Non-Invasive Diagnostics For Cardiovascular Flows: Medical Applications*

The human cardiovascular system displays complex flow patterns, which become abnormal in disease states. Imaging methods such as ultrasound and phase-contrast MRI offer opportunities and present challenges in patient care when measuring these flows. This presentation will discuss advancements that enhance cardiovascular flow measurements for diagnosing and monitoring diseases. These techniques have been tested in pre-clinical and clinical studies, revealing correlations between flow pressure, shear, strain, energy losses, and disease conditions. We developed methods to improve the accuracy of our tools, addressing limitations such as image noise, resolution, and velocity reconstruction. We employ data augmentation techniques that adhere to flow physics principles, phase unwrapping, denoising methods, and bias error models. Moreover, we have created ways to evaluate relative pressure and wall shear stress and use data fusion to increase accuracy and resolution. Finally, we demonstrate these capabilities in clinical applications, including aneurysms, congenital heart disease, and heart failure. These advancements have the potential to improve diagnosis and patient management significantly.