Short CV

Dr. Nina Doorn holds a Bachelor’s degree in Technical Medicine and a Master’s degree in Biomedical Engineering from the University of Twente, the Netherlands. She recently obtained her PhD in Computational Neuroscience at the same institution, which was awarded cum laude.

Her research focuses on developing computational models to interpret electrophysiological measurements from the brain and in vitro neuronal systems, with the aim of uncovering disease mechanisms from multiscale data. She combines detailed biophysical models of neurons and networks with machine learning–based inference methods to extract mechanistic insight from complex recordings. Through this work, she aims to advance our understanding of neurological disorders and support the development of targeted, data-driven interventions.

Title of the talk

Unveiling the neural origins of emergent network activity using biophysical modeling and machine learning.

Abstract

In this talk, Dr. Nina Doorn will present a computational modeling framework to uncover disease mechanisms from neuronal network activity. The presentation will introduce how human induced pluripotent stem cell (hiPSC)-derived neuronal networks measured with multi-electrode arrays (MEAs) provide rich, patient-specific activity data, but translating these patterns into underlying molecular mechanisms remains a key challenge.

The talk will focus on a biophysically detailed network model combined with simulation-based inference (SBI), a machine-learning approach that enables efficient estimation of model parameters from experimental recordings. This framework allows identification of ion-channel properties, synaptic dynamics, and connectivity changes that best explain observed activity patterns, while also quantifying uncertainty in these estimates.

Finally, the talk will highlight how this approach recovers known disease mechanisms in patient-derived networks, detects drug-induced perturbations, and provides a systematic and scalable route to mechanistic insight. This work demonstrates how combining modeling and machine learning can bridge the gap between complex electrophysiological data and interpretable biological mechanisms, enabling more targeted experiments and advancing neuroengineering applications.