Comparison between routine cylindrical cerebral aneurysm volume approximation and three-dimensional volume measurements in experimental aneurysms

Aneurysm volume is routinely approximated calculating cylindrical volumes. Exact aneurysm volume assessment is crucial for liquid polymer embolization. The aim of this study was to compare simple cylindrical volume approximations with direct multiplanar reconstruction (MPR) segmentational volumetry in a saccular/complex experimental rabbit bifurcation aneurysm model.

In 12 female New Zealand white rabbits, saccular, broad-based, bilobular, and bisaccular aneurysms (three of each) were created using the rabbit venous pouch bifurcation model. Contrast-enhanced magnetic resonance angiography (CE-MRA) was performed, and maximal intensity projection (MIP) reconstructions as well as an MPR dataset were acquired. Aneurysm width and length were measured in MIP images, and the volume was approximated calculating cylindrical volumes. Three-dimensional (3D) segmentational volumetry using the MPR dataset was performed in a semi-automated manner.

Maximal intensity projection cylindrical volumes ranged from 53·6 to 503·5 mm(3) (mean 186·5±118 mm(3)). Multiplanar reconstruction segmentation-based volumes ranged from 74·7 to 581·0 mm(3) (mean 202·2±133 mm(3)). The mean relative difference between MIP cylindrical and MPR segmentation volume calculation was 24·7% (range -77·5 to +50·8%). Only 4 of 12 MPR segmentational volumes were within a 10% range of results calculated for MIP cylindrical volume, and 3 of those were in broad-based aneurysms.

This descriptive study demonstrates that estimated MIP cylindrical volumes differ from those measured by MPR segmentation volumetry. With the increasing acquisition of 3D data as 3D-MRA and the increasing need for exact volume determination, studies on the accuracy of computational segmentational volumetry of CE-MRA are necessary.