Michelle Chong | Estimation algorithms for neural mass models of epilepsy
INS Seminar Room, Campus Timone, Red Wing, 5th Floor.
Abstract :
Neural mass models describe the electrical activity of populations of neurons, which includes the subpopulations of excitatory neurons and inhibitory interneurons. These models have been shown to produce realistic behaviour as seen in the electroencephalogram (EEG) under epileptic conditions (Wendling et. al. 2007, Lytton 2008). Additionally, their states (mean membrane potential and firing rate) and parameters (the synaptic gain of neural populations) provide an avenue for seizure detection and prediction. This is achieved via algorithms to estimate the model’s unmeasured state and parameters from EEG measurements in real time.
In this talk, I will present a collection of estimation algorithms we have designed for a class of neural mass models. Time-domain as well as joint time-frequency domain-based algorithms will be presented. These algorithms are applicable to well-known neural mass models, such as the models by Jansen and Fit (1993) and Wendling et. al. (2007). We provide convergence guarantees for all our algorithms, in the presence of measurement noise. The efficacy of our algorithms are illustrated on a neural mass model by Jansen and Fit (1993).
Joint work with Romain Postoyan (CNRS, France), Dragan Nesic (University of Melbourne), Levin Kuhlmann (University of Melbourne), Maria Sandsten (Lund University), and Anders Rantzer (Lund University).
Short Bio:
Michelle Chong is an Assistant Professor at the Control Systems Technology section, Department of Mechanical Engineering, Eindhoven University of Technology, the Netherlands. Previously, she was a postdoctoral researcher at the Division of Decision and Control Systems, KTH Royal Institute of Technology, and in the Department of Automatic Control, Lund University, Sweden, from 2015-17 and 2018-19, respectively. She was awarded the American Australian Association’s ConocoPhilips postdoctoral fellowship in 2013 to the University of California, Santa Barbara in Prof. Joao Hespanha’s group working on designing secure cyber-physical systems against adversarial attacks. She obtained her PhD in mathematical control theory with applications in neuroscience from the Department of Electrical and Electronic Engineering at the University of Melbourne in December 2013. Her research interest lies in the broad area of mathematical control theory, in particular estimation, control and optimisation algorithms, with applications in neuroscience, cyber security and power systems.
Website: