Vessel size estimation in peripheral artery interventions: are angiographic measurements reliable?

To describe and evaluate a new technique for angiographic measurement of arterial diameters and to make a comparison with other common techniques.

The Angiometer has a radiopaque metal body with six cylindrical segments of varying diameter. Arterial diameter is estimated by visually relating the vessel to these segments. The measurements (method A) were compared: to measuring in relation to a ruler on the table-top by means of compasses (B), to an automated measurement technique (quantitative angiography) utilizing calibration to a catheter of known diameter (C), and to a computer-based distance measurement using the ruler on the table-top as reference (D). Twenty-five patients were studied. Each method was used twice by each of two independent observers. A phantom with four drilled holes, filled with contrast agent, was also studied.

The coefficient of variation was highest for method C and lowest for method A. Correlations between readings were highest for method A and lowest for method C. Mean diameter values were highest for method A and lowest for methods B and D. In the phantom experiments, sizes were overestimated by between 5% and 9% with method A, and underestimated by between 4% and 27% with the other methods.

In terms of reproducibility and accuracy, the proposed method compares favorably with alternative methods.