Abstract
This preliminary report explores insights about mechanisms of autism spectrum disorders (ASDs) gained from a mathematical model of interconnected clusters of excitatory neurons. This local, integrate-and-fire, spiking neural network is subjected to random noise from other populations of excitatory and inhibitory neurons outside the network. Its emergent behavior in response to different ratios of inhibitory to excitatory noise mimics several aspects of ASDs. The spiking neural network functions according to rules of classical neurophysiology, involving resting membrane potential, threshold potential, excitatory or inhibitory post-synaptic potentials, action potential, and refractory period. Key parameters include the incremental change in membrane potential, p , from external noise and the ratio, q , of inhibitory to excitatory external signals. With low levels of p and with q = 1 the network is quiet. Below a critical threshold level of q < 1 there is sudden development of chaotic activation of all nodes in the network, simulating autistic symptoms. Transition to chaotic activation occurs at a distinct threshold of out-of-network disinhibition, q* , which depends upon the background noise level, p , according to a nonlinear function. This state function represents the autism spectrum and explains several features of ASDs—including the existence and cause of the spectrum, the tendency of patients to seek out quiet environments, and the clinical rationale for GABA boosting drugs to treat ASDs. The model provides a new and independent test of the excitatory/inhibitory mechanism, namely that loss of input from inhibitory GABAergic neurons, for any of a variety of reasons, is a final common abnormality in autism spectrum disorders. Links between this simple biologically plausible model and criticality theory are discussed.
Keywords
action potential, ADHD, arbaclofen, ASD, baclofen, clonazepam, criticality, disinhibition, excitatory, GABA, GABA/glutamate ratio, inhibitory, integrate-and-fire, neural network, neurophysiology, noise, performance, phase transition, phenobarbital, spiking neural network, synaptic pruning, threshold, seizure, spectrum, white matter connectivity
Date of this Version
10-14-2025
Recommended Citation
Babbs, Charles F., "Insights from a simple mathematical model of autism spectrum disorders" (2025). Weldon School of Biomedical Engineering Faculty Working Papers. Paper 40.
https://docs.lib.purdue.edu/bmewp/40