Motion artifact reduction in self-gated CMR 4D flow imaging under exercise stress

Free-breathing self-gated CMR 4D flow imaging using traditional Compressed Sensing (CS) methods invariably contains motion artifacts due to the inaccuracy of self-gating signal. Self-gating signal degrades even further in the case of exercise stress imaging due to excessive movement of the subject. We propose Compressive recovery with Outlier Rejection (CORe) to reduce the motion artifacts. Using data from a 2D digital phantom and 4D flow data under rest and stress conditions, we demonstrate that CORe is effective in suppressing motion artifacts while maintaining agreement with 2D-PC based flow quantification.