Abstract
Neuropsychiatric disorders have complex causes and exhibit considerable individual variability as they develop over time. This suggests the need for a shift from a focus on observable clinical symptoms to a personalized trajectory monitoring paradigm that incorporates brain function checkups into routine primary care to allow detection of risk prior to the emergence of distress and impairment. A dynamical systems model of brain function enables quantitative snapshots of neural circuit function to be derived from electrophysiological measurements. Latent neurodynamical features can then be combined with personal and clinical data, enabling personalized neuropsychiatric trajectory monitoring. We present a general framework with recommended methods from dynamical systems theory to extract dynamical information from readily available EEG measurements. The dynamical features can then be incorporated into machine or statistical learning methods, where additional personal characteristics, experiences, and clinical data can be integrated to create risk prediction models for psychiatric conditions.