Multiple Aneurysms AnaTomy CHallenge 2018 (MATCH): uncertainty quantification of geometric rupture risk parameters

Area of Science:

• Neurosurgery
• Medical Imaging
• Computational Anatomy

Background:

• Geometric parameters are used to predict cerebral aneurysm rupture risk.
• Accurate assessment of these parameters is influenced by imaging resolution and reconstruction methods.
• Investigating reconstruction variability is key for reliable rupture risk prediction in clinical settings.

Purpose of the Study:

• To quantify the uncertainty of geometric parameters used for cerebral aneurysm rupture risk assessment.
• To evaluate the impact of reconstruction procedures on the reliability of these parameters.

Main Methods:

• 26 research groups reconstructed five cerebral aneurysms.
• 40 dimensional and non-dimensional geometric parameters were computed.
• Uncertainties were calculated, and linear regression analysis was performed.

Main Results:

• Relative uncertainties varied widely, from 3.9% to 179.8%.
• Non-dimensional parameters (isoperimetric ratio, convexity ratio, ellipticity index) had the lowest uncertainties (<6%).
• Curvature parameters exhibited the highest uncertainties (>80%), while 1D parameters were more certain than 2D/3D size parameters.

Conclusions:

• Uncertainty quantification is vital for the clinical application of aneurysm rupture risk models.
• The study's findings on parameter uncertainty can guide the development of more robust predictive models.
• Standardizing reconstruction procedures is essential for improving the accuracy and reliability of geometric parameters in clinical practice.