Advanced three-dimensional quantitative coronary angiographic assessment of bifurcation lesions: methodology and phantom validation

Abstract

Aims: Validation of new three-dimensional (3-D) bifurcation quantitative coronary angiography (QCA) software.

Methods and results: Cardiovascular Angiography Analysis System (CAAS 5v10) allows 3-D angiographic reconstructions based on two or more 2-D projection images. Measurements for minimal lumen diameter (MLD), reference vessel diameter (RVD), percent diameter stenosis (DS) and bifurcation angle (BA) were validated against precision manufactured phantom bifurcations. Length measurements were validated against angiographic measurement catheters inserted into a plexiglas bifurcation phantom. In 3-D reconstructions based on two 2-D images, acquired at variable rotation and angulation, accuracy and precision (mean difference ± SD) of the 11-segment model for MLD, RVD and DS were 0.013±0.131 mm, -0.052±0.039 mm and -1.08±5.13%, respectively; inter-observer variability was 0.141 mm, 0.058 mm and 5.42%, respectively. Adding the antero-posterior (optimal) projection to these basic reconstructions resulted in reduced variability (0.101 mm, 0.041 mm and 3.93% for MLD, RVD and DS, p<0.01 for all) and showed a trend towards improved precision (0.109 mm, 0.031 mm and 4.26%, respectively, p>0.05 for all). In basic reconstructions, accuracy and precision for BA was -1.3±5.0°, whereas inter-observer variability was 7.5°; respective measures for length were 0.15±0.26 mm and 0.54 mm. Adding the antero-posterior projection resulted in decreased precision (0.47 mm, p<0.01) and increased variability (1.03 mm, p<0.01) for length measurements; precision (5.4°) and variability (7.9°) for BA did not change significantly (p>0.30).

Conclusions: Advances in the methodology of 3-D reconstruction and quantitative analysis for bifurcation lesions translated into highly accurate, precise and reproducible measures of diameter, length and BA.