Global optimization strategies to simplify cardiovascular diagnostics

Designing and leveraging ‘high throughput’ aspects of efficient engineering models can increase their relevance in understanding biological systems. This thesis explores the use of design of experiments, optimization, and perturbation analysis to fit models to two complex cardiovascular problems: one computational (Case A: Living Heart Project®) and one experimental (Case B: FlowMAP™). A) The Living Heart Project® considers the physiological optimization and calibration of an electro-mechanical computational heart model designed for virtual patient testing. Given several interdependent parameters that interact nonlinearly, understanding how to best fit the model to specific individuals and disease states is challenging. This thesis discusses strategies of calibration and considers methods of increasing software and hardware efficiency to enable efficient, multivariate optimization for complex models of living systems. B) FlowMAP™ is a platform technology developed in a research setting that allows detailed measurement of an individual´s clotting and bleeding risk by constructing an experimentally derived ´model´ of an individual´s blood state. It is a microfluidic-based technology that, having completed preliminary laboratory testing, now requires translation for clinical diagnostic applications such as improved, personalized management of conditions such as heart attack and stroke. This study describes the manufacturing and scale up of the technology, to permit precisely controlled measurement of blood clotting and enable assessment and prediction of a patient’s clotting and bleeding risk. While each scenario is distinct, together they demonstrate a relevant cross-functional work path that combines computational and experimental models to reduce the translational gap that exists between research environments, market application, and clinical viability