Use this URL to cite or link to this record in EThOS:
Title: Automated classification of human epileptic spikes for the purpose of modelling bold changes using simultaneous intracranial EEG-fMRI
Author: Sharma, Niraj
ISNI:       0000 0004 7231 8587
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Mapping the BOLD correlates of interictal epileptiform discharges (IEDs) using EEG-fMRI can provide a unique insight into the region(s) responsible for their generation. Scalp EEG-fMRI studies have shown to provide added clinical value in the localisation of the epileptogenic zone in patients with pharmacoresistant epilepsy undergoing presurgical evaluation. However, scalp EEG has limited sensitivity in detecting IEDs as only a small percentage of the underlying electrical activity is recorded. Intracranial EEG (icEEG) provides a higher sensitivity of detecting underlying IEDs compared to scalp EEG due to the electrodes being closer to their generators. Recent safety and feasibility studies have allowed the acquisition of simultaneous icEEG-fMRI circumventing the lack of whole brain coverage of icEEG. Therefore, icEEG-fMRI has the potential to provide unprecedented insight in the relationship between the region(s) generating IEDs and the epileptogenic zone. However, one of the main challenges associated with icEEG-fMRI data is the difficulty of forming a parsimonious model of potential BOLD changes from the complex spatio-temporal dynamics of icEEG IEDs. The aim of this thesis is to provide a solution for a more consistent and less biased marking of icEEG IEDs using an automated neuronal spike classification algorithm, Wave_clus (WC), for the purpose of producing more biological meaningful IED-related BOLD maps. Adapting the icEEG IED dataset to Wave_clus was the first problem tackled which involved developing a new algorithm that identified the peak of the spiky component of an IED and defining an optimal IED classification epoch time-window. The two chapters that followed involved assessing the performance of WC as an icEEG IED classifier. First, I assessed the performance by comparing WC IED classification to the classification of multiple EEG reviewers using a novel validation scheme. This was determined by analysing whether WC-human agreement variability falls within inter-reviewer agreement variability and comparing the individual IED class labels visually and quantitatively. In this regard WC performance was found to be indistinguishable to that of EEG reviewers. Second I assessed the performance of WC by comparing the IED-related BOLD maps obtained using WC to those obtained using the visual/conventional approach. I found that WC was able to produce more biologically meaningful IED-related BOLD maps indicating that this approach can be used to further explore the region(s) responsible for generating IEDs in patients that have undergone icEEG-fMRI.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available