Reconstruction algorithm for near-infrared imaging in turbid media by means of time-domain data

Light propagation in highly scattering media as a result of the injection of short laser pulses is described by the diffusion model, with absorption and scattering being considered as parameter functions. Different boundary conditions are discussed. The resulting parabolic differential equation with boundary conditions of the third kind has been integrated by the finite-element method, which allows different geometries and various embedded objects to be taken into account. For image reconstruction we introduce an iterative method based on the finite-element method forward model and on an optimization strategy that uses the full information contained in the time-resolved measurements. The algorithm includes a regularization strategy so that it is specifically fit for solving the ill-posed problem. Furthermore, it is shown that considerable improvements in reconstruction results can be achieved by adaptation of the detector-source arrangement. (C) 1997 Optical Society of America.