Recently, the plaque characterization field was explored with the use of the substrate (frequency domain analysis) rather than the envelope (amplitude or gray-scale imaging) of the intravascular ultrasound (IVUS) radiofrequency data. However, there is no data about the agreement of quantitative outcome between the two methods. The aim of this study was to assess the correlation and agreement between quantitative coronary ultrasound and the geometrical measurements provided by the spectral analysis of ultrasound radiofrequency data [IVUS-Virtual Histology (IVUS-VH), Volcano Therapeutics). Twenty-five patients were included in this study. The IVUS catheter used was a commercially available mechanical sector scanner (Ultracross 2.9 Fr 30 MHz catheter, Boston Scientific) covered with an outer sheath. IVUS-VH significantly underestimated lumen [relative difference (RD)=14.8+/-5.6; P<0.001], vessel (RD=14.1+/-4.8; P<0.001), and plaque (RD=11.5+/-10.8; P<0.001) cross-sectional areas (CSAs). Nevertheless, when adjusted for the ultrasound propagation delay caused by the sheath, relative differences of measurements were remarkably low (0.49%+/-6.3%, P=0.64 for lumen; 2.33%+/-4.6%, P=0.007 for vessel; and 4.2%+/-10.4%, P=0.005 for plaque CSA). These data suggest that the volumetric output of the IVUS-VH software underestimates measurements when acquired with a 30 MHz catheter. However, after applying a mathematical adjustment method for the ultrasound propagation delay caused by the outer sheath of the 30 MHz catheter, relative differences of direct measurements were negligible. These results suggest that ultrasound radiofrequency data analysis could provide, aside from precise compositional data, an accurate geometrical output.
Copyright (c) 2005 Wiley-Liss, Inc.