The conductivity of brain inner tisssue is the key factor in EEG/MEG research. The brain tissue is inhomogeneous, and specially white matter exhibits serious anisotropic conductivity. The traditional methods such as electrical impedance tomography(EIT) is hard to solve the anisotropic conductivity problem. This paper put forward a new method to covert diffusion tensor to conductivity tensor based on diffusion tensor imaging (DTI) technology. This method incorporates the parameter of ions'concentration based on Stokes-Einstein and Nernst-Einstein equations. At last the method was testified on a human subject DTI dataset. The accuracy of this method is mainly determined by the parameter of diffusion-weighted time.