Superposition of virtual sources: accelerated modeling of light distribution in heterogeneous tissues

We present a new method of called 'Superposition of Virtual Sources' (SVS), that combines the speed of diffusion theory with the ability of modeling complex geometries. Modeling the light distribution in tissues under the existing methods, Monte Carlo simulation is known for its accuracy and flexibility in treating complex geometries. Under the existing methods, Monte carlo simulation is known for its accuracy and flexibility in treating complex geometries. However, these advantages are countered by very extensive computation times. Analytical solutions found from diffusion theory can be calculated very fast but are limited to special simple geometries. The SVS method, which models a sphere with 57 virtual sources, is found to be in good agreement with the other two methods. However, SVS excels in computation time. The Monte Carlo simulation took 5 days, the diffusion theory analytical solution was evaluated with a Mathematica routine in 2 hours, while the SVS-solution was generated in 5 seconds (all on a sun sparc10-workstation).