Electrogastrography (EGG) is obtaining stomach myoelectrical activity non-invasively. Today, reflux is common between disorders of the digestive system. The invasiveness of methods used in the diagnosis of reflux is the biggest disadvantage for diagnosis. In this study, features helping non-invasive diagnosis of reflux disease are intended to obtaine using Electrogastrogram signals. Records were done as hunger and satiety in two ways from patients and healty individuals. Distinctive features was obtained using Singular Spectrum Analysis and Power Spectral Densities. 6 features were extracted from signals. Distinctive offFeautres were examined statistically. Finally, The features have been examined among groups and features that can be successful for classification have been determined.