Generalized self‐navigated motion detection technique: Preliminary investigation in abdominal imaging

Patient motion remains a primary obstacle to diagnostic image quality, especially in the abdomen, despite the existence of various motion artifact reduction techniques. This work presents a self-navigated motion detection method that can be generalized for most pulse sequences and k-space trajectories. Motion information is extracted directly from raw MR data, requiring no additional gradient or RF pulses, no physiologic monitoring equipment, and minimal--if any--impact on scan time. Initial feasibility results with a two-dimensional gradient echo sequence are shown in phantom studies and in vivo volunteer abdominal studies, demonstrating the sensitivity of the method to both respiratory motion and cardiovascular pulsatility. Prospectively gated images were acquired using the self-navigated data to synchronize image acquisition with motion. These preliminary results suggest that the self-navigated method is a promising technique for reducing motion artifacts in clinical abdominal and cardiac applications.