Quantification of absolute luminal diameter by computer-analyzed digital subtraction angiography: an assessment in human coronary arteries.

Determination of absolute lumen diameters has been shown to be useful in predicting the functional importance of a coronary stenosis. In this study, both single-plane and orthogonal biplane digital subtraction angiograms were obtained in human cadaver coronary arteries. A single absolute diameter was calculated at the site of greatest narrowing in 20 segments by two automated computerized algorithms. Minimum and maximum diameters at the site of the stenosis were measured from pathologic sections prepared after pressure fixation. Method 1, which determines the edges by means of the first derivative of the videodensity curve, derived absolute diameters that fell between the pathologic minimum and maximum in 10 of 20 segments. Method 2, which determines the edges by an average of the first and second derivatives of the videodensity change, derived absolute diameters that fell between the pathologic minimum and maximum diameters in 15 of 20 segments. Method 1 correlated well with the maximum pathologic diameter (r = .76) and less well with the minrmum pathologic diameter (r = .67). Method 2 correlated very well with the maximum pathologic diameter (r = .79) and also correlated well with the minimum pathologic diameter (r = .74). As would be expected, the computerized algorithms tended to overestimate the minimum pathologic diameter and to underestimate the maximum pathologic diameter. In six segments, two orthogonal views were analyzed; no further accuracy was discernible over single-plane determinations. Thus quantitative coronary angiography by digital subtraction angiography is sufficiently accurate to be of use in the measurement of the severity of a coronary stenosis.