Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.746230
Title: Monte Carlo methods in quantitative photoacoustic tomography
Author: Hochuli, R.
ISNI:       0000 0004 7230 5356
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:
Quantitative photoacoustic tomography (QPAT) is a hybrid biomedical imaging technique that derives its specificity from the wavelength-dependent absorption of near-infrared/visible laser light, and its sensitivity from ultrasonic waves. This promising technique has the potential to reveal more than just structural information, it can also probe tissue function. Specifically, QPAT has the capability to estimate concentrations of endogenous chromophores, such as the concentrations of oxygenated and deoxygenated haemoglobin (from which blood oxygenation can be calculated), as well as the concentrations of exogenous chromophore, e.g. near-infrared dyes or metallic nanoparticles. This process is complicated by the fact that a photoacoustic image is not directly related to the tissue properties via the absorption coefficient, but is proportional to the wavelength-dependent absorption coefficient times the internal light fluence, which is also wavelength-dependent and is in general unknown. This thesis tackles this issue from two angles; firstly, the question of whether certain experimental conditions allow the impact of the fluence to be neglected by assuming it is constant with wavelength, a `linear inversion', is addressed. It is demonstrated that a linear inversion is appropriate only for certain bands of illumination wavelengths and for limited depth. Where this assumption is not accurate, an alternative approach is proposed, whereby the fluence inside the tissue is modelled using a novel Monte Carlo model of light transport. This model calculates the angle-dependent radiance distribution by storing the field in Fourier harmonics, in 2D, or spherical harmonics, in 3D. This thesis demonstrates that a key advantage of computing the radiance in this way is that it simplifies the computation of functional gradients when the estimation of the absorption and scattering coefficients is cast as a nonlinear least-squares problem. Using this approach, it is demonstrated in 2D that the estimation of the absorption coefficient can be performed to a useful level of accuracy, despite the limited accuracy in reconstruction of the scattering coefficient.
Supervisor: Cox, B. T. ; Beard, P. Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.746230  DOI: Not available
Share: