Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.746373
Title: Source modelling of the human hippocampus for MEG
Author: Meyer, S. S.
ISNI:       0000 0004 7231 392X
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
Magnetoencephalography (MEG) is a neuroimaging technique which gives direct non-invasive measurements of neuronal activity with high temporal resolution. Given its increasing use in cognitive and clinical research, it is important to characterize, and ideally improve upon, its advantages and limitations. For example, it is conventionally assumed to be insensitive to deep structures because of their distance from the sensors. Consequently, knowledge about their signal contribution is limited. One deep structure of particular interest is the hippocampus which plays a key role in memory and learning, and in organising temporal flow of information across regions. A large body of rodent studies have demonstrated quantifiable oscillatory underpinnings of these functions, now waiting to be addressed in humans. Due to its high temporal resolution, MEG is ideally suited for doing so but faces technical challenges. Firstly, the source-to-sensor distance is large, making it difficult to obtain sufficiently high signal-to-noise ratio (SNR) data. Secondly, most generative models (which describe the relationship between sensors and signal) include only the cortical surface. Thirdly, errors in co-registering data to an anatomical image easily obstruct or blur hippocampal sources. This thesis tested the hypotheses that a) identification and optimisation of acquisition parameters which improve the SNR, b) inclusion of the hippocampus in the generative model, and c) minimisation of co-registration error, together enable reliable inferences about hippocampal activity from MEG data. We found the most important empirical factor in detecting hippocampal activity using the extended generative model to be co-registration error; that this can be minimised using flexible head-casts; and that combining anatomical modelling, head-casts, and a spatial memory task, allows hippocampal activity to be reliably observed. Hence the work confirmed the overall hypothesis to be valid. Additionally, simulation results revealed that for a new generation of MEG sensors, ~5-fold sensitivity improvements can be obtained but critically depend on low sensor location errors. These findings set down a new basis for time-resolved examination of hippocampal function.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.746373  DOI: Not available
Share: