction between a fixed anatomical network (the structural connectome) and dynamic brain activity. Earlier work has shown that loss of consciousness is associated with increased similarity between patterns of functional connectivity and anatomical connections, but those investigations typically (a) treat structure-function correspondence with symmetric correlation or distance metrics that do not reflect the directional constraint from structure to function, and (b) operate at a single spatial scale despite the brain's hierarchical organisation. The authors adopt the mathematical framework of connectome harmonic decomposition (CHD), a graph-spectral generalisation of the Fourier transform, which re-expresses spatial patterns of fMRI activity in terms of eigenmodes of a representative human structural connectome. Low spatial-frequency (coarse) harmonics indicate activity closely constrained by structure, whereas high spatial-frequency (fine) harmonics indicate divergence from structural constraints. This paper uses CHD to probe whether multi-scale, structure-informed harmonic patterns generalise as signatures of consciousness across different perturbations: transient pharmacological unconsciousness induced by propofol, chronic disorders of consciousness (DOC) after brain injury, and altered (but not absent) conscious states induced by sub-anaesthetic ketamine and the serotonergic psychedelic LSD. The stated aim is to identify connectome-harmonic signatures that (i) distinguish conscious from unconscious states regardless of how unconsciousness is induced, (ii) characterise psychedelic-altered states, and (iii) link those signatures to pharmacology and subjective phenomenology, thereby providing a distributed, multi-scale alternative to the traditional localisation-centric neuroimaging perspective.