Research / Computational methods / 06 / 06

Uncertainty quantification

A model used for diagnosis has to state how much its answer moves when its inputs are uncertain, and which inputs are responsible.

Sensitivity analysis
Global sensitivity of a computed diagnostic index to physiological input parameters.

What moves the answer

We applied global sensitivity analysis to a calibrated one-dimensional coronary model under realistic pulsatile conditions. Intramyocardial pressure, myocardial contractility and systolic duration dominate the variation in computed fractional flow reserve; most other physiological parameters contribute little.

The index is otherwise robust, but its scatter grows systematically as the narrowing becomes more severe — which is the regime where the clinical decision is closest to the threshold. Geometry, boundary conditions and blood viscosity each contribute in different proportions depending on the vessel.

Selected work

Under review

Uncertainty quantification of fractional flow reserve under pulsatile physiology

Annals of Biomedical Engineering

2016

Uncertainty quantification in coronary blood flow simulations: impact of geometry, boundary conditions and blood viscosity

Journal of Biomechanics 49(12):2540–2547