Assessment of a Hyperelastic Model for the Mechanical Characterisation of Intracranial Aneurysm Walls
摘要
Intracranial aneurysms arise from the progressive deterioration of the biomechanical integrity of arterial walls. Currently, there is no available method that can predict the risk of aneurysm wall rupture based on a quantitative analysis of its mechanical characteristics. This work forms part of a broader research project aimed at developing a personalised tool to support clinicians. This tool will help improve the management of unruptured intracranial aneurysms through knowledge of the in vivo mechanical properties of patients’ aneurysmal tissue. In line with this objective, an original arterial wall deformation device was designed and tested using a non-destructive experimental method on polymer-based arterial phantoms. In parallel, a numerical model was developed to complement the experimental study and to deepen the understanding of the interaction between the deformation device and the aneurysm wall.
ObjectiveAn initial assessment, carried out using a simplified numerical model and compared with experimental results, has already been conducted. The current objective is to increase the biofidelity of the model and to further evaluate this framework.
MethodsThe deformation prescribed on the polymer phantom arteries by the device was measured using Digital Image Correlation. The fluid–structure interaction between the device, the fluids and the arterial wall was simulated using the Finite Element Method. The model was made more realistic through the use of a geometrically biofidelic artery, a fluid reproducing blood viscosity, and hyperelastic parameters for the vascular wall properties. The evaluation process involved extracting and interpolating the numerical results, and then comparing the computed strain and stress fields with the experimental measurements.
ResultsThe strain fields obtained from the linear elastic and hyperelastic models were compared with experimental data for validation. Analysis of the stress fields, however, revealed that the hyperelastic model allowed a more accurate validation of the numerical model. The maximum error associated with the maximum stress reaches 27.7 % for the linear elastic model, compared to only 1.7 % for the hyperelastic model.
ConclusionsThe reliability of the model was confirmed on several artery geometries. It could be further applied to more complex experimental setups, especially by including a realistic pulsatile blood flow.