Intravascular ultrasound, computed tomography, magnetic resonance imaging, and direct magnification radiography for in vitro evaluation of peripheral artery morphology and dimensions

Abstract

Objective: The aim of this in vitro study was to validate five imaging modalities (12.5- and 20-MHz intravascular ultrasound, computed tomography, spin- and gradient-echo magnetic resonance imaging) for their potential to depict peripheral vessel wall calcifications, to estimate calcified plaque area and to measure luminal areas using direct magnification radiography as a reference. Methods: Forty-three postmortem, human iliac segments were consecutively examined by intravascular ultrasound (12.5- and 20-MHz mechanical system), computed tomography and magnetic resonance (MR) imaging. In each specimen, the presence of calcified regions was assessed by independent observers. Luminal area and plaque area were quantitatively estimated. All specimens were finally examined using direct magnification radiography as a reference. Results: The sensitivity was 73% using a 20-MHz probe for the identification of calcification (vs. 59% with a 12.5-MHz transducer), specificity was 97% (with 12.5-MHz 96%). Positive predictive value with 20-MHz was 90% (with 12.5-MHz 91%); negative predictive value was 87% (81% with 12.5-MHz) and accuracy 87% with 20-MHz and 84% with 12.5-MHz. With magnetic resonance imaging, the sensitivity of the gradient-echo technique to identify calcified areas was 94% (vs. 75% with the spin-echo technique) and specificity was 86% (vs. 93% with spin-echo technique). Computed tomography reached a sensitivity level of 88%, specificity was 88%. No significant difference was found for luminal area estimated by 20-MHz intravascular ultrasound and direct magnification radiography. Computed tomography, spin-echo MR and gradient-echo MR were significantly different from reference measurement ( p < 0.005). Plaque area estimated by direct magnification was 8.3 ± 6.5 mm 2 (mean and standard deviation); computed tomography, spin-echo MR and gradient-echo MR showed significant overestimation ( p < 0.008) in comparison to the reference. The correct identification of regional vessel wall calcifications by intravascular ultrasound strongly depended on both morphology and plaque thickness. Conclusion: In an in vitro setting, intravascular ultrasound, computed tomography and MR imaging of peripheral vessel wall calcifications may lead to misinterpretation.