Quantitative assessment of coronary atherosclerotic plaque profile by morphometric analysis of angiographic images.

Coronary plaque morphology (simple or complex) has a recognized clinical, pathophysiologic, and prognostic importance, but its routine use is hampered by the subjective and qualitative nature of the assessment. The aim of this study was to assess the feasibility and accuracy of mathematical, objective, operator-independent assessment of plaque morphology, by means of algorithms based on vector and fractal analysis, previously developed for describing continuous curves. Twenty coronary angiograms were analyzed: 10 with "simple-type" stenosis (defined according to Ambrose type I pattern, Group I) and 10 with "complex-type" stenosis (Ambrose type II pattern, Group II). Five morphometric parameters effectively separated simple and complex-type stenosis: peaks/cm (Group I = 0.24 +/- 0.31 vs Group II = 2.46 +/- 1.83, P < .01); summed maximum error/cm (I = 0.94 +/- 1.19 vs II = 4.53 +/- 0.56, P < .01); integrated error/cm (I = 0.22 +/- 0.27 vs II = 2.25 +/- 1.68, P < .01); number of features/cm (I = 0.02 +/- 0.05 vs II = 1.19 +/- 0.65, P < .01); standard deviation curvature (I = 0.07 +/- 0.01 vs II = 0.22 +/- 0.21, P = 0.06). The scaled edge-length ratio-the only parameter based on fractal analysis-was not different between the two groups (I = 1.03 +/- 0.01 vs II = 1.13 +/- 0.10, P = ns). Hence, a quantitative description of plaque morphology is feasible and allows an effective discrimination of simple and complex plaque profiles with morphometric parameters based on vector analysis.