From Neuron to Network to Neurology: Deciphering multiscale electrophysiological measurements through biophysical modeling

The human brain is a multiscale marvel: billions of neurons, trillions of synapses, and countless molecular components working together to generate thought, learning, and behaviour. Yet that same complexity becomes a challenge when something goes wrong. A change in a single gene can alter a protein, disturb how neurons fire, reshape network dynamics, and ultimately appear as neurological symptoms—but tracing this cascade across scales is notoriously difficult.
This thesis explores how we can decode the brain’s electrical activity across levels of organization, from single neurons to networks to the entire brain. Electrophysiology offers rich readouts—EEG in patients and multi-electrode array (MEA) recordings in laboratory-grown neuronal networks—but these measurements often show patterns without revealing the mechanisms that produced them. To bridge that gap, I developed and applied biophysical computational models: detailed mathematical descriptions of neuronal networks that allow us to simulate the effect of specific molecular changes on electrophysiological network recordings.
Using patient-stem-cell-derived neurons and mechanistic network simulations, this work links electrical phenotypes to underlying cellular and synaptic changes—illustrated through studies of the genetic epilepsy Dravet Syndrome, the fragmented bursting phenomenon observed in neurodevelopmental disorders, and the burst-suppression EEG pattern seen in unconscious patients.
By combining human electrophysiology with biophysical modeling and automated inference techniques, this thesis aims to turn electrical recordings into biological insight—helping to connect Neuron to Network to Neurology, and bringing us closer to explanations that can ultimately support diagnosis, prognosis, and treatment.