Automatic respiratory gating for perfusion quantification of DCEUS

Abstract

A novel automatic respiratory gating (ARG) algorithm for ultrasound is presented. The ARG algorithm was applied on dual contrast imaging (contrast/tissue) loops in order to remove respiratory motion artifacts from dynamic contrast enhanced ultrasound (DCEUS) of liver lesion images. The ARG algorithm is fully automated and it only requires the operator to define the breathing cycle phase to be extracted. The ARG algorithm was applied on the tissue side of the dual contrast acquisition. Motion was detected using frame subtraction; frequency domain analysis identified the contribution of respiration in the motion. A respiratory kinetics curve was extracted based on the motion detection and frequency analyses. Gating was performed by detecting the signal envelope of the respiratory kinetics curve and applying a threshold in order to remove frames that are out of synchronization with the trigger frame. DCEUS images from fourteen liver metastasis patients were extracted with and without the use of ARG. The lesion DCEUS data were fitted onto the lognormal indicator dilution model and the goodness of fit was evaluated. Significant improvement to the goodness of fit of the liver lesion DCEUS data onto the lognormal model was demonstrated with the use of ARG.