Gas hydrates buried in the shallow sediments below the ocean cause amplitude attenuation, wavelet phase distortion, and frequency loss in the seismic reflection data beneath them. This can make reservoir evaluation based on the deeper seismic image difficult. The distortion can be compensated by using pre-stack inverse Q-filtering, or by a more sophisticated Q-migration. While inverse Q-filtering (amplitude and phase) is a traditional method to correct these problems, it is not sufficient to compensate for localized image distortion resulting from complex overburden structures. A dataset in Alaminos Canyon, Gulf of Mexico was used to perform a 3D Q-tomographic inversion. The derived Q-model of the gas hydrates was used in Q-Reverse Time Migration (Q-RTM) (Zhang et al, 2010) to improve the imaging around complex geological structures. The results show that Q-tomography can generate a reliable Q-model from the surface seismic data, and that Q-RTM can reduce the energy loss and image distortion caused by the shallow gas hydrates. Altogether, Q-tomography and Q-RTM provide an improved structural definition and more uniform amplitude distribution for a more accurate and interpretable seismic section.