Physical explanation of the SLIPI technique by the large scatterer approximation of the RTE

Abstract

Visualizing the interior of a turbid scattering media by means light-based methods is not a straightforward task because of multiple light scattering, which generates image blur. To overcome this issue, a technique called Structured Laser Illumination Planar Imaging (SLIPI) was developed within the field of spray imaging. The method is based on a ‘light coding’ strategy to distinguish between directly and multiply scattered light, allowing the intensity from the latter to be suppressed by means of data post-processing. Recently, the performance of the SLIPI technique was investigated, during which deviations from theoretical predictions were discovered. In this paper, we aim to explain the origin of these deviations, and to achieve this end, we have performed several SLIPI measurements under well-controlled conditions. Our experimental results are compared with a theoretical model that is based on the large scatterer approximation of the Radiative Transfer Equation but modified according to certain constraints. Specifically, our model is designed to (1) ignore all off-axis intensity contributions, (2) to treat unperturbed- and forward-scattered light equally and (3) to accept light to scatter within a narrow forward-cone as we believe these are the rules governing the SLIPI technique. The comparison conclusively shows that optical measurements based on scattering and/or attenuation in turbid media can be subject to significant errors if not all aspects of light-matter interactions are considered. Our results indicate, as were expected, that forward-scattering can lead to deviations between experiments and theoretical predictions, especially when probing relatively large particles. Yet, the model also suggests that the spatial frequency of the superimposed ‘light code’ as well as the spreading of the light-probe are important factors one also needs to consider. The observed deviations from theoretical predictions could, however, potentially be exploited to assess particle size.