Porcine In Vivo Validation of a Virtual Contrast Model: The Influence of Contrast Agent Properties and Vessel Flow Rates.

T.W Peach,Y Ventikos,J.V Byrne,Z You

doi:10.3174/ajnr.a4884

Abstract

Accurately and efficiently modeling the transport of angiographic contrast currently offers the best method of verifying computational fluid dynamics simulations and, with it, progress toward the lofty goal of prediction of aneurysm treatment outcome a priori. This study specifically examines the influence of estimated flow rate and contrast properties on such in silico predictions of aneurysm contrast residence and decay. Four experimental sidewall aneurysms were created in swine, with aneurysm contrast flow patterns and decay rates observed under angiography. A simplified computational fluid dynamics model of the experimental aneurysm was constructed from 3D angiography and contrast residence predicted a priori. The relative influence of a number of estimated model parameters (contrast viscosity, contrast density, and blood flow rate) on contrast residence was then investigated with further simulations. Contrast infiltration and washout pattern were accurately predicted by the a priori computational fluid dynamics model; however, the contrast decay rate was underestimated by ∼25%. This error was attributed to the estimated parent vessel flow rate alone, and the effects of contrast viscosity and density on the decay rate were found to be inconsequential. A linear correlation between the parent vessel flow rate and the corresponding contrast decay rate was observed. In experimental sidewall aneurysms, contrast fluid properties (viscosity and density) were shown to have a negligible effect on variation in the modeled contrast decay rate. A strong linear correlation was observed between parent vessel flow rate and contrast decay over a physiologically reasonable range of flow rates.

Highlights

BACKGROUND AND PURPOSEAccurately and efficiently modeling the transport of angiographic contrast currently offers the best method of verifying computational fluid dynamics simulations and, with it, progress toward the lofty goal of prediction of aneurysm treatment outcome a priori
Contrast infiltration and washout pattern were accurately predicted by the a priori computational fluid dynamics model; the contrast decay rate was underestimated by ϳ25%
The complex hemodynamic environment in and around a cerebral aneurysm has been linked to the initiation, development, and possible rupture of the lesion.[1,2,3]

Summary

MATERIALS AND METHODS

Experimental Setup Experimental aneurysms were created in a porcine animal model by using 2 female Large White pigs of 60 – 65 kg and adopting the method described by German and Black.[18]. The equivalent mixture of blood and the contrast agent used experimentally (iohexol, Omnipaque 240; GE Healthcare, Piscataway, New Jersey) would result in an effective increase in the Re of 23% Both simulations (high Re and low Re) were run with the same parameters and postprocessed in the same manner as described in the previous section. It was assumed that the range of mean flow rates considered (115–345 mL/min) encompassed that likely to be seen experimentally Both simulations with altered flow rate (Q) (low Q and high Q) were run with solver settings and postprocessing procedures identical to simulations discussed previously. The velocity of the contrast-front was recorded for 2 consecutive 0.133second intervals in each experimental aneurysm, which resulted in 8 separate estimates of the instantaneous parent vessel flow rate

RESULTS

CONCLUSIONS

DISCUSSION

Study Limitations