Abstract
The brain’s ability to detect unexpected events, deviance detection (DD), is critical for survival. While DD has been computationally explained by synaptic plasticity, the role of neuromodulators such as acetylcholine (ACh) remains less understood. Here, we examine how ACh modulates DD. Using a cholinergic-sensitive Hodgkin-Huxley network of 200 neurons arranged in 2D space and stimulated via five spatially distinct inputs (A–E), we implemented an oddball paradigm with three conditions: standard (80% A, 20% B), deviant (20% A, 80% B), and a multi-standard control (20% each of A–E). ACh levels were modeled by varying the conductance of a slow K⁺ current, reflecting cholinergic modulation of the M-current. In the absence of ACh, the network already exhibited DD, responding more strongly to deviant A compared to control A. Notably, introducing a small amount of ACh amplified DD, while further increases suppressed it. Maximal DD occurred when strong spike frequency adaptation to standard B reduced competition, and enhanced phase-locking synchronized the network’s response to deviant A. These findings reveal how neuromodulation can shape context-sensitive neural computation, optimizing detection of salient events through a dynamic balance of suppression and synchronization.