Locating of macro-inhomogeneity in random turbid medium by using artificial neural networks

摘要

Nowadays the artificial neural network (ANN), an effective powerful technique that is able denoting complex input and output relationships, is widely used in different biomedical applications. In present Letter the applying of ANN for the determination of characteristics of random highly scattering medium (like bio-tissue) is considered. Spatial distribution of the backscattered light calculated by Monte Carlo method is used to train ANN for single and multiply scattering regimes. The potential opportunities of use of ANN for image reconstruction of an absorbing macro inhomogeneity located in topical layers of random scattering medium are presented. This is especially of high priority because of new diagnostics/treatment developing that is based on the applying gold nano-particles for labeling cancer cells.