A bstract Brain injury is the leading cause of death and disability worldwide every year. Estimating the location and degree of brain injury has always been an important factor which it can save people from death. Researchers have been pursuing this goal through different methods. Computational models of head are valuable tools to predict human brain mechanical behavior in health and diseases. One of the most significant challenge in the Computational models of head is brain tissue examination in terms of “anisotropy” behavior as it can change all modeling procedures. In this study, an anisotropy of various components of the brain tissue has been investigated using the state-of-the-art “digital image correlation” method. For this aim, the lateral strains were earned in two different directions for various component of brain tissue. The brain stem components were examined and were proved to be anisotropic. Moreover, the corona radiata and a combination of white and gray matter, the reference behavior of whole brain in the related literature, were analyzed. Experimental tests revealed the medulla oblongata, pons and midbrain, all parts of the brain stem, to be transversely isotropic and corona radiate is an isotopic matter. Furthermore, the combination of gray and white matter depends on the feature of specimen. Moreover, in this study, importance of using an accurate method for extracting the mechanical properties of the brain tissue was examined. For this reason Effect of experimental boundary conditions in experiments and heterogeneous of deformation and stress were considered. The deficiency of the common method for calculating material parameters was investigated and finally, the brain tissue parameters for different constitutive models were computed using particle swarm optimization algorithm and complete modeling of boundary conditions of tests performed on the brain. Keywords: Mechanical behavior of brain tissue, Digital image corroletion, Particle swarm optimization, Corona radita, Brain stem