We propose here a prospective study for the automatic quantification of neurotransmission defects using information fusion. The process uses two registered clinical acquisitions of a same patient: a SPECT (Single Photon Emission Computed Tomography) acquisition providing functional information and a MR (Magnetic Resonance) image as anatomical reference. The first step of the quantification process is a data fusion algorithm for brain structures location in MR images. The segmented structures will be used as a mask to compute quantification indexes in the SPECT image. The segmentation step has already been validated. We now plan to assess the reliability of the whole process in studies concerning a wide range of pathologies.