Circuit Topology Shapes the Mapping between Excitatory-Inhibitory Input Composition and Aperiodic Spectral Slope

Abstract

Aperiodic ([Formula: see text]) spectral components are widely interpreted as markers of excitation–inhibition (E/I) balance, but this interpretation spans levels from synaptic currents and microcircuit spiking to population field potentials. Whether macroscopic spectral slope can index microcircuit-level E/I input independently of circuit architecture remains unresolved. This study introduces a connectome- and morphology-constrained framework that embeds electron-microscopy–reconstructed cortical microcircuits in near-critical propagation dynamics and generates field potentials under fixed synaptic, membrane, and forward-modeling parameters. By linking controlled synaptic E/I input, microcircuit spiking propagation, and macroscopic spectra, the framework directly tests the explanatory level of the aperiodic exponent. Across matched E/I regimes, the direction and magnitude of the E/I input–[Formula: see text] relationship differed between biological and topology-randomized networks, yielding a significant E/I input × topology interaction. To test whether this architecture-dependent mapping extends to biological systems, paired cortical ECoG and thalamic EThG recordings from juvenile pigs were analyzed. Under the same systemic propofol perturbation, exponent changes were stronger in thalamus than cortex and remained so after accounting for global state shifts, baseline spectra, and distance from criticality. These findings identify circuit architecture as a determinant of how microcircuit E/I input is transformed into macroscopic aperiodic spectra, defining the level-specific interpretability of spectral slope.