Transient chaos and periodic structures in a model of neuronal early afterdepolarization

The presence of chaos is ubiquitous in mathematical models of neuroscience. In experimental neural systems, chaos was convincingly demonstrated in membranes, neurons, and small networks. However, its effects on the brain have long been debated. In this work, we use a three-dimensional map-based membrane potential model, the logistic KTz, to study chaos in single and coupled neurons. We first obtain an alternative phase diagram for the model using the interspike interval (ISI), evidencing a region of slow spikes (SS), missing from the original diagram of the KTz model. A large chaotic region is found inside the SS phase. Embedded in chaos are several self-similar periodic structures, such as shrimp-shaped domains and other structures. Sampling the behavior of neurons in this diagram, we detect a novel type of action potential, the neuronal early afterdepolarization (nEAD). EADs are pathological oscillations during the action potential, commonly found in cardiac cells and believed to be chaotic and responsible for generating arrhythmias in the heart. nEAD was found experimentally in neurons in a type of epilepsy. We study two chemically coupled neurons with this behavior. We identify and characterize transient chaos in their interaction. A phase diagram for this system presents a novel type of self-similar periodic structures, where the structures appear "chopped" in pieces.