Observability issues in networked clocks with applications to epilepsy

Inspired by the epileptic seizure prediction problem, this paper investigates what we can expect to be able to observe from electroencephalography (EEG) measurements of brain dynamics. Earlier research efforts have unsuccessfully attempted to track synchrony from EEG, viewing the brain matter as a coupled oscillator with the seizure state resulting from global synchronization. Using a model of networked pendulum clocks, we find that observing sufficient information to track the build up of synchrony is an ill-posed problem unless measurements are taken from a very localized region of brain tissue.