Modeling of Light Propagation

Monte Carlo modeling is a very useful technique for modeling light in biological tissues. It is based on random walks that photon make as they travel through tissue.

Our imaging techniques (LOT, camera imaging, two photon microscopy) are focusing on imaging smaller volumes (brain) or superficial tissues (skin). Therefore accurate and computationally efficient forward models are essential for reliable imaging.

In this ‘mesoscopic’ regime, we have developed Monte Carlo models that allow accurate predictions of light propagation. Our implementation can generate a map of the probability of the photon visiting each position within the tissue : the “sensitivity function” J (fig1). We have included in our forward model the possibility to exploit the assumption of reciprocity, significantly improving the time performance of our simulation. The second part of our modeling research is to use these forward models to do image reconstruction and corrections for our different imaging techniques.

Fig1 : Sensitivity function J in an homogeneous medium for a source detector separation of 1.4mm using Monte Carlo models. The map of J on the right exploits the assumption of reciprocity and clearly show a high signal-to-noise compared to a generalized forward model on the left.