Here we present an Algebraic Reconstruction Technique (ART) for solving the identification problem in Single Photon Emission Computed Tomography (SPECT). Traditional reconstruction for SPECT is done by finding the radiation source, nevertheless the attenuation of the surrounding tissue affects the data. In this context, ballistic and first scattering information are used to recover source and attenuation simultaneously. Both measurements are related with the Attenuated Radon Transform and a Klein-Nishina angular type dependency is considered for the scattering. The proposed ART algorithm allow us to obtain good reconstructions of both objects in a few number of iterations.