Clinical validation of a novel software for quantitative analysis of coronary intravascular ultrasound.

Abstract

CAAS IntraVascular (CAAS-IV) is a recently released software that provides options for sophisticated quantitative coronary ultrasound (QCU) analysis. The aim of this study was to validate CAAS-IV for QCU in diseased human coronaries. Ten preprocedural and 5 postprocedural IVUS studies were derived from daily practice. Intraobserver, interobserver, intersoftware (CAAS-IV vs. Curad) and interplatform (CAAS-IV vs. Volcano console) variability were assessed for cross-sectional area (CSA) measurements. Interobserver and intersoftware comparisons were made for volume measurements. Measurements of lumen, EEM, plaque and stent CSA demonstrated small differences in the intraobserver (0.0±3.7%, -0.7±2.8%, -0.5±7.0% and -0.9±3.4%), interobserver (0.1±4.4%, 0.1±3.4%, -0.5±8.2% and -0.8±4.3%), intersoftware (-0.3±4.5%, 0.2±2.4%, 0.4±6.8% and -0.5±3.2%) and interplatform (0.7±7.9%, 0.9±4.0%, -1.1±12%, -1.8±3.6%) comparisons. For lumen, EEM, plaque and stent volume, the interobserver (-2.1±9.3%, 0.9±5.6%, 3.4±7.2% and -0.2±3.6%) and intersoftware (-2.2±6.2%, -2.6±6.1%, -2.7±12% and -4.1±3.2%) differences were substantially larger. Excluding large side-branches and calcifications, post-hoc measurements of lumen, EEM, plaque and stent volume showed small differences in the interobserver (-0.3±3.2%, 0.9±2.4%, 2.9±4.4% and -1.3±1.8%) and intersoftware (0.5±2.5%, -1.2±1.7%, -3.4±5.1% and -1.5±2.2%) comparisons. Analysis time for entire pullbacks was reduced by 19.2 [14.9-30.0]% using CAAS-IV (p<0.01). CAAS-IV demonstrated reliable QCU with excellent agreement with previously validated software and the IVUS imaging console. Precision and reproducibility of measurements were high, proving CAAS-IV to be a valid option for QCU analysis in clinical practice and research. Interactive contour editing reduced analysis time by 20%.