|Authors||K. Valen-Sendstad, K. Mardal and D. A. Steinman|
|Title||Revisiting 'Turbulence' in Cerebral Aneurysms|
|Afilliation||Scientific Computing, , Scientific Computing|
|Project(s)||Center for Biomedical Computing (SFF)|
|Year of Publication||2012|
One of the commonly made assumptions in Computational Fluid Dynamics (CFD) is that cerebral blood flow is laminar, i.e., viscous forces are dominating and the flow is smooth. The numerical methods that are used, are chosen or tailored accordingly, i.e., to converge to such a laminar flow solution with the minimal amount of work1-2. The flow is modeled with a temporal resolution far below the clinically reported frequencies3 and thus any potential flow disturbances, if physically present, are being numerically suppressed. Since no potential disturbances are resolved, no disturbances are observed and the assumption of laminar flow is therefore a self-fulfilling prophecy. Even though recent laminar CFD studies can discriminate between ruptured and unruptured aneurysms better than, e.g. aneurysm size, the link between the proposed hemodynamic agonists and the mechanisms of rupture is unclear. However, the literature contains much evidence which supports that certain aneurysms exhibit energetic high frequency flow fluctuations. For example, high frequency flow fluctuations are known to produce sound, and clinicians have reported such sound or 'bruits' from aneurysms during craniotomy.3 The predominant frequencies were at 460 Hz, which is consistent with the energy peaks at higher frequencies recorded acoustically on the eyes in patients with aneurysms.4 Unstable flow has also been reported in glass model studies of aneurysms.5 This evidence seems to have been ignored, or at least has not received much attention in computational modeling studies. The goal of the present study was to investigate if such high-frequency flow fluctuations in intracranial aneurysms might be a common occurrence, by using thousands of time steps in contrast to, e.g., one hundred2.