3D simulation of PET brain images using segmented MRI data and positron tomograph characteristics

A 3D simulation procedure has been developed to generate simulated PET brain images from MRI data. MRI slices were segmented into gray matter, white matter, CSF structures, and assigned with radionuclide distributions. Projections through these regions were generated according to physical characteristics of a positron tomograph, including 3D sampling and resolution, attenuation, scatter, randoms, and counting statistics. The projection data were then reconstructed by filtered backprojection. The procedure was validated with a cold spot phantom. Simulated PET images for cerebral blood flow or metabolism are presented, along with a brief discussion of some applications and limitations.