Reproducibility of image-based analysis of cerebral aneurysm geometry and hemodynamics: An in-vitro study of magnetic resonance imaging, computed tomography, and three-dimensional rotational angiography

摘要

Background and Study Aims Image-based computational fluid dynamics (CFD) provides a means for analysis of biofluid mechanical parameters of cerebral aneurysms. This may enable patient-specific rupture risk analysis and facilitate treatment decisions. Application of different imaging methods may, however, alter the geometrical basis of these studies. The present study compares geometry and hemodynamics of an aneurysm phantom model acquired by means of magnetic resonance imaging (MRI), computed tomography (CT), and rotational angiography (3DRA). Materials and Methods The phantom model of a basilaris artery aneurysm was fabricated based on data generated by CT angiography. This model underwent imaging by means of CT, MRI, and 3DRA. We compared the geometrical reconstructions using the original dataset with those obtained from CT, MRI, and 3DRA. Similarly, CFD analyses were performed using the four reconstructions (3DRA, MRI, CT, and original dataset). Results MRI and the 3DRA-based reconstructions yield mean reconstruction errors of 0.097 mm and 0.1 mm, which are by a factor of 2.5 better than the CT reconstruction. The maximal error for the aneurysm radius (7.11 mm) measurement was found in the 3DRA reconstruction and was 3.8% (0.28 mm). A comparison of calculated time-averaged wall shear stress (WSS) shows good correlations for the entire surface and, separately, for the surface of the aneurysmal sack. The maximal error of 8% of the mean WSS calculation of the whole surface was found for the CT reconstruction. The calculations of the aneurysmal sack mean WSS from the MRI reconstruction were estimated to have a maximal error of 7%. Conclusion All three imaging techniques (CT, MRI, 3DRA) adequately reproduce aneurysm geometry and allow meaningful CFD analyses.