Angiographic views used for percutaneous coronary interventions: a three-dimensional analysis of physician-determined vs. computer-generated views.

The goal of this study was to determine the severity of vessel foreshortening in standard angiographic views used during percutaneous coronary intervention (PCI). Coronary angiography is limited by its two-dimensional (2D) representation of three-dimensional (3D) structures. Vessel foreshortening in angiographic images may cause errors in the assessment of lesions or the selection and placement of stents. To date, no technique has existed to quantify these 2D limitations or the performance of physicians in selecting angiographic views. Stent deployment was performed in 156 vessel segments in 149 patients. Using 3D reconstruction models of each patient's coronary tree, vessel foreshortening was measured in the actual working view used for stent deployment. A computer-generated optimal view was then identified for each vessel segment and compared to the working view. Vessel foreshortening ranged from 0 to 50% in the 156 working views used for stent deployment and varied by coronary artery and by vessel segment within each artery. In general, views of the mid circumflex artery were the most foreshortened and views of the right coronary artery were the least foreshortened. Expert-recommended views frequently resulted in more foreshortening than computer-generated optimal views, which had only 0.5% +/- 1.2% foreshortening with < 2% overlap for the same 156 segments. Optimal views differed from the operator-selected working views by > or = 10 degrees in over 90% of vessels and frequently occurred in entirely different imaging quadrants. Vessel foreshortening occurs frequently in standard angiographic projections during stent deployment. If unrecognized by the operator, vessel foreshortening may result in suboptimal clinical results. Modifications to expert-recommended views using 3D reconstruction may improve visualization and the accuracy of stent deployment. These results highlight the limitations of 2D angiography and support the development of real-time 3D techniques to improve visualization during PCI.