Patient=Specific Modeling and Uncertainty Quantification using a Data-Driven Surrogate Model

Abstract

A patient-specific simulation has been exapnding its horizon to clinical applications. Particulary in the field of computational hemodynamics, it is an effective way to grasp blood flow condtions for an indivudal patient. Further advancement of the patient-specific simlauiton can be acheived by predinction of the blood flow after the srugery. The meical data used for the pateinet-specifc simulation contain uncertainties, which affect the simulation results by propagating through mathematical models and the simulation. Thus, quantifying an impact of uncertainties in medical images on simulated quantities is an essential task to obtain reliable results. Since uncertainty quantification requires a large number of case studies to investigate the effects of uncertainties in a probabilistic manner, a surrogate model was developed using a machine-learning techque with data driven by the blood flow simulation. In the session, uncertainty quqntification is presented by the risk predcition of cerebral hyperperfusion syndrome (CHS) with exmples of acutal patient cases.