Use this URL to cite or link to this record in EThOS:
Title: Improvements in reconstruction algorithms for electrical impedance tomography of brain function
Author: Pérez-Juste Abascal, Juan Felipe
ISNI:       0000 0001 3483 8918
Awarding Body: University of London
Current Institution: University College London (University of London)
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Electrical Impedance Tomography (EFT) is a relatively new medical imaging method which, by injecting current and measuring voltage, estimates a volume conductivity map of the sub ject. It has the potential to become a portable noninvasive imaging technique of particular use in imaging brain function. A good estimate of the modelling parameters is essential for abso lute image reconstruction. While biological tissue like bone and white matter is anisotropic, clinical applications have assumed isotropic conductivity and adopted linear reconstruction of time difference data, as this is less affected by systematic errors. In previous studies, measured scalp impedance changes during evoked response on adults and on neonates were consistent, yet data had a low signal-to-noise ratio and image localisation using truncated singular-value- decomposition and a fixed truncation level was unsuccessful. There were four main goals in this thesis. The first goal was to examine ways of optimising linear reconstruction. This was attempted by comparison of standard methods for selecting the truncation level with modelling of the covariance of the noise. When examined on data from simulation, a head-shaped saline tank, and scalp neonatal evoked responses, there was no significant difference among selection methods, yet modelling a general covariance of the noise led to a significant improvement for simulated data. The second goal was to reduce the noise by applying Principal Component Analysis (PCA), for the case of EIT images collected during cortical evoked responses in the neonate. PCA significantly improved the SNR by 15dB on both tank and neonatal data. The third goal was to study the possibility of the recovery of a piecewise linear anisotropic tensor with known eigenvectors. It was possible to recover three smooth eigenvalues for simulated conductivity distributions with eigenvectors generally orientated and for a conductivity tensor estimated from diffusion weighted MRI. The fourth goal was to develop a method for incor porating anisotropy in a forward numerical model for EIT of the head and assess the resulting improvement in image quality. Neglecting anisotropy of the scalp, skull, and brain, yielded a 50% error in the forward solution and 24mm localisation error for a linearised inverse solution. This suggests that use of anisotropy is likely to improve EIT image quality.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available