Reduced order modelling applied to parametric EIT and EEG for comprehensive characterisation of the electrical conductivities of head tissues

Walker, Matthew 2024. Reduced order modelling applied to parametric EIT and EEG for comprehensive characterisation of the electrical conductivities of head tissues. PhD Thesis, Cardiff University. Item availability restricted.

Preview	PDF - Accepted Post-Print Version Available under License Creative Commons Attribution. Download (7MB) | Preview
	PDF (Cardiff University Electronic Publication Form) - Supplemental Material Restricted to Repository staff only Download (140kB)

Abstract

Electromagnetic source imaging (ESI) techniques based on electroencephalography (EEG) and magnetoencephalography (MEG) are used to better understand the function of the human brain and neurological conditions through monitoring electrical activity. ESI methods depend on models of the head encoding realistic and personalised anatomy and electrical conductivity characteristics. The anatomy can be extracted reliably from standard imaging techniques such as magnetic resonance imaging (MRI). However, extracting the individual conductivity of head tissues poses a unique challenge. This has been partially addressed by the development of parametric electrical impedance tomography (pEIT) and EEG/MEG-based calibrations, which aim to estimate these conductivities. Unfortunately, these techniques suffer from a heavy computational cost due to high-dimensional systems of equations needing to be solved. This thesis explores the application of a dimensionality reduction technique called reduced order modelling (ROM) to this problem, with the objective of accurate estimation of the conductivity of tissues in the head. For pEIT, the main results obtained were a substantial increase in speed and therefore practical accuracy while simultaneously unlocking, for the first time, the new ability to estimate the conductivity of deeper head tissues. An ROM-based calibration method utilising EEG data is also developed in this thesis, where the tuned conductivities of head tissues directly result in a large improvement in source localisation accuracy. The new frameworks developed in this work could have a far reaching impact in the field of EEG, providing new capabilities for researchers and clinicians.

Item Type:	Thesis (PhD)
Date Type:	Completion
Status:	Unpublished
Schools:	Physics and Astronomy
Subjects:	Q Science > QC Physics
Uncontrolled Keywords:	electroencephalography, electrical impedance tomography, reduced order modelling
Funders:	EPSRC
Date of First Compliant Deposit:	13 January 2025
Last Modified:	13 Jan 2025 12:38
URI:	https://orca.cardiff.ac.uk/id/eprint/175201

Actions (repository staff only)

Edit Item