Improved Translation Correction Applied to PROPELLER Using Expansion-matching Filter
Gaddipati A, Hartley M, Gui D, Chang S
PROPELLER is robust to rigid body motion with the capability of estimating rotation and translation shifts of each blade. This paper presents a translation correction algorithm using expansion-matching (EXM) filter, which estimates the template position more accurately, thus yielding images with sharper edges and less motion-derived ghosting artifacts in both phantom and volunteer studies. The EXM filter was also designed to adapt to the noise level in images, resulting in accurate motion estimation even in noise conditions.