Skip directly to content

How Specific is Enough?: Computational Sensitivity Analysis for Patient-Specific Medicine

NSF Award:

CAREER: Quantifying and Controlling Error and Uncertainty in Computational Inverse Problems  (University of Utah)

Congressional Districts:
Research Areas:

Advances in computational hardware and medical devices have prompted an interest in the medical community for patient-specific quantitative investigation of disease, with the goal of increasing efficacy of treatment. The accessibility of patient-specific data has aided both in the improved understanding of pathophysiology and in developing effective disease treatments. Researchers hope that numerical simulation of biological systems will be able to illustrate a connection between patient-specific data and the understanding of disease, and they have focused on heart disorders for their investigation.

Computational simulations are capable of solving complex mathematical systems with high numerical accuracy. With such accuracy, the confidence in these computational approximations has increased over the years. Researchers can apply quantitative methods to approximate physical systems. The sensitivity of the system must take into account uncertainties in data (i.e. initial conditions, boundary conditions, geometry), which has led researchers to wonder, how specific must the data be in order to achieve both diagnostic and prognostic medical benefit?

In the driving application area of this study, electrophysiology, medical techniques currently used are not able to non-invasively pinpoint the specific location within the heart that is functioning abnormally.  Using quantitative computational systems, researchers hope to develop a localization procedure that is more accurate and can take into account uncertainty and variability. In the case of electrocardiography, the simulation must account for variability in organ conductivity and heart position in the patient.

Researchers have adapted mathematical and computational techniques that have been applied to bioengineering. By collaborating with bioengineering experts they have been able to better quantify the levels of uncertainty that occur in computational simulations. Their research brings them one step closer to knowing how specific the data must be to achieve the best diagnostic results.


  • Image Currently Not Available
Permission Granted

Recent Award Highlights

the grain weevil maintains a symbiotic relationship with bacteria

The origins of insect bacteria

Life inside an insect makes some microbes switch to a mutualistic relationship with their host

Research Areas: Biology Locations: Utah
Diagram of wiring within a large cache bank

Lowering Energy Consumption of Large Data Caches

New techniques to reduce the energy consumption associated with computer caches

Research Areas: Computing, Earth & Environment Locations: Utah