Time-Resolved Spectral Optical Breast Tomography

Abstract

Abstract : The research carried out during the current reporting period involved: (a) extending the cumulant solution of radiative transfer to planar geometries and comparing with Monte Carlo simulations; (b)developing and enhancing the 3D tomographic image reconstruction algorithm by using the cumulant transport model; (c) developing the criterion of the optimal regularization for image reconstruction dependent on the noise presented in measurements; and (d) deriving the nonlinear correction factor for optical imaging due to multiple passages of an absorption inhomogeneity by a photon and providing a measure of the efficacy of linear inversion schemes. The cumulant transport model was found to provide a more accurate model than the conventional diffusion model for image reconstruction in turbid media such as human breasts. The proper modeling of the noise and the appropriate regularization improves the quality of image reconstruction. The nonlinear effect of the multiple passages of an absorption site by a photon on optical imaging only becomes appreciable when the size of the inhomogeneity reaches about ten times of the transport mean free path or larger in human tissues. The theoretical formalism and computer algorithm for 3D tomographic image reconstruction shows with simulated data the potential to provide fast 3D images of the scattering and absorption objects at various depths in turbid media.