This proposal, Interactive Real-time fMRI at High Field with Automatic Classification of Activation Patterns is submitted in response to RFA EB-03-007. The objective of this application is to further develop highly sensitive multi-echo pulse sequences with parallel imaging data acquisition and compensation of susceptibility artifacts at high magnetic field (4 Tesla), and to design real-time data analysis methods that allow automatic classification of brain activation patterns in terms of underlying cognitive processes. Feasibility of this methodology has already been shown in pilot studies at 1.5 and 4 Tesla. A secondary goal is to validate this methodology with interactive experimental paradigms by characterizing neural correlates of cognitive and emotional processes in individual subjects. This novel real-time fMRI methodology is applicable to a wide range of neuroscience research and to a variety of neuro-psychiatric conditions in which pathology of the amygdala is implicated, such as unipolar and bipolar depression, anxiety, schizophrenia, psychopathy and social phobia.