The long term objective of this research is to develop clinically practical visualization and image-analysis tools for 3D ultrasound to diagnose arrhythmias and valvular disfunction, to study wall motion and wall thickness, and to monitor fluctuations in cardiac function and volume during surgery and intensive care. The proposed approach is twofold. (l) To explore computer graphics volumetric and surface rendering techniques that preserve the content of the data for the human observer. Preliminary experiments are underway using data from the existing video output of the 3D ultrasound machine. The next generation of 3D ultrasound machine will provide the entire raw data set for off-line graphical rendering. (2) To develop a practical and reliable means of automatically identifying anatomical structures in the 3D ultrasound data set. One such structure, the ventricular wall, would allow the automated calculation of ventricular volume versus time, producing a """"""""volumetricardiogram"""""""" (VCG). Proposed validation of the VCG includes in vitro studies with a pumped cadaveric heart and in vivo studies using canine preparations.