Accuracy of pre-interventional computed tomography angiography post-processing software and extravascularly calibrated devices to determine vessel diameters: comparison with an intravascularly located calibrated catheter

Background: Accurate vessel sizing might affect treatment outcome of endovascular therapy. Purpose: To compare accuracy of peripheral vessel diameter measurements using pre-interventional computed tomography angiography post processing software (CTA-PPS) and extravascularly located calibrated devices used during digital subtraction angiography (DSA) with an intravascular scaled catheter (SC). Material and Methods: In 33 patients (28 men, mean age = 72 ± 11 years) a SC was used during DSA of the femoro-popliteal territory. Simultaneously, one scaled radiopaque tape (SRT) was affixed to the lateral thigh, one scaled radiopaque ruler (SRR) was positioned on the angiography table. For each patient, diameters of five anatomic landmarks were measured on DSA images after calibration using different scaled devices and CTA-PPS. Diameters were compared to SC (reference) and between groups of non-obese (NOB) and obese (OB) patients. Results: In total, 660 measurements were performed. Compared to the reference, SRT overestimated the diameter by 1.2% (range = -10–12, standard deviation [SD] = 4.1%, intraclass correlation coefficient [ICC] = 0.992, 95% confidence interval [CI] = 0.989–0.992, P = 0.01), the SRR and CTA-PPS underestimated it by 21.3% (range = 1–47, SD = 9.4%, ICC = 0.864, 95% CI = 0.11–0.963, P = 0.08) and 3.2% (range = 17–38, SD = 9.7%, ICC = 0.976, 95% CI = 0.964–0.983, P = 0.01), respectively. Underestimation using the SRR was greatest in the proximal superficial-femoral artery (31%) and lowest at the P2 level of the popliteal artery (15%). In the NOB group, diameter overestimation of the SRT was 0.8% (range = 4–7, SD = 4.2%, B = 0.071, 95% CI = 0.293–0.435, P = 0.08) compared to the OB group of 1.6% (range = −7–4, SD = 2.9%, B = 0.010, 95% CI = 0.474–0.454, P = 0.96). Diameter underestimation of the SRR was 17.3% (range = 13–21, SD = 3.1%, B = 0.946, 95% CI = 0.486–1.405, P = 0.002) in the NOB group, 23.3% (range = 11–36, SD = 6.6%, B = 0.870, 95% CI = 0.268–1.472, P = 0.007) in the OB group. Conclusion: For calibrated measurements SRT and CTA-PPS prove accurate compared to the reference, while SRR does not. Obesity has a significant impact on underestimation of diameter if SRR is used.