Skip to content

Immune-Mediated Changes in Conscious Experience: A Computational Model of Cytokine-Driven E/I Dysregulation

Aditi Doddavaram

DOI: 10.22541/au.177385901.15539620/v1 (opens in new tab)

Study at a glance

AI-extracted from the abstract
Characteristics Theoretical or philosophical paper
Key points Argues that inflammation acts as a global uncertainty signal disrupting the self-model, and that a neural mass model with added cytokine parameter shows shallower attractor basins, increased temporal autocorrelation, and under-damped recovery, consistent with temporal blurring during sickness.

Abstract

The sense of a unified minimal self is a dynamical achievement of the brain, arising from the precise coordination of predictive processing and E/I interactions. While the phenomenological shifts of sickness response are well-documented, a unified computational mechanism linking sickness behaviour and destabilization of consciousness does not yet exist. This paper proposes that systemic immune response leading to inflammation acts as a global uncertainty signal that disrupts the self-model by leading to reduced precision and increased autocorrelation, specifically, that pro-inflammatory cytokines interfere with GABA-ergic signalling and interoceptive inference. To test this hypothesis, we employ a Wilson-Cowan neural mass model with an added cytokine parameter and test it under different cytokine loads to observe the changes in E/I dynamics. The model illustrated shallower attractor basins, characterised by increased temporal autocorrelation and under-damped recovery from stochastic perturbations. This model illustrates how immune-modulated inhibition could produce dynamical regimes consistent with temporal blurring observed during sickness. By reframing sickness behaviour as a reactive reconfiguration of self-model stability, this paper helps to distinguish modulatory immune effects from structural neuropathology and helps with an understanding of dissociative states in acute and chronic inflammatory conditions.

Explore topics