The study is devoted to upgrading the computational grid when solving direct and inverse diffraction problems on an inhomogeneous body. With this improvement, it was possible to restore inhomogeneity without significantly increasing the frequency of the source of the external field. The results of the inhomogeneous reconstruction were obtained for different bodies and source parameters. The purpose of this work is to apply the new computation grid to the body under study. We increase the number of cells in the computation grid without changing the size of the cells, thus building a new mathematical model of the problem. Thanks to these changes, we are less dependent on the choice of frequencies. The last chapter contains an example of solving the problem at a fixed frequency with a special computational grid. We will also show a solution to the problem with additional noise. The numerical results of reconstruction for the inhomogeneity parameters of the body are presented. The results of the work can be applied to create methods for determining inhomogeneities in human tissues during a medical diagnostic. Due to the enormous computational complexity of the problem, parallel algorithms based on the use of the MPI library were applied for this task.