The purpose of this study was to investigate the toolbox based on the attenuation correction algorithm proposed by Thomas Beyer at el. in 2008 which seems to be useful for PET image reconstruction using MRI data in the brain. We retrospectively included six patients who had MRI examination and the FDG PET/CT scan. Histogram-matching algorithm was applied to transform the MR images to the so-called “pseudo-CT” images which are simulated as attenuation correction map for PET image reconstruction. The two model corrected PET images were visually analyzed and SUV were compared. The results showed better image quality and good correlation of SUV analysis on both supra-tentorial brain images than on infra-tentorial brain images. It is therefore the intensity transformation method seems to provide a simple way to transform the MR images to pseudo-CT for the attenuation correction of PET images in PET/MR, but needs more technical implementation to solve some pitfalls in the future.