Comparative in-vitro validation of eight first- and second-generation quantitative coronary angiography systems

It is known that first-generation quantitative coronary angiography (QCA) systems overestimate small vessel sizes owing to the point-spread function of the respective X-ray imaging chain. With second-generation systems new algorithms were introduced to correct for this source of error.

To evaluate the efficiency of the modified contour detection algorithms.

Six second-generation QCA systems (CMS, QANSAD, AWOS, CAAS II, Cardio 500, and Angioimage) were validated and compared with first-generation systems (CAAS and ARTREK). By using an arterial phantom consisting of stenotic and nonstenotic glass tubes (of diameters 0.5-5.0 mm) the accuracy and precision of each analysis system, as well as their additional accuracy and precision values for phantom diameters < or = 1.0 mm were determined.

All systems had high accuracy and precision values, but first-generation systems overestimated small vessel diameters. With second-generation systems a significantly improved accuracy in the submillimeter range (an accuracy within +/-0.028 mm) was obtained. This improvement was accompanied by a moderate reduction in precision in the submillimeter range.

The new algorithms of the second-generation QCA systems allow accurate and reliable measurements of small coronary dimensions and, therefore, precise analysis of coronary stenoses of moderate-to-high grade seems feasible with the improved accuracy of the new systems.