Loss of consciousness as a collapse of neural propagation geometry. A geometro-diffusion theory approach

Abstract

General anesthesia provides a reversible means of investigating the dynamical conditions required for consciousness. Within the Geometro-Diffusion Theory (GDT), loss of consciousness occurs when neural signal propagation no longer generates a sufficiently differentiated geometry whose relational information can be intrinsically accessed by the system. We investigate this mechanism through anomalous diffusion on anatomical brain networks, using the cat cortex as a representative example. We show that temporal nonlocality contracts the repertoire of weighted shortest communication paths, whereas spatial nonlocality expands it. Despite their opposite effects on functional connectivity, both regimes reduce geometric differentiation, as revealed by the intrinsic measures of GDT. A coupled model combining these mechanisms is embedded in a reversible anesthetic-burden model describing the progression from normal diffusion to maximum anesthetic effect and subsequent recovery. At maximum burden, regional and inter-region measures reveal a marked loss of differentiated and readable geometric information, which within GDT signifies loss of consciousness. Anesthetic clearance restores these measures to their normal values and re-establishes the geometric conditions for consciousness. Our results identify anesthetic-induced unconsciousness as a reversible collapse of neural propagation geometry.