| Photoacoustic tomography (PAT) is an emerging hybrid imaging modality that combines the advantages of both optical and ultrasonic imaging principles. It has demonstrated great promises for important biomedical imaging applications. Image reconstruction in PAT requires the solution of an inverse source problem, where the source represents the optical energy absorption distribution in the object.;In this thesis, we revealed a mathematical relationship between measured data and Fourier components of the object function. We developed and implemented various time-domain iterative reconstruction algorithms for PAT. We introduced a new data-space preconditioning concept and systematically investigate its effects. We investigated data sufficiency conditions and data redundancy in PAT, and developed image reconstruction algorithms from reduced PAT data for various measurement geometries. Finally, we proposed a heuristic iterative method that can simultaneously estimate both the acoustic speed and optical absorption distributions from PAT data alone. We utilized a multi-directional search method in combination with the standard conjugate gradient method to iteratively solve this non-linear optimization problem. |
