— We propose here a new method for the automatic quantification of neurotransmission using data fusion. The process uses a Magnetic Resonance image as anatomical reference and a Single Photon Emission Computed Tomography image of the same patient for functional information. Once both acquisitions are coregistered, the process is divided in two steps: first, structures of interest are automatically segmented using a data fusion-based algorithm, then the masks obtained are used to compute functional indexes in the SPECT. Both steps of the process are now validated and we plan to apply it as a tool for differential diagnosis in a wide range of pathologies.