Detection of tumor-like inclusions embedded within human liver tissue using a short-pulsed near-infrared laser beam: Parallel simulations with radiative transfer equation

Abstract

An efficient solution to detect tumor-like inclusions embedded within a human liver tissue model is presented, using illumination by a short-pulsed laser beam. Light propagation was accurately solved using the time-dependent radiative transfer equation, with multithreaded parallel computing. A modified finite volume method based on unstructured grids and the fourth-order Runge–Kutta approach was employed to solve the equation in the (2-D/3-D) spatial and time domains. The normalization technique applied to the Henyey–Greenstein phase function was adopted to ensure numerical stability for values of the anisotropy factor that were close to unity. The presence of one or two circular/spherical inclusions was analyzed from the time and spatially resolved reflectance on the medium bounding surface. The results allowed a minimal size and a maximum distance for the detection of the inclusion to be highlighted.