Functional neuroimaging is a powerful biological tool to investigate the regions of the brain responsible for performing diflerent mental functions. The traditional anterpretation method is statistical analysis in the spatial domain, which is computationally expensive. In this paper we present a hybrid wavelet/neural network scheme to analyze functional brain images. Features are extracted in the Wavelet domain and then fed t o a neural network for detection. The proposed method is examined by exploring the diflerences between Positron Emission Tomography (PET) images acquired under different ezperimental conditions during a tone recognition task. The performance shows its potential in the fast developing functional neuroimaging area.