Near-infrared optical imaging of dense tissues is one of the new emerging molecularly based imaging technologies being developed for medical diagnostics. Propagation of light in turbid media, which is the principle behind the optical imaging modality, is mathematically represented by the diffusion approximation to the radiative transport equation. Numerical techniques such as the finite difference approach (using MUDPACK solver) are well developed to model the non-linear diffusion equation in 2D and 3D solid geometries such as slabs. However, the complex geometry of the breast tissues can be efficiently discretized using finite elements, and the forward problem can be modeled using the Galerkin approximation of the diffusion equation. Here, the finite element approach was applied to simulate the forward problem of the diffusion equation in fluorescence-enhanced frequency-domain photon migration system using tetrahedral elements for solid 3D slab geometry. The simulations were performed using twodifferent modes of source application, which include the isotropic point source and the planar source. Experimentally these sources are reproduced using a single 1000 mm optical fiber for a point source and point collection of light (termed as the singlepixel system) and an expanded laser beam for area detection of fluorescence in response to plane illumination of excitation light (termed as the multi-pixel system). Experiments were performed using the single-pixel and multi-pixel frequency-domain photon migration equipment in the Photon Migration Laboratory using indocyanine green (ICG) as the exogenous contrast agent added to 1 percent intralipid solution that closely simulates the breast tissue optical properties. The simulated phase and amplituderatio obtained using the finite element approach were analyzed and compared well with the experimental data. With proper prediction of measurements, the inverse imaging problem can be attacked. This work is supported by the National Institutes of Health(R01CA67176).
展开▼
