IMPACT: Image‐based physiological artifacts estimation and correction technique for functional MRI

Abstract

Functional MRI (fMRI) signal variation induced by respiratory and cardiac motion affects the activation signal and limits the accuracy of analysis. Current physiological motion correction methods require either synchronization with external monitoring of respiration and heartbeat, specialized pulse sequence design, or k‐space data. The IMage‐based Physiological Artifacts estimation and Correction Technique (IMPACT), which is free from these constraints, is described. When images are acquired fast enough to sample physiological motion without aliasing, respiratory and cardiac signals can be directly estimated from magnitude images. Physiological artifacts are removed by reordering images according to the estimated respiratory and cardiac phases and then subtracting the Fourier‐fitted variation from magnitude images. Compared with the k‐space–based method, this method can efficiently and effectively reduce the overall signal fluctuation in the brain and increase the activated area. With this new technique, physiological artifacts can be reduced using traditional fMRI pulse sequences, and existing data can be corrected and reanalyzed without additional experiments. Magn Reson Med 46:344–353, 2001.