Validation of a Mathematical Model for Rupture Status of Spherical Intracranial Aneurysms

Area of Science:

• Biomedical Engineering
• Medical Physics
• Computational Mechanics

Background:

• Accurate mathematical models of intracranial aneurysm (IA) mechanics are crucial for predicting rupture risk.
• A previously developed model for spherical IAs predicts rupture at a wall-thickness-to-IA-radius ratio (WTR) of 6.1 × 10⁻³.
• Clinical validation of this IA rupture model is lacking.

Purpose of the Study:

• To assess the accuracy and clinical utility of a mathematical model for spherical intracranial aneurysm (IA) rupture mechanics.
• To validate the model's prediction of IA rupture using clinical data.
• To evaluate the model's performance in discriminating between ruptured and unruptured IAs.

Main Methods:

• Radiologic images of 80 intracranial aneurysms (IAs) were analyzed to measure dimensions (height, width, neck diameter, vessel radius).
• Geometric modeling was employed to estimate IA wall thickness, calculating the wall-thickness-to-IA-radius ratio (WTR).
• Receiver operating characteristic (ROC) curves, likelihood ratios (LR), and logistic regression were used to analyze WTR, aspect ratio (AR), bottleneck factor (BF), and size ratio (SR) in relation to rupture status, with and without post-rupture dimensional adjustments.

Main Results:

• Analysis included 52 unruptured and 28 ruptured spherical IAs.
• ROC curve analysis showed comparable areas under the curve for WTR, AR, BF, and SR (0.636–0.773).
• Logistic regression indicated a strong association between decreased WTR and rupture. The 50% rupture probability WTR was calculated as 7.9 × 10⁻³ (without post-rupture adjustments) and 6.2 × 10⁻³ (with adjustments), closely aligning with the model's prediction.

Conclusions:

• The mathematical model for IA rupture mechanics demonstrates reasonable agreement with clinical data.
• The model effectively discriminates between ruptured and unruptured aneurysms, performing similarly to simpler parameters.
• Biomathematical models, despite simplifying assumptions, offer valuable insights into aneurysmal lesion behavior and rupture probability.