Application of RBF Neural Networks for Solving the Inverse Light Scattering Problem of Platelets

In this paper, the possibility of using Radial Basis Function (RBF) neural networks to solve the inverse problem of light scattering for platelets is considered. The database of theoretical signals has been generated by the discrete dipole approximation (DDA) method and has been used to train the predictive model. The method is tested on model and experimental data. The main features of the RBF neural network solution method have been considered and compared with the DATABASE method. The error of solving the inverse problem for the proposed neural network model was 6.4-11 % with a solution time of 8ms.