In vivo assessment of a digital angiographic method to measure absolute coronary artery diameters

Several techniques exist for the quantification of absolute coronary artery diameters using radiologic methods. An in vivo assessment of a quantitative technique based on direct digitally acquired images was performed by imaging inflated angioplasty balloons (n = 25), balloon catheter shafts (n = 16) and coronary guidewires (n = 20) at the time of coronary angioplasty. After this, the actual size of the objects was determined with a micrometer. Diameters measured by the quantitative digital method had an excellent correlation with the actual diameters (digital diameter = 0.80 [actual diameter] + 0.32; n = 61; r = 0.97; standard error of the estimate = 0.26 mm; p less than 0.001). Moreover, the correlation between interobserver and intraobserver measurements was excellent (r = 0.99 for both, standard error of the estimate = 0.16 mm and 0.09 mm, respectively). However, there was a consistent error present that was related to the size of the object measured. Objects less than 0.5 mm were consistently overestimated and objects greater than 1 mm were usually underestimated by the digital technique, although the actual magnitude of the error was small. Objects less than 0.5 mm in diameter were overestimated by 0.41 +/- 0.11 mm and objects greater than 1 mm were underestimated by 0.23 +/- 0.19 mm. Based on an analysis of the error present, correction algorithms were formulated and tested prospectively using an additional 29 object measurements. This resulted in an improvement in the quantification of the diameters with a smaller magnitude of error. This in vivo assessment suggests that the rapid online assessment of absolute coronary artery diameters is possible, but also demonstrates important errors inherent in this method.