Electrical impedance tomography of human brain function
Electroencephalography (EEG) has been used for over 70 years to record the electrical signals of the brain. Electrical impedance tomography (EIT) is a more recent imaging technique which when applied to brain function and structure has the potential to provide a rapid portable bedside neuroimaging device. The purpose of this work has been to investigate several applications of EIT and EEG in the imaging of brain function. EEG does not always give the required spatial information, especially if the current generator is in the deep brain structures such as the hypothalamus. Dipole source localisation has become a common research tool that can be used estimate the current sources that are responsible for the EEG signals recorded on the scalp. Using this method, the accuracy and ease of use for four commercially available headnets was assessed. No headnet performed better at localisation, with all localising one dipole well, and two or three dipoles poorly. EIT has the potential to image the impedance changes that occur during neuronal depolarisation. Modelling work has been carried out to predict the size of these impedance changes and this thesis presents some work carried out in an attempt to record these changes in human subjects. The levels of noise at present are too great to record the impedance changes, but suggestion for improving the signal to noise ratio are given. Previous work on EIT and fMRI studies has shown that there are changes in blood volume (and as a result changes in impedance) after interictal spike activity. The impedance changes relating to the blood flow response to interictal epileptiform activity were recorded using EEG-correlated continuous EIT acquisition from scalp electrodes from patients on telemetry. Despite averaging up to 900 spikes, there was no recordable change in impedance after the interictal activity. Bioimpedance changes also occur due to pathological conditions. Multifrequcncy HIT makes use of the differences in impedance properties between healthy and ischaemic or tumour tissue, in an attempt to image these conditions. Data were collected from patients with tumours and other conditions and healthy volunteers, and the raw data and images compared. No differences were seen in the raw data between the different patients groups thought changes were seen in individual patients. . These results will inform the design of an EIT system which operates at a lower frequency band where the largest changes in impedance are seen.