Use this URL to cite or link to this record in EThOS:
Title: Towards inferring evoked neural activity from haemodynamic changes: nonlinear dynamic modelling of the relationship between stimulus and neural activity
Author: Lefebvre, Veronique
ISNI:       0000 0001 3607 7286
Awarding Body: University of Sheffield
Current Institution: University of Sheffield
Date of Award: 2008
Availability of Full Text:
Access from EThOS:
The blood oxygen level dependent (BOLD) signal acquired by functional magnetic resonance imaging (fMRI) is increasingly used to study neural activation, as it offers excellent spatial and temporal resolutions for a non-invasive technique. However the BOLD signal is indirectly linked to the evoked changes in neural activity. Biophysical frameworks have been developed to relate the stimulus sequence to the BOLD signal through changes in neural activity and haemodynamic variables. The present thesis proposes aJnonlinear dynamic model of the relationship between stimulus and neural activity to be incorporated in the aforementioned frameworks, which until now lacked an accurate representation ofneural activity changes. The first part of the thesis introduces the reader to neuroscience and more specifically to the recent investigations underpinning the BOLD signal. A review of existing neural activity models is then provided to demonstrate the need to find the right balance between simplicity and realism to developing a suitable model in the context of fMRI data interpretation. A simple nonlinear dynamic model describing the amplitude of neural -responses to a stimulus pulse train is presented. Although this model satisfactorily fits the 'profiles of neural activity measurements it was limited to predict only the amplitude information which may not be sufficient in characterising the neural activity changes. A second model is thus proposed, capable of capturing the entire time series of local field potential recordings. Although more complex, this second model is based on the architecture of sensory pathways. While its parameters do not represent physical quantities they may have physiological implications. This model IS successfully usedto infer neural activity changes from blood flow measurements. Both models can predict neural responses profile reliably, for regular and random stimulation pulse trains. The second model can however accept more diverse stimuli types and may he easily extended due to its modular structure. Finally, as weIl as ' being a necessary tool in fMRI interpretation, it is postulated that the model may be used in physiological investigations and for the development of further biophysical models.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available