Infants with single ventricle physiology generally undergo three palliative surgeries starting with stage-one, in which a systemic-to-pulmonary connection is established via a shunt. Mortality is the highest among stage-one patients (up to 23%) due to sub-optimal oxygen delivery, ventricle volume overload, myocardial ischemia, and high risk of shunt blockage. The clinical objective of the present study is to simulate the stage-one circulation, analyze possible surgical options, optimize current surgical methods, and explore a novel alternative surgical option. Simulating the stage-one circulation in single ventricle repair requires a set of numerical tools that are developed in the first part of this dissertation. First, an implicit and modular multidomain framework with excellent stability and convergence properties is introduced that allows multiscale simulation of the circulatory system. Second, a stabilized formulation is presented for treating backflow at Neumann boundaries that is inexpensive, stable, simple, and minimally intrusive, and offers a promising alternative to previous methods. Third, an efficient pre-conditioner for coupled boundary conditions and an efficient iterative algorithm for solving system of equations governing incompressible flows are introduced. Fourth, a scalable parallel data structure is introduced for performing algebraic operations in iterative solvers efficiently. Fifth, an Eulerian formulation is proposed for calculating residence time that lacks mesh dependency and avoids the high computational cost of Lagrangian particle-based approaches. These tools are applicable to other cardiac mechanics and CFD simulations as well. In second part of this dissertation, single ventricle physiology is studied using the tools presented in the first part. First, a multiscale model of single ventricle physiology is simulated and the shunt geometry is optimized to maximize oxygen delivery and improve performance. Second, surgical scenarios single and multiple systemic-to-pulmonary connections are compared, revealing higher thrombotic risk and lower oxygen delivery in the presence of multiple connections. Third, a novel stage one palliative surgery, which provides an alternative source of blood flow in case of shunt blockage and may ultimately reduce the number of open chest surgeries from three to two, is proposed and tested using multiscale modeling. Results reveal the proposed surgical method, the Assisted Bidirectional Glenn, can deliver more oxygen at a reduced heart load with only a modest increase in venous return pressure

Kawasaki disease (KD) is the leading cause of acquired heart disease in children and can result in life-threatening coronary artery aneurysms in up to 25 % of patients. These aneurysms put patients at risk of thrombus formation, myocardial infarction, and sudden death. Clinicians must therefore decide which patients should be treated with anticoagulant medication, and/or surgical or percutaneous intervention. Current recommendations regarding initiation of anticoagulant therapy are based on anatomy alone with historical data suggesting that patients with aneurysms [Formula: see text]8 mm are at greatest risk of thrombosis. Given the multitude of variables that influence thrombus formation, we postulated that hemodynamic data derived from patient-specific simulations would more accurately predict risk of thrombosis than maximum diameter alone. Patient-specific blood flow simulations were performed on five KD patients with aneurysms and one KD patient with normal coronary arteries. Key hemodynamic and geometric parameters, including wall shear stress, particle residence time, and shape indices, were extracted from the models and simulations and compared with clinical outcomes. Preliminary fluid structure interaction simulations with radial expansion were performed, revealing modest differences in wall shear stress compared to the rigid wall case. Simulations provide compelling evidence that hemodynamic parameters may be a more accurate predictor of thrombotic risk than aneurysm diameter alone and motivate the need for follow-up studies with a larger cohort. These results suggest that a clinical index incorporating hemodynamic information be used in the future to select patients for anticoagulant therapy.