Adaptive Curve Region based Motion Estimation and Motion Visualization of Cardiac Ultrasound Imaging

This paper presents a novel method for motion estimation and visualization of cardiac ultrasound images. The motion vectors are derived from an adaptive curve region based matching algorithm. The size of curve regions can be adjusted adaptively based on ultrasound system parameters and position information of the region´s center which is located on the scan lines as sampling points. Then the Curve Minimum Mean of Absolute Difference (CMMAD) method is proposed to the curve region to estimate the motion vector field of two successive images. After obtained a vector field of the sampling points, we employ the Thin-Plate Spline (TPS) transformation function to recover the non-rigid motion and interpolation motion vectors for each pixel in the region of interest (ROI) of the ultrasound image. Finally we apply the Unsteady Flow Line Integral Convolution (UFLIC) to the vector field for motion visualization.