Clinical Evaluation of 64-Slice CT Assessment of Global Left Ventricular Function Using Automated Cardiac Phase Selection

Abstract

Background Left ventricular (LV) function provides prognostic information regarding the morbidity and mortality of patients. An automated cardiac phase selection algorithm has the potential to support the assessment of LV function with computed tomography (CT). This algorithm is clinically evaluated for 64-slice cardiac CT. Methods and Results Examinations of twenty consecutive patients were selected. Electrocardiogram gated contrast-enhanced CT was performed. Reconstructions were performed using an automated and a manual method, followed by the determination of the global LV function. Significances were tested using 2-sided Student's t-tests. Reduction in post processing time and storage capacity were estimated. A slightly smaller mean end-systolic volume was found with the automated method (52±18 ml vs 54±17 ml, p=0.02, r=0.99). The mean LV ejection fraction was slightly larger with the automated method (65±8% vs 64±8%, p=0.004, r=0.99). The estimated reduction in post processing time was maximal 5 min per patient with a potential 80% data storage reduction. Conclusions Results of the automated phase selection algorithm are similar to the manual method. The automated tool reduces post processing time, reconstruction time and transfer time. (Circ J 2008; 72: 641 - 646)