Artikel
Functional characterization of cerebral aneurysms. Influence of data acquisition by means of different imaging modalities
Suche in Medline nach
Autoren
Veröffentlicht: | 20. Mai 2009 |
---|
Gliederung
Text
Objective: Functional characterization of cerebral aneurysm by analysing biofluidmechanical parameters may help to perform patient-specific aneurysm rupture risk analysis and facilitate treatment decisions. Combined computational fluid dynamics (CFD) with non-invasive 3D imaging methods provide a means for quantitative analysis of biofluidmechanical parameters (wall shear stress (WSS), aneurysm size, surface curvature radius) in vivo. However, this analysis depends on the imaging method used which thereby may alter the assessment of the geometrical and hemodynamic parameters.
Methods: This study compares geometry and calculated WSS values based on magnetic resonance imaging (MRI), computed tomography (CT) and 3D-rotational angiography (3DRA). All imaging methods were applied to a threefold upscaled silicon model of a basilar tip aneurysm. This model served as the geometrical gold standard. The use of the upscaled model made it possible to generate a high resolution raw data set from CT with a voxel size of 0.13x0.13x0.3 mm3, from MRI with a voxel size of 0.16x0.16x0.16 mm3, and from 3DRA with a voxel size of 0.408x0.408x0.408 mm3.
Results: Comparing surface (Hausdorff) distances to the original geometry, MRI reconstruction with 0.097±0.07 mm (mean ± sd) is 2.5 times better than the CT based reconstruction with 0.242±0.131 mm and similar to 3DRA reconstruction with 0.10±0.086 mm. Analysis of curvature radius on the aneurysm surface yielded values (mean ± sd) of 5.08±2.44, 4.84±1.75. 4.93±1.65, and 4.82±2.39 for the original geometry, MRI reconstruction, CT reconstruction, and 3DRA reconstruction respectively. Calculated values for mean WSS revealed relatively large differences between the gold standard and both non-invasive imaging modalities. The mean WSS was 24% lower than the gold standard for the CT-based geometry and 10% lower for the MR-based geometry. The 3DRA based geometry resulted in a mean error of 3% for mean WSS. Nevertheless, the distribution of WSS showed similar patterns for all three imaging modalities.
Conclusions: All imaging modalities can be used to generate very similar distribution patterns of functional parameters in aneurysms. 3DRA is the best modality. Being readily available and non-invasive, CT-angiography is still the modality of choice in functional characterization of cerebral aneurysms.