An efficient analysis of temporal and spatial pulse propagation through a turbid medium

Abstract

Optical imaging through a highly scattering medium has become an active field of study because of its potential applications in medical diagnosis and microscopic medical imaging. At the forefront of medical imaging are methods utilizing the spatial filtering capability of the confocal microscope to reject scattered light and the intrinsic rejection of two-photon fluorescence excitation. Recent work has concentrated on the elimination of unwanted stray photons by using fast detector gating systems, working in the temporal domain.
The predominant model to describe the temporal broadening of the propagating pulse is the diffusion equation. Because of the inherent difficulty in fitting the boundary conditions for samples with complex geometry, the technique is limited to simple imaging systems. This approach is difficult to apply to scatter-imaging applications using focused femtosecond light pulses. Furthermore, the technique cannot account for medium anisotropy and is valid only in cases where the photons have undergone numerous scattering events.
In this paper, we present a Monte Carlo approach to describe the temporal propagation of focused gaussian pulses through turbid media of different depths and anisotropies. This is an alternative paradigm to diffusion theory and is more robust to anisotropy and complex sample geometries but is computationally intensive. This problem is addressed by treating the scattering medium as a linear system, which broadens the input signal because of its inherent impulse response. The scattered output temporal pulse is then determined by mere convolution of the input pulse with the response of the medium considerably reducing computational time. The protocol is then applied to spatial imaging of reflecting cylindrical samples embedded in the turbid medium.