Biomedical Engineering Reference
In-Depth Information
4 Discussion
This work deals with the problem of haemodynamics modelling and visualisation
in an anatomically realistic arterial tree. We aim to use a 1D model to capture the
main haemodynamic features that might be useful in endovascular interventions.
To that end we modeled a particular clinical procedure, i.e. balloon occlusion test of
a VA, and presented the preliminary results. We showed that the perfusion of an
(ischemic) brain region may be compensated via the collateral arteries.
However, the fact that each individual may have vascular variations deviating
from the “standard” or “generic” morphology demands that each procedure should
be treated individually. Furthermore, the haemodynamic effects of endo-devices
may be transient (such as temporary balloon occlusion) or permanent (such as
complete occlusion of a saccular aneurysm using coils). Thus, a computer model
must be made patient specific to be applied to a clinical scenario.
As pointed out in the Introduction section, a flow solver should be highly
efficient to be clinically relevant. Among the various blood flow modelling
techniques, a 1D model can achieve a high computational efficiency as well as a
reasonable geometric accuracy. However, the hyperbolic nature of the governing
equations makes the 1D model prone to numerical oscillations. Also, the computa-
tion time (i.e., minutes for a cardiac cycle) of the 1D model presented in this work is
not suitable for real-time applications. To that end a distributed model such as the
one proposed in [
13
] may serve the purpose. The cost of the computing efficiency of
model [
13
], however, is that the spatial variations along a vessel are ignored.
Nevertheless, the presented 1D model is a powerful tool in the scenarios where
computational time is not critical but the vasculature is too large for 3D models.
Moreover, we showed that the presented vascular construction method can be
quickly adapted to a variety of vascular anatomies, hence may aid haemodynamic
analysis in vascular interventions after rigorous validation.
5 Conclusion
In this work we introduced a computational approach to model blood flow in an
anatomically accurate arterial tree that spans the whole range of arteries involved in
neuro-interventions. Blood flow modelling was performed by using a 1D equation
system. The influence of endovascular devices on the arterial system was analyzed
by changing the configuration of the arterial tree. Some preliminary results such as
the flow distribution in the arterial tree and the simulation of a balloon-occlusion
procedure were presented. This numerical technique can be a powerful tool for
computing and visualizing blood flow in a variety of vascular structures.
Acknowledgment This project was partly funded by @neurIST, a European aneurysm project,
which we gratefully acknowledge.
