Objective Contrast-enhanced ultrasound (CEUS) cines of tumors can be quantitatively analyzed to measure perfusion using the time-intensity-curve (TIC) on a pixel-by-pixel basis.Using a rat ENU breast tumor model, Linear Discriminant Analysis of TIC parameters showed that pixel-based variables (variance) created 5 times greater separation of benign from malignant tumors than ROI-based analysis (means) [1].However,CEUS cines are acquired freehand causing non-uniform in-plane and out-of-plane motion from frame to frame ,particularly when imaging focal liver lesions (FLLs) [2].The challenge of pixel-by-pixel analysis is the elimination of motion and out-of-plane frames that occur on 3-minute long cines that increase variance and decrease statistical power.We aimed to determine if motion correction improved the predictive value of automated TIC analysis of benign from malignant FLLs.